Donate
Independent, objective, nonpartisan research

R 796DRR

Authors

R 796DRR

Tagged with:

Publication PDFs

Database

This is the content currently stored in the post and postmeta tables.

View live version

object(Timber\Post)#3742 (44) { ["ImageClass"]=> string(12) "Timber\Image" ["PostClass"]=> string(11) "Timber\Post" ["TermClass"]=> string(11) "Timber\Term" ["object_type"]=> string(4) "post" ["custom"]=> array(5) { ["_wp_attached_file"]=> string(12) "R_796DRR.pdf" ["wpmf_size"]=> string(6) "487956" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(231434) "The Distribution of Income in California Deborah Reed, Melissa Glenn Haber, Laura Mameesh July 1996 Copyright © 1996 Public Policy Institute of California, San Francisco, CA. All rights reserved. PPIC permits short sections of text, not to exceed three paragraphs, to be quoted without written permission, provided that full attribution is given to the source and the above copyright notice is included. Foreword This report on income distribution is the first research publication of the Public Policy Institute of California (PPIC). In developing the initial research agenda for the institute, we focused on fundamental changes that are sweeping the state. Of the many possibilities, one area that clearly deserves a place on our list is the dramatically changing nature of the state’s economy. California has emerged from its deepest recession since the 1930s. The economy is expanding steadily, with hundreds of thousands of jobs being created annually. In the bloom of this recovery, it makes sense to step back and look at the changes in the distribution of income in California that have occurred in recent years and over the last several decades. Recent efforts to measure and explain changes in income distribution in the United States have generated considerable interest and debate. This report is the first in a series that will replicate for California much of iii the national-level analysis. The authors document, for the first time, the annual changes in income distribution in the state from the late 1960s through 1994. Subsequent reports will examine the causes of increasing inequality, exploring the role of such factors as technological change, international competition, immigration, deunionization, and the shifting demographics of California. We trust that these reports not only will improve our understanding of the economic changes under way in California but will signal PPIC’s commitment to high-quality research and analyses useful to policy audiences. The authors express their appreciation to Sheldon Danziger from the University of Michigan and Lynn Karoly from RAND for their timely and extensive comments on an earlier draft. Lori Dair and John Ellwood were essential to the production of this report. Patricia Bedrosian, Jerry Lubenow, and Joyce Peterson provided considerable editorial assistance. The study has benefited from the efforts of Janet DeLand, Rod Pedersen, Eileen Roush, Peg Schumacher, Michael Shires, Karen Steeber, Michael Teitz, Paul Tractenberg, and numerous colleagues at PPIC. While this report reflects the contributions of many people, the authors are solely responsible for its content. David W. Lyon President and CEO Public Policy Institute of California iv Summary In recent years, increasing inequality in the distribution of income has been a subject of considerable public concern, political attention, and academic research. Income inequality is a measure of how equally the income pie is divided among all members of society. In other words, it is a measure of relative income, gauging, for example, how well the poor are doing economically compared with the rich. In the United States, income inequality remained stable in the three decades that followed World War II, as rich and poor alike benefited from the nation’s growing affluence. By the 1960s, Americans had come to accept as an article of faith President John Kennedy’s assertion that a rising tide would lift all boats. However, since the early 1970s the gap separating the rich and the poor has grown wider. While national studies have documented a growth in income inequality throughout the 1970s and 1980s, relatively little research has been done on income distribution in California. Such research is crucial v to the reasoned resolution of a broad range of state issues such as tax policy, public education, the minimum wage, and welfare reform that both affect and are affected by the distribution of income. The well-being of California’s population is a major research theme of the Public Policy Institute of California. This report is the first in a series that aims to identify state-specific policy strategies to promote equity as well as growth in the state’s economy. This initial study documents trends in income distribution in California from 1967 to 1994 and compares them to trends in other states, other regions, and the nation as a whole. Successive studies will investigate the underlying causes of the trends and will examine the relationship between public policy and the distribution of income. In this summary, we discuss the major findings of the study that, we believe, will be of interest to general and policy audiences concerned with important state issues. The body of the report and the appendices describe in greater detail the study’s results, approaches, measures, and data sources. We have striven to make the discussion in all parts of the report accessible to all interested audiences. Summing Up the Picture of California Income Inequality Income inequality has increased steadily in California over the last three decades. Until the late 1980s, the trend in California was remarkably similar to the national trend but, since then, inequality has risen much faster in the state than in the nation. This change has held for adjusted household incomes and for male earnings but not for female earnings. vi In both California and the nation, the increasing inequality results from income growth at the top of the distribution and decline in incomes at the very bottom. However, the recent divergence in inequality trends between California and the nation does not arise from faster growth at the top in California: In fact, income growth at all levels has been slower in California. Instead, the greater increase in the state results from a precipitous drop in income at the mid-to-lowest levels of the distribution. Rapid growth in income inequality has coincided with business cycle recessions, with those at the lower levels especially hard hit during recessions. A crucial difference has been that in the nation at large, incomes of people at those levels rebounded more during business cycle upswings than they did in California. However, the inequality gap between the nation and California began to widen as early as 1987, even before the recent, deep recession. These results suggest that in the interest of equity and economic growth in the state, it is essential that future research identify the forces that have made people at the lower end of the distribution lose so much ground and examine what happened in California even before the most recent recession. More on the Study’s Major Findings In this study, we used five summary measures of inequality, 26 definitions of income, and two data series (the Current Population Survey and the Census) to analyze California income levels and trends and to compare them with national and regional levels and trends. Our major findings are summarized below. vii Income Inequality Has Increased Substantially in California Figure S.1 illustrates how much the distribution of annual earnings has widened among male workers in California. The middle line of the graph shows the percentage change in real, inflation-adjusted median male earnings since 1967. The lower line of the figure shows the decline of male earnings at the 20th percentile, the income level that separates the bottom 20 percent of earners from the top 80 percent. The upper line of the figure shows the growing earnings at the 80th percentile. As shown in Figure S.1, the median of male earnings fell 20 percent between 1967 and 1994. This 20 percent decline represents a drop in median male earnings from $31,252 to $25,000 in real 1994 dollars. At Percent change since 1967 30 California 20 10 0 –10 –20 –30 –40 –50 1967 20th Median 80th 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Sample includes civilians age 18 and older who received wage and salary income. Statistics reported in this figure are adjusted for inflation. Figure S.1—Percentage Change in Real Annual Earnings for Males in California, by Income Percentile, 1967–1994 viii the 20th percentile, male earnings fell 40 percent from $17,316 in 1967 to $10,400 in 1994. In 1967, a man at the 80th percentile earned $44,345, about two and a half times what a man at the 20th percentile earned. By 1994, male earnings at the 80th percentile had increased 13 percent to $50,000, about five times what a man at the 20th percentile earned in that year. This comparison of the earnings of men in the upper-middle to the lower-middle of the distribution—the 80/20 ratio— is one simple measure of inequality. By this measure, male earnings inequality increased by 88 percent between 1967 and 1994 in California. Although the 80/20 ratio is an intuitive measure of inequality, it captures only two points in the distribution of income. Other measures of inequality are preferable because they summarize the entire distribution of income including the extreme top and bottom. One such measure is the coefficient of variation (CV). By this measure, male earnings inequality increased 41 percent between 1967 and 1994. Even by 1989, before the most recent recession, inequality had increased 35 percent since 1967. Income Inequality in California Matched That of the Nation Until the Late 1980s Inequality in household income has also grown. As Figure S.2 shows, household income inequality was similar in California and the nation for most of the years studied: It fluctuated in the 1970s, increasing during recessions and declining in recovery. It shot up dramatically during the recession of the early 1980s and never returned to pre-recession levels. It then remained fairly stable at new, higher levels through the mid-1980s. ix Index of inequality 0.81 0.79 0.77 0.75 0.73 0.71 0.69 0.67 0.65 0.63 1967 CA U.S. 1971 1975 1979 1983 1987 1991 1994 SOURCE: Authors’ calculations from the March CPS. NOTE: Household income is adjusted for household size and weighted by person. Statistics in this figure are not affected by inflation. The index of inequality used is the coefficient of variation (CV). The CV is the standard deviation of income divided by the mean of income. Figure S.2—Household Income Inequality in California and the Nation, 1967–1994 The trend in California’s income inequality began to diverge from the national trend in 1987. Inequality in household income started to increase faster in the state than in the nation, with especially rapid increases during the most recent recession. The period beginning in the late 1980s stands out as the only time when California has had substantially higher household income inequality than the nation for so many consecutive years. This divergence is also found in male earnings but not in female earnings. The fast-rising trend of inequality in California is also markedly visible when compared with other states. In 1969, 20 states had higher household income and male earnings inequality. By 1989, only five x states had higher household income inequality and only two had higher male earnings inequality. Income in California Has Grown More Slowly at the Top and Declined More Rapidly at the Bottom The sharp divergence between the state and the nation is not the result of higher income growth for the rich in California. As Figure S.3 shows, household income grew more in the nation than in the state. Between 1969 and 1989, two peak years of the business cycle, household income at the 90th percentile grew by 42 percent in the nation and 31 percent in the state. Instead, the divergence results from a greater income decline at the bottom. At the 10th percentile, while income in the nation grew by 7 percent, it actually fell by 7 percent in California. Incorporating data from the most recent recession shows an even more dramatic difference in growth between the nation and the state, as seen in the lower panel of the figure. Between the business cycle troughs of 1976 and 1994, income levels at the median and below fell in California, but in the United States they fell only at the 10th and 20th percentiles. Moreover, the decline in income at the 10th percentile in the United States was not nearly so drastic as in the state: Nationally, income fell by 8 percent, but in California, it fell by a remarkable 30 percent. Rapid Growth in Income Inequality Coincided with Business Cycle Recessions In both California and the nation, rapid growth in inequality coincided with recessions. The most noticeable increases in household income inequality, for example, occurred during the recessions of the early 1970s, early 1980s, and early 1990s. To illustrate, inequality in xi Percent change Percentage Change Between 1969 and 1989 Business Cycle Peaks 42 40 38 CA 30 U.S. 34 31 31 27 26 23 20 19 21 17 16 11 10 7 12 3 0 –10 –7 –5 10th 20th 30th 40th 50th 60th 70th 80th 90th Percentiles Percent change Percentage Change Between 1976 and 1994 Business Cycle Troughs 30 25 21 20 13 16 15 18 10 8 9 45 5 0 –10 –20 –30 –1 –8 –14 –22 –30 –10 –3 CA U.S. –40 10th 20th 30th 40th 50th 60th 70th 80th 90th Percentiles SOURCE: Based on authors’ calculations from the March CPS. NOTES: Household income is adjusted for household size and weighted by persons. Statistics in this figure are adjusted for inflation. Figure S.3—Percentage Change in Household Income Between Selected Years xii adjusted household income increased by 9 percent in California during the 1979–1982 recession, but it increased by only 3 percent during the economic growth of the next seven years. The relationship between the business cycle and inequality is particularly strong for male annual earnings in California: Inequality increased by 13 percent between 1979 and 1982 but by only 1 percent between 1982 and 1989. The recessions of the early 1970s and 1990s hit California harder than the nation (as is reflected in the larger increases in inequality shown in Figure S.2). While much of the rapid rise in inequality since 1981 occurred during the recession of the early 1990s, not all of the difference between California and the nation can be attributed to the strength of the recession in the state. The California growth trend in inequality began to outpace the national trend even before the start of the recession. Considering the Implications While inequality can increase because of the unequal sharing of income growth, it is particularly disturbing when it arises because of a decline in the income of poor individuals and households. This is the pattern that has characterized the increasing inequality in California over the last three decades. It is important to note, however, that the results of the study do not indicate that people who were poor in the past have gotten poorer—nor, conversely, that none have prospered. People who were in the 20th percentile in 1967 could have been in the 80th percentile in 1994. The data for this analysis are cross-sectional (snapshots of those in income groups in each year), not longitudinal, and therefore do not follow the fortunes of specific families or individuals over time. What the analysis xiii does tell us is that the poor in 1994 were considerably worse off than the poor in 1967. Moreover, as income falls at the bottom of the distribution, a greater percentage of people fall below the official poverty line (or any other absolute level of need). In other words, more Californians are poor today than were poor in the late 1960s. Given the similar trends in income inequality in California and the nation, it seems likely that the same forces are at work in both. Research on the underlying causes at the national level suggests a combination of factors: labor market trends influenced by technological change, international competition, immigration, and deunionization; and demographic trends in marriage and female employment. If these same forces explain the rise in inequality in the state, however, the recent sharp divergence suggests possible differential effects of those forces in California. Some Americans believe that differences in income arise primarily from individual choices, preferences, abilities, investments, and productivity, and that income inequality is a product of an economy that values hard work and talent. Other Americans believe that income differences reflect the unequal distribution of economic opportunity in our society, and that the opportunity to succeed is elusive for those who do not belong to privileged groups. The first viewpoint implies that public policy can affect inequality only by redistributing income; the second implies that policy can reduce inequality by promoting opportunity. Research on the determinants of the income distribution and the extent to which policy provides or restricts economic opportunity will suggest avenues for improving opportunities for the lessadvantaged. xiv Continuing growth in inequality is not inevitable. It is evident that government policies do affect the distribution of income, although the mechanisms are not fully understood. The challenge for future research is to examine the underlying forces behind the recent growth in inequality and to identify state policies that can promote equity and opportunity, as well as efficiency, in the California economy. xv Contents Foreword ..................................... Summary..................................... Figures ...................................... Tables ....................................... iii v xxi xxv 1. INTRODUCTION ........................... Trends in Income Inequality ..................... Nature of the Study ........................... 2. TRENDS IN THE DISTRIBUTION OF HOUSEHOLD INCOME ................................. What Is Household Income? ..................... Adjusted Household Income Is Sensitive to the Business Cycle ................................. The Distribution of Household Income Has Widened, Especially During Recessions .................. Other Summary Measures Also Show Rising Inequality .... Census Data Also Show Rising Household Income Inequality .............................. Household Income Inequality Rose Faster in California Than in Other Regions and States ................... Adjusted Family Income Shows Rising Inequality ........ 1 1 4 6 7 8 11 17 21 24 26 xvii 3. TRENDS IN THE DISTRIBUTION OF LABOR INCOME ................................. What Is Labor Income? ......................... Trends in the Distribution of Labor Income Among Males .. The Widening Distribution of Male Annual Earnings .... Summary Measures Show Rising Inequality in Male Annual Earnings .......................... Census Data Show Rising Inequality of Male Annual Earnings ............................... The Widening Distribution of Male Hourly Wages ..... Summary Measures Show Rising Inequality in Male Wages................................. Male Labor Income Inequality Rose Faster in California Than in Other Regions and States ............... Other Definitions of Male Labor Income Show Rising Inequality .............................. Trends in the Distribution of Labor Income Among Females ................................ The Narrowing, Then Widening, Distribution of Female Annual Earnings .......................... Measures of Inequality Show Falling, Then Rising, Inequality in Female Annual Earnings ............ Census Data Show a Fall and Then a Rise in Inequality of Female Annual Earnings ..................... The Widening Distribution of Female Hourly Wages .... Measures of Inequality Show Rising Inequality in Female Hourly Wages ........................... Female Labor Income Inequality Is Similar in California to Other Regions and States .................... Other Definitions of Female Labor Income Show Falling, Then Rising, Inequality ..................... 4. CONCLUSIONS AND IMPLICATIONS FOR POLICY AND FUTURE RESEARCH ..................... Public Policy and the Distribution of Income ........... Labor Market Explanations for Rising Earnings Inequality ... Demographic Explanations for Rising Family Income Inequality .............................. 30 31 33 34 37 37 40 42 43 46 47 47 50 50 53 56 56 58 60 61 62 64 xviii Additional Measurement Issues .................... 66 The Challenge for the State ...................... 67 Appendix A. Notes on Data and Methodology ................... 69 B. Using the Current Population Survey to Represent California .................................. 82 C. Trends in the Distributions of Alternative Measures of Income ................................... 87 D. Supplementary Statistics ........................ 115 Bibliography .................................. 135 xix Figures S.1. Percentage Change in Real Annual Earnings for Males in California, by Income Percentile, 1967–1994 ........ viii S.2. Household Income Inequality in California and the Nation, 1967–1994 ........................ x S.3. Percentage Change in Household Income Between Selected Years ............................ xii 2.1. Trends in the Unemployment Rate and Median Real Adjusted Household Income, 1967–1994 .......... 9 2.2. Percentage Change in Real Adjusted Household Income, by Income Percentile, 1967–1994 ............... 13 2.3. Summary Measures of Inequality for Real Adjusted Household Income, 1967–1994................. 20 2.4. Summary Measures of Inequality for Real Adjusted Family Income, 1967–1994 ................... 28 3.1. Percentage Change in Real Annual Earnings for Males, by Income Percentile, 1967–1994 ............... 35 3.2. Summary Measures of Inequality for Male Annual Earnings, 1967–1994 ....................... 38 xxi 3.3. Percentage Change in Real Hourly Wages for Males, by Income Percentile, 1975–1994 ................. 41 3.4. Summary Measures of Inequality for Male Hourly Wages, 1975–1994 ......................... 44 3.5. Percentage Change in Real Annual Earnings for Females, by Income Percentile, 1967–1994 ............... 49 3.6. Summary Measures of Inequality for Female Annual Earnings, 1967–1994 ....................... 51 3.7. Percentage Change in Real Hourly Wages for Females, by Income Percentile, 1975–1994 ............... 54 3.8. Summary Measures of Inequality for Female Hourly Wages, 1975–1994 ......................... 57 C.1. Income Type 2: Summary Measures of Inequality for Unadjusted Household Income Among Persons, 1967–1994 .............................. 95 C.2. Income Type 3: Summary Measures of Inequality for Adjusted Household Income Among Households, 1967–1994 .............................. 96 C.3. Income Type 4: Summary Measures of Inequality for Unadjusted Household Income Among Households, 1967–1994 .............................. 97 C.4. Income Type 6: Summary Measures of Inequality for Unadjusted Family Income Among Persons, 1967–1994 .............................. 98 C.5. Income Type 7: Summary Measures of Inequality for Adjusted Family Income Among Families, 1967–1994 .............................. 99 C.6. Income Type 8: Summary Measures of Inequality for Unadjusted Family Income Among Families, 1967–1994 .............................. 100 C.7. Income Type 9: Summary Measures of Inequality for Adjusted Primary Family Income Among Persons, 1967–1994 .............................. 101 xxii C.8. Income Type 10: Summary Measures of Inequality for Unadjusted Primary Family Income Among Persons, 1967–1994 .............................. 102 C.9. Income Type 11: Summary Measures of Inequality for Adjusted Primary Family Income Among Families, 1967–1994 .............................. 103 C.10. Income Type 12: Summary Measures of Inequality for Unadjusted Primary Family Income Among Families, 1967–1994 .............................. 104 C.11. Income Type 15: Summary Measures of Inequality for Annual Earnings for Males Ages 18 to 55, 1967–1994 .. 105 C.12. Income Type 16: Summary Measures of Inequality for Hourly Wages for Males Ages 18 to 55, 1975–1994 .... 106 C.13. Income Type 17: Summary Measures of Inequality for Annual Earnings for All Male Workers, 1967–1994 .... 107 C.14. Income Type 18: Summary Measures of Inequality for Hourly Wages for All Male Workers, 1975–1994...... 108 C.15. Income Type 19: Summary Measures of Inequality for Annual Salary for Males, 1967–1994.............. 109 C.16. Income Type 22: Summary Measures of Inequality for Annual Earnings for Females Ages 18 to 55, 1967–1994 .............................. 110 C.17. Income Type 23: Summary Measures of Inequality for Hourly Wages for Females Ages 18 to 55, 1975–1994 ... 111 C.18. Income Type 24: Summary Measures of Inequality for Annual Earnings for All Female Workers, 1967–1994 ... 112 C.19. Income Type 25: Summary Measures of Inequality for Hourly Wages for All Female Workers, 1975–1994 .... 113 C.20. Income Type 26: Summary Measures of Inequality for Annual Salary for Females, 1967–1994 ............ 114 xxiii Tables 2.1. Percentage Change in Real Adjusted Household Income Between Selected Years, by Income Percentile: CPS .... 2.2. Percentage Change in Real Adjusted Household Income Between Selected Years, by Income Percentile: Census .. 2.3. Levels and Trends in the Coefficient of Variation for Adjusted Household Income: CPS and Census ....... 2.4. Regional Trends in the Coefficient of Variation for Real Adjusted Household Income, 1969–1994 .......... 2.5. Percentage Change in Real Adjusted Family Income Between Selected Years, by Income Percentile ........ 3.1. Percentage Change in Real Annual Earnings for Males Between Selected Years, by Income Percentile: CPS .... 3.2. Percentage Change in Real Annual Earnings for Males, by Income Percentile: Census .................. 3.3. Levels and Trends in the Coefficient of Variation for Male Annual Earnings: CPS and Census ........... 3.4. Percentage Change in Real Hourly Wages for Males Between Selected Years, by Income Percentile ........ 16 22 23 25 27 36 39 40 43 xxv 3.5. Regional Trends in the Coefficient of Variation for Real Hourly Wages Among Males, 1979–1994 .......... 3.6. Percentage Change in Real Annual Earnings for Females Between Selected Years, by Income Percentile ........ 3.7. Percentage Change in Real Annual Earnings and Hourly Wages for Females, by Income Percentile: Census ..... 3.8. Levels and Trends in the Coefficient of Variation for Female Annual Earnings: CPS and Census ......... 3.9. Percentage Change in Real Hourly Wages for Females Between Selected Years, by Income Percentile ........ 3.10. Regional Trends in the Coefficient of Variation for Real Hourly Wages Among Females, 1979–1994 ......... A.1. Price and Cost of Living Adjustments, California and the United States, 1967–1994 .................... B.1. Percentage of Population in Each Category: Census and CPS .................................. C.1. Alternative Measures of Household and Family Income................................. C.2. Alternative Measures of Labor Income ............ C.3. Income Type 2: Percentage Change in Real Unadjusted Household Income Among Persons Between Selected Years, by Income Percentile ................... C.4. Income Type 3: Percentage Change in Real Adjusted Household Income Among Households Between Selected Years, by Income Percentile .............. C.5. Income Type 4: Percentage Change in Real Unadjusted Household Income Among Households Between Selected Years, by Income Percentile .............. C.6. Income Type 6: Percentage Change in Real Unadjusted Family Income Among Persons Between Selected Years, by Income Percentile ........................ 45 48 52 53 55 58 80 86 90 92 95 96 97 98 xxvi C.7. Income Type 7: Percentage Change in Real Adjusted Family Income Among Families Between Selected Years, by Income Percentile ........................ 99 C.8. Income Type 8: Percentage Change in Real Unadjusted Family Income Among Families Between Selected Years, by Income Percentile ........................ 100 C.9. Income Type 9: Percentage Change in Real Adjusted Primary Family Income Among Persons Between Selected Years, by Income Percentile .............. 101 C.10. Income Type 10: Percentage Change in Real Unadjusted Primary Family Income Among Persons Between Selected Years, by Income Percentile ........ 102 C.11. Income Type 11: Percentage Change in Real Adjusted Primary Family Income Among Families Between Selected Years, by Income Percentile .............. 103 C.12. Income Type 12: Percentage Change in Real Unadjusted Primary Family Income Among Families Between Selected Years, by Income Percentile ........ 104 C.13. Income Type 15: Percentage Change in Real Annual Earnings for Males Ages 18 to 55 Between Selected Years, by Income Percentile ................... 105 C.14. Income Type 16: Percentage Change in Real Hourly Wages for Males Ages 18 to 55 Between Selected Years, by Income Percentile ........................ 106 C.15. Income Type 17: Percentage Change in Real Annual Earnings for All Male Workers Between Selected Years, by Income Percentile ........................ 107 C.16. Income Type 18: Percentage Change in Real Hourly Wages for All Male Workers Between Selected Years, by Income Percentile ........................ 108 C.17. Income Type 19: Percentage Change in Real Annual Salary for Males Between Selected Years, by Income Percentile ............................... 109 xxvii C.18. Income Type 22: Percentage Change in Real Annual Earnings for Females Ages 18 to 55 Between Selected Years, by Income Percentile ................... 110 C.19. Income Type 23: Percentage Change in Real Hourly Wages for Females Ages 18 to 55 Between Selected Years, by Income Percentile ................... 111 C.20. Income Type 24: Percentage Change in Real Annual Earnings for All Female Workers Between Selected Years, by Income Percentile ........................ 112 C.21. Income Type 25: Percentage Change in Real Hourly Wages for All Female Workers Between Selected Years, by Income Percentile ........................ 113 C.22. Income Type 26: Percentage Change in Real Annual Salary for Females Between Selected Years, by Income Percentile ............................... 114 D.1. Deciles of Nominal Adjusted Household Income, California ............................... 116 D.2. Deciles of Nominal Adjusted Household Income, United States .................................. 117 D.3. Deciles of Nominal Annual Earnings Among Male Workers, California ........................ 118 D.4. Deciles of Nominal Annual Earnings Among Male Workers, United States ...................... 119 D.5. Deciles of Nominal Hourly Wages Among Male Workers, California ........................ 120 D.6. Deciles of Nominal Hourly Wages Among Male Workers, United States ...................... 121 D.7. Deciles of Nominal Annual Earnings Among Female Workers, California ........................ 122 D.8. Deciles of Nominal Annual Earnings Among Female Workers, United States ...................... 123 D.9. Deciles of Nominal Hourly Wages Among Female Workers, California ........................ 124 xxviii D.10. Deciles of Nominal Hourly Wages Among Female Workers, United States ...................... 125 D.11. Regional Trends in the Coefficient of Variation for Real Annual Earnings Among Males, 1969–1994 ......... 126 D.12. Regional Trends in the Coefficient of Variation for Real Annual Earnings Among Females, 1969–1994 ....... 127 D.13. State Rankings for Adjusted Household Income Inequality Based on the Coefficient of Variation: Census ................................. 128 D.14. State Rankings for Male Annual Earnings Inequality Based on the Coefficient of Variation: Census ....... 130 D.15. State Rankings for Female Annual Earnings Inequality Based on the Coefficient of Variation: Census ....... 132 xxix 1. Introduction A fuller understanding of state-level trends in the distribution of income is essential for California. The recent trends not only will determine the need for strengthened state policies to aid low-income families but will affect the likely success of those policies. This report provides a comprehensive description of the widening distribution of income in California. Its findings reveal a general pattern of increasing income inequality in the state, explained by a dramatic decline in income for the poor and near-poor accompanied by income growth for the rich. Subsequent studies by the Public Policy Institute of California will examine the underlying causes of the trends and will explore the relationships between state policy and income inequality. Trends in Income Inequality The upward trend in income inequality in the United States throughout the 1970s and 1980s stands in marked contrast with the 1 trends in the distribution of income from the Great Depression to the late 1960s. Jeffrey Williamson and Peter Lindert (1980) report a dramatic decline in income inequality between the Depression and the end of World War II. From the late 1940s until the late 1970s, inequality fluctuated within a relatively narrow band. This long period of stability in income inequality led to speculation that, with the exception of short-term fluctuations, the distribution of economic wellbeing would remain constant (Blinder, 1980). Tracking changes in inequality, wrote one researcher, was like “watching grass grow” (Aaron, 1978, p. 17). The conventional wisdom was too optimistic. In the early 1980s, Census Bureau reports provided some of the earliest indications of a growing inequality among families. Census Bureau income statistics revealed that family income inequality had reached a postwar low in the late 1960s but had climbed almost constantly from that time. Since the early 1980s, family income inequality has remained higher than in any previous year since the end of the Second World War.1 In recent years, numerous studies have documented the widening distribution of family income and male earnings in the United States. We summarize this work here. Sheldon Danziger and Peter Gottschalk (1995) report that the gap in income between families near the top of the income distribution and those near the bottom has increased, in both recession and recovery, since the recession of the early 1970s. The years 1983 to 1989 stand out as an anomalous period that recorded growth in mean family income along with rising income inequality. ____________ 1U.S. Bureau of the Census, Current Population Reports, P-60 series, various issues. 2 In their comprehensive 1992 review article, Frank Levy and Richard Murnane conclude that the 1970s were a period of either stability or gradual growth in male annual earnings inequality and that the 1980s were a period of rapid increase. Lynn Karoly (1993) shows that this rise in income inequality is explained by a decline in the income of poor families and workers and by growth in the income of the rich. While much attention has been focused on the trends in income inequality at the national level, relatively few studies have investigated income distribution in the state of California. There are many reasons to expect that the trends in the distribution of income in California will differ from those of the nation. Income inequality measures for the country as a whole aggregate regional diversity in economic and demographic trends. California is distinctive in its industrial base, trading partners, racial and ethnic composition, patterns of domestic and international migration, and in the age and education of its workforce. Previous research on the distribution of income in California shows that the state has experienced a rise in income disparity. Jay Chamberlain and Phil Spillberg (1991) report a rising concentration of adjusted gross income in the 1980s: Between 1980 and 1988, the proportion of the total after-tax adjusted income received by the top 20 percent of taxpayers increased from 52 to 57 percent; the proportion received by the top one percent increased from 10 to 16 percent. Karoly (1995) finds that the ratio of the income of wealthy families at the 90th percentile to the income of poor families at the 10th percentile—the 90/10 ratio— 3 increased by 74 percent between 1973 and 1993 in California.2 This rise in inequality was due to growth in the incomes of the rich and a substantial decline in the incomes of the poor. Moreover, the rise in inequality in California was larger than in the nation as a whole, where the 90/10 ratio increased by 54 percent. Research that compares California to other regions of the country is less conclusive. On the one hand, Robert Topel (1994) finds that the western region of the nation, dominated in population by California, experienced the largest increase of any region in male wage inequality between 1972 and 1990. On the other hand, Thomas Husted (1991) shows that between 1981 and 1987, the percentage increase in the Gini coefficient (one index of inequality) was higher in California than in the nation as a whole, but 24 states had larger percentage increases. Nature of the Study This study contributes to the existing research on the distribution of income in California by providing a comprehensive description of state trends and by comparing these to trends of the nation, other regions, and other states. To document the trends in income inequality thoroughly, the study uses five measures of inequality and 26 definitions of income. Data for this analysis come from the annual March file of the Current Population Survey and the decennial Census of Population and ____________ 2The 10th percentile is defined as the level of income that divides the bottom 10 percent from the top 90 percent; similarly, 90 percent of people have incomes below the 90th percentile, whereas only 10 percent have incomes above. 4 Housing.3 The analysis covers the entire period spanned by available public-use files of the Current Population Survey: 1967 to 1994. This study measures the trends for two main types of income: Household income provides a picture of general economic well-being because it includes all sources of money income and it is measured for all people regardless of work status. Labor income, the largest component of household income, measures earnings from work. Labor income reflects changes in the economy and is not directly influenced by changes in household structure. The next two chapters describe results of the study. Chapter 2 describes trends in the distribution of household income and Chapter 3 describes trends in the distribution of male and female labor income. Each chapter analyzes the California experience in relation to that of other regions and states. Chapter 4 presents our conclusions, discusses the relationship between public policy and the distribution of income, and outlines possible explanations for the rise of income inequality in California. Readers interested in greater technical details of the study are directed to the appendices: Appendix A describes the datasets used in the study. Appendix B discusses the representativeness of the California subsample of the Current Population Survey. Appendix C reports on trends in the distribution of alternative measures of income. Appendix D provides supplementary statistics on the distributions of income measures discussed in the text. ____________ 31970 Public Use Sample, 1 percent, and the 1980 and 1990 Public Use Micro Sample, 5 percent. 5 2. Trends in the Distribution of Household Income Household income is a measure of economic well-being that explicitly accounts for income-sharing among members of the same household. It is the most comprehensive measure of income in this study because it is measured for all people regardless of age and work status, and it incorporates income from all reported sources. As this chapter demonstrates, the distribution of household income in California has widened considerably over the past three decades, especially during business cycle recessions. Summary measures of inequality show that the increasing trend in household income inequality was similar for California and the nation until the late 1980s. Since then, the rise in inequality has been much greater in California. Compared to other states, California had one of the highest levels of inequality, even before the recent recession. 6 What Is Household Income? Household income is defined as the sum of income from all sources for all persons living in the same household unit. Because households with many persons require more resources than small households to maintain the same level of consumption, we adjust household income based on the number of household residents.1 We evaluate the distribution of adjusted household income across people, rather than across household units, by assigning to each person the adjusted income of his or her household. This method treats each person equally, rather than implicitly giving less weight in the calculation to people in large households.2 All income statistics reported in this study are adjusted to real 1994 dollars based on the consumer price index computed by the Bureau of Labor Statistics.3 Recent studies suggest that the official consumer price index may exaggerate inflation, thus understating growth and overstating ____________ 1We calculate adjusted household income by dividing total household income by the square root of the number of household residents. Karoly and Burtless (1995) suggest this adjustment factor because it is close to the adjustment for family size implicit in the official poverty thresholds. This adjustment takes into account “economies of scale” made possible through the sharing of common resources in large households. For example, the adjustment implies that a household with four people will require twice, rather than four times, the income of a single person to maintain the same level of consumption. We make the same adjustments to family income based on family size. Median levels of adjusted household and family income reported in the text are multiplied by two to represent income levels for households and families of four persons. For comparison, we also measure changes in the distribution of unadjusted household and family income (see Appendix C). 2Using this method, 50 percent of people live in households with adjusted incomes lower than the median, as opposed to 50 percent of households falling below the median. Similarly, we evaluate the distribution of adjusted family income across people as opposed to family units. For comparison, we measure trends in the distributions of household income across households and family income across families (see Appendix C). 3We use the CPI-U-X1 and allow for differences in the rate of inflation in California and the United States. See Appendix A for details. 7 decline. However, although the consumer price index affects estimated growth trends, the summary measures of inequality used in this report are based on relative income (e.g., the income of the rich relative to the income of the poor) and are not affected by inflation adjustments. The Current Population Survey and the Census report pre-tax money income including wages and salary, farm income, selfemployment income, interest and dividends, welfare receipts, and Social Security and retirement benefits. The income measures are imperfect indices of economic well-being because the data do not include information on tax payments, non-monetary transfers (e.g., housing subsidies, health benefits, food stamps), the return to investments such as owner-occupied housing, or measures of accumulated wealth. However, studies that have used more comprehensive measures of income have found trends in income inequality similar to those for pre-tax money income. (See Appendix C for a review of such studies.) Adjusted Household Income Is Sensitive to the Business Cycle Because the business cycle plays a strong role in the distributional trends we describe, we begin by showing business cycle fluctuations as measured by unemployment and associated fluctuations in household income. Figure 2.1 shows how strongly fluctuations in adjusted household income are related to the business cycle. The upper panel of the figure displays the unemployment rate in California and the United States from 1967 to 1994. Rising rates of unemployment characterize the periods of recession of the early 1970s, mid-1970s, early 1980s, and 8 Median adjusted household income ($) Unemployment rate (%) 10 9 8 7 6 5 4 3 2 1 0 1967 55,000 CA U.S. 1971 1975 1979 1983 1987 1991 1994 50,000 45,000 40,000 35,000 CA U.S. 30,000 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Real median adjusted household income is based on authors’ calculations from the March CPS. Unemployment rates are from the Bureau of Labor Statistics. NOTES: Median incomes have been converted to real 1994 dollars. The U.S. median has been adjusted to reflect the higher cost of living in California. Comparison of median income in California to median income in the United States should be made with caution because of measurement problems in the cost of living index (as described in Appendix A). The trend in median household income is sensitive to the consumer price index. Household income is adjusted for the number of people living in the household. Reported median household income is calibrated to represent a household of four people. Median household income in California in 1988 may not be comparable to other years due to changes in the CPS (as described in Appendix A). Figure 2.1—Trends in the Unemployment Rate and Median Real Adjusted Household Income, 1967–1994 9 early 1990s for both California and the nation. With the exception of the 1980s, unemployment has been higher in California than in the nation, particularly during the recessions of the early 1970s and early 1990s. Median adjusted household income (the lower panel of Figure 2.1) shows a positive growth trend through the mid-1980s for both California and the nation, with higher overall growth in the nation.4 Declines in median household income generally occurred only during periods of recession. However, median household income in California began to stagnate as early as 1987, even before the most recent recession. The greater decline in median household income and the higher unemployment rate in California indicate the stronger effect of the early 1990s recession in the state. (The dip in median household income in 1988 is probably explained by changes in sampling in the Current Population Survey. Results for 1988 are reported in this study but conclusions are not based on statistics specific to 1988.)5 The medians in Figure 2.1 are adjusted both for inflation and for the higher cost of living in California. The position of the Californian median relative to the national median is a function of the adjustments ____________ 4If no adjustments were made to household income for household size and if the distribution were evaluated at the household level and not the person level, the median in the United States would be less than 1 percent higher than the median in California in 1994. The median of unadjusted household income (weighted at the household level) increased in the United States relative to California from 1967 to 1978; after 1978, the relative growth of the U.S. median fluctuated with no clear trend. However, the median of adjusted household income (weighted by persons), the median reported in the text, is a preferred measure of economic well-being because it accounts for the greater resource needs of large households and it applies the same weight to people in large households as to people in small households. 5See Appendix A for further discussion of sampling and other data issues. 10 made for differences in the cost of living.6 For example, in 1994 the cost of living estimate was 9 percent higher in California than in the nation. The cost of living adjustments applied in this figure are calculated from Bureau of Labor Statistics data.7 Because cost of living estimates are imprecise, comparison of the California median to the national median should be made with caution. Although Figure 2.1 shows that the U.S. median adjusted household income was about $3,000 below that of California in 1967 and was about $4,000 above that of California in 1994, this result could be different if a more accurate cost of living index were available. However, the faster rise in the U.S. median does not depend on the cost of living adjustment. The median adjusted household income statistics in Figure 2.1 are the only statistics in this report that are affected by the cost of living adjustment. The Distribution of Household Income Has Widened, Especially During Recessions The most significant widening of the distribution of adjusted household income occurred during periods of recession, particularly in the early 1980s and early 1990s. Overall, the gap between the incomes of people in rich and poor households increased not only because incomes at the top of the distribution rose but also because incomes at the bottom of the distribution fell. ____________ 6For example, if no adjustments were made for cost of living, the median of adjusted household income in the United States would be about half a percent lower than the median in California in 1994. See Appendix D for the nominal value of income at each decile without adjustments for cost of living. 7See Appendix A for the calculation of cost of living adjustment for 1967–1994. 11 A straightforward way to investigate the changing shape of the distribution of income is to examine the relative income positions of low-, middle-, and high-income people. Figure 2.2 illustrates the income trends at the 10th, 20th, 50th (median), 80th, and 90th percentiles of adjusted household income.8 The figure shows the percentage change in income since 1967: For example, the highest point on the graph for California shows that people in the 90th percentile in 1987 had income slightly more than 40 percent higher than people in the 90th percentile in 1967. Although the reported statistics are standardized to the base year of 1967, the figure is not meant to imply that there was no household income inequality in 1967.9 Instead, the figure graphically represents the widening of the distribution and corresponding increases in inequality from its 1967 levels. The absolute decline of income levels for households near the bottom of the distribution in California is a striking feature of the figure. During the 1970s, the income received by households at the 10th and 20th percentiles of the distribution in California fluctuated mildly but showed little overall growth. During the recession of the early 1980s, the ____________ 8People in the 10th percentile have incomes higher than only 10 percent of the population; those in the 90th percentile have incomes higher than 90 percent of the population. People in the 10th and 20th percentiles are in the lower and lower-middle ranks of the income distribution; the median (or 50th percentile) describes the income level of people in the middle of the income distribution; the 80th and 90th percentiles indicate the income levels of people in the upper-middle and upper ranks of the income distribution. 9The information in the figure can be used to calculate the percentage change between any two years by using the following calculation: Add 100 to the values displayed on the figure, take the ratio, subtract 1, and multiply by 100. For example, between the business cycle peaks in 1979 and 1989, adjusted household income at the 10th percentile fell by 8 percent in California (100/108.5 – 1) * 100. 12 Percent change since 1967 Percent change since 1967 60 California 50 10th 40 20th Median 30 80th 90th 20 10 0 –10 –20 –30 1967 1971 1975 1979 1983 1987 1991 1994 60 50 United States 40 30 20 10 0 10th 20th –10 Median 80th –20 90th –30 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Statistics reported in this figure are sensitive to the consumer price index. Adjusted household income in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 2.2—Percentage Change in Real Adjusted Household Income, by Income Percentile, 1967–1994 13 income of households at the 20th percentile fell by 13 percent in California. Before recovering fully from this recession, income at the 20th percentile in California fell again by another 20 percent during the most recent recession. The income decline for households at the 10th percentile was even greater, with income plunging 15 percent and 23 percent during the two recessions. The national distribution of adjusted household income shows the same pattern of sharp decline at the bottom during recessions. However, the decline during recessions was greater in California than in the nation as a whole, especially for low-income households, and the growth during the recovery of the 1980s was weaker in California than in the United States. The trends depicted in Figure 2.2 can easily be misinterpreted. The figure shows that Californians at the 10th percentile in 1994 received 24 percent less income than Californians at the 10th percentile in 1967. The cross-sectional data used in this report do not track the same people over the years. The figure, therefore, does not show that the income of specific people at the 10th percentile declined by 24 percent. The distinction is often subtle. When we say that “the poor got poorer,” we mean that the people who were poor in 1994 were poorer than the people who were poor in 1967, but not that the same people who were poor in 1967 were even poorer in 1994.10 This interpretation issue is ____________ 10As an analogy, imagine a class of nine third-graders lined up in order of height. The height of the fifth child in the line is the median height. Now suppose that four shorter children enter the line, making the total 13. The median child is now the seventh child in the line—the child who was third in the original line. The new median height is lower than the previous median, but no child can be said to have “experienced a decline in height.” 14 particularly important in California where there is a high degree of mobility into and out of the state. Figure 2.2 clearly shows the widening gap between the upper (80th and 90th) and lower (10th and 20th ) percentiles. The income trends displayed in Figure 2.2 demonstrate the strong relationship between business cycle conditions and the widening distribution of household income. For both California and the United States, the recessions in the early 1980s and early 1990s stand out as periods when the distribution of household income widened rapidly, with precipitous drops in income levels at the lower percentiles of the distribution and small shorter-lived declines at the upper percentiles. In California, the widening of the distribution is more substantial than in the nation, showing a larger increase in inequality. Because of this relationship between the business cycle and income inequality, it is important to focus on years in similar points of the business cycle when describing the long-run trends in the distribution of income. Comparing the distributions of adjusted household income in 1967 and 1994, for example, is likely to exaggerate the trends in inequality growth because the economy was strong in 1967 and weak in 1994. To avoid such distortion, Table 2.1 summarizes the trends in Figure 2.2 for selected years at similar points in the business cycle. The first row of the table shows the absolute decline of adjusted household income for the lower-middle of the distribution (the 20th percentile) in California. Between the two major business cycle peaks spanned by our study, 1969 and 1989, the income level of households at the 20th percentile declined by 5 percent. The median of household income grew 16 percent over 15 Table 2.1 Percentage Change in Real Adjusted Household Income Between Selected Years, by Income Percentile: CPS Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th 2 –6 –5 –22 Median 14 2 16 –3 80th 20 5 26 15 Change in 80/20 ratio (%) +18 +12 +32 +48 United States 20th 10 1 11 –1 Median 18 8 27 8 80th 22 14 38 21 Change in 80/20 ratio (%) +10 +13 +24 +22 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. In this and following tables the percentage change between 1969 and 1989 is not equal to the sum of the changes between 1969–1979 and 1979–1989, because the change over the 1980s is calculated from the base year of 1979 and not 1969. For example, if income grew 100 percent between 1969 and 1979 from $10 to $20, and then grew by another 100 percent between 1979 and 1989 to $40, the overall change from 1969 to 1989 would be 400, not 200, percent. the same period. For the upper-middle of the distribution (the 80th percentile), household income grew 26 percent. Even over the 1980s, household income grew more in the United States than in California at each of these percentiles. The widening of the distribution of adjusted household income described in Figure 2.2 can be summarized by the ratio of income at the top of the distribution to income at the bottom. The ratio of the income of the 90th percentile to the income of the 10th percentile, the 90/10 ratio, is often used as a measure of inequality. We use the 80/20 ratio 16 instead, in order to focus on the widening of the middle of the distribution. Table 2.1 illustrates how seriously inequality has grown in California and the nation: The 80/20 ratio increased by 32 percent in California and by 24 percent in the United States between 1969 and 1989. The 80/20 ratios in Table 2.1 suggest that California had much faster growth in inequality than the nation during the 1970s. However, as the next section will show, this finding is not confirmed by other measures of inequality that take into account the entire distribution. This study emphasizes trends in the income distribution up until 1989 because it is impossible to determine whether later changes reflect short-run fluctuations due to the severity of the recent recession or a continuing trend of rapidly rising inequality. Nevertheless, changes between 1976 and 1994 are reported to demonstrate how seriously a deep recession, like the one of the early 1990s, can affect income inequality. As the fourth column in Table 2.1 displays, incorporating the most recent recession reflects the same pattern of income growth as shown up until 1989, but an even bleaker picture emerges, especially in California. Other Summary Measures Also Show Rising Inequality The percentile graphs in the previous section show the widening of the distribution of adjusted household income relative to 1967, but they do not provide an absolute measure of income inequality that would allow us to compare inequality in California and the nation. Since summary measures describe inequality with a single statistic, they make it 17 possible to rank the level of inequality in two different distributions of income (e.g., in the United States and in California). Moreover, the summary measures used in this report are independent of the consumer price index. Even if the consumer price index overstates inflation, the magnitude of the reported summary measures is not affected. The 80/20 ratio reported in the previous section is one summary measure of income inequality, but it suffers from the drawback that it evaluates only two positions in the distribution. There are numerous summary measures of income inequality that evaluate income throughout the distribution, including the extreme top and bottom. This study reports four commonly used and easily calculated measures: the coefficient of variation (CV), Theil’s entropy (ENTROPY), mean log-deviation (MLD), and the variance of the natural logarithm of income (VLN).11 These four were chosen in part to allow for comparability with other studies, particularly Karoly’s (1993) work on income inequality in the United States. There is no a priori best measure of inequality. All four measures agree on what it means to have a perfectly equal society: Each measure is scaled to equal zero when all members of society have the same amount of income. However, the measures do not agree on how to quantify deviations from perfect equality. For example, the VLN is more responsive to reductions in income near the bottom of the distribution: In an economy where nine people have $10 dollars each and one person has $8, the VLN measure will show higher inequality than if the ____________ 11The CV is the standard deviation of income divided by the mean of income. The ENTROPY measure is the mean of [y/mean(y) * ln(y/mean(y))], where y is income. The MLD is the natural logarithm of the mean of income minus the mean of the natural logarithm of income. VLN is the variance of ln(y). 18 anomalous person had $12, even though the deviation is $2 in both cases. This effect reflects the idea that downward deviations from equity have more negative consequences than upward ones. The CV treats upward and downward deviations the same—the CV measure would have the same value if the anomalous person has $8 or $12. If income grows across the distribution, but grows faster for the rich, then the CV will register a greater change in inequality than the VLN will. The MLD and ENTROPY measures emphasize the bottom of the distribution more than the CV but less than the VLN. That is, the VLN is the most responsive to changes at the bottom of the income distribution, followed by the MLD, ENTROPY, and CV, in that order. Because summary measures respond differently to income disparity, they may produce different rankings of income distributions. It is possible to find that income inequality is higher in the United States by some measures and higher in California by others. Similarly, the summary measures may show different time trends for income inequality in California. For this reason, using several summary measures of inequality and comparing the results across measures provide a fuller picture of the trends in income inequality. Figure 2.3 illustrates the inequality of adjusted household income in California and the United States using the four measures of inequality. Overall, the measures show that the level of household income inequality in California was quite similar to that of the nation until the late 1980s. In both California and the United States, the main patterns in income inequality are consistent with a rapid rise in inequality during recession periods. The recession of the early 1980s was a period of dramatic increase in household inequality in both California and the nation: 19 .80 CV .78 .76 CA .74 U.S. .72 .30 ENTROPY .28 CA U.S. .26 .70 .24 .68 .66 .22 .64 .20 .62 .60 .18 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 .49 MLD .44 CA U.S. .39 2.15 1.95 1.75 1.55 VLN CA U.S. .34 1.35 1.15 .29 .95 .24 .75 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Statistics reported in this figure are not sensitive to the consumer price index. Adjusted household income in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 2.3—Summary Measures of Inequality for Real Adjusted Household Income, 1967–1994 20 Between 1979 and 1982, the CV grew by 9 percent in California and by 8 percent in the United States. Over the same period, the VLN grew by 23 percent in California and by 29 percent in the nation, reflecting the greater sensitivity of this measure to the declining income near the bottom of the distribution. Adjusted household income inequality in the United States began another steep increase at the beginning of the most recent recession. In California, the increase began earlier, with small increases as early as 1987 preceding more drastic increases in the early 1990s. The late 1980s and early 1990s stand out as the only period over the last three decades that California has maintained a substantially higher level of adjusted household income inequality than the United States for several consecutive years. This is consistent with the more severe decline in adjusted household income at the median and lower percentiles in California during the most recent recession, as shown in Figure 2.2. Census Data Also Show Rising Household Income Inequality This report focuses primarily on income data from the Current Population Survey (CPS) because its annual data provide a fuller picture of the distribution trends than the decennial Census. Furthermore, the income data in the CPS are likely to be more accurate than the income data in the Census. For example, in 1990, the Census asked respondents about eight specific types of income. In the same year, the CPS asked about more than 20 types of income.12 Results from Census data are ____________ 12In addition, the CPS is conducted by phone by trained survey-takers whereas the Census is taken by mail. For a further discussion of these two datasets, see Appendix A. 21 provided here to address the concern that the CPS may not adequately represent California.13 The Census data do in fact confirm the trends in the distribution of adjusted household income as measured by the CPS.14 Table 2.2, based on Census data, shows that the growth in the upper-middle of the distribution (the 80th percentile) exceeded the growth in the lower-middle of the distribution (the 20th percentile). Table 2.2 Percentage Change in Real Adjusted Household Income Between Selected Years, by Income Percentile: Census Business Cycle Peaks 1969–1979 1979–1989 1969–1989 California 20th Median 80th Change in 80/20 ratio (%) 5 –4 14 0 85 +12 +10 1 14 24 +24 United States 20th Median 80th Change in 80/20 ratio (%) 13 4 18 8 20 14 +6 +9 18 28 36 +15 SOURCE: Based on authors’ calculations from the decennial Census. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. ____________ 13For further discussion of the representativeness of the CPS for California, see Appendix B. 14For a comparison of the income levels at each decile for the Census and CPS, see Appendix D. 22 The Census data also show an absolute decline in adjusted household income at the 20th percentile in California during the 1980s, but the decline is slightly smaller than measured in the CPS, as shown in Table 2.1. Like those calculated from the CPS, the 80/20 ratios show that the widening of the distribution was more pronounced in California than it was in the United States. Table 2.3 shows the levels and trends in the CV using the Census and the CPS. Both datasets show similar levels of adjusted household income inequality in California and the nation and a greater upward trend in inequality in California. The most significant difference between the datasets is that the Census suggests that inequality increased between 1969 and 1979 in California, whereas the CPS data indicate that the increase began after 1979. Table 2.3 Levels and Trends in the Coefficient of Variation for Adjusted Household Income: CPS and Census CV: Level 1969 1979 1989 California CPS Census 0.65 0.66 0.65 0.70 0.74 0.75 United States CPS Census 0.65 0.67 0.64 0.66 0.72 0.73 CV: Percent change 1969–1979 0 5 –1 –1 1979–1989 13 8 12 11 1969–1989 13 13 11 9 SOURCE: Based on authors’ calculations from the March CPS and the decennial census. NOTE: Statistics reported in this table are not sensitive to the consumer price index. 23 Household Income Inequality Rose Faster in California Than in Other Regions and States The large sample size of the Census allows for measurement of income inequality at the state level, even for small states. In 1969, 20 states had higher adjusted household income inequality than California did, as measured by the CV. By 1989, California was the sixth highest state. Over the period 1969 to 1989, only Michigan experienced higher percentage growth than California in adjusted household income inequality (see Appendix D for full state rankings). A limitation of the Census data is that they cannot be used to measure the dramatic increase in California income inequality that occurred during the deep recession of the early 1990s. Fortunately, data for the CPS do cover this period. Because of sample-size limitations, however, the CPS data can only be used to compare California to regions, not to other states. Relative to the other regions of the country, California has experienced higher growth in adjusted household income inequality since 1969. Table 2.4 reports trends in the level and growth of the CV for ten regions of the country: California plus nine geographically defined regions (California is also included as part of the Pacific region).15 The ____________ 15The nine Census regions are New England (Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Connecticut); Middle Atlantic (New York, New Jersey, Pennsylvania); East North Central (Ohio, Indiana, Illinois, Michigan, Wisconsin); West North Central (Minnesota, Iowa, Missouri, North Dakota, South Dakota, Nebraska, Kansas); South Atlantic (Delaware, Maryland, District of Columbia, Virginia, West Virginia, North Carolina, South Carolina, Georgia, Florida); East South Central (Kentucky, Tennessee, Alabama, Mississippi); West South Central (Arkansas, Louisiana, Oklahoma, Texas); Mountain (Montana, Idaho, Wyoming, Colorado, New Mexico, Arizona, Utah, Nevada); and Pacific (Washington, Oregon, California, Alaska, Hawaii). 24 Table 2.4 Regional Trends in the Coefficient of Variation for Real Adjusted Household Income, 1969–1994 Region CV (Rank) Percentage Change in CV (Rank) 1969 1979 1989 1994 1969–1979 1979–1989 1989–1994 California 0.65 0.65 0.74 0.79 0 13 7 (4) (4) (3) (1) (4) (2) (1) New England 0.55 0.60 0.64 0.68 10 6 6 (10) (10) (10) (10) (1) (10) (2) Mid Atlantic 0.64 0.63 0.71 0.74 –2 13 3 (6) (7) (6) (5) (8) (1) (4) E. N. Central 0.59 0.60 0.67 0.69 1 12 3 (9) (9) (9) (8) (3) (5) (5) W. N. Central 0.62 0.61 0.67 0.68 –2 11 0 (8) (8) (8) (9) (9) (7) (9) S. Atlantic 0.67 0.66 0.72 0.73 –1 9 0 (3) (3) (4) (6) (7) (9) (8) E. S. Central 0.72 0.67 0.75 0.74 –6 11 –1 (1) (2) (2) (4) (10) (8) (10) W. S. Central 0.69 0.69 0.78 0.78 0 13 0 (2) (1) (1) (2) (5) (3) (7) Mountain 0.62 0.63 0.71 0.72 2 12 2 (7) (6) (7) (7) (2) (6) (6) Pacific 0.65 0.64 0.72 0.76 –1 12 6 (5) (5) (5) (3) (6) (4) (3) SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are not sensitive to the consumer price index. size of California makes it reasonable to consider the state as its own region: California has more residents than do half of the nine regions. In 1969, California had the fourth-highest level of income inequality of the ten regions. Between 1979 and 1989, California had the secondhighest level of growth in income inequality; between 1989 and 1994, California had the highest. In 1994, California had the highest level of inequality of adjusted household income of the ten regions. 25 Adjusted Family Income Shows Rising Inequality Thus far, we have focused on the distribution trends in adjusted household income among persons. Focusing on household income implicitly assumes income-sharing among household residents regardless of relationship. Many researchers examine the distribution of family income rather than household income. For completeness, we also examined the trends in the distribution of adjusted family income among persons, implicitly assuming that there is no income-sharing among residents of the same household who are not related by blood, marriage, or adoption.16 As shown below, the trend in the distribution of family income exhibits the same widening as the distribution of household income, but the rise in inequality is even more pronounced for adjusted family income. The Census Bureau defines a “family” as the head of household and at least one resident relative: single people living alone and subfamilies unrelated to their household head are not included in the Census Bureau’s sample of families. We use a more comprehensive definition: Single persons living alone are included as their own family; people who do not live alone but are not related to the head of their household are included as separate families.17 This comprehensive definition is preferred to the Census Bureau definition because it includes the entire sample population.18 ____________ 16Karoly and Burtless (1995) suggest an alternative to this assumption: an adjustment for family size that also allows for some sharing among residents of the same household who are not related. 17Separate family-level observations are constructed for each single person and for each secondary family within a household. 18See Appendix C for distribution trends using the Census Bureau definition of primary family. 26 Table 2.5 shows the change in income levels at the 20th, median, and 80th percentiles of the distribution of adjusted family income among persons. Adjusted family income at the 20th percentile fell 11 percent in California between 1969 and 1989. During the same period, income grew by 12 percent at the median and by 24 percent at the 80th percentile. This widening of the distribution led to a 39 percent increase in the 80/20 ratio over the period. In the nation, family income growth was higher at each percentile and the 80/20 ratio increased by 27 percent. Relative to the results for adjusted household income, the growth in the 80/20 ratio for adjusted family income is higher in every period. Figure 2.4 illustrates the rise in inequality for adjusted family income, using the four summary measures discussed above. Adjusted Table 2.5 Percentage Change in Real Adjusted Family Income Between Selected Years, by Income Percentile California 20th Median 80th Change in 80/20 ratio (%) Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 –2 –9 –11 –27 10 1 12 –7 18 5 24 14 +20 +16 +39 +56 United States 20th Median 80th Change in 80/20 ratio (%) 8 17 21 +11 –1 7 –6 7 25 7 13 36 19 +14 +27 +26 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. 27 .83 CV .81 .79 CA .77 U.S. .33 ENTROPY .31 CA .29 U.S. .75 .27 .73 .25 .71 .69 .23 .67 .21 .65 .63 .19 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 .55 MLD .50 CA U.S. .45 .40 .35 2.75 2.55 2.35 2.15 1.95 1.75 1.55 VLN CA U.S. 1.35 .30 1.15 .25 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Statistics reported in this figure are not sensitive to the consumer price index. Adjusted family income in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 2.4—Summary Measures of Inequality for Real Adjusted Family Income, 1967–1994 28 family income inequality exhibits many of the same trends as does adjusted household income inequality. Inequality has increased in both California and the nation since the recession of the early 1980s. Inequality has increased more rapidly in the state than in the nation since the late 1980s, but the faster growth of inequality in California began even earlier for family income than for household income. Levels of adjusted family income inequality were similar to those of adjusted household income inequality in the late 1960s. By the 1990s, measures of inequality were considerably higher for families than for households, especially using those measures that emphasize income at the bottom of the distribution. 29 3. Trends in the Distribution of Labor Income This chapter examines trends in the distribution of labor income, the largest component of household income. There are a number of reasons for looking at labor income inequality as well as household income inequality. Trends in adjusted household income inequality are complicated by societal changes in family size and marriage behavior.1 In contrast, labor income inequality measures the disparity of income of individuals rather than families, and it is not directly affected by changes in household structure. While adjusted household income may be a better indicator of general economic well-being, labor income provides a clearer picture of changes in the economy. ____________ 1The increase in female-headed households has affected the distribution of adjusted household income, as have the falling marriage rates for men. In the past, the wives of low-income men were more likely to have earnings than the wives of high-income men. In recent years, there has been an increase in the number of families with two professional-level salary earners. The increasing correlation of husbands’ and wives’ earnings also has affected the distribution of adjusted household income. 30 As measured by annual earnings and hourly wages, the trends in labor income reveal many of the same patterns found for household income. Since the early 1980s, both California and the nation have experienced growth in labor income inequality. This result holds true for multiple definitions of labor income and measures of inequality. As was true for household income, the rising inequality of male annual earnings began in the early 1970s. In contrast, the inequality of female annual earnings declined substantially during that decade and did not begin to rise until the early 1980s. Since 1975, male hourly wages at the top of the distribution have shown slow growth. For low-wage male workers, hourly wages have declined considerably. For female workers, in contrast, hourly wages have grown near the top of the distribution. Near the bottom of the distribution, female wages declined by a small amount in California and grew by a small amount in the nation. What Is Labor Income? Labor income is income from work. Labor income comprises income from wages, salary, self-employment, and one’s own farm. For people who receive income from their own farm or from selfemployment, however, reports of “earnings” often include income from previous capital investments such as ownership of the farm or business.2 This income from capital is not part of labor income. Therefore, the data sample used to study labor income excludes workers who report that ____________ 2For example, a restaurant owner might show higher net income if she owns, rather than rents, her stoves. 31 their primary occupation was “self-employed”3 and workers who receive a substantial income from self-employment or from their own farm.4 After making these sample exclusions, we compute annual earnings as the sum of earnings from wages and salaries plus income from self-employment and farms. To ensure that the findings do not depend on sample exclusions, we compare these results to distribution trends for total earnings among all adult workers regardless of self-employment or farm owner status. We also measure the trends in the distribution of income from wages and salary for all adults with income from these sources. Finally, we examine trends for a subsample of workers between ages 18 and 55 to remove any effects of early retirement. As is customary, all samples are limited to civilians age 18 and older who are not students and who report some earnings.5 The data on annual earnings include only pre-tax monetary compensation. A brief discussion of the effect of non-monetary ____________ 3People who are self-employed in incorporated businesses are not identified in the CPS before 1975. To maintain the same sample definition throughout all years, these people are not excluded from the sample in any year. 4The sample excludes people who report more income from their farm or business than from wages and salaries and excludes any person reporting an absolute value of more than $2,000 in income in 1994 dollars from their own farm or self-employment. Some wage and salary workers included in the sample receive a small amount of income from farms and self-employment. This income was included in annual earnings to improve estimates of hourly wages, because estimates of annual hours of work include hours worked in the farm or business. The measure of annual earnings used in this study is similar to that of Karoly (1993). Although both studies exclude people who classify themselves as “self-employed,” our study additionally excludes people who receive substantial income from self-employment or farms. Also, Karoly does not include even small amounts of income from farms or self-employment in annual earnings. The results reported here for the nation are similar to those of Karoly. 5The sample also excludes people who report that their primary position was “without pay.” 32 compensation and taxes on the distribution of income can be found in Appendix C. In this chapter, we evaluate trends in both annual earnings and hourly wages. Neither measure by itself allows for a complete understanding of changes in labor income inequality: The distribution of hourly wages gives little indication of total annual earnings; the distribution of annual earnings is confounded by differences in hours of work. Using both of these measures, along with household income as discussed in the previous chapter, provides a more complete picture of income inequality in California. We examine the trends in income inequality for male and female workers separately because of the recent significant changes in the labor force participation of women. Over the years of the study, women’s labor force participation rate has increased from 51 percent to 61 percent in California; between 1975 and 1994, the average hours worked per year among adult women in the labor force increased from 780 to 980.6 Trends in the Distribution of Labor Income Among Males Inequality in male annual earnings and hourly wages is rising. There has been a slow growth in annual earnings and hourly wages near the top of the distribution and a substantial decline near the bottom—and even at the median after 1986 in California. In addition, California had one ____________ 6Labor market participation rates are based on the authors’ calculations from the CPS. The sample includes civilian women age 18 and older. The CPS began including information on hours of work only in 1975. Annual hours are calculated as the product of annual weeks of work and usual hours worked per week of work. 33 of the highest increases of any state in male annual earnings inequality between 1969 and 1989. The Widening Distribution of Male Annual Earnings Figure 3.1 illustrates the trends in annual earnings between 1967 and 1994 for the 10th, 20th, median, 80th, and 90th percentiles of male workers for California and the nation. The figure shows much the same pattern as observed for adjusted household income: The distribution of male annual earnings widened over the past three decades, with the most noticeable increases occurring during periods of recession in the early and mid 1970s, the early 1980s, and the early 1990s. During each recession, male annual earnings fell drastically for the lower and lower-middle positions of the distributions in California and the nation. The decline in male annual earnings was greater in California than in the nation because of slower growth in recovery periods and more rapid decline in the recessions of the early 1970s and early 1990s. For example, in California, men at the 20th percentile in 1971 had 16 percent lower annual earnings than men at the 20th percentile in 1969. For the nation, the decline was 8 percent. The recession of the early 1990s hit California even harder. Between 1989 and 1993, the 20th percentile fell 14 percent in the nation but 27 percent in California. While the trends in the distribution of male earnings are similar to those of adjusted household income in their overall shape, male earnings exhibit much slower growth. Table 3.1 summarizes the trends in male earnings for comparable years in the business cycle. The 80th percentile 34 Percent change since 1967 Percent change since 1967 30 California 20 10 0 –10 –20 10th –30 20th Median 80th –40 90th –50 1967 1971 1975 30 United States 20 1979 1983 1987 1991 1994 10 0 –10 –20 10th 20th –30 Median 80th –40 90th –50 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Sample includes civilians age 18 and older who received wage and salary income. Sample excludes students, those self-employed who are not in incorporated businesses, workers whose primary position is unpaid, workers who receive more farm or self-employment income than wage and salary income, and workers who receive more than $2,000 (real 1994 dollars) from farm or self-employment income. Annual earnings are computed as the sum of earnings from wages, salaries, self-employment and farms. Statistics reported in this figure are sensitive to the consumer price index. Real annual earnings in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.1—Percentage Change in Real Annual Earnings for Males, by Income Percentile, 1967–1994 35 Table 3.1 Percentage Change in Real Annual Earnings for Males Between Selected Years, by Income Percentile: CPS Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –15 –2 11 +30 –21 –12 –5 +21 –33 –27 –13 –20 62 +57 +41 United States 20th Median 80th Change in 80/20 ratio (%) –4 5 11 +16 –14 –7 4 +22 –18 –14 –2 –13 16 4 +42 +21 SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. of male earnings climbed only 6 percent in California between 1969 and 1989, in contrast with a 26 percent growth in adjusted household income. After 1979, the nation displayed a decline in the median of male earnings, particularly during recessions. The median in California fell throughout the period of the study, dropping 13 percent between 1969 and 1989. Surprisingly, the most recent decline in median male earnings began as early as 1987 in California, three years before the most recent recession. Despite the slow growth at the top of the distribution of male annual earnings in California, the 80/20 ratio increased a staggering 57 percent between 1969 and 1989 because of the drastic decline in earnings at the 36 20th percentile. The national 80/20 ratio also shows a remarkable (though smaller) increase of 42 percent. Summary Measures Show Rising Inequality in Male Annual Earnings As Figure 3.2 shows, the summary measures of inequality (introduced in Chapter 2) demonstrate the increasing trend in male annual earnings inequality over the last three decades. Male annual earnings show a clear pattern: Inequality rose sharply during recessions and remained at new, higher levels during recovery periods. In many cases, in fact, inequality continued to increase even during periods of growth.7 Beginning in the 1970s, male earnings inequality was consistently higher in California than in the nation, except for a brief period in the mid 1980s. The measures show that the gap between California and the United States began to widen noticeably as early as 1987. Although inequality did increase in the nation during the most recent recession, levels of inequality in California continue to be considerably higher. Census Data Show Rising Inequality of Male Annual Earnings The Census results, shown in Table 3.2, confirm the large decline in earnings near the bottom of the distribution and the slow growth in ____________ 7All four measures (and especially the VLN) appear to show a decrease in inequality in the early 1980s because of the large spike in inequality between 1979 and 1982. The cause of this spike is clear in Figure 3.1: Earnings fell sharply for the bottom of the distribution between 1979 and 1982 and then showed some compensating recovery in the next few years. If the spike is ignored, the continuing upward pattern of increasing inequality is clear. 37 .80 CV .75 CA U.S. .70 .65 .31 ENTROPY .29 CA .27 U.S. .25 .23 .21 .19 .60 .17 .55 .15 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 .40 MLD .38 .36 CA U.S. .34 .32 1.4 VLN 1.3 CA 1.2 U.S. 1.1 .30 1.0 .28 .9 .26 .24 .8 .22 .7 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are not sensitive to the consumer price index. Real annual earnings in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.2—Summary Measures of Inequality for Male Annual Earnings, 1967–1994 38 Table 3.2 Percentage Change in Real Annual Earnings for Males, by Income Percentile: Census Annual Earnings 1969–1979 1979–1989 1969–1989 California 20th Median 80th Change in 80/20 ratio (%) United States 20th Median 80th Change in 80/20 ratio (%) –18 –3 10 +35 –6 6 11 +18 –18 –33 –12 –14 –7 3 +13 +53 –9 –14 –9 –4 –1 9 +8 +28 SOURCE: Based on authors’ calculations from the decennial Census. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. earnings near the top found by the CPS. The main difference between the Census and CPS results is that over the period 1969 to 1989, the national 80/20 ratio increased by 28 percent according to the Census and by 42 percent according to the CPS. For California, the results are closer: a 53 percent increase according to the Census and a 57 percent increase according to the CPS. The Census data show slower growth but higher levels of inequality. For example, the coefficient of variation for male annual earnings was 0.63 in California in 1969 as calculated from the Census (see Table 3.3). 39 Table 3.3 Levels and Trends in the Coefficient of Variation for Male Annual Earnings: CPS and Census California CPS Census United States CPS Census CV: Level 1969 1979 1989 0.56 0.63 0.56 0.63 0.65 0.72 0.62 0.68 0.75 0.76 0.70 0.71 CV: Percent change 1969–1979 1979–1989 1969–1989 15 14 32 14 11 9 5 13 4 20 26 14 SOURCE: Based on authors’ calculations from the March CPS and the decennial Census. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are not sensitive to the consumer price index. For that same year, the CV calculated from the CPS was 0.56. Despite these differences, both datasets show a substantial rise in inequality in California that exceeded the rise in the United States. The Widening Distribution of Male Hourly Wages We examine trends in hourly wages, in addition to annual earnings, to observe changes in salary separate from changes in hours worked. Hourly wages are calculated by dividing annual earnings by annual hours; annual hours are the product of weeks worked and usual hours worked per week of work.8 Figure 3.3 shows the widening distribution ____________ 8Hourly wages are measured imprecisely because they are calculated from annual data. This imprecision leads to extreme values in some years (e.g., some years have several observations with an hourly wage of less than $1). To avoid fluctuation in the summary measures of inequality due to extreme values, hourly wages were top-coded at 97 percent and bottom-coded at 3 percent in all years. 40 Percent change since 1975 Percent change since 1975 10 California 0 –10 10th 20th –20 Median 80th 90th –30 –40 1975 1978 1981 1984 1987 1990 1994 10 United States 0 –10 10th 20th –20 Median 80th 90th –30 –40 1975 1978 1981 1984 1987 1990 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are sensitive to the consumer price index. Real hourly wages in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.3—Percentage Change in Real Hourly Wages for Males, by Income Percentile, 1975–1994 41 of male hourly wages. (The figure begins in 1975 because information on hours worked per week is not available in earlier years of the CPS.9) The most striking features of Figure 3.3 are the slow growth at the top and the decline at the bottom of the distribution of male hourly wages in both California and the United States. Over most of the period, inequality increased even though wages at the 90th percentile were never more than 10 percent higher than they had been in 1975. Table 3.4 summarizes these trends. Between 1979 and 1989 in California, male hourly wages fell by 21 percent at the 20th percentile and fell by 14 percent at the median. Even at the 80th percentile, wages fell 2 percent. The nation exhibited a similar pattern but with smaller declines. Because of the faster decline in wages at the 20th percentile in California, the 80/20 ratio increased by 24 percent in the state compared to 17 percent in the nation. Summary Measures Show Rising Inequality in Male Wages The summary measures of inequality shown in Figure 3.4 confirm that male wage inequality has risen steadily and significantly in California since 1977. As with male earnings inequality, male hourly wage inequality was higher in California than in the nation for most of the years of the study. Since the late 1980s, however, male wage inequality ____________ 9Before the 1976 survey, the CPS did not ask about hours of work in a usual week in the previous year. This information is needed to calculate hourly wages from annual earnings in the previous year. Hourly wages were not computed from the Census data because the 1970 Census survey did not ask about hours of work per week in the previous year. 42 Table 3.4 Percentage Change in Real Hourly Wages for Males Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1979–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –21 –14 –2 +24 –30 –22 –2 +40 United States 20th Median 80th Change in 80/20 ratio (%) –12 –6 3 +17 –19 –13 1 +25 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. grew more rapidly in California, widening the difference in inequality between California and the nation. Male Labor Income Inequality Rose Faster in California Than in Other Regions and States Relative to the other regions of the country, California experienced high growth in male hourly wage inequality. Table 3.5 shows the CV for 43 .67 CV .65 .63 CA U.S. .61 .59 .57 .55 .53 .51 .49 1975 1979 1983 1987 .21 ENTROPY .20 .19 CA .18 U.S. .17 .16 .15 .14 .13 .12 .11 1991 1994 1975 1979 1983 1987 1991 1994 .22 MLD .21 .20 CA U.S. .19 .18 .46 VLN .44 .42 CA U.S. .40 .38 .17 .36 .16 .34 .15 .32 .14 .30 .13 .28 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are not sensitive to the consumer price index. Real hourly wages in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.4—Summary Measures of Inequality for Male Hourly Wages, 1975–1994 44 Table 3.5 Regional Trends in the Coefficient of Variation for Real Hourly Wages Among Males, 1979–1994 Region CV (Rank) Percentage Change in CV (Rank) 1979 1989 1994 1979–1989 1989–1994 California 0.52 0.61 0.66 (3) (2) (1) New England 0.52 0.54 0.56 (6) (9) (10) Mid Atlantic 0.48 0.57 0.60 (9) (6) (5) E. N. Central 0.45 0.53 0.58 (10) (10) (8) W. N. Central 0.50 0.57 0.57 (8) (7) (9) S. Atlantic 0.54 0.58 0.63 (1) (4) (4) E. S. Central 0.52 0.57 0.60 (4) (5) (6) W. S. Central 0.54 0.63 0.64 (2) (1) (2) Mountain 0.52 0.56 0.59 (5) (8) (7) Pacific 0.51 0.59 0.64 (7) (3) (3) 18 (3) 5 (10) 18 (1) 18 (2) 14 (6) 8 (9) 10 (7) 16 (4) 8 (8) 16 (5) 7 (4) 3 (8) 6 (5) 10 (1) 1 (10) 8 (3) 5 (6) 2 (9) 5 (7) 8 (2) SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are not sensitive to the consumer price index. California and the nine geographic regions, including California’s own Pacific region. In 1979, the CV of male hourly wages in California was the third-highest among the ten regions; in 1989, it was second; by 45 1994, inequality was higher in California than in any other region.10 Appendix D presents the same analysis for male annual earnings. Comparing male annual earnings trends between the states further emphasizes the growth in inequality in California. The Census data show that 20 states had greater inequality than California in 1969, as measured by the CV. Between 1969 and 1989, California was tied with Indiana and Ohio for the fastest percentage growth in male earnings inequality in the country. In 1989, only two states had higher levels of male earnings inequality. Appendix D reports the CV for all 50 states in 1969 and 1989. Other Definitions of Male Labor Income Show Rising Inequality The labor income results reported in the previous sections are based on a data sample that excludes workers who are primarily self-employed or farm owners. This sample definition is preferable because of the difficulty in separating capital income from labor income for people who work in their own business or farm. However, including these workers and looking at the sum of income from wages, salary, self-employment, and farms does not alter the basic trends of decline near the bottom of the distribution, slow growth near the top, and rising inequality. For California between 1969 and 1989, the decline in male annual earnings at the 20th percentile was 30 percent among all workers, compared to 33 percent in the restricted sample of wage and salary workers (shown in Table 3.1). The growth in male earnings at the 80th percentile was about ____________ 10This result is consistent with the regional inequality trends found by Karoly and Klerman (1994). 46 5 percent in both samples. Over the same period, the growth in the CV was 25 percent for all workers and 32 percent for wage and salary workers. The same general trends also hold for the distributions of hourly wages among all workers, of annual earnings and hourly wages among wage and salary workers ages 18 to 55, and of wage and salary income. See Appendix C for further details. Trends in the Distribution of Labor Income Among Females The trends in inequality of female annual earnings are quite different from those of male annual earnings. The distribution of female annual earnings narrowed during the 1970s, when women’s incomes rose substantially near the bottom of the distribution. In the 1980s, the declining inequality of female annual earnings either slowed or reversed itself, depending on which measure of inequality is used. In contrast, all the measures show that the inequality of hourly wages among women increased during the 1980s. The difference between the trends in annual earnings and hourly wages suggests that some of the increase in female earnings, especially in the lower ranks of the distribution, was due to increased hours of work. The levels and trends of female labor income inequality, however, were nearly identical in California and the nation, even over the last decade. The Narrowing, Then Widening, Distribution of Female Annual Earnings In contrast to male annual earnings and adjusted household income, female annual earnings inequality actually declined between 1967 and 47 the early 1980s. As Figure 3.5 shows, growth was fastest among women at the bottom of the distribution during the 1970s in both California and the nation: The upper lines on the figure represent earnings growth at the 10th and 20th percentiles. The relative gains of the lowest-earning women were short-lived, however. In the 1980s, annual earnings began to grow for women in the upper half of the distribution, and, interestingly, did not show the same tendency as male annual earnings to fall during recessions. Earnings at the lower percentiles did continue to grow but fell during recessions, especially in California. Table 3.6 allows us to see these trends clearly. The income at the 20th percentile increased over the 1970s, growing 61 percent in Table 3.6 Percentage Change in Real Annual Earnings for Females Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) 61 22 10 –31 13 82 15 6 29 27 22 35 30 +8 –26 +13 United States 20th Median 80th Change in 80/20 ratio (%) 44 20 16 –20 20 72 11 33 20 39 +0 –20 43 28 33 –7 SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. 48 Percent change since 1967 Percent change since 1967 160 140 California 120 10th 20th 100 Median 80th 80 90th 60 40 20 0 1967 1971 1975 160 140 United States 120 10th 100 20th Median 80 80th 90th 60 40 20 0 1967 1971 1975 1979 1979 1983 1983 1987 1987 1991 1994 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are sensitive to the consumer price index. Real annual earnings in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.5—Percentage Change in Real Annual Earnings for Females, by Income Percentile, 1967–1994 49 California and 44 percent in the nation. Female earnings at the 80th percentile grew at a slower pace, so that the 80/20 ratio fell by 31 percent in California and by 20 percent in the nation over that decade. In California in the 1980s, the 80th percentile grew faster than the 20th, leading to an 8 percent increase in the 80/20 ratio. In the nation, the 80/20 ratio was the same in 1989 as in 1979. Measures of Inequality Show Falling, Then Rising, Inequality in Female Annual Earnings The summary measures in Figure 3.6 all show that female earnings inequality fell during the 1970s and increased beginning in the early 1980s. After 1986, the trends in inequality are dependent on which measure is used. The CV shows a continuing increase throughout the 1980s, whereas the other measures exhibit fluctuations without clear trends.11 Unlike trends in inequality for male earnings and household income, trends in female earnings inequality were virtually identical in California and the nation, even in the late 1980s. Inequality did rise more sharply in California in the early 1990s, but the difference between the state and the nation had narrowed substantially by 1994. Census Data Show a Fall and Then a Rise in Inequality of Female Annual Earnings Table 3.7 shows the trends in the distribution of female annual earnings calculated from the Census data. The picture of growth is ____________ 11The fact that the measures do not agree on whether inequality was higher in the 1990s than in the early 1970s reflects the tremendous growth in the income of the lowest-earning women. The two measures that put more weight on the bottom of the distribution—the MLD and the VLN—do not show as steep an increase in inequality as the other measures do. 50 .84 CV .33 ENTROPY .82 .32 CA CA U.S. .31 U.S. .80 .30 .78 .29 .76 .28 .74 .27 .72 .26 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 .52 1.7 MLD VLN .50 1.6 CA CA .48 U.S. U.S. 1.5 .46 1.4 .44 1.3 .42 1.2 .40 .38 1.1 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are not sensitive to the consumer price index. Real annual earnings in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.6—Summary Measures of Inequality for Female Annual Earnings, 1967–1994 51 Table 3.7 Percentage Change in Real Annual Earnings and Hourly Wages for Females, by Income Percentile: Census California 20th Median 80th Change in 80/20 ratio (%) Annual Earnings 1969–1979 1979–1989 1969–1989 52 13 12 14 9 18 72 27 29 –28 +4 –25 United States 20th Median 80th Change in 80/20 ratio (%) 30 15 11 –15 22 58 12 28 19 33 –2 –16 SOURCE: Based on authors’ calculations from the decennial Census. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. similar to that found using the CPS, shown in Table 3.6. In California, the 80/20 ratio decreased in the 1970s and then increased by a small amount in the 1980s. The overall decline in the 80/20 ratio between 1969 and 1989 was nearly identical in the CPS and the Census (26 percent versus 25 percent for California). The Census data do not show as much growth in female earnings as the CPS data do. Table 3.8 depicts the very similar levels of female annual earnings inequality between the CPS and the Census data, as measured by the CV. In addition, the changes in the CV are also similar—the main 52 Table 3.8 Levels and Trends in the Coefficient of Variation for Female Annual Earnings: CPS and Census California CPS Census United States CPS Census CV: Level 1969 1979 1989 0.78 0.77 0.77 0.77 0.76 0.78 0.74 0.77 0.80 0.84 0.79 0.78 CV: Percent change 1969–1979 1979–1989 1969–1989 –3 4 2 1 –3 –1 8 62 9 31 SOURCE: Based on authors’ calculations from the March CPS and the decennial Census. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are not sensitive to the consumer price index. difference is that the Census shows higher inequality growth in California than the CPS does. The Widening Distribution of Female Hourly Wages The trends in the distribution of female annual earnings reflect the increase in hours worked by women in the labor market. Among women who work, average hours increased 26 percent between 1975 and 1994 in California. Examining the trends in the distribution of hourly wages removes the effect of hours of work. These trends are portrayed in Figure 3.7. Like the distribution of female annual earnings, the distribution of female hourly wages narrowed between 1975 and 1979 as wages for the lowest-paid women rose quickly. During the recession of the early 53 Percent change since 1975 Percent change since 1975 30 California 20 10 0 10th 20th Median 80th 90th –10 1975 1978 1981 1984 30 United States 10th 20 20th Median 80th 90th 10 1987 1990 1994 0 –10 1975 1978 1981 1984 1987 1990 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are sensitive to the consumer price index. Real hourly wage in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.7—Percentage Change in Real Hourly Wages for Females, by Income Percentile, 1975–1994 54 1980s, however, wages fell throughout the distribution. When female wages began to rise again, they grew in a familiar pattern: Wages grew fastest at the upper percentiles. In California, female hourly wages fell at the 10th percentile between 1985 and 1994. In contrast to male wages, female wages near the top of the distribution in California grew over the 1980s, and even in the early 1990s. Moreover, female wages did not show the same strong influence of recessions. In addition, female wages grew only slightly faster in the nation than in California. As Table 3.9 shows, wages at the median increased by 2 percent in California and by 8 percent in the nation between 1979 and 1989. At the 20th percentile, female wages fell 9 Table 3.9 Percentage Change in Real Hourly Wages for Females Between Selected Years, by Income Percentile Business Cycle Peaks 1979–1989 California 20th Median 80th Change in 80/20 ratio (%) –9 2 15 +26 Recessions 1976–1994 –8 8 18 +28 United States 20th Median 80th Change in 80/20 ratio (%) –5 8 16 +22 –2 10 22 +24 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. 55 percent in California compared with 5 percent in the nation over the same period. At the upper end of the distribution, the increase was nearly identical in California and the United States. Measures of Inequality Show Rising Inequality in Female Hourly Wages Female hourly wages show a clear upward trend in inequality beginning in the early 1980s, as depicted in Figure 3.8. This is in contrast to the results for female annual earnings, which depend on the measure of inequality used. Like female annual earnings, hourly wage inequality in California tracks closely with that of the nation. Female Labor Income Inequality Is Similar in California to Other Regions and States Table 3.10 compares female wage inequality in California and in the regions of the country. It confirms the finding, shown in Figure 3.8, that, in contradistinction to household and male labor income, female wages did not show higher levels of inequality in California than in the nation. In the business cycle peak of 1979 and in the business cycle trough of 1994, California’s level of female earnings inequality was firmly in the middle of the regions. Even its apparent high ranking in 1989 is somewhat misleading: Five of the regions had levels of inequality nearly identical to California’s in that year. Appendix D presents the same analysis for female annual earnings. Compared with other states, California had a moderate level of female earnings inequality in 1989 (as measured by the CV)—16 states had higher levels. However, between 1969 and 1989, 39 states experienced larger declines in inequality than California did. 56 .66 CV .64 .62 CA U.S. .60 .58 .56 .54 .52 .19 ENTROPY .18 CA .17 U.S. .16 .15 .14 .13 .50 .12 .48 .11 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 .21 MLD .20 .19 CA .18 U.S. .44 VLN .42 .40 CA .38 U.S. .17 .36 .16 .34 .15 .32 .14 .30 .13 .28 .12 .26 .11 .24 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are not sensitive to the consumer price index. Real hourly wages in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.8—Summary Measures of Inequality for Female Hourly Wages, 1975–1994 57 Table 3.10 Regional Trends in the Coefficient of Variation for Real Hourly Wages Among Females, 1979–1994 CV (Rank) Percentage Change in CV (Rank) Region 1979 1989 1994 1979–1989 1989–1994 California New England Mid Atlantic E. N. Central W. N. Central S. Atlantic E. S. Central W. S. Central Mountain Pacific 0.50 (5) 0.48 (9) 0.49 (6) 0.49 (7) 0.48 (10) 0.50 (4) 0.49 (8) 0.51 (2) 0.55 (1) 0.51 (3) 0.57 (2) 0.53 (10) 0.57 (1) 0.55 (8) 0.57 (5) 0.56 (6) 0.55 (7) 0.57 (3) 0.54 (9) 0.57 (4) 0.59 (5) 0.57 (10) 0.60 (2) 0.60 (4) 0.57 (9) 0.61 (1) 0.58 (8) 0.59 (6) 0.60 (3) 0.59 (7) 15 (3) 10 (9) 16 (2) 13 (5) 19 (1) 12 (8) 14 (4) 12 (7) –1 (10) 12 (6) 4 (8) 6 (4) 5 (5) 9 (3) 0 (10) 9 (2) 5 (6) 4 (9) 10 (1) 4 (7) SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are not sensitive to the consumer price index. Other Definitions of Female Labor Income Show Falling, Then Rising, Inequality As was true for male labor income, alternative sample definitions do not alter the basic trends in female labor income reported in the previous sections. Female annual earnings grew rapidly near the bottom of the 58 distribution and inequality declined until the early 1980s. There is one notable difference in results between the sample of all workers and the restricted sample of wage and salary workers: Annual earnings inequality among all female workers followed an increasing trend after 1983 for all the summary measures of inequality. This is in contrast to the trends in inequality after 1983 for female wage and salary workers, which depended on the summary measure used (as shown in Figure 3.6). See Appendix C for further results using the alternative definitions of female annual earnings and hourly wages. 59 4. Conclusions and Implications for Policy and Future Research Our study finds a large increase in income inequality in California over the last three decades for both household income and male earnings. This rise in income inequality is explained by a dramatic decline in income at the lower and lower-middle ranks of the distribution, and a simultaneous growth in income in the upper ranks. The trends in income inequality show a strong relationship to the business cycle: Inequality grew fastest during the recessions of the early 1970s, early 1980s, and early 1990s. Until the late 1980s, the levels and trends in income inequality in California and the nation were similar. Since that time, inequality has grown faster in California than in the United States. Moreover, compared to the nation, California has experienced slower income growth throughout the distribution. 60 Provocative as these findings are, measuring the trends in the distribution of income is only the first step in understanding them and their implications for policy. As California designs programs to promote equity, it will benefit from research on the relationship between existing state policies and income inequality, as well as from a better understanding of the causes of rising inequality in the state. In addition, the income trends measured in this study provide an incomplete picture of the distribution of economic well-being. Distribution trends for morecomprehensive definitions of income (e.g., accounting for taxes and nonmonetary compensation) and the issue of income variability remain to be studied. Public Policy and the Distribution of Income Continued growth in income inequality is not inevitable. As a society, we face a choice as to whether we will act to reverse the trend in growing income inequality. Many policy mechanisms already exist for reducing inequality. Progressive taxes, for example, directly redistribute income. Quality public schools and access to higher education provide an opportunity for people of all income levels to invest in themselves and improve their future incomes. Research on the role of the existing policy mechanisms, as well as identification of new policy options, is essential for understanding how California state policy influences the distribution of income. Some Americans believe that differences in income arise primarily from individual choices, preferences, abilities, investments, and productivity, and that income inequality is a product of an economy that values hard work and talent. Other Americans believe that income 61 differences reflect the unequal distribution of economic opportunity in our society, and that the opportunity to succeed is elusive for those who do not belong to privileged groups. The first viewpoint implies that public policy can affect inequality only by redistributing income; the second implies that policy can reduce inequality by promoting opportunity. Research on the determinants of income distribution and the extent to which policy provides or restricts economic opportunity will suggest avenues for improving opportunities for the less-advantaged. If California seeks to reduce income inequality, the state will benefit from research that identifies policy options that promote equity as well as efficiency in our economy. Labor Market Explanations for Rising Earnings Inequality The similar trends in California and the nation suggest that the same forces that explain the widening of the income distribution in the United States account for the growth in income inequality in California. At the national level, the rise in male earnings inequality has been explained by a combination of factors. Economists agree that changes in the supply and demand of labor have favored skilled workers over less-skilled workers. The underlying forces that have led to these labor market trends include technological change, international competition, immigration, and deunionization.1 However, the effect of each of these forces on the distribution of earnings in California’s distinct economy remains to be studied. ____________ 1Cassidy (1995) provides a straightforward summary of these explanations for rising earnings inequality. 62 Many economists believe that technological change has benefited educated workers who are able to implement new technology and has harmed less-educated workers who may be replaced by mechanized production.2 The effect of technological innovation on California workers may be more pronounced than in the nation. On one hand, the state has a higher percentage of people with at least some college education (23 percent) than the national average (19 percent). On the other hand, the school dropout rate is 14 percent in California, 3 percentage points above the national average of 11 percent.3 There is less agreement on the role of international competition in explaining the rise in income inequality. The cost of a low-skilled workforce is higher in the United States than in other countries, particularly developing countries. Thus, the United States increasingly imports manufactured products and textiles, lowering the labor market demand for low-skilled U.S. workers.4 International competition may have played a different role in the state because California has more Pacific-region trade than the rest of the nation and because a slightly smaller percentage of the state’s workforce is in manufacturing (15 percent compared to the national 16 percent).5 Growth in immigration may have contributed to the rise in income inequality. Immigrants can adversely affect the wage distribution by ____________ 2Krueger (1993) finds evidence that supports this theory. 3Population statistics based on the 1990 Census as reported in U.S. Bureau of the Census (1994), Table 236. 4Borjas, Freeman, and Katz (1992) have found evidence that trade patterns account for a substantial part of the wage losses of high school dropouts. 5Workforce statistics for 1993 reported in U.S. Bureau of the Census (1994), Table 655. 63 raising the number of low-wage workers. Furthermore, by increasing the competition for low-skill employment, immigration can lead to a reduction in the wages offered to natives with low skills.6 The impact of immigration on California is likely to be greater than in the nation, since California has the largest foreign immigration of any state. The decline in the power of unions has reduced the bargaining power of labor with the likely effect of lowering the wages of labor relative to that of management.7 The decline of unions is frequently offered as an explanation for the more rapid growth in earnings inequality in the United States than in other industrialized countries. The sharp rise in income inequality in California beginning in the late 1980s is probably explained, in part, by the same forces that caused the strong recession of the early 1990s. In addition to cuts in defense spending, suggested causes of the severe recession in the state include a decline in residential building, a fall in commercial aircraft orders, and a reduction in spending relative to income.8 The effect of each of these factors on the distribution of income remains to be studied. Demographic Explanations for Rising Family Income Inequality In addition to the economic and labor market forces that explain the increase in earnings inequality in the nation, trends in marriage and ____________ 6Butcher and Card (1991) and Borjas, Freeman, and Katz (1992) find evidence of an effect of immigration on wage inequality. 7Freeman (1993) reports evidence that the decline in unions lowered the wages of blue-collar workers relative to wages of white-collar workers. 8These factors are discussed in a study by the Center for the Continuing Study of the California Economy (1994). 64 female labor force participation may contribute to the rise in household and family income inequality. Declines in the percentage of people who are married may explain a portion of the rise in family income inequality. The growing share of families that rely on the earnings of single mothers has increased the number of low-income families.9 In addition, low-income men are less likely to be married than high-income men and are thus less likely to have a spouse who contributes to family income.10 Trends in marriage behavior may have a larger effect in California than in the nation. Compared to the national average, California had a lower rate of marriage and a higher rate of divorce between 1980 and 1992.11 The growth in the female labor force participation has an undetermined effect on the distribution of family income. As the percentage of women with earnings increased, earnings inequality among women fell. In addition, the rising earnings of married women have increased family income and reduced inequality among married-couple families.12 At the same time, however, the increased contribution of the earnings of wives has further polarized the incomes of single people relative to those of married couples. Furthermore, the correlation of the earnings of husbands and wives has increased: The wives of men with high earnings tend to earn more than the wives of men with low ____________ 9Danziger and Gottschalk (1995) find that the rise in female headship increased the poverty rate by 1.6 percentage points between 1973 and 1991 (Table 5.3, p. 102). 10Burtless (1996) makes this observation. 11Marriage statistics reported in U.S. Bureau of the Census (1994), Table 146. 12Cancian, Danziger, and Gottschalk (1993) find that changes in the earnings of married women reduced income inequality among married-couple families between 1968 and 1988. 65 earnings.13 The effect of female labor force participation may have been different in California because the increase in the average hours worked by women has been smaller than in the nation.14 Additional Measurement Issues There are a number of measurement issues we could not explore with Census Bureau income data that are important for a more complete understanding of the recent trends in income inequality and their implications for public policy in California. The income data used in this study do not account for the effect of taxes and non-monetary compensation (e.g., housing subsidies, health insurance). While national studies show that using more comprehensive measures of income does not substantially change income inequality trends,15 the effect may be different in California. For example, the percentage of people in California without health insurance was 19.3 in 1992, compared to a national average of 14.7 percent.16 The statistics reported in this study describe the distribution of income in each year. Because a person is likely to occupy different places in the distribution of income during his or her lifetime, the distribution ____________ 13Karoly and Burtless (1995) show the rising correlation of earnings between husbands and wives. 14Mean annual hours worked increased from 362 to 576 (59 percent) in California and from 348 to 597 (72 percent) in the United States between 1975 and 1994. These statistics include women who do not work in the labor market (zero hours). Statistics are based on the authors’ calculations from the March CPS. 15See Appendix C for a brief review of the literature on the distribution trends of more comprehensive measures of income. 16Health insurance statistics are reported in U.S. Bureau of the Census (1994), Table 165. 66 of annual income may not accurately reflect the level of inequality in lifetime income. Research at the national level suggests that economic mobility, the changing of positions within the distribution, has remained stable or declined in recent decades.17 However, income variability, the year-to-year fluctuations in income, appears to explain a substantial portion of the increase in male earnings inequality.18 Income mobility and variability remain to be studied in California. The Challenge for the State The combination of the sharp rise in household income inequality in California that began even before the most recent recession, the stagnation and decline of male wages, and the decline of household income for the lower and lower-middle ranks of the distribution pose a challenge to public policy in California. Can state policy help to meet the needs of low-income residents of the state and promote economic equity while not sacrificing economic growth? The answer to this question depends on the causes of recent trends and the policy options for the state. Future reports in this series will address these issues. ____________ 17Hungerford (1993) estimates income mobility in the United States in the 1970s and 1980s. 18Gottschalk and Moffitt (1994) find that one-third to one-half of the increase in the variance of earnings among white males from the 1970s to the 1980s can be explained by increased earnings instability. 67 Appendix A Notes on Data and Methodology This appendix addresses several limitations of the income data and the adjustments for price inflation and cost of living. When applicable, we describe our methodology for reducing the effect of these limitations on the estimated trends in income inequality. Income Data Income data for this study come from two national household surveys collected by the U.S. Bureau of the Census: the decennial Census of Population and Housing (1970, 1980, and 1990)1 and the ____________ 11970 Public Use Sample, 1 percent; 1980 and 1990 Public Use Micro Sample, 5 percent. 69 March Annual Demographic File of the Current Population Survey (public-use files, survey years 1968–1995).2 The Current Population Survey (CPS) and the Census report pre- tax, money income, which includes wages, salary, farm income, self- employment income, Social Security, railroad retirement, Supplemental Social Security, public assistance, welfare, interest, dividends, income from estates and trusts, net rental income, veterans’ payments, unemployment and workers’ compensation, private and government pensions, alimony, child support, regular contributions from persons not living in the same household, and other periodic income. Capital gains are not included. Current Population Survey The March file of the CPS, an annual survey of civilian households, provides detailed demographic information, including income received, for about 5,000 households in California and 50,000 households in the nation.3 The main benefit of using the CPS to study income ____________ 2Each survey has income information from the previous year. This study covers income years 1967–1994. Uniform series data files for CPS survey years 1964–1967, created under the direction of Robert Mare and Christopher Winship, are available from the University of Wisconsin. We chose not to use these files because of possible compatibility problems with the public-use files. We found much smaller average household sizes in the Mare-Winship files relative to the public-use files (e.g., the MareWinship file for 1967 had an average household size of 2.4 persons and the public-use file for 1968 had an average household size of 3.2 persons). The increase in household size leads to a sizable drop in adjusted household income between 1966 and 1967, whereas unadjusted household income shows a slight increase. We interpret the change in household size as evidence of a problem with the data and therefore we report statistics beginning with the 1968 public-use files. 3The March file of the CPS also includes Armed Forces personnel living with civilians. Our measures of household and family income include these households and families. Samples of workers do not include military personnel. About 3,000–4,000 male workers and 2,000–3,000 female workers are in the California sample. The national samples have about ten times as many workers as the California samples. 70 distribution is that it contains annual data, making it possible to observe short-term departures from long-term trends. As this study shows, the business cycle fluctuations observable in those data have strong effects on the distribution of income. The CPS allows us to use comparison years at the same stages of the business cycle when examining inequality changes over time. Over the period of the study, several changes were made in the design of the CPS, which could affect the comparability of the surveys across years. Survey changes that affect the distribution of income will result in one-time jumps in the measures of inequality but not in a pattern of changes across several years. We have confidence in the measured distribution trends discussed in the text because none of these results relies on a change that occurred in a single year. Each decade, the Census Bureau changes the sample design of the survey using population estimates from the most recent Census. The Census Bureau randomly selects a new sample of geographic areas called Primary Sampling Units (PSUs); in California, the PSUs are generally counties. To make the sample representative of all parts of the state, the PSUs are selected from groups of counties with similar population characteristics. Thus, even when the PSUs change, estimates of the income distribution should not be affected because each new PSU should be similar to the one it replaced. All significant sample design changes for this study occurred in 1972–1973 and 1985–1986 (the 1995–1996 redesign was implemented after the March 1995 survey).4 ____________ 4The Census Bureau does rotate “Enumeration Districts” within the Primary Sampling Units. However, substitutions in Enumeration Districts are chosen based on 71 The Census Bureau constructs sample weights such that the CPS sample will represent the national population. The sample weights are based on information from the decennial Census. In survey years 1973, 1982, and 1994, the Census Bureau revised the sample weights to reflect new population estimates from the 1970, 1980, and 1990 Census. Karoly (1993) compares income inequality in the original release of the 1980 CPS and in a reissue of the same survey using the new sample weights. She finds that the change in sampling weights had little effect on the increase in inequality between income years 1978 and 1979. The Census Bureau has changed the survey procedure with respect to Hispanics. In 1976, an additional sample of 2,000 Hispanic households was added to the March CPS to increase Hispanic representation. These households were chosen randomly from Hispanic households interviewed in the November CPS. The addition of these households could affect the measured distribution of income between income years 1974 and 1975. In 1984, the sample weighting procedure was changed to incorporate Hispanics explicitly. This change increased the estimated number of Hispanics and may have affected the distribution of income. In 1994, the Census Bureau automated the CPS survey questionnaire and introduced new sample weights. The Census Bureau (1996) reported that these changes may have increased measured income inequality. Finally, one specific problem with the CPS occurred in a single year of the survey. The median reported income received in 1988 shows an ____________________________________________________ similarity of population characteristics and geographic proximity and should not affect population and income statistics. 72 anomalous decrease in California. Karoly (1995) also reports a decline in income in California in 1988 based on the CPS data. Measures of income in California from other sources do not suggest a dip in 1988. For example, Department of Commerce data show the level of per capita income in 1988 about midway between 1987 and 1989 (California Statistical Abstract, 1995, Table D-7). The probable cause of this reported aberration lies not in California but in Washington. In 1989, funding for the CPS was cut and the sample for California fell to fewer than 3,000 households. The smaller sample size led to a higher sampling error for the 1988 data than in other years and may have affected the representativeness of the sample in that year.5 For this reason, we do not rely heavily on our results for 1988 in the California data. Funding was restored the following year and the California sample size returned to almost 5,000 in 1990. Census of Population and Housing We use the Public Use Sample of the Census to investigate the distribution of income and earnings in 1969, 1979, and 1989. One advantage of using the Census is that its larger sample size leads to more precise statistical estimates. In addition, the Census is designed to survey the entire population and therefore is representative of each state (with the important exception of undercount problems). The main limitation of the Census is that with only three years of data, we cannot distinguish long-run trends from short-run business cycle effects. However, because ____________ 5The sample weights were adjusted to reflect the smaller sample size, but the representativeness of the sample may still have been affected because the sample was not cut randomly, but only in Los Angeles. 73 the Census years are all business cycle peaks, changes in the distribution of income as reported in the Census are likely to represent trends rather than cyclical fluctuations. Although we expect to observe similar trends in the distribution of income using the CPS and the Census, income data from the two sources are not identical. For example, in 1990, the Census asked respondents about eight specific types of income. In the same year, the CPS asked about more than 20 types of income, making it less likely that a respondent will omit a source of income than when answering the Census questions. Also, the CPS is collected over the phone by trained survey-takers, who help improve the survey accuracy relative to the Census, which is done by mail. For this reason, we expect the CPS to reflect more accurately the sum of income from all sources as well as earnings and wages. The changes in the Census survey procedures were not as significant as changes in the CPS for measuring the trends in the distribution of income. It is worth noting that the Public Use Sample in 1970 (1 percent of the population) was much smaller than the Public Use Microdata Samples in 1980 and 1990 (5 percent of the population). Also, income in 1970 was reported as a range (e.g., $100–$199); we used the median of the ranges in our calculations. We did not calculate hourly wages with the Census data because the 1970 Census asks about hours of work in the previous week, as opposed to in a usual week in the previous year,6 and these hours are reported as a range (e.g., 1 to 14 hours). ____________ 6For this same reason, we do not calculate hourly wages in the CPS before 1975. 74 Top-Codes Both the CPS and the decennial Census restrict responses to income questions to a certain range. Responses outside of the range are “topcoded”: reported at the range cutoff points. For instance, from 1967 to 1975, sampled households with income above $50,000 were reported as income at $50,000 in the CPS. The range for reporting incomes changed over time. Increasing the magnitude of the top-code can increase measured income inequality even when the true underlying distribution of income does not change. To limit biases in our measures of inequality due to changing top-codes, we standardized the percentage top-coded across every year for each type of income for both surveys. Similarly, we recoded the same percentage in California and the United States. For example, the highest percentage of persons affected by the top-code of household income in the CPS was 98.8 percent (in 1975 in California). We recoded household income in every year of the CPS so that 98.8 percent of people were top-coded in both the state and the nation. Despite this recode, the top-coding can still affect estimates of trends in income distribution. The recode consistently top-codes total household income but not its component parts. In some cases, a person will have one component of income top-coded so that the sum of household income is affected by this top-code even when his or her household income is below the top of the range for household income. For example, a person with a salary of over $50,000 in 1980 will have that component of his income top-coded. This top-coding will affect the sum of income in his household, even though household income was not top-coded in that year. The measure of annual earnings used in the text 75 is not affected by this problem because we only need to recode based on a single component: income from wages and salary. The household size adjustment results in an additional problem with top-coding. A household with income top-coded at $50,000 in 1970 may not be in the top of the distribution of adjusted household income in that year if several people share that income. Top-coding will also dampen the magnitude of levels of inequality by masking the distribution of income above the cutoff points. As a result, an increasing concentration of income among the super-rich (the top 1 percent of income recipients) will not register in our measures of income inequality. Similarly, if the spread of income above the top-code is greater in some areas of the country than in others, the top-code will affect our comparisons of California to other states and regions. Thus, although we have recoded the data for consistent top-codes, the trends in adjusted household income are still affected by top-coding. Top-coding changed in each Census and in the CPS in years 1976, 1981, 1982, 1985, and 1989. Imputation Procedures In both the CPS and the Census, some respondents do not answer some of the income questions or answer inconsistently. When this happens, the Census Bureau uses a “hot deck” procedure to impute the missing income information from another person or household with similar characteristics. The hot deck procedure has changed over time in the CPS and in every year of the Census. In the CPS in 1976, education was added to the list of items used to define a hot deck match and the procedure was changed so that earnings, 76 weeks of work, and hours per week are supplied by the same matched observation. Juhn, Murphy, and Pierce (1993) show that these changes lowered estimates of hourly wage inequality. The trends in the distribution of hourly wages in this study begin with the 1976 survey and therefore are not affected by this change. In 1989, the CPS hot deck procedure was changed so that all income items are supplied by the same matched observation. In addition, the processing system was updated and more sources of income were added to the questionnaire. These changes led to an increase in aggregate income. To allow for comparisons with earlier survey years, the 1988 survey was reissued using the new 1989 processing system. Although we report income statistics for income year 1987 only from the reissue of the March 1988 CPS, all statistics in this study were also calculated with the original 1988 survey. A comparison of the results based on the two surveys shows that the new processing system reduces income inequality for all income metrics and all inequality measures, but the change is small. For example, the coefficient of variation of adjusted household income based on the original 1988 survey was 110.9; it was 110.7 based on the reissue with the new processing system. (See Appendix D for the decile levels of income in both issues of the 1988 survey. The original survey is labeled 1987a; the reissue, which we used, is labeled 1987.) Consumer Price Index and Cost of Living Adjustment All income statistics reported in this study have been adjusted to 1994 dollars based on the consumer price index computed by the Bureau of Labor Statistics (BLS). The consumer price index for California is 77 calculated by the California Department of Finance based on the population-weighted sum of the consumer price indices for San Francisco and Los Angeles (and San Diego between 1965 and 1986). The consumer price index used in this report is based on all urban consumers (CPI-U). In 1983, the method for calculating the CPI-U was changed to include a rental equivalence measure for owner-occupied housing. At the national level, the consumer price index was reissued for the years 1967 to 1982 to reflect this change (CPI-U-X1). The CPI-U-X1 series is the preferred price index because the CPI-U overstated inflation during the 1970s due to housing cost estimation procedures; after 1982, the CPI-U is the same as the CPI-U-X1. Because the CPI-U-X1 series is not available at the level of metropolitan areas before 1983, however, the California price index is based on the CPI-U. To construct a CPI-U-X1 series for California, we assumed that the ratio of (CPI-U)/(CPI-U-X1) in the national statistics is the same for the California statistics. Using this assumption and the CPI-U and CPI-U-X1 series for the nation and the CPI-U series for California, we computed an estimate of the CPI-U-X1 for California. The consumer price index provided by the BLS does not adjust for cost of living differences among regions. If the cost of living is higher in California than the national average, a higher income in California will have less purchasing power than a lower income elsewhere in the nation. Because of the difficulty in measuring the regional cost of living, the BLS stopped reporting this statistic in 1981. It is possible to create a cost of living series using the 1981 estimate of a 8.4 percent7 higher cost of ____________ 7The BLS reported a cost of living index for 24 standard metropolitan statistical areas (SMSAs) in 1981. Using the BLS index, McMahon (1991) calculated an index of 78 living in California and adjusting by the California consumer price index to create a yearly cost of living estimate. However, this estimate may not be accurate enough to allow reliable income comparisons between California and the nation. Median household income, reported in Figure 2.1, has been adjusted in this manner. All other statistics and figures reported in the text are insensitive to the cost of living adjustments. The price and cost of living adjustments for conversion to 1994 California dollars are summarized in Table A.1. The first column shows the CPI-U for California as reported by the California Department of Finance. The second column shows the CPI-U and the fourth column shows the CPI-U-X1 for the nation, as calculated by the BLS. The third column shows our calculation of a CPI-U-X1 for California using the assumptions described above. The fifth column converts Column 3 so that the 1994 value is equal to 1. The sixth column converts Column 4 to reflect the higher cost of living in California. As described above, the series was calculated by making the cost of living 8.41 percent higher in California than in the nation in 1981. To convert income data to 1994 California dollars, we multiply California data by Column 5 and national data by Column 6. The “Ideal” Data Several improvements in the quality of the data and the accuracy of the analysis could be made if California were to collect state-level data for ____________________________________________________ 108.41 in California (where the population-weighted average for the United States is 100). 79 Table A.1 Price and Cost of Living Adjustments, California and United States, 1967–1994 Year 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 12 CPI-U CA U.S. 33.0 34.4 36.1 37.9 39.3 40.6 43.0 47.4 52.3 55.6 59.5 64.4 71.3 82.4 91.4 97.3 98.9 103.8 108.6 112.0 116.6 121.9 128.0 135.0 140.6 145.6 149.4 151.5 33.4 34.8 36.7 38.8 40.5 41.8 44.4 49.3 53.8 56.9 60.6 65.2 72.6 82.4 90.9 96.5 99.6 103.9 107.6 109.6 113.6 118.3 124.0 130.7 136.2 140.3 144.5 148.2 34 CPI-U-X1 CA U.S. 35.9 37.3 38.8 40.3 41.8 43.1 45.7 49.9 54.6 58.0 62.1 66.7 72.7 82.3 90.6 96.4 98.9 103.8 108.6 112.0 116.6 121.9 128.0 135.0 140.6 145.6 149.4 151.5 36.3 37.7 39.4 41.3 43.1 44.4 47.2 51.9 56.2 59.4 63.2 67.5 74.0 82.3 90.1 95.6 99.6 103.9 107.6 109.6 113.6 118.3 124.0 130.7 136.2 140.3 144.5 148.2 56 COLA CA U.S. 4.22 4.50 4.07 4.33 3.91 4.15 3.76 3.96 3.62 3.79 3.51 3.68 3.31 3.46 3.04 3.15 2.77 2.91 2.61 2.75 2.44 2.58 2.27 2.42 2.08 2.21 1.84 1.98 1.67 1.81 1.57 1.71 1.53 1.64 1.46 1.57 1.40 1.52 1.35 1.49 1.30 1.44 1.24 1.38 1.18 1.32 1.12 1.25 1.08 1.20 1.04 1.16 1.01 1.13 1.00 1.10 SOURCES: Column 1, California Department of Finance; Columns 2 and 4, U.S. Bureau of Labor Statistics; Columns 3, 5, and 6, authors’ calculations. the study of the economy, including income inequality. Ideally, the state dataset would use a sample representative of the population of California that would be big enough to look at subregions and groups within the population. The dataset would be consistently collected over several 80 years and would include a panel component (i.e., would interview the same people over time). It would add to the value of the survey to have accurate inflation and cost of living estimates for the state. 81 Appendix B Using the Current Population Survey to Represent California1 The weighting procedure in the Current Population Survey (CPS) makes the survey representative of the nation as a whole. In calculating the March file weights, the Census Bureau does not attempt to correct for population distributions within states (e.g., California’s distinct racial distribution). Therefore, the California subsample of the March CPS may not accurately represent the population of the state. Before beginning our analysis of the distribution of income in California based ____________ 1The information on Census Bureau weighting procedures comes from the U.S. Department of Commerce and the Bureau of the Census (1978) and subsequent publications regarding redesign and revision of the CPS. 82 on the CPS data, we first evaluated the ability of the CPS to represent California.2 After conducting the March survey, the Census Bureau calculates a weight for each observation in the sample. The weight is based on a combination of factors, including adjustments to make the survey population match the national population’s distributions of age, sex, and race (with full interactions: age within sex within race). In survey years before 1978, the national sample weights did not specifically take into account the total number of people within each state. Since that time, several changes have been implemented in the calculation of the weights so that estimates of state populations based on the CPS sample are consistent with estimates of state total populations from other sources. However, the estimates of state populations within sex, age, and racial groups are not adjusted by the state-specific weights. Compared to the nation, California has had very different distributions of these characteristics, especially race. We suspected that adjusting the sample to match the U.S. distributions might severely affect the sample distributions for California. Furthermore, the Census Bureau bases the national weights on the decennial Census. Between Census updates, the weights are based on calculations of the population change. California’s distinct population trends provide another reason to suspect that the sample distributions would not accurately reflect the changes in California’s population. ____________ 2In addition to weighting procedures, sample design issues (e.g., changes in Primary Sampling Units) can also affect the representativeness of the CPS at the state level. The CPS sample design is discussed in Appendix A. 83 The Census Bureau does calculate labor force estimates at the state level from the CPS. For some states, this requires a state supplemental sample. However, the Census Bureau deemed the California subsample in the CPS large enough to make a state supplement for California unnecessary. Thus, we anticipated that the potential problem for our study was not sample size but rather whether the population distributions would be representative of the state. Although the Census Bureau believes that the California sample is large enough, it does construct state-specific weights (based on each state’s population distribution of race by residence) when it calculates statespecific labor force statistics from the CPS. The Census Bureau’s statespecific sample weights are not calculated for the March demographic survey used in this study (and are not provided in the public-use data). However, if the California subsample is found to be not representative of the state, state-specific weights for California could be constructed for the March survey using estimates of the California population by age, sex, and race (available from the Department of Finance). To determine whether a reweight of the California data was necessary, we evaluated the California subsample of the CPS on the distributions of residence, sex, age, and race—the same characteristics that the Census Bureau uses to weight the national sample. The CPS distributions were compared to the Census distributions for the years 1970, 1980, and 1990.3 Although the Census is also a national survey, it ____________ 3We expected that the representativeness of the CPS would be particularly poor in Census years. The Census Bureau recalculates the national weights based on each Census and applies updates of them in subsequent years—new weights were introduced in 1973, 1982, and 1994. Therefore, the weights in Census years are based on updates of tenyear-old population estimates. 84 is designed to survey the entire population in each state. The 5 percent sample of the Census used in this study is randomly chosen from the national population and is therefore representative at the state level (except for undercount problems). Table B.1 reports the distributions of farm households, sex, age, race, and ethnicity for California from the Census and the CPS.4 Judging by the similarity of the distributions, we concluded that the California subsample of the national CPS appears to represent the California population accurately with respect to these characteristics. Thus, we used the national sample weights in our calculations and did not reweight at the California level. The race and ethnicity distributions do show an interesting difference between the CPS and the Census. In 1980 and 1990, the distributions of race in the CPS do not match well with the distributions of race in the Census. However, when race and ethnicity are combined, the distributions from the two surveys match closely. This pattern suggests that people respond differently to race questions in the CPS and in the Census. Hispanic respondents in the Census are much more likely to record race as “other” than Hispanic respondents in the CPS. Although the statistics reported in Table B.1 certainly suggest that the California subsample of the CPS can be used to represent the state, further research is required to verify that the interacted distributions (sex within age within race) and the intercensal distributions are representative. Such verification is beyond the scope of this report. ____________ 4Deciles of the income distributions in the CPS and the Census also match fairly closely. See Appendix D. 85 Table B.1 Percentage of Population in Each Category: Census and CPS Characteristic 1970 1980 Census CPS Census CPS 1990 Census CPS Farm % non-farm household 90 n/a 99 99 Sex % male 48 48 49 49 Age in years 0–4 8 9 8 8 5–9 10 11 7 7 10–14 10 10 8 7 15–19 9 999 20–24 8 8 9 10 25–29 7 799 30–34 6 699 35–39 6 676 40–44 6 655 45–49 6 755 50–54 5 555 55–59 5 555 60–64 4 344 65–69 3 344 70–74 2 233 75–79 2 222 80+ 2 2 2 2 Hispanic Mexican n/a n/a 15 15 Other Hispanic n/a n/a 4 3 Not Hispanic n/a n/a 80 82 Race White 90 90 77 86 Black 7 688 Native American n/a n/a n/a n/a Asian n/a n/a n/a n/a Other 3 4 16 6 Race and ethnicity White, non-Hispanic n/a n/a 67 68 Black, non-Hispanic n/a n/a 7 8 Asian, non-Hispanic n/a n/a 5 n/a Hispanic n/a n/a 19 18 Other, non-Hispanic n/a n/a 1 6 99 49 8 8 7 7 8 9 10 8 7 6 5 4 4 4 3 2 2 21 5 74 69 7 1 10 13 57 7 9 26 1 99 49 9 8 7 7 8 9 9 8 7 5 4 4 4 4 3 2 2 21 4 75 83 7 1 9 1 58 7 9 25 1 SOURCES: Authors’ calculations from the March CPS and the Census. NOTE: Percentages may not add to 100 due to rounding. 86 Appendix C Trends in the Distributions of Alternative Measures of Income The first section of this appendix reviews the literature on the trends in the distributions of income for measures that account for taxes and non-monetary compensation and transfers. The second section presents results for alternative measures of money income not discussed fully in the text. Income Other Than Money Income The Current Population Survey (CPS) and the Census measure only pre-tax money income. When taxes and non-monetary transfers (e.g., health insurance, housing subsidies) are incorporated in the income measure, the decline in annual earnings is often diminished and the level of income inequality is generally lower. However, national research 87 shows that the growth in income inequality remains at levels similar to pre-tax money income. Pre-tax money income measures are imperfect indices of economic well-being. Money income is not reduced for payments such as personal taxes, Social Security, and union dues. Money income does not include non-monetary compensation such as health insurance, employer contributions to retirement programs, and room and board. Money income also does not include non-monetary transfers such as Medi-Cal, housing subsidies, food stamps, and energy assistance. Money income does not include the return to non-financial investments, such as owneroccupied housing. Studies that adjust money income for tax payments have reported rising inequality trends similar to those found for pre-tax income. Chamberlain and Spillberg (1991) report that the share of pre-tax adjusted gross income going to the top 20 percent of the distribution increased 9.5 percent between 1980 and 1988; the share of after-tax income increased 9.3 percent. Moreover, Gramlich, Kasten, and Sammartino (1993) and Pechman (1990) find that for the nation during the 1980s, inequality in after-tax income increased even more than inequality in pre-tax income. National studies that attempt to account for non-monetary benefits and taxes find that the trends in money income inequality are confirmed. The U.S. House of Representatives Committee on Ways and Means (1989) reports similar trends in the quintile shares of money income and more comprehensive income (after-tax income, including food and housing benefits). For example, the share of money income received by the poorest 20 percent of families fell by 9.8 percent between 1979 and 88 1987; their share of comprehensive income fell by 9.2 percent. Levy (1987) finds that the level of income inequality is lower when taxes, Medicare, Medicaid, food stamps, and fringe benefits are included in income, but that the trends in income inequality in Census and CPS data are essentially the same. This result may be different in California, which has lower health insurance rates than the rest of the country. Consumption data provide an alternative measure of economic wellbeing. Cutler and Katz (1991, 1992) find that changes in the distribution of expenditures parallel changes in the distribution of money income during the 1980s. Alternative Measures of Money Income Chapter 2 describes trends in the distribution of adjusted household income among persons. This measure of income was chosen because it allows for income-sharing among members of the same household, accounts for the greater income needs of large households, and counts each person equally regardless of household size. With CPS data it is possible to create alternative measures of income that vary the incomepooling unit (e.g., income-sharing within the family versus within the household), the size adjustment, and the unit of analysis (e.g., each person counts as a unit versus each household counts as a unit). Table C.1 lists the 12 types of household and family income examined in this study. All 12 measures use the sum of income received from all reported sources. Chapter 3 describes the trends in the distributions of annual earnings and hourly wages among people who are primarily employees (i.e., people who receive most of their earnings from wages and salary as 89 Table C.1 Alternative Measures of Household and Family Income Income Measure Income Pooling Unit of Size Location of Analysis Adjustment Results 1. Adjusted household income Household among persons residents Person n Chapter 2 2. Unadjusted household income among persons Household residents Person None Appendix C 3. Adjusted household income Household among households residents Household n Appendix C 4. Unadjusted household income among households Household residents Household None Appendix C 5. Adjusted family income among persons Family members Person n Chapter 2 6. Unadjusted family income among persons Family members Person None Appendix C 7. Adjusted family income among families Family members Family n Appendix C 8. Unadjusted family income among families Family members Family None Appendix C 9. Adjusted primary family income among persons Primary family members Person n Appendix C 10. Unadjusted primary family Primary family income among persons members Person None Appendix C 11. Adjusted primary family income among families Primary family members Family n Appendix C 12. Unadjusted primary family Primary family income among families members Family None Appendix C NOTES: A “family” includes all people living in the same household who are related by blood, marriage, or adoption. Separate family observations are created for single people and secondary families (families not related to their head of household). A “primary family” includes the head of household and relatives. Single people and secondary families are excluded. There is only one primary family per household. opposed to farm ownership and self-employment). The sample was restricted to employees because income from self-employment and one’s own farm often includes not only income from labor but also income from capital investments. To ensure that the measured trends were not a 90 result of limiting the sample to employees, the study also examined the distributions of annual earnings and hourly wages without this sample restriction. In addition, we examined the distributions limited to workers ages 18 to 55 (to remove any effects of early retirement) and the distributions of income from wages and salary only (for comparison to earlier national studies). Table C.2 summarizes the seven measures of labor income examined in this study. The trends in each income measure were estimated for males and females separately. The alternative measures generally display similar results to those discussed in the text. For family income and male earnings and hourly wages, we find the same five results: Inequality has increased in California since the early 1970s, the level and trends in inequality were similar in California and the nation until the late 1980s when inequality in California grew more rapidly, inequality increased most rapidly during recessions, income in the lower percentiles declined, and income growth was slower in California than in the nation. There are few exceptions to these trends. When family income is weighted at the family level, the VLN measure of inequality shows that California had higher inequality as early as 1979. When male annual earnings includes all workers, the VLN measure shows higher inequality in the United States than in California before 1975 and no substantial difference between the United States and California in the 1990s. For male income from wages and salary, the level of the VLN measure of inequality essentially recovers to pre-recession levels after the recession of the early 1980s. Despite these differences, the basic trends remain fairly consistent and the measures of income discussed in the text are preferred (as discussed above). 91 Table C.2 Alternative Measures of Labor Income Income Measure 13. Male annual earnings among workers 20. Female annual earnings among workers Sample Includes Source of Location of Anyone Who: Earnings Unit Ages Results Receives earnings primarily from wages and salary All earnings Annual 18 and Chapter 3 over 14. Male hourly wages among workers 21 Female hourly wages among workers Receives earnings primarily from wages and salary All earnings Hourly 18 and Chapter 3 over 15. Male annual earnings, workers ages 18 to 55 22. Female annual earnings, workers ages 18 to 55 Receives earnings primarily from wages and salary All earnings Annual 18 to Appendix C 55 16. Male hourly wages, workers ages 18 to 55 23. Female hourly wages, workers ages 18 to 55 Receives earnings primarily from wages and salary All earnings Hourly 18 to Appendix C 55 17. Male annual Receives All earnings Annual 18 and Appendix C earnings among all income over workers from wages, 24. Female annual salary, self- earnings among all employment, workers or own farm 92 Table C.2—continued Income Measure Sample Includes Source of Location of Anyone Who: Earnings Unit Ages Results 18. Male hourly wages Receives All earnings Hourly 18 and Appendix C among all workers income over 25. Female hourly from wages, wages among all salary, self- workers employment, or own farm 19. Male annual wages Receives any Earnings Annual 18 and Appendix C and salary income from from wages over 26. Female annual wages and and salary wages and salary salary NOTES: The income category “Receives earnings primarily from wages and salary” excludes people who report more income from their farm or business than from wages and salaries and excludes any person reporting an absolute value of more than $2,000 in 1994 dollars in income from their own farm or self-employment. Some wage and salary workers included in the sample receive a small amount of income from farms and self-employment. This income was included in annual earnings to improve estimates of hourly wages because estimates of annual hours of work include hours worked in the farm or business. The income category “All earnings” includes earnings from wages, salary, self-employment, or own farm. Hourly wages are not calculated for income types 19 and 26 because hours of work includes hours worked in selfemployment or own farm. All samples exclude military personnel, students, people with earnings less than or equal to zero, people under age 18, and workers whose primary occupation is “without pay.” The trends in the distribution of household income show more sensitivity to the adjustments for household size and weighting by persons. When each household is counted as a single unit (as opposed to each person) or no adjustments are made for household size, the upward trend in the VLN is less clear. Household income growth is generally, but not always, higher in the United States than in California. When household income is weighted at the household level, income at the 20th percentile does not decline between 1969 and 1989, but it does decline 93 between 1976 and 1994. Although these differences are notable, unadjusted and unweighted household income does not reflect economic well-being as accurately as adjusted household income (weighted at the person level) because it does not adjust for the greater needs of large households and it gives less weight to people in large households. For all the measures of female annual earnings, inequality declined until the early 1980s, the level and trends in inequality were similar in California and the nation, and income in the lower percentiles increased. When female annual earnings are measured for all female workers regardless of self-employment or own farm status, the decline in inequality does not begin until after 1975 and all measures of inequality show a rising trend after 1983. This is in contrast to the trends discussed in the text for female annual earnings among workers who receive earnings primarily from wages and salary: The trends in inequality after the early 1980s depended on which measure of inequality was used. For the measures of female hourly wages, inequality has increased since the early 1980s, the level and trends in inequality were similar in California and the nation, and wages in the lower percentiles fell. There were no substantial exceptions to this pattern. For annual earnings and hourly wages among both males and females, the trends in the distributions remain nearly identical when the age range is restricted to ages 18 to 55. Tables C.3 through C.22 and their associated figures provide summary statistics for the trends in the distributions of each alternative measure of income that is not described fully in the text. 94 Income Type 2: Unadjusted Household Income, Weighted by Persons Table C.3 Percentage Change in Real Unadjusted Household Income Among Persons Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –7 4 13 +21 –2 –8 –20 2 7 –2 7 21 14 +9 +32 +42 United States 20th Median 80th Change in 80/20 ratio (%) –1 9 14 +15 –1 3 11 +13 –2 –7 12 2 27 15 +29 +24 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .77 CV .75 CA .73 U.S. 2.35 VLN 2.15 CA 1.95 U.S. .71 1.75 .69 1.55 .67 1.35 .65 1.15 .63 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.1—Summary Measures of Inequality for Unadjusted Household Income Among Persons, 1967–1994 95 Income Type 3: Adjusted Household Income, Weighted at the Household Level Table C.4 Percentage Change in Real Adjusted Household Income Among Households Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) 0 11 18 +18 88 8 19 10 30 +2 +20 –11 3 17 +31 United States 20th Median 80th Change in 80/20 ratio (%) 13 15 18 +5 5 18 10 26 14 36 +9 +15 5 9 20 +15 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .83 CV .81 CA .79 U.S. .77 2.15 1.95 1.75 VLN CA U.S. .75 1.55 .73 1.35 .71 1.15 .69 .67 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.2—Summary Measures of Inequality for Adjusted Household Income Among Households, 1967–1994 96 Income Type 4: Unadjusted Household Income, Weighted at the Household Level Table C.5 Percentage Change in Real Unadjusted Household Income Among Households Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –3 2 10 +13 14 10 –9 684 10 21 18 –3 +10 +29 United States 20th Median 80th Change in 80/20 ratio (%) 3 5 11 +8 470 5 11 2 11 23 14 +7 +15 +15 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .81 CV .79 CA U.S. .77 2.35 2.15 1.95 VLN CA U.S. .75 1.75 .73 1.55 .71 1.35 .69 1.15 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.3—Summary Measures of Inequality for Unadjusted Household Income Among Households, 1967–1994 97 Income Type 6: Unadjusted Family Income, Weighted by Persons Table C.6 Percentage Change in Real Unadjusted Family Income Among Persons Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –15 1 10 +29 –4 –2 6 +10 –18 –25 0 –8 16 12 +42 +50 United States 20th Median 80th Change in 80/20 ratio (%) –4 7 12 +17 –5 1 10 +15 –9 –14 8 –2 23 13 +35 +31 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .81 CV .79 2.95 VLN 2.75 .77 CA U.S. .75 .73 .71 .69 .67 2.55 2.35 2.15 1.95 1.75 1.55 1.35 CA U.S. .65 1.15 .63 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.4—Summary Measures of Inequality for Unadjusted Family Income Among Persons, 1967–1994 98 Income Type 7: Adjusted Family Income, Weighted at the Family Level Table C.7 Percentage Change in Real Adjusted Family Income Among Families Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) 0 6 15 +15 –1 –1 6 12 7 23 +9 +25 –16 –2 13 +34 United States 20th 12 2 14 Median 13 8 21 80th 17 13 33 Change in 80/20 ratio (%) +5 +11 +16 0 6 17 +17 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .86 CV .84 3.3 VLN 3.1 .82 CA U.S. .80 .78 .76 .74 .72 2.9 CA 2.7 U.S. 2.5 2.3 2.1 1.9 1.7 .70 1.5 .68 1.3 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.5—Summary Measures of Inequality for Adjusted Family Income Among Families, 1967–1994 99 Income Type 8: Unadjusted Family Income, Weighted at the Family Level Table C.8 Percentage Change in Real Unadjusted Family Income Among Families Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –12 –8 5 +19 4 –9 3 –5 5 11 +2 +22 –15 –5 12 +32 United States 20th Median 80th Change in 80/20 ratio (%) 2 1 10 +7 02 23 8 19 +8 +17 –8 –5 10 +20 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .90 CV .88 .86 CA .84 U.S. .82 3.95 3.45 2.95 VLN CA U.S. .80 2.45 .78 .76 1.95 .74 1.45 .72 .70 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.6—Summary Measures of Inequality for Unadjusted Family Income Among Families, 1967–1994 100 Income Type 9: Adjusted Primary Family Income, Weighted by Persons Table C.9 Percentage Change in Real Adjusted Primary Family Income Among Persons Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th 2 –13 –11 –28 Median 14 0 14 –5 80th 20 5 25 15 Change in 80/20 ratio (%) +18 +20 +42 +59 United States 20th 10 –2 9 –4 Median 18 7 27 9 80th 22 13 38 21 Change in 80/20 ratio (%) +10 +15 +27 +26 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .80 CV .78 .76 CA .74 U.S. .72 2.4 VLN 2.2 CA 2.0 U.S. 1.8 .70 1.6 .68 1.4 .66 1.2 .64 .62 1.0 .60 .8 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.7—Summary Measures of Inequality for Adjusted Primary Family Income Among Persons, 1967–1994 101 Income Type 10: Unadjusted Primary Family Income, Weighted by Persons Table C.10 Percentage Change in Real Unadjusted Primary Family Income Among Persons Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th –15 –4 –18 –25 Median 1 –2 0 –8 80th 10 6 16 12 Change in 80/20 ratio (%) +29 +10 +42 +50 United States 20th –4 –5 –9 –14 Median 7 1 8 –2 80th 12 10 23 13 Change in 80/20 ratio (%) +17 +15 +35 +31 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .85 CV 3.3 VLN .80 CA U.S. 2.8 CA U.S. .75 2.3 .70 1.8 .65 1.3 .60 .8 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.8—Summary Measures of Inequality for Unadjusted Primary Family Income Among Persons, 1967–1994 102 Income Type 11: Adjusted Primary Family Income, Weighted at the Family Level Table C.11 Percentage Change in Real Adjusted Primary Family Income Among Families Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th 1 –8 –8 –23 Median 11 2 14 1 80th 19 6 27 16 Change in 80/20 ratio (%) +19 +16 +37 +51 United States 20th 10 –1 10 –1 Median 16 8 25 8 80th 19 15 36 21 Change in 80/20 ratio (%) +8 +15 +24 +23 SOURCE; Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .80 CV .78 .76 CA .74 U.S. 2.2 VLN 2.0 CA 1.8 U.S. .72 1.6 .70 .68 1.4 .66 1.2 .64 1.0 .62 .60 .8 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.9—Summary Measures of Inequality for Adjusted Primary Family Income Among Families, 1967–1994 103 Income Type 12: Unadjusted Primary Family Income, Weighted at the Family Level Table C.12 Percentage Change in Real Unadjusted Primary Family Income Among Families Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –3 7 14 +18 –5 –8 19 8 23 +13 +33 –20 0 16 +44 United States 20th Median 80th Change in 80/20 ratio (%) 5 11 14 +9 –3 2 4 16 13 29 +16 +26 –4 4 17 +22 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .78 .76 .74 CA U.S. .72 .70 .68 .66 .64 .62 2.4 VLN 2.2 CA 2.0 U.S. 1.8 1.6 1.4 1.2 1.0 .60 .8 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.10—Summary Measures of Inequality for Unadjusted Primary Family Income Among Families, 1967–1994 104 Income Type 15: Annual Earnings Among Male Workers Ages 18 to 55 Table C.13 Percentage Change in Real Annual Earnings for Males Ages 18 to 55 Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –22 –6 10 +40 –22 –39 –12 –17 –5 4 +22 +71 –30 –23 2 +46 United States 20th Median 80th Change in 80/20 ratio (%) –7 3 11 +19 –17 –22 –8 –5 4 16 +25 +49 –20 –13 2 +27 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .80 CV .75 CA U.S. .70 .65 .60 1.25 1.15 1.05 VLN CA U.S. .95 .85 .75 .55 .65 .50 .55 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.11—Summary Measures of Inequality for Annual Earnings for Males Ages 18 to 55, 1967–1994 105 Income Type 16: Hourly Wages Among Male Workers Ages 18 to 55 Table C.14 Percentage Change in Real Hourly Wages for Males Ages 18 to 55 Between Selected Years, by Income Percentile Business Cycle Peaks 1979–1989 Recessions 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –21 –14 –5 +20 –31 –23 –6 +36 United States 20th Median 80th Change in 80/20 ratio (%) –13 –8 1 +16 –20 –13 –2 +22 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .65 CV .63 CA .61 U.S. .59 .45 VLN .43 CA .41 U.S. .39 .57 .37 .55 .35 .53 .33 .51 .31 .49 .29 .47 .27 .45 .25 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.12—Summary Measures of Inequality for Hourly Wages for Males Ages 18 to 55, 1975–1994 106 Income Type 17: Annual Earnings Among All Male Workers Table C.15 Percentage Change in Real Annual Earnings for All Male Workers Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th –13 –19 –30 –31 median –5 –10 –14 –20 80th 6 –2 4 –2 Change in 80/20 ratio (%) +22 +20 +47 +41 United States 20th median 80th Change in 80/20 ratio (%) –5 3 10 +15 –14 –5 4 +21 –18 –15 –2 –12 15 3 +40 +21 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .85 CV .80 CA U.S. .75 .70 .65 1.5 VLN 1.4 CA U.S. 1.3 1.2 1.1 1.0 .60 .9 .55 .8 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.13—Summary Measures of Inequality for Annual Earnings for All Male Workers, 1967–1994 107 Income Type 18: Hourly Wages Among All Male Workers Table C.16 Percentage Change in Real Hourly Wages for All Male Workers Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1979–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) United States 20th Median 80th Change in 80/20 ratio (%) –19 –12 –4 +19 –10 –5 2 +13 –30 –21 –4 +37 –16 –13 0 +20 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .69 CV .67 CA .65 U.S. .63 .53 VLN CA .48 U.S. .61 .43 .59 .57 .38 .55 .53 .33 .51 .49 .28 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.14—Summary Measures of Inequality for Hourly Wages for All Male Workers, 1975–1994 108 Income Type 19: Annual Income from Wages and Salary Among Male Workers Table C.17 Percentage Change in Real Annual Salary for Males Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –3 –4 6 +9 –14 –16 –12 –15 –1 6 +15 +26 –14 –23 4 +21 United States 20th Median 80th Change in 80/20 ratio (%) –5 5 8 +14 –5 –10 –6 –1 6 15 +12 +28 –3 –8 6 +10 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .85 CV .80 CA U.S. 1.65 VLN 1.55 CA 1.45 U.S. .75 1.35 1.25 .70 `` 1.15 .65 1.05 .60 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.15—Summary Measures of Inequality for Annual Salary for Males, 1967–1994 109 Income Type 22: Annual Earnings Among Female Workers Ages 18 to 55 Table C.18 Percentage Change in Real Annual Earnings for Females Ages 18 to 55 Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) 62 22 11 –31 13 83 6 29 22 36 +8 –26 25 29 30 +5 United States 20th Median 80th Change in 80/20 ratio (%) 48 19 16 –22 24 85 19 42 24 44 0 –22 56 25 34 –14 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .84 CV .82 .80 CA U.S. 1.7 VLN 1.6 1.5 CA U.S. .78 1.4 .78 1.3 .74 1.2 .72 1.1 .70 1.0 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.16—Summary Measures of Inequality for Annual Earnings for Females Ages 18 to 55, 1967–1994 110 Income Type 23: Hourly Wages Among Female Workers Ages 18 to 55 Table C.19 Percentage Change in Real Hourly Wages for Females Ages 18 to 55 Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1979–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –9 2 15 +27 –7 10 21 +31 United States 20th Median 80th Change in 80/20 ratio (%) –4 8 16 +21 –1 11 22 +24 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .64 CV .62 CA .60 U.S. .58 .43 VLN .41 CA .39 U.S. .37 .35 .56 .33 .31 .54 .29 .52 .27 .50 .25 .48 .23 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.17—Summary Measures of Inequality for Hourly Wages for Females Ages 18 to 55, 1975–1994 111 Income Type 24: Annual Earnings Among All Female Workers Table C.20 Percentage Change in Real Annual Earnings for All Female Workers Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) 77 22 10 –38 12 7 22 +10 98 30 35 –32 28 27 28 0 United States 20th Median 80th Change in 80/20 ratio (%) 42 21 16 –18 30 85 51 14 38 24 20 39 34 –8 –25 –11 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .84 CV .72 .80 .78 2.0 VLN 1.9 CA CA U.S. 1.8 U.S. 1.7 1.6 .76 1.5 1.4 .74 1.3 .72 1.2 .70 1.1 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.18—Summary Measures of Inequality for Annual Earnings for All Female Workers, 1967–1994 112 Income Type 25: Hourly Wages Among All Female Workers Table C.21 Percentage Change in Real Hourly Wages for All Female Workers Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1979–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –9 2 15 +26 –8 7 18 +28 United States 20th Median 80th Change in 80/20 ratio (%) –5 10 16 +23 –1 11 22 +23 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .66 CV .64 CA .62 U.S. .60 .54 VLN .49 CA U.S. .44 .58 .39 .56 .54 .34 .52 .29 .50 .48 .24 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.19—Summary Measures of Inequality for Hourly Wages for All Female Workers, 1975–1994 113 Income Type 26: Annual Income from Wages and Salary Among Female Workers Table C.22 Percentage Change in Real Annual Salary for Females Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th 53 23 89 Median 22 10 34 80th 12 18 33 Change in 80/20 ratio (%) –26 –4 –30 28 27 25 –2 United States 20th 36 20 64 50 Median 13 15 30 27 80th 12 24 39 30 Change in 80/20 ratio (%) –18 +4 –15 –14 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .88 CV .86 1.80 VLN 1.75 1.70 1.65 .84 1.60 .82 1.55 1.50 .80 1.45 CA .78 U.S. 1.40 1.35 CA U.S. .76 1.30 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.20—Summary Measures of Inequality for Annual Salary for Females, 1967–1994 114 Appendix D Supplementary Statistics This appendix provides additional statistics on the trends in the distributions of income for those measures of income discussed in the main text. The appendix contains several tables. Tables D.1 through D.10 show deciles of the distributions of adjusted household income, male annual earnings, male hourly wages, female annual earnings, and female hourly wages. Reported decile levels are in nominal terms. The price index from Table A.1 is provided for cost of living and inflation adjustments (multiply the income level by the price index). Then, Tables D.11 and D.12 show regional comparisons of the coefficient of variation for male and female annual earnings. Finally, Tables D.13 through D.15 give state comparisons of the coefficient of variation for adjusted household income, male annual earnings, and female annual earnings. 115 Table D.1 Deciles of Nominal Adjusted Household Income, California Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 3499 5281 1968 3914 5676 1969 4061 5997 1969 (C) 3667 5831 1970 4175 6203 1971 4159 6179 1972 4213 6197 1973 4654 7008 1974 5074 7419 1975 5266 8088 1976 6174 8894 1977 6692 9442 1978 7188 10347 1979 7690 11439 1979 (C) 7446 11553 1980 8653 12732 1981 8516 12874 1982 8697 13187 1983 8490 13594 1984 9903 15561 1985 10141 15970 1986 10543 16270 1987a 11132 17711 1987 11320 17792 1988 10825 17669 1989 12462 18878 1989 (C) 13101 20500 1990 12132 19401 1991 11790 18288 1992 11787 18771 1993 11257 17615 1994 11205 18002 6720 7989 9236 10651 12408 14901 19249 7246 8665 10003 11621 13583 16151 21194 7627 9090 10695 12465 14578 17315 22437 7647 9263 10854 12656 14779 17718 22900 7800 9491 11291 13145 15309 18185 23079 8030 9785 11485 13266 15448 18976 24496 8195 10213 12310 14533 17184 20417 26103 9248 11489 13781 16075 18567 21705 28150 9855 12262 14633 17017 19610 23145 29210 10512 13343 15792 18473 21717 26122 33852 11525 14195 17070 20046 23365 27636 35836 12278 15500 18712 22007 25788 30645 39637 13881 17124 20743 24223 28765 34064 44608 15561 19078 22841 26709 32373 38951 49068 15530 19420 23210 27492 32436 39332 50926 16946 21016 25495 30023 35586 43226 54066 17139 21451 26840 31797 38266 46669 60000 17588 22643 27853 33964 41121 50260 64044 18224 24066 29087 35233 42658 51945 67108 21220 26308 31441 37827 46015 55472 72588 22018 27314 33606 40559 48734 59935 78713 22731 29082 35929 43431 51897 63873 82836 24027 30758 36958 44456 53882 67134 88947 24013 30781 37034 44434 53788 66402 87723 24144 31056 36850 45152 55029 69489 91351 25999 33511 40876 49150 58213 72184 96805 27944 35330 43070 51962 62367 77000 102615 26180 34007 41117 49901 61079 75996 100742 25297 33529 41709 50295 61040 76628 102252 25718 34583 41924 51974 63210 78740 103648 24265 31757 39969 50999 64284 81672 107405 25963 33305 43330 55040 66503 82618 110106 4.22 4.07 3.91 3.91 3.76 3.62 3.51 3.31 3.04 2.77 2.61 2.44 2.27 2.08 2.08 1.84 1.67 1.57 1.53 1.46 1.40 1.35 1.30 1.30 1.24 1.18 1.18 1.12 1.08 1.04 1.01 1.00 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTE: Household income is adjusted for household size: Reported deciles are calibrated to represent a household of four. 116 Table D.2 Deciles of Nominal Adjusted Household Income, United States Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 2773 1968 3089 1969 3380 1969 (C) 3060 1970 3511 1971 3695 1972 3929 1973 4319 1974 4683 1975 4935 1976 5373 1977 5784 1978 6306 1979 6938 1979 (C) 6710 1980 7320 1981 7734 1982 7651 1983 7953 1984 8616 1985 9084 1986 9375 1987a 9728 1987 9865 1988 10395 1989 11385 1989(C) 11314 1990 11507 1991 11535 1992 11641 1993 11629 1994 12343 4294 4761 5220 5050 5465 5658 6060 6713 7202 7539 8246 8903 9814 10807 10760 11559 12185 12550 12837 14013 14787 15503 16109 16269 17046 18245 18779 18796 18931 19142 19316 20332 5602 6209 6784 6736 7072 7323 8010 8851 9476 9970 10783 11749 13119 14475 14432 15422 16366 17069 17657 19254 20139 21222 22170 22382 23349 25060 25470 25512 26021 26256 26383 27900 6767 7442 8169 8198 8546 8940 9811 10740 11556 12225 13362 14505 16033 17882 17947 19255 20516 21458 22321 24125 25481 26813 28001 28300 29565 31519 32000 32193 32972 33499 33802 35094 7965 8703 9579 9650 10019 10512 11540 12595 13604 14583 15916 17285 19025 21220 21311 22909 24601 25900 27043 29194 30740 32372 34173 34471 35843 38355 38682 39019 40070 41121 41413 43070 9166 10028 11077 11201 11664 12242 13452 14710 15821 16977 18566 20261 22341 24935 25020 26892 29095 30701 32260 34857 36731 38620 40629 40860 42955 45643 46052 46446 48084 49384 50242 52192 10679 11688 12888 13106 13625 14271 15718 17232 18455 19843 21721 23778 26206 29240 29456 31756 34428 36466 38378 41505 43735 46086 48432 48852 51327 54520 55007 55629 57254 58902 60558 62897 12761 14005 15337 15704 16260 17124 18734 20530 22037 23760 25729 28336 31209 35033 35348 38110 41359 44202 46739 50590 53358 56253 59224 59469 62401 66719 67199 68278 70090 72078 75345 77440 16395 4.50 17802 4.33 19639 4.15 20350 4.15 20791 3.96 21954 3.79 24186 3.68 26047 3.46 28093 3.15 30262 2.91 32682 2.75 35940 2.58 39817 2.42 44182 2.21 45368 2.21 48264 1.98 52786 1.81 57358 1.71 60170 1.64 65506 1.57 69485 1.52 73222 1.49 76889 1.44 77557 1.44 82020 1.38 87567 1.32 88912 1.32 89170 1.25 91503 1.20 94658 1.16 99102 1.13 101849 1.10 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTE: Household income is adjusted for household size: Reported deciles are calibrated to represent a household of four. 117 Table D.3 Deciles of Nominal Annual Earnings Among Male Workers, California Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 2200 4099 1968 2434 4496 1969 2342 4499 1969 (C) 2150 4550 1970 2502 4503 1971 1997 4095 1972 2231 4498 1973 2484 4796 1974 2554 5006 1975 2797 5106 1976 2716 5490 1977 3004 6009 1978 3506 6793 1979 3991 7184 1979 (C) 3805 7005 1980 4003 7406 1981 3886 7511 1982 4008 7514 1983 4856 8989 1984 4501 8403 1985 5018 9032 1986 5190 9222 1987a 5411 9953 1987 5603 9705 1988 5488 9977 1989 5822 9970 1989(C) 6000 10136 1990 5988 9980 1991 6014 10023 1992 5997 9995 1993 4980 9960 1994 6000 10400 5499 6499 5994 6994 6006 7346 6050 7350 6081 7505 6081 7590 6447 7997 6702 8392 6862 8510 7358 9411 7986 9982 8253 10516 9490 11987 9977 12555 10005 12505 10408 13530 11016 14266 10520 14327 12465 15981 12034 16925 12545 17061 12974 17964 13739 18020 13307 18010 14168 17959 13957 18418 15000 19142 14272 18962 14032 19316 14992 19990 13546 17928 15000 19200 7399 7993 8369 8450 8665 8988 9496 9991 10012 11012 11979 12981 14385 15433 15435 16713 17847 18034 19976 20807 21813 22954 23512 23012 22948 23927 24000 24042 25057 24987 23904 25000 8130 8992 9508 9650 9890 9987 10995 11689 12014 13015 14175 15022 16622 17959 18205 20015 21031 22042 23971 25008 26093 27537 28515 28015 27936 29410 29971 29941 30069 30984 29880 30772 9098 10498 9991 11689 10608 12010 10750 12050 11007 12808 11485 13183 12095 14294 12988 14986 14017 16019 15017 17520 16091 18717 17526 20030 19127 22234 20952 24943 21005 25005 23018 26020 24869 29043 26049 31059 27367 32960 29009 34953 30107 36128 31937 37925 33991 40021 33418 40021 32925 39909 34894 41872 35000 41000 34931 44100 36083 45103 38481 45977 36852 45816 40000 50000 12997 14303 15012 15050 15510 16279 17893 18483 20024 21424 22989 25037 27471 30930 30505 34026 36053 40076 41949 43014 46164 48903 50026 50026 50884 54833 52032 58883 59136 60969 59760 65000 4.22 4.07 3.91 3.91 3.76 3.62 3.51 3.31 3.04 2.77 2.61 2.44 2.27 2.08 2.08 1.84 1.67 1.57 1.53 1.46 1.40 1.35 1.30 1.30 1.24 1.18 1.18 1.12 1.08 1.04 1.01 1.00 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. 118 Table D.4 Deciles of Nominal Annual Earnings Among Male Workers, United States Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 2001 1968 2101 1969 2275 1969 (C) 2150 1970 2210 1971 2100 1972 2403 1973 2597 1974 2623 1975 2745 1976 2895 1977 3206 1978 3661 1979 3997 1979 (C) 3905 1980 4003 1981 3999 1982 3911 1983 4006 1984 4211 1985 4809 1986 4992 1987a 4982 1987 4982 1988 5388 1989 6024 1989 (C) 6000 1990 5979 1991 5810 1992 5780 1993 6026 1994 6580 3609 4891 5604 4001 5002 6002 4091 5503 6603 4050 5450 6550 4160 5601 6803 4131 5694 7001 4596 6072 7510 5003 6664 8005 5025 7009 8511 5133 7218 8985 5590 7794 9750 6011 8279 10287 6691 8988 11083 7352 9994 11992 7125 10005 12005 7506 10186 13011 7838 10998 13997 7521 10730 14039 7956 11115 14921 8341 12031 15506 9017 12524 16031 9225 12979 16972 9868 13741 17618 9888 13552 17538 9978 14392 17961 10543 15061 19077 10906 15000 19000 10603 14948 18935 10017 15026 19334 10436 15052 20070 11047 15064 20086 12000 16000 20800 6504 7204 7002 7803 7503 8465 7550 8550 7806 8982 8002 9203 8911 10013 9561 11007 10013 11515 10482 11980 11478 13027 12022 14426 13476 15579 14862 16989 15005 17025 15513 18215 16854 19996 17047 20055 18025 21029 19048 23059 20039 24047 20149 24960 21245 25411 20926 25012 21952 26641 23093 28113 22802 27000 23419 27904 24042 29050 25087 30105 25107 30129 25000 30772 8005 8960 9675 9850 10001 10402 11415 12108 13017 13976 14974 16430 17976 19987 20005 21017 22995 24066 25034 27069 28054 29951 29895 29895 31132 33134 31500 33385 35061 35122 35652 37964 9406 10003 11005 11050 11701 12002 13017 14409 15020 15973 17649 19091 20772 22985 23005 25020 26994 28077 30041 32081 34066 34943 35874 35874 37917 40162 38000 39863 41071 43150 44189 46000 11507 12004 13806 14050 14502 15003 16521 18012 19025 19966 21961 24044 25267 28082 29005 30024 33983 36100 38052 40102 42082 44927 46317 45839 48893 50202 48098 51323 53092 55318 58249 60000 4.50 4.33 4.15 4.15 3.96 3.79 3.68 3.46 3.15 2.91 2.75 2.58 2.42 2.21 2.21 1.98 1.81 1.71 1.64 1.57 1.52 1.49 1.44 1.44 1.38 1.32 1.32 1.25 1.20 1.16 1.13 1.10 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. 119 Table D.5 Deciles of Nominal Hourly Wages Among Male Workers, California Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1975 2.41 3.24 1976 2.56 3.52 1977 2.70 3.76 1978 2.88 3.92 1979 3.17 4.32 1980 3.46 4.81 1981 3.49 4.81 1982 3.76 5.01 1983 3.84 5.27 1984 3.85 5.37 1985 3.94 5.44 1986 4.08 5.76 1987a 4.15 5.77 1987 4.17 5.77 1988 4.32 5.82 1989 4.44 5.98 1990 4.57 6.11 1991 4.73 6.26 1992 4.81 6.53 1993 4.64 6.17 1994 4.81 6.41 4.09 4.88 5.73 6.35 7.22 8.42 4.32 5.23 6.08 6.91 7.86 9.05 4.67 5.54 6.46 7.31 8.35 9.63 4.99 5.95 6.99 7.99 9.12 10.57 5.37 6.57 7.67 8.76 10.00 11.51 5.77 7.11 8.39 9.62 11.07 12.83 6.08 7.50 8.81 10.30 11.94 13.70 6.42 7.86 9.63 11.13 12.81 15.07 6.81 8.34 9.70 11.52 13.28 15.61 6.93 8.66 10.13 12.02 13.77 16.22 7.06 9.12 10.72 12.54 14.62 17.37 7.20 9.12 11.20 13.21 15.35 17.75 7.42 9.38 11.42 13.68 15.87 18.76 7.50 9.38 11.17 13.47 15.49 18.28 7.48 9.40 11.17 13.43 15.71 19.19 7.67 9.59 11.56 14.24 16.49 19.82 7.70 9.60 12.00 14.39 16.79 19.96 8.19 10.12 12.30 14.55 17.35 21.03 8.24 10.25 12.88 15.38 18.26 21.97 7.76 9.85 11.97 14.59 17.62 21.55 8.33 10.10 12.39 15.38 18.75 23.08 10.38 11.04 11.80 13.09 14.39 15.93 17.33 19.27 19.63 20.20 21.44 22.36 24.05 23.09 24.55 24.92 26.32 27.34 28.83 28.73 29.62 2.77 2.61 2.44 2.27 2.08 1.84 1.67 1.57 1.53 1.46 1.40 1.35 1.30 1.30 1.24 1.18 1.12 1.08 1.04 1.01 1.00 SOURCE: Authors’ calculations from the March CPS. 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTES: Hourly wages not calculated before 1975 because earlier CPSs did not ask respondents about their hours of work in a usual week in the previous year (annual earnings refers to earnings in the previous year). Hourly wages not calculated for the Census because the 1970 Census also did not ask about weekly hours of work in the previous year. 120 Table D.6 Deciles of Nominal Hourly Wages Among Male Workers, United States Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987a 1987 1988 1989 1990 1991 1992 1993 1994 2.36 3.17 3.85 4.57 5.27 5.93 6.72 7.68 9.60 2.91 2.46 3.35 4.08 4.80 5.60 6.37 7.20 8.35 10.24 2.75 2.61 3.56 4.38 5.17 5.99 6.86 7.78 9.02 11.13 2.58 2.84 3.78 4.69 5.52 6.40 7.36 8.45 9.60 12.00 2.42 3.07 4.16 5.09 6.03 7.08 8.17 9.37 10.68 13.15 2.21 3.30 4.47 5.50 6.54 7.70 8.85 10.10 11.84 14.43 1.98 3.47 4.81 5.77 7.02 8.19 9.61 11.06 12.83 15.86 1.81 3.55 4.82 6.05 7.26 8.68 10.03 11.75 13.76 17.08 1.71 3.57 4.90 6.26 7.51 8.83 10.35 12.04 14.35 17.81 1.64 3.67 5.06 6.42 7.71 9.40 11.03 12.66 14.94 18.80 1.57 3.85 5.30 6.74 8.19 9.63 11.48 13.21 15.56 19.27 1.52 3.99 5.46 6.91 8.40 9.98 11.76 13.82 16.22 20.45 1.49 3.85 5.44 6.87 8.32 9.76 11.56 13.82 16.18 20.21 1.44 4.00 5.70 7.13 8.62 10.12 11.98 14.37 16.77 20.91 1.44 4.23 5.79 7.42 8.97 10.55 12.47 14.47 17.27 21.93 1.38 4.46 6.11 7.72 9.41 11.10 13.03 15.45 18.34 23.17 1.32 4.57 6.23 7.73 9.58 11.24 13.18 15.33 18.69 23.96 1.25 4.67 6.26 8.01 9.63 11.56 13.58 16.11 19.26 24.56 1.20 4.82 6.51 8.20 9.91 12.01 14.23 16.84 19.68 25.57 1.16 4.83 6.53 8.21 9.95 12.05 14.28 16.90 20.21 26.34 1.13 5.00 6.77 8.55 10.22 12.16 14.42 17.31 20.98 27.47 1.10 SOURCE: Authors’ calculations from the March CPS. 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTES: Hourly wages not calculated before 1975 because earlier CPSs did not ask respondents about their hours of work in a usual week in the previous year (annual earnings refers to earnings in the previous year). Hourly wages not calculated for the Census because the 1970 Census also did not ask about weekly hours of work in the previous year. 121 Table D.7 Deciles of Nominal Annual Earnings Among Female Workers, California Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 366 1968 400 1969 375 1969 (C) 450 1970 450 1971 399 1972 520 1973 599 1974 601 1975 701 1976 799 1977 929 1978 999 1979 1297 1979 (C) 1305 1980 1601 1981 1602 1982 2004 1983 1998 1984 2001 1985 2007 1986 1996 1987a 2401 1987 2501 1988 2993 1989 3274 1989 (C) 3000 1990 2994 1991 3157 1992 2998 1993 2789 1994 3000 900 1700 2400 3068 3999 4799 5799 6999 4.22 999 1698 2498 3397 4196 5015 6011 7493 4.07 993 1699 2502 3503 4333 5304 6265 7771 3.91 1050 1950 2950 3750 4650 5550 6550 8050 3.91 1006 1871 3002 3903 4815 5904 7005 8223 3.76 1011 1804 2796 3941 4994 5992 7040 8988 3.62 1299 2234 3287 4498 5498 6697 7797 9796 3.51 1399 2318 3312 4396 5695 6994 8117 10091 3.31 1464 2467 3504 4601 5857 7008 8589 11013 3.04 1511 2729 3954 5050 6507 8009 9570 12013 2.77 1996 2995 4193 5450 6988 8485 9982 12877 2.61 2003 3294 4807 6009 7511 9013 11016 14021 2.44 2399 3896 5095 6793 7992 9989 11987 14984 2.27 2993 4590 5986 7981 9379 10975 12970 17161 2.08 3005 4505 6005 7875 9505 11005 13465 17365 2.08 3493 5004 7005 9007 10508 12510 15011 19815 1.84 3505 5308 7395 9546 11577 13812 16149 21031 1.67 3866 6011 8015 10019 12263 15029 18034 23044 1.57 4195 6193 8984 10987 12984 15781 19177 24970 1.53 4258 6442 9137 11604 14004 16983 20006 26008 1.46 4215 7025 9706 12143 15053 18064 22078 28602 1.40 4890 7171 9980 12974 15968 18962 22954 29941 1.35 5003 7804 10506 14005 17009 20011 24513 31016 1.30 5003 7671 10405 13209 16753 20011 24776 31016 1.30 5522 8082 11195 14467 17959 21566 25941 33923 1.24 5982 8773 11964 14954 17945 22930 27915 35891 1.18 6000 9393 12000 15800 19500 23421 28050 36000 1.18 5988 8982 11976 14970 19960 23952 28943 36927 1.12 6014 9622 12829 16574 20046 25057 30069 38438 1.08 6497 9995 12993 16991 20989 25487 30984 39980 1.04 5976 9362 12948 16932 20916 24900 31872 42828 1.01 6000 10000 13000 18000 23000 28000 34000 43000 1.00 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. 122 Table D.8 Deciles of Nominal Annual Earnings Among Female Workers, United States Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 300 1968 350 1969 390 1969 (C) 450 1970 400 1971 450 1972 501 1973 500 1974 575 1975 653 1976 735 1977 801 1978 939 1979 1039 1979 (C) 1195 1980 1154 1981 1300 1982 1474 1983 1502 1984 1540 1985 1703 1986 1797 1987a 1993 1987 1993 1988 2130 1989 2465 1989 (C) 2500 1990 2491 1991 2719 1992 2950 1993 3013 1994 3000 750 850 925 1050 1000 1040 1172 1201 1352 1547 1757 2004 2001 2498 2565 2856 2999 3169 3455 3609 3908 3994 4504 4599 4989 5020 5219 5406 5810 6021 6026 6288 1301 1500 1578 1750 1700 1800 2003 2001 2303 2496 2822 3005 3396 3997 4005 4323 4879 5014 5464 5922 6012 6330 6975 6975 7284 8032 8000 8013 8834 9031 9039 9625 2001 2748 2156 3001 2371 3107 2550 3250 2500 3345 2712 3554 3004 3845 3002 4003 3259 4296 3604 4776 3993 5021 4276 5510 4894 5992 5496 6995 5505 7005 6005 7599 6499 8098 7019 9025 7510 9613 8020 10025 8260 10520 8892 10982 9487 11958 9565 11958 9978 12473 10442 13053 10500 13110 10962 13952 11565 14585 12042 15052 12051 15566 12500 16000 3302 3581 3912 4050 4100 4371 4706 5003 5207 5890 6189 6913 7490 8195 8165 9101 9998 10930 11716 12031 13025 13897 14768 14748 14967 16065 16000 16942 18031 19066 19164 20000 4003 4201 4647 4850 5001 5201 5607 6004 6408 6988 7587 8015 8988 9994 10005 11009 11997 13036 14019 15038 15630 16673 17680 17738 18459 20081 19917 20330 21638 23080 23099 24000 4953 6004 5002 6452 5527 7003 5750 7150 6001 7601 6284 8002 6904 8511 7205 9006 7910 9895 8486 10694 9014 11480 10018 12190 10631 13482 11992 14990 12005 15005 13011 16613 14497 17996 15765 20055 17023 21433 18046 23059 19338 25049 19968 25958 20926 27520 21126 27404 22950 29517 24097 31126 24000 30130 24914 31890 26045 34560 28098 35508 29024 37661 30000 39200 4.50 4.33 4.15 4.15 3.96 3.79 3.68 3.46 3.15 2.91 2.75 2.58 2.42 2.21 2.21 1.98 1.81 1.71 1.64 1.57 1.52 1.49 1.44 1.44 1.38 1.32 1.32 1.25 1.20 1.16 1.13 1.10 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. 123 Table D.9 Deciles of Nominal Hourly Wages Among Female Workers, California Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1975 1.64 2.19 2.59 3.03 3.49 3.99 4.55 5.29 6.67 2.77 1976 1.86 2.40 2.75 3.24 3.74 4.22 4.80 5.61 6.93 2.61 1977 1.96 2.50 3.00 3.51 3.97 4.51 5.14 6.05 7.54 2.44 1978 2.12 2.79 3.29 3.75 4.32 4.80 5.40 6.38 7.97 2.27 1979 2.49 3.12 3.63 4.21 4.80 5.40 6.21 7.13 9.11 2.08 1980 2.69 3.37 4.07 4.74 5.29 6.00 6.90 8.18 10.13 1.84 1981 2.89 3.61 4.38 5.06 5.78 6.48 7.45 8.82 11.07 1.67 1982 3.01 3.85 4.67 5.32 6.25 7.16 8.19 9.63 12.04 1.57 1983 3.08 4.00 4.82 5.71 6.45 7.37 8.61 10.08 12.64 1.53 1984 3.13 4.15 5.00 5.96 6.84 7.85 9.14 11.04 13.55 1.46 1985 3.35 4.34 5.31 6.27 7.24 8.44 9.76 11.67 14.47 1.40 1986 3.20 4.24 5.33 6.40 7.49 8.73 10.00 12.00 15.21 1.35 1987a 3.46 4.69 5.77 6.83 8.00 9.31 10.63 12.83 16.35 1.30 1987 3.50 4.69 5.77 6.73 7.87 9.24 10.63 12.71 16.03 1.30 1988 3.74 4.88 5.99 7.14 8.39 9.59 11.51 13.43 16.79 1.24 1989 3.83 4.98 6.23 7.48 8.63 10.20 11.98 14.38 18.33 1.18 1990 3.91 5.23 6.39 7.68 9.21 10.59 12.48 14.97 19.19 1.12 1991 4.25 5.57 6.79 8.25 9.64 11.38 13.23 15.69 20.05 1.08 1992 4.32 5.77 7.08 8.65 10.09 11.66 13.45 16.34 20.82 1.04 1993 4.15 5.58 7.11 8.62 10.02 11.92 14.11 17.24 22.03 1.01 1994 4.25 5.77 7.21 8.88 10.58 12.31 14.42 17.31 22.70 1.00 SOURCE: Authors’ calculations from the March CPS. 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTES: Hourly wages not calculated before 1975 because earlier CPSs did not ask respondents about their hours of work in a usual week in the previous year (annual earnings refers to earnings in the previous year). Hourly wages not calculated for the Census because the 1970 Census also did not ask about weekly hours of work in the previous year. 124 Table D.10 Deciles of Nominal Hourly Wages Among Female Workers, United States Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1975 1.54 2.00 2.38 2.70 3.08 3.50 4.04 4.80 5.87 2.91 1976 1.68 2.20 2.50 2.89 3.33 3.83 4.32 5.05 6.30 2.75 1977 1.78 2.34 2.71 3.13 3.58 4.04 4.68 5.48 6.84 2.58 1978 1.92 2.50 2.92 3.36 3.84 4.32 4.95 5.76 7.24 2.42 1979 2.22 2.88 3.30 3.75 4.26 4.80 5.50 6.44 8.17 2.21 1980 2.41 3.10 3.59 4.09 4.70 5.29 6.01 7.15 8.95 1.98 1981 2.56 3.35 3.85 4.44 5.01 5.77 6.67 7.74 9.77 1.81 1982 2.75 3.50 4.11 4.82 5.41 6.27 7.23 8.56 10.63 1.71 1983 2.86 3.61 4.33 5.01 5.78 6.59 7.70 9.06 11.53 1.64 1984 2.89 3.76 4.47 5.19 6.01 6.91 8.09 9.64 12.05 1.57 1985 3.01 3.85 4.70 5.45 6.26 7.23 8.56 10.12 12.77 1.52 1986 3.00 3.90 4.80 5.67 6.54 7.68 8.89 10.56 13.44 1.49 1987a 3.06 4.01 4.81 5.71 6.68 7.72 9.13 10.72 13.78 1.44 1987 3.19 4.15 4.98 5.92 6.95 8.00 9.49 11.21 14.37 1.44 1988 3.30 4.32 5.28 6.24 7.20 8.50 9.98 11.99 14.97 1.38 1989 3.40 4.59 5.58 6.61 7.72 8.93 10.55 12.55 16.16 1.32 1990 3.65 4.79 5.75 6.79 7.97 9.34 10.95 13.08 16.77 1.25 1991 3.85 5.01 6.01 7.16 8.35 9.63 11.56 13.76 17.53 1.20 1992 3.99 5.14 6.27 7.37 8.68 10.13 12.06 14.47 18.53 1.16 1993 4.02 5.24 6.40 7.73 9.01 10.51 12.31 15.02 19.31 1.13 1994 4.17 5.40 6.50 7.69 9.13 10.72 12.60 15.38 20.13 1.10 SOURCE: Authors’ calculations from the March CPS. 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTES: Hourly wages not calculated before 1975 because earlier CPSs did not ask respondents about their hours of work in a usual week in the previous year (annual earnings refers to earnings in the previous year). Hourly wages not calculated for the Census because the 1970 Census also did not ask about weekly hours of work in the previous year. 125 Table D.11 Regional Trends in the Coefficient of Variation for Real Annual Earnings Among Males, 1969–1994 CV (Rank) Percentage Change in CV (Rank) Region 1969 1979 1989 1994 1969–1979 1979–1989 1989–1994 California 0.56 0.65 0.75 0.78 15 14 4 (4) (2) (2) (1) (3) (6) (8) New England 0.51 0.63 0.65 0.69 25 2 7 (9) (4) (10) (10) (1) (10) (7) Mid Atlantic 0.53 0.58 0.67 0.72 9 16 8 (7) (9) (8) (7) (6) (3) (4) E. N. Central 0.50 0.54 0.65 0.70 8 19 8 (10) (10) (9) (8) (8) (1) (3) W. N. Central 0.52 0.60 0.70 0.69 16 15 0 (8) (8) (5) (9) (2) (4) (10) S. Atlantic 0.62 0.66 0.70 0.76 6 6 9 (1) (1) (4) (4) (9) (9) (1) E. S. Central 0.62 0.62 0.69 0.74 0 11 7 (2) (6) (6) (5) (10) (8) (5) W. S. Central 0.60 0.65 0.77 0.77 9 19 0 (3) (3) (1) (2) (7) (2) (9) Mountain 0.54 0.61 0.68 0.73 13 11 8 (6) (7) (7) (6) (4) (7) (2) Pacific 0.56 0.62 0.71 0.76 11 15 7 (5) (5) (3) (3) (5) (5) (6) SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are not sensitive to the consumer price index. 126 Table D.12 Regional Trends in the Coefficient of Variation for Real Annual Earnings Among Females, 1969–1994 Region California New England Mid Atlantic E. N. Central W. N. Central S. Atlantic E. S. Central W. S. Central Mountain Pacific 1969 0.78 (3) 0.71 (10) 0.72 (9) 0.76 (5) 0.76 (6) 0.77 (4) 0.75 (8) 0.81 (1) 0.76 (7) 0.80 (2) CV (Rank) 1979 1989 0.76 0.80 (3) (2) 0.73 0.76 (5) (7) 0.72 0.78 (8) (6) 0.72 0.78 (7) (5) 0.71 0.80 (10) (3) 0.72 0.76 (9) (8) 0.73 0.73 (6) (10) 0.75 0.81 (4) (1) 0.79 0.74 (1) (9) 0.76 0.79 (2) (4) 1994 0.81 (7) 0.76 (10) 0.80 (8) 0.84 (1) 0.80 (9) 0.81 (5) 0.81 (3) 0.81 (4) 0.84 (2) 0.81 (6) Percentage Change in CV (Rank) 1969–1979 1979–1989 1989–1994 –3 4 1 (4) (6) (7) 240 (2) (7) (10) 083 (3) (2) (5) –5 8 8 (7) (3) (3) –7 12 0 (10) (1) (9) –6 6 7 (8) (5) (4) –3 1 11 (5) (9) (2) –6 7 1 (9) (4) (8) 4 –6 12 (1) (10) (1) –4 4 2 (6) (8) (6) SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this figure are not sensitive to the consumer price index. 127 Table D.13 State Rankings for Adjusted Household Income Inequality Based on the Coefficient of Variation: Census State Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma 1969 CV Rank 0.71 4 0.65 15 0.65 15 0.70 5 0.63 21 0.62 23 0.57 42 0.60 28 0.69 6 0.69 6 0.59 34 0.60 28 0.60 28 0.56 47 0.60 28 0.62 23 0.68 9 0.73 2 0.57 42 0.61 26 0.57 42 0.57 42 0.59 34 0.78 1 0.65 15 0.61 26 0.63 21 0.59 34 0.55 50 0.58 38 0.72 3 0.65 15 0.64 20 0.62 23 0.57 42 0.65 15 1989 CV Rank 0.71 6 0.62 37 0.70 8 0.70 8 0.71 6 0.66 20 0.62 37 0.61 43 0.69 13 0.70 8 0.61 43 0.64 27 0.67 16 0.61 43 0.62 37 0.65 22 0.70 8 0.76 1 0.60 46 0.63 32 0.63 32 0.65 22 0.62 37 0.76 1 0.67 16 0.65 22 0.62 37 0.64 27 0.57 50 0.63 32 0.74 4 0.73 5 0.66 20 0.64 27 0.64 27 0.70 8 Change Percent Rank 0 43 –4 49 8 12 1 42 13 2 6 21 7 15 3 32 0 43 2 38 4 28 7 15 13 2 99 3 32 5 25 3 32 5 25 6 21 3 32 10 7 15 1 6 21 –3 48 3 32 7 15 –1 46 8 12 4 28 7 15 2 38 12 4 2 38 2 38 11 6 7 15 128 Table D.13—continued 1969 1989 Change State CV Rank CV Rank Percent Rank Oregon Pennsylvania Rhode Island South Carolina South Dakota Tennessee Texas Utah Vermont Virginia Washington West Virginia Wisconsin Wyoming 0.60 28 0.58 38 0.60 28 0.69 6 0.68 9 0.67 12 0.68 9 0.56 47 0.58 38 0.67 12 0.58 38 0.66 14 0.56 47 0.59 34 0.65 22 0.65 22 0.62 37 0.67 16 0.64 27 0.69 13 0.75 3 0.60 46 0.60 46 0.67 16 0.63 32 0.69 13 0.60 46 0.63 32 9 12 3 –2 –6 4 10 8 4 0 9 5 7 6 19 4 32 46 50 28 7 12 28 43 9 25 15 21 SOURCE: Authors’ calculations from the 1970 and 1990 Census. NOTES: Ties in rank are reported with the highest common rank. For example, if two states are tied for first, both states are reported with rank 1 and the next highest state is reported with rank 3. The CV values for California reported in this table are lower than reported in the text due to top-coding differences. A greater amount of topcoding was required to achieve consistency across all states than was required for consistency between California and the nation as a whole. For consistent comparison across states, adjusted household income was top-coded at 4 percent in each state. The top-code used in the text figures was 2.42 percent. The CV values reported in the text are more accurate for California; the values above are more accurate for comparison with other states. 129 Table D.14 State Rankings for Male Annual Earnings Inequality Based on the Coefficient of Variation: Census State Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma 1969 CV Rank 0.60 7 0.57 15 0.56 18 0.61 6 0.55 21 0.54 26 0.53 30 0.55 21 0.63 2 0.62 3 0.53 30 0.53 30 0.51 40 0.48 49 0.52 35 0.55 21 0.57 15 0.60 7 0.52 35 0.56 18 0.52 35 0.50 43 0.52 35 0.67 1 0.56 18 0.53 30 0.55 21 0.54 26 0.49 46 0.54 26 0.62 3 0.55 21 0.59 10 0.58 13 0.48 49 0.57 15 1989 CV Rank 0.63 15 0.63 15 0.67 3 0.64 8 0.67 3 0.64 8 0.59 35 0.58 42 0.66 5 0.63 15 0.60 30 0.61 25 0.61 25 0.58 42 0.58 42 0.60 30 0.64 8 0.66 5 0.56 47 0.59 35 0.58 42 0.60 30 0.59 35 0.66 5 0.63 15 0.64 8 0.60 30 0.62 22 0.56 47 0.61 25 0.69 1 0.63 15 0.61 25 0.64 8 0.59 35 0.64 8 Change Percent Rank 5 40 10 31 20 6 4 43 22 1 18 12 11 30 5 40 4 43 2 48 13 22 16 14 19 9 22 1 12 24 9 33 12 24 10 31 8 36 5 40 12 24 21 4 15 16 –3 50 12 24 20 6 8 36 16 14 14 14 13 22 12 24 15 16 3 47 9 33 22 1 14 14 130 Table D.14—continued 1969 1989 Change State CV Rank CV Rank Percent Rank Oregon 0.51 40 Pennsylvania 0.50 43 Rhode Island 0.52 35 South Carolina 0.58 13 South Dakota 0.59 10 Tennessee 0.59 10 Texas 0.60 7 Utah 0.50 43 Vermont 0.51 40 Virginia 0.62 3 Washington 0.49 46 West Virginia 0.54 26 Wisconsin 0.49 46 Wyoming 0.53 30 0.62 22 0.59 35 0.57 46 0.61 25 0.62 22 0.63 15 0.68 2 0.60 30 0.56 47 0.63 15 0.59 35 0.64 8 0.56 47 0.59 35 20 19 9 4 4 8 14 19 8 2 21 18 15 12 6 9 33 43 43 36 19 9 36 48 4 12 16 24 SOURCE: Authors’ calculations from the 1970 and 1990 Census. NOTES: Ties in rank are reported with the highest common rank. For example, if two states are tied for first, both states are reported with rank 1 and the next highest state is reported with rank 3. The CV values for California reported in this table are lower than reported in the text due to top-coding differences. A greater amount of top-coding was required to achieve consistency across all states than was required for consistency between California and the nation as a whole. For consistent comparison across states, adjusted household income was top-coded at 4 percent in each state. The top-code used in the text figures was 0.93 percent. The CV values reported in the text are more accurate for California; the values above are more accurate for comparison with other states. 131 Table D.15 State Rankings for Female Annual Earnings Inequality Based on the Coefficient of Variation: Census State Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma 1969 CV Rank 0.73 18 0.79 5 0.74 15 0.69 32 0.70 29 0.73 18 0.65 46 0.69 32 0.72 22 0.70 29 0.67 39 0.79 5 0.65 46 0.68 36 0.75 10 0.73 18 0.69 32 0.77 7 0.71 25 0.69 32 0.67 39 0.71 25 0.72 22 0.76 9 0.68 36 0.77 7 0.75 10 0.66 42 0.66 42 0.66 42 0.80 2 0.65 46 0.65 46 0.81 1 0.70 29 0.75 10 1989 CV Rank 0.70 13 0.69 17 0.70 13 0.68 24 0.69 17 0.69 17 0.63 45 0.65 39 0.67 31 0.68 24 0.62 49 0.74 2 0.67 31 0.69 17 0.67 31 0.68 24 0.71 9 0.72 5 0.65 39 0.63 45 0.63 45 0.73 4 0.66 37 0.70 13 0.69 17 0.72 5 0.68 24 0.66 37 0.62 49 0.65 39 0.74 2 0.68 24 0.63 45 0.70 13 0.69 17 0.71 9 Change Percent Rank –5 23 –13 48 –7 37 –2 12 –1 9 –6 29 –2 12 –6 29 –7 37 –3 15 –6 29 –6 29 41 16 –10 46 –7 37 33 –6 29 –8 42 –9 44 –5 23 33 –8 42 –7 37 16 –6 29 –9 44 –1 9 –5 23 –1 9 –6 29 41 –3 15 –14 49 –2 12 –4 21 132 Table D.15—continued 1969 1989 Change State CV Rank CV Rank Percent Rank Oregon 0.75 10 Pennsylvania 0.66 42 Rhode Island 0.67 39 South Carolina 0.65 46 South Dakota 0.80 2 Tennessee 0.68 36 Texas 0.73 18 Utah 0.74 15 Vermont 0.72 22 Virginia 0.71 25 Washington 0.74 15 West Virginia 0.75 10 Wisconsin 0.71 25 Wyoming 0.80 2 0.71 9 0.68 24 0.64 43 0.65 39 0.67 31 0.67 31 0.71 9 0.72 5 0.64 43 0.68 24 0.69 17 0.72 5 0.67 31 0.75 1 –5 2 –5 0 –17 –3 –3 –3 –11 –3 –7 –4 –5 –6 23 5 23 8 50 15 15 15 47 15 37 21 23 29 SOURCE: Authors’ calculations from the 1970 and 1990 Census. NOTES: Ties in rank are reported with the highest common rank. For example, if two states are tied for first, both states are reported with rank 1 and the next highest state is reported with rank 3. The CV values for California reported in this table are lower than reported in the text due to top-coding differences. A greater amount of top-coding was required to achieve consistency across all states than was required for consistency between California and the nation as a whole. For consistent comparison across states, adjusted household income was top-coded at 4 percent in each state. The top-code used in the text figures was 0.13 percent. The CV values reported in the text are more accurate for California; the values above are more accurate for comparison with other states. 133 Bibliography Aaron, Henry (1978), Politics and the Professors: The Great Society in Perspective, Brookings Institution, Washington, D.C. Blinder, Alan (1980), “The Levels and Distribution of Economic WellBeing: Introduction and Preview,” in Martin Feldstein (ed.), The American Economy in Transition, University of Chicago Press, Chicago. Borjas, George, Richard Freeman, and Larry Katz (1992), “On the Labor Market Effects of Immigration and Trade,” in George Borjas and Richard Freeman (eds.), Immigration and the Work Force: Economic Consequences for the United States and Source Areas, University of Chicago Press, Chicago. Burtless, Gary (1996), “Trends in the Level and Distribution of U.S. Living Standards, 1973–1993,” unpublished manuscript, Brookings Institution, Washington, D.C. Butcher, Kristine, and David Card (1991), “Immigration and Wages: Evidence from the 1980s,” American Economic Review, Vol. 81, No. 2, May 1991, pp. 292–296. California Department of Finance (1995), Statistical Abstract 1995, Sacramento, California. 135 Cancian, Maria, Sheldon Danziger, and Peter Gottschalk (1993), “Working Wives and Family Income Inequality Among Married Couples,” in Sheldon Danziger and Peter Gottschalk (eds.), Uneven Tides: The Rising Inequality in America, Russell Sage Foundation, New York, pp. 195–221. Cassidy, J. (1995), “Who Killed the Middle Class?” The New Yorker, October 16, pp. 113–124. Center for the Continuing Study of the California Economy (1994), California Economic Growth—1994 Edition, Palo Alto, California. Chamberlain, Jay D., and Phil Spillberg (1991), “Trends in the Distribution of Income and Personal Income Tax Burden for California Taxpayers,” Occasional Papers Series, California Franchise Tax Board, FTB/OPS 91-01, Sacramento, California. Cutler, David M., and Larry F. Katz (1991), “Macroeconomic Performance and the Disadvantaged,” Brookings Papers on Economic Activity, Vol. 2, pp. 1–61. Cutler, David M., and Larry F. Katz (1992), “Rising Inequality? Changes in the Distribution of Income and Consumption in the 1980s,” American Economic Review, Vol. 82, No. 2, pp. 546–551. Danziger, Sheldon, and Peter Gottschalk (1995), America Unequal, Russell Sage Foundation, New York. Freeman, Richard (1993), “How Much Has Deunionization Contributed to the Rise in Male Earnings Inequality?” in Sheldon Danziger and Peter Gottschalk (eds.), Uneven Tides: The Rising Inequality in America, Russell Sage Foundation, New York, pp. 133– 163. Gottschalk, Peter, and Robert Moffitt (1994), “The Growth of Earnings Instability in the U.S. Labor Market,” Brookings Papers on Economic Activity, Vol. 2, pp. 217–272. Gramlich, Edward M., Robert Kasten, and Frank Sammartino (1993), “Growing Inequality in the 1980s: The Role of Federal Taxes and Cash Transfers,” in Sheldon Danziger and Peter Gottschalk (eds.), 136 Uneven Tides: Rising Inequality in America, Russell Sage Foundation, New York, pp. 225–249. Hungerford, Thomas (1993), “U.S. Income Mobility in the Seventies and Eighties,” Review of Income and Wealth, Vol. 39, No. 4, pp. 403– 417. Husted, Thomas (1991), “Changes in State Income Inequality from 1981 to 1987,” The Review of Regional Studies, Vol. 21, No. 3, pp. 249–260. Juhn, Chinhui, Kevin M. Murphy, and Brooks Pierce (1993), “Wage Inequality and the Rise in the Return to Skill,” Journal of Political Economy, Vol. 101, June, pp. 410–442. Karoly, Lynn (1993), “The Trend in Inequality Among Families, Individuals, and Workers in the United States: A Twenty-Five Year Perspective,” in Sheldon Danziger and Peter Gottschalk (eds.), Uneven Tides: Rising Inequality in America, Russell Sage Foundation, New York, pp. 19–97. Karoly, Lynn (1995), unpublished testimony prepared for the California Legislature Assembly Committee on Labor and Employment, October 23. Karoly, Lynn, and Jacob Alex Klerman (1994), “Using Regional Data to Reexamine the Contribution of Demographic and Sectoral Changes to Increasing U.S. Wage Inequality,” in J. H. Bergstrand, T. F. Cosimano, J. W. Houck, and R. G. Sheehan (eds.), The Changing Distribution of Income in an Open U.S. Economy, North-Holland, Amsterdam, pp. 183–216. Karoly, Lynn, and Gary Burtless (1995), “Demographic Change, Rising Earnings Inequality, and the Distribution of Personal Well-Being, 1959–1989,” Demography, Vol. 32, No. 3, pp. 379–405. Krueger, Alan (1993), “How Computers Have Changed the Wage Structure: Evidence from Microdata, 1984–89,” Quarterly Journal of Economics, Vol. 108, February, pp. 33–60. Levy, Frank (1987), Dollars and Dreams: The Changing American Income Distribution, Russell Sage Foundation, New York. 137 Levy, Frank, and Richard J. Murnane (1992), “U.S. Earnings Levels and Earnings Inequality: A Review of Recent Trends and Proposed Explanations,” Journal of Economic Literature, Vol. 30, September, pp. 1333–1381. McMahon, Walter (1991), “Geographical Cost of Living Differences: An Update,” AREUEA Journal, Vol. 19, No. 3, pp. 426–450. Pechman, Joseph A. (1990), “The Future of the Income Tax,” American Economic Review, Vol. 80, No. 1, pp. 1–20. Topel, Robert (1994), “Regional Labor Markets and the Determinants of Wage Inequality,” American Economic Review, Vol. 84, No. 2, pp. 17– 22. U.S. Bureau of the Census (1990), Current Population Reports, Series P-60, Washington, D.C. U.S. Bureau of the Census (1994), Statistical Abstract of the United States, Washington, D.C. U.S. Bureau of the Census (1996), Current Population Reports, P-60-191, Washington, D.C. U.S. Bureau of Labor Statistics (1982), “Family Budgets for 1981: Final Report,” Monthly Labor Review, U.S. Department of Labor, Washington, D.C. U.S. Department of Commerce and the U.S. Bureau of the Census (1978), The Current Population Survey: Design and Methodology, January, Washington, D.C. U.S. House of Representatives, Committee on Ways and Means (1989), “Trends in Family Income and Income Inequality,” Background Material and Data on Programs Within the Jurisdiction of the Committee on Ways and Means, U.S. Government Printing Office, Washington, D.C. Williamson, Jeffrey, and Peter Lindert (1980), American Inequality: A Macroeconomic History, Academic Press, New York. 138 About the Authors DEBORAH S. REED Deborah Reed, a research fellow at the Public Policy Institute of California, is a specialist in labor economics and development resources. Her research interests include labor markets, public policy, and poverty in the United States and Brazil. Reed recently completed a post-doctoral fellowship at the Population Studies Center at the University of Michigan. A recipient of fellowships from the Mellon Foundation and Yale University, she has served as a consultant for the World Bank in addition to her teaching and research activities. Reed received an A.B. (1989) in economics from the University of California, Berkeley, and an M.A. (1990) in economics from Yale University. She will receive her Ph.D. in economics from Yale University in 1996. MELISSA GLENN HABER Melissa Glenn Haber is a research assistant at the Public Policy Institute of California. Before joining PPIC, Haber was a research associate at the Family Welfare Research Group where she developed a cost-benefit model to evaluate a state-wide teen pregnancy prevention program funded by the California Office of Family Planning. While at the Research Group, she also produced a report summarizing the cost, causes, and incidence of adolescent childbearing in the United States and in California. Haber received a B.A. (1991) in comparative religion from Harvard University and a Master’s degree in Public Policy (1995) from the University of California, Berkeley. LAURA A. MAMEESH Laura Mameesh is a research assistant at the Public Policy Institute of California. She has worked for Los Angeles Mayor Richard Riordan on a business tax relief policy aimed at retaining targeted industries in Los Angeles. Her analysis included calculating the cost to the city of providing tax relief as well as researching the cost of doing business in Los Angeles compared with competitor cities. Prior to her work in Los Angeles, Mameesh worked at the Law & Economics Consulting Group on anti-trust cases. Mameesh received her B.A. (1990) in economics from Mills College and is near completion of a Master’s degree in Public Policy from the University of Southern California." } ["___content":protected]=> string(102) "

R 796DRR

" ["_permalink":protected]=> string(83) "https://www.ppic.org/publication/the-distribution-of-income-in-california/r_796drr/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(8122) ["ID"]=> int(8122) ["post_author"]=> string(1) "1" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-05-20 02:35:02" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(3231) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(8) "R 796DRR" ["post_type"]=> string(10) "attachment" ["slug"]=> string(8) "r_796drr" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(12) "R_796DRR.pdf" ["wpmf_size"]=> string(6) "487956" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(231434) "The Distribution of Income in California Deborah Reed, Melissa Glenn Haber, Laura Mameesh July 1996 Copyright © 1996 Public Policy Institute of California, San Francisco, CA. All rights reserved. PPIC permits short sections of text, not to exceed three paragraphs, to be quoted without written permission, provided that full attribution is given to the source and the above copyright notice is included. Foreword This report on income distribution is the first research publication of the Public Policy Institute of California (PPIC). In developing the initial research agenda for the institute, we focused on fundamental changes that are sweeping the state. Of the many possibilities, one area that clearly deserves a place on our list is the dramatically changing nature of the state’s economy. California has emerged from its deepest recession since the 1930s. The economy is expanding steadily, with hundreds of thousands of jobs being created annually. In the bloom of this recovery, it makes sense to step back and look at the changes in the distribution of income in California that have occurred in recent years and over the last several decades. Recent efforts to measure and explain changes in income distribution in the United States have generated considerable interest and debate. This report is the first in a series that will replicate for California much of iii the national-level analysis. The authors document, for the first time, the annual changes in income distribution in the state from the late 1960s through 1994. Subsequent reports will examine the causes of increasing inequality, exploring the role of such factors as technological change, international competition, immigration, deunionization, and the shifting demographics of California. We trust that these reports not only will improve our understanding of the economic changes under way in California but will signal PPIC’s commitment to high-quality research and analyses useful to policy audiences. The authors express their appreciation to Sheldon Danziger from the University of Michigan and Lynn Karoly from RAND for their timely and extensive comments on an earlier draft. Lori Dair and John Ellwood were essential to the production of this report. Patricia Bedrosian, Jerry Lubenow, and Joyce Peterson provided considerable editorial assistance. The study has benefited from the efforts of Janet DeLand, Rod Pedersen, Eileen Roush, Peg Schumacher, Michael Shires, Karen Steeber, Michael Teitz, Paul Tractenberg, and numerous colleagues at PPIC. While this report reflects the contributions of many people, the authors are solely responsible for its content. David W. Lyon President and CEO Public Policy Institute of California iv Summary In recent years, increasing inequality in the distribution of income has been a subject of considerable public concern, political attention, and academic research. Income inequality is a measure of how equally the income pie is divided among all members of society. In other words, it is a measure of relative income, gauging, for example, how well the poor are doing economically compared with the rich. In the United States, income inequality remained stable in the three decades that followed World War II, as rich and poor alike benefited from the nation’s growing affluence. By the 1960s, Americans had come to accept as an article of faith President John Kennedy’s assertion that a rising tide would lift all boats. However, since the early 1970s the gap separating the rich and the poor has grown wider. While national studies have documented a growth in income inequality throughout the 1970s and 1980s, relatively little research has been done on income distribution in California. Such research is crucial v to the reasoned resolution of a broad range of state issues such as tax policy, public education, the minimum wage, and welfare reform that both affect and are affected by the distribution of income. The well-being of California’s population is a major research theme of the Public Policy Institute of California. This report is the first in a series that aims to identify state-specific policy strategies to promote equity as well as growth in the state’s economy. This initial study documents trends in income distribution in California from 1967 to 1994 and compares them to trends in other states, other regions, and the nation as a whole. Successive studies will investigate the underlying causes of the trends and will examine the relationship between public policy and the distribution of income. In this summary, we discuss the major findings of the study that, we believe, will be of interest to general and policy audiences concerned with important state issues. The body of the report and the appendices describe in greater detail the study’s results, approaches, measures, and data sources. We have striven to make the discussion in all parts of the report accessible to all interested audiences. Summing Up the Picture of California Income Inequality Income inequality has increased steadily in California over the last three decades. Until the late 1980s, the trend in California was remarkably similar to the national trend but, since then, inequality has risen much faster in the state than in the nation. This change has held for adjusted household incomes and for male earnings but not for female earnings. vi In both California and the nation, the increasing inequality results from income growth at the top of the distribution and decline in incomes at the very bottom. However, the recent divergence in inequality trends between California and the nation does not arise from faster growth at the top in California: In fact, income growth at all levels has been slower in California. Instead, the greater increase in the state results from a precipitous drop in income at the mid-to-lowest levels of the distribution. Rapid growth in income inequality has coincided with business cycle recessions, with those at the lower levels especially hard hit during recessions. A crucial difference has been that in the nation at large, incomes of people at those levels rebounded more during business cycle upswings than they did in California. However, the inequality gap between the nation and California began to widen as early as 1987, even before the recent, deep recession. These results suggest that in the interest of equity and economic growth in the state, it is essential that future research identify the forces that have made people at the lower end of the distribution lose so much ground and examine what happened in California even before the most recent recession. More on the Study’s Major Findings In this study, we used five summary measures of inequality, 26 definitions of income, and two data series (the Current Population Survey and the Census) to analyze California income levels and trends and to compare them with national and regional levels and trends. Our major findings are summarized below. vii Income Inequality Has Increased Substantially in California Figure S.1 illustrates how much the distribution of annual earnings has widened among male workers in California. The middle line of the graph shows the percentage change in real, inflation-adjusted median male earnings since 1967. The lower line of the figure shows the decline of male earnings at the 20th percentile, the income level that separates the bottom 20 percent of earners from the top 80 percent. The upper line of the figure shows the growing earnings at the 80th percentile. As shown in Figure S.1, the median of male earnings fell 20 percent between 1967 and 1994. This 20 percent decline represents a drop in median male earnings from $31,252 to $25,000 in real 1994 dollars. At Percent change since 1967 30 California 20 10 0 –10 –20 –30 –40 –50 1967 20th Median 80th 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Sample includes civilians age 18 and older who received wage and salary income. Statistics reported in this figure are adjusted for inflation. Figure S.1—Percentage Change in Real Annual Earnings for Males in California, by Income Percentile, 1967–1994 viii the 20th percentile, male earnings fell 40 percent from $17,316 in 1967 to $10,400 in 1994. In 1967, a man at the 80th percentile earned $44,345, about two and a half times what a man at the 20th percentile earned. By 1994, male earnings at the 80th percentile had increased 13 percent to $50,000, about five times what a man at the 20th percentile earned in that year. This comparison of the earnings of men in the upper-middle to the lower-middle of the distribution—the 80/20 ratio— is one simple measure of inequality. By this measure, male earnings inequality increased by 88 percent between 1967 and 1994 in California. Although the 80/20 ratio is an intuitive measure of inequality, it captures only two points in the distribution of income. Other measures of inequality are preferable because they summarize the entire distribution of income including the extreme top and bottom. One such measure is the coefficient of variation (CV). By this measure, male earnings inequality increased 41 percent between 1967 and 1994. Even by 1989, before the most recent recession, inequality had increased 35 percent since 1967. Income Inequality in California Matched That of the Nation Until the Late 1980s Inequality in household income has also grown. As Figure S.2 shows, household income inequality was similar in California and the nation for most of the years studied: It fluctuated in the 1970s, increasing during recessions and declining in recovery. It shot up dramatically during the recession of the early 1980s and never returned to pre-recession levels. It then remained fairly stable at new, higher levels through the mid-1980s. ix Index of inequality 0.81 0.79 0.77 0.75 0.73 0.71 0.69 0.67 0.65 0.63 1967 CA U.S. 1971 1975 1979 1983 1987 1991 1994 SOURCE: Authors’ calculations from the March CPS. NOTE: Household income is adjusted for household size and weighted by person. Statistics in this figure are not affected by inflation. The index of inequality used is the coefficient of variation (CV). The CV is the standard deviation of income divided by the mean of income. Figure S.2—Household Income Inequality in California and the Nation, 1967–1994 The trend in California’s income inequality began to diverge from the national trend in 1987. Inequality in household income started to increase faster in the state than in the nation, with especially rapid increases during the most recent recession. The period beginning in the late 1980s stands out as the only time when California has had substantially higher household income inequality than the nation for so many consecutive years. This divergence is also found in male earnings but not in female earnings. The fast-rising trend of inequality in California is also markedly visible when compared with other states. In 1969, 20 states had higher household income and male earnings inequality. By 1989, only five x states had higher household income inequality and only two had higher male earnings inequality. Income in California Has Grown More Slowly at the Top and Declined More Rapidly at the Bottom The sharp divergence between the state and the nation is not the result of higher income growth for the rich in California. As Figure S.3 shows, household income grew more in the nation than in the state. Between 1969 and 1989, two peak years of the business cycle, household income at the 90th percentile grew by 42 percent in the nation and 31 percent in the state. Instead, the divergence results from a greater income decline at the bottom. At the 10th percentile, while income in the nation grew by 7 percent, it actually fell by 7 percent in California. Incorporating data from the most recent recession shows an even more dramatic difference in growth between the nation and the state, as seen in the lower panel of the figure. Between the business cycle troughs of 1976 and 1994, income levels at the median and below fell in California, but in the United States they fell only at the 10th and 20th percentiles. Moreover, the decline in income at the 10th percentile in the United States was not nearly so drastic as in the state: Nationally, income fell by 8 percent, but in California, it fell by a remarkable 30 percent. Rapid Growth in Income Inequality Coincided with Business Cycle Recessions In both California and the nation, rapid growth in inequality coincided with recessions. The most noticeable increases in household income inequality, for example, occurred during the recessions of the early 1970s, early 1980s, and early 1990s. To illustrate, inequality in xi Percent change Percentage Change Between 1969 and 1989 Business Cycle Peaks 42 40 38 CA 30 U.S. 34 31 31 27 26 23 20 19 21 17 16 11 10 7 12 3 0 –10 –7 –5 10th 20th 30th 40th 50th 60th 70th 80th 90th Percentiles Percent change Percentage Change Between 1976 and 1994 Business Cycle Troughs 30 25 21 20 13 16 15 18 10 8 9 45 5 0 –10 –20 –30 –1 –8 –14 –22 –30 –10 –3 CA U.S. –40 10th 20th 30th 40th 50th 60th 70th 80th 90th Percentiles SOURCE: Based on authors’ calculations from the March CPS. NOTES: Household income is adjusted for household size and weighted by persons. Statistics in this figure are adjusted for inflation. Figure S.3—Percentage Change in Household Income Between Selected Years xii adjusted household income increased by 9 percent in California during the 1979–1982 recession, but it increased by only 3 percent during the economic growth of the next seven years. The relationship between the business cycle and inequality is particularly strong for male annual earnings in California: Inequality increased by 13 percent between 1979 and 1982 but by only 1 percent between 1982 and 1989. The recessions of the early 1970s and 1990s hit California harder than the nation (as is reflected in the larger increases in inequality shown in Figure S.2). While much of the rapid rise in inequality since 1981 occurred during the recession of the early 1990s, not all of the difference between California and the nation can be attributed to the strength of the recession in the state. The California growth trend in inequality began to outpace the national trend even before the start of the recession. Considering the Implications While inequality can increase because of the unequal sharing of income growth, it is particularly disturbing when it arises because of a decline in the income of poor individuals and households. This is the pattern that has characterized the increasing inequality in California over the last three decades. It is important to note, however, that the results of the study do not indicate that people who were poor in the past have gotten poorer—nor, conversely, that none have prospered. People who were in the 20th percentile in 1967 could have been in the 80th percentile in 1994. The data for this analysis are cross-sectional (snapshots of those in income groups in each year), not longitudinal, and therefore do not follow the fortunes of specific families or individuals over time. What the analysis xiii does tell us is that the poor in 1994 were considerably worse off than the poor in 1967. Moreover, as income falls at the bottom of the distribution, a greater percentage of people fall below the official poverty line (or any other absolute level of need). In other words, more Californians are poor today than were poor in the late 1960s. Given the similar trends in income inequality in California and the nation, it seems likely that the same forces are at work in both. Research on the underlying causes at the national level suggests a combination of factors: labor market trends influenced by technological change, international competition, immigration, and deunionization; and demographic trends in marriage and female employment. If these same forces explain the rise in inequality in the state, however, the recent sharp divergence suggests possible differential effects of those forces in California. Some Americans believe that differences in income arise primarily from individual choices, preferences, abilities, investments, and productivity, and that income inequality is a product of an economy that values hard work and talent. Other Americans believe that income differences reflect the unequal distribution of economic opportunity in our society, and that the opportunity to succeed is elusive for those who do not belong to privileged groups. The first viewpoint implies that public policy can affect inequality only by redistributing income; the second implies that policy can reduce inequality by promoting opportunity. Research on the determinants of the income distribution and the extent to which policy provides or restricts economic opportunity will suggest avenues for improving opportunities for the lessadvantaged. xiv Continuing growth in inequality is not inevitable. It is evident that government policies do affect the distribution of income, although the mechanisms are not fully understood. The challenge for future research is to examine the underlying forces behind the recent growth in inequality and to identify state policies that can promote equity and opportunity, as well as efficiency, in the California economy. xv Contents Foreword ..................................... Summary..................................... Figures ...................................... Tables ....................................... iii v xxi xxv 1. INTRODUCTION ........................... Trends in Income Inequality ..................... Nature of the Study ........................... 2. TRENDS IN THE DISTRIBUTION OF HOUSEHOLD INCOME ................................. What Is Household Income? ..................... Adjusted Household Income Is Sensitive to the Business Cycle ................................. The Distribution of Household Income Has Widened, Especially During Recessions .................. Other Summary Measures Also Show Rising Inequality .... Census Data Also Show Rising Household Income Inequality .............................. Household Income Inequality Rose Faster in California Than in Other Regions and States ................... Adjusted Family Income Shows Rising Inequality ........ 1 1 4 6 7 8 11 17 21 24 26 xvii 3. TRENDS IN THE DISTRIBUTION OF LABOR INCOME ................................. What Is Labor Income? ......................... Trends in the Distribution of Labor Income Among Males .. The Widening Distribution of Male Annual Earnings .... Summary Measures Show Rising Inequality in Male Annual Earnings .......................... Census Data Show Rising Inequality of Male Annual Earnings ............................... The Widening Distribution of Male Hourly Wages ..... Summary Measures Show Rising Inequality in Male Wages................................. Male Labor Income Inequality Rose Faster in California Than in Other Regions and States ............... Other Definitions of Male Labor Income Show Rising Inequality .............................. Trends in the Distribution of Labor Income Among Females ................................ The Narrowing, Then Widening, Distribution of Female Annual Earnings .......................... Measures of Inequality Show Falling, Then Rising, Inequality in Female Annual Earnings ............ Census Data Show a Fall and Then a Rise in Inequality of Female Annual Earnings ..................... The Widening Distribution of Female Hourly Wages .... Measures of Inequality Show Rising Inequality in Female Hourly Wages ........................... Female Labor Income Inequality Is Similar in California to Other Regions and States .................... Other Definitions of Female Labor Income Show Falling, Then Rising, Inequality ..................... 4. CONCLUSIONS AND IMPLICATIONS FOR POLICY AND FUTURE RESEARCH ..................... Public Policy and the Distribution of Income ........... Labor Market Explanations for Rising Earnings Inequality ... Demographic Explanations for Rising Family Income Inequality .............................. 30 31 33 34 37 37 40 42 43 46 47 47 50 50 53 56 56 58 60 61 62 64 xviii Additional Measurement Issues .................... 66 The Challenge for the State ...................... 67 Appendix A. Notes on Data and Methodology ................... 69 B. Using the Current Population Survey to Represent California .................................. 82 C. Trends in the Distributions of Alternative Measures of Income ................................... 87 D. Supplementary Statistics ........................ 115 Bibliography .................................. 135 xix Figures S.1. Percentage Change in Real Annual Earnings for Males in California, by Income Percentile, 1967–1994 ........ viii S.2. Household Income Inequality in California and the Nation, 1967–1994 ........................ x S.3. Percentage Change in Household Income Between Selected Years ............................ xii 2.1. Trends in the Unemployment Rate and Median Real Adjusted Household Income, 1967–1994 .......... 9 2.2. Percentage Change in Real Adjusted Household Income, by Income Percentile, 1967–1994 ............... 13 2.3. Summary Measures of Inequality for Real Adjusted Household Income, 1967–1994................. 20 2.4. Summary Measures of Inequality for Real Adjusted Family Income, 1967–1994 ................... 28 3.1. Percentage Change in Real Annual Earnings for Males, by Income Percentile, 1967–1994 ............... 35 3.2. Summary Measures of Inequality for Male Annual Earnings, 1967–1994 ....................... 38 xxi 3.3. Percentage Change in Real Hourly Wages for Males, by Income Percentile, 1975–1994 ................. 41 3.4. Summary Measures of Inequality for Male Hourly Wages, 1975–1994 ......................... 44 3.5. Percentage Change in Real Annual Earnings for Females, by Income Percentile, 1967–1994 ............... 49 3.6. Summary Measures of Inequality for Female Annual Earnings, 1967–1994 ....................... 51 3.7. Percentage Change in Real Hourly Wages for Females, by Income Percentile, 1975–1994 ............... 54 3.8. Summary Measures of Inequality for Female Hourly Wages, 1975–1994 ......................... 57 C.1. Income Type 2: Summary Measures of Inequality for Unadjusted Household Income Among Persons, 1967–1994 .............................. 95 C.2. Income Type 3: Summary Measures of Inequality for Adjusted Household Income Among Households, 1967–1994 .............................. 96 C.3. Income Type 4: Summary Measures of Inequality for Unadjusted Household Income Among Households, 1967–1994 .............................. 97 C.4. Income Type 6: Summary Measures of Inequality for Unadjusted Family Income Among Persons, 1967–1994 .............................. 98 C.5. Income Type 7: Summary Measures of Inequality for Adjusted Family Income Among Families, 1967–1994 .............................. 99 C.6. Income Type 8: Summary Measures of Inequality for Unadjusted Family Income Among Families, 1967–1994 .............................. 100 C.7. Income Type 9: Summary Measures of Inequality for Adjusted Primary Family Income Among Persons, 1967–1994 .............................. 101 xxii C.8. Income Type 10: Summary Measures of Inequality for Unadjusted Primary Family Income Among Persons, 1967–1994 .............................. 102 C.9. Income Type 11: Summary Measures of Inequality for Adjusted Primary Family Income Among Families, 1967–1994 .............................. 103 C.10. Income Type 12: Summary Measures of Inequality for Unadjusted Primary Family Income Among Families, 1967–1994 .............................. 104 C.11. Income Type 15: Summary Measures of Inequality for Annual Earnings for Males Ages 18 to 55, 1967–1994 .. 105 C.12. Income Type 16: Summary Measures of Inequality for Hourly Wages for Males Ages 18 to 55, 1975–1994 .... 106 C.13. Income Type 17: Summary Measures of Inequality for Annual Earnings for All Male Workers, 1967–1994 .... 107 C.14. Income Type 18: Summary Measures of Inequality for Hourly Wages for All Male Workers, 1975–1994...... 108 C.15. Income Type 19: Summary Measures of Inequality for Annual Salary for Males, 1967–1994.............. 109 C.16. Income Type 22: Summary Measures of Inequality for Annual Earnings for Females Ages 18 to 55, 1967–1994 .............................. 110 C.17. Income Type 23: Summary Measures of Inequality for Hourly Wages for Females Ages 18 to 55, 1975–1994 ... 111 C.18. Income Type 24: Summary Measures of Inequality for Annual Earnings for All Female Workers, 1967–1994 ... 112 C.19. Income Type 25: Summary Measures of Inequality for Hourly Wages for All Female Workers, 1975–1994 .... 113 C.20. Income Type 26: Summary Measures of Inequality for Annual Salary for Females, 1967–1994 ............ 114 xxiii Tables 2.1. Percentage Change in Real Adjusted Household Income Between Selected Years, by Income Percentile: CPS .... 2.2. Percentage Change in Real Adjusted Household Income Between Selected Years, by Income Percentile: Census .. 2.3. Levels and Trends in the Coefficient of Variation for Adjusted Household Income: CPS and Census ....... 2.4. Regional Trends in the Coefficient of Variation for Real Adjusted Household Income, 1969–1994 .......... 2.5. Percentage Change in Real Adjusted Family Income Between Selected Years, by Income Percentile ........ 3.1. Percentage Change in Real Annual Earnings for Males Between Selected Years, by Income Percentile: CPS .... 3.2. Percentage Change in Real Annual Earnings for Males, by Income Percentile: Census .................. 3.3. Levels and Trends in the Coefficient of Variation for Male Annual Earnings: CPS and Census ........... 3.4. Percentage Change in Real Hourly Wages for Males Between Selected Years, by Income Percentile ........ 16 22 23 25 27 36 39 40 43 xxv 3.5. Regional Trends in the Coefficient of Variation for Real Hourly Wages Among Males, 1979–1994 .......... 3.6. Percentage Change in Real Annual Earnings for Females Between Selected Years, by Income Percentile ........ 3.7. Percentage Change in Real Annual Earnings and Hourly Wages for Females, by Income Percentile: Census ..... 3.8. Levels and Trends in the Coefficient of Variation for Female Annual Earnings: CPS and Census ......... 3.9. Percentage Change in Real Hourly Wages for Females Between Selected Years, by Income Percentile ........ 3.10. Regional Trends in the Coefficient of Variation for Real Hourly Wages Among Females, 1979–1994 ......... A.1. Price and Cost of Living Adjustments, California and the United States, 1967–1994 .................... B.1. Percentage of Population in Each Category: Census and CPS .................................. C.1. Alternative Measures of Household and Family Income................................. C.2. Alternative Measures of Labor Income ............ C.3. Income Type 2: Percentage Change in Real Unadjusted Household Income Among Persons Between Selected Years, by Income Percentile ................... C.4. Income Type 3: Percentage Change in Real Adjusted Household Income Among Households Between Selected Years, by Income Percentile .............. C.5. Income Type 4: Percentage Change in Real Unadjusted Household Income Among Households Between Selected Years, by Income Percentile .............. C.6. Income Type 6: Percentage Change in Real Unadjusted Family Income Among Persons Between Selected Years, by Income Percentile ........................ 45 48 52 53 55 58 80 86 90 92 95 96 97 98 xxvi C.7. Income Type 7: Percentage Change in Real Adjusted Family Income Among Families Between Selected Years, by Income Percentile ........................ 99 C.8. Income Type 8: Percentage Change in Real Unadjusted Family Income Among Families Between Selected Years, by Income Percentile ........................ 100 C.9. Income Type 9: Percentage Change in Real Adjusted Primary Family Income Among Persons Between Selected Years, by Income Percentile .............. 101 C.10. Income Type 10: Percentage Change in Real Unadjusted Primary Family Income Among Persons Between Selected Years, by Income Percentile ........ 102 C.11. Income Type 11: Percentage Change in Real Adjusted Primary Family Income Among Families Between Selected Years, by Income Percentile .............. 103 C.12. Income Type 12: Percentage Change in Real Unadjusted Primary Family Income Among Families Between Selected Years, by Income Percentile ........ 104 C.13. Income Type 15: Percentage Change in Real Annual Earnings for Males Ages 18 to 55 Between Selected Years, by Income Percentile ................... 105 C.14. Income Type 16: Percentage Change in Real Hourly Wages for Males Ages 18 to 55 Between Selected Years, by Income Percentile ........................ 106 C.15. Income Type 17: Percentage Change in Real Annual Earnings for All Male Workers Between Selected Years, by Income Percentile ........................ 107 C.16. Income Type 18: Percentage Change in Real Hourly Wages for All Male Workers Between Selected Years, by Income Percentile ........................ 108 C.17. Income Type 19: Percentage Change in Real Annual Salary for Males Between Selected Years, by Income Percentile ............................... 109 xxvii C.18. Income Type 22: Percentage Change in Real Annual Earnings for Females Ages 18 to 55 Between Selected Years, by Income Percentile ................... 110 C.19. Income Type 23: Percentage Change in Real Hourly Wages for Females Ages 18 to 55 Between Selected Years, by Income Percentile ................... 111 C.20. Income Type 24: Percentage Change in Real Annual Earnings for All Female Workers Between Selected Years, by Income Percentile ........................ 112 C.21. Income Type 25: Percentage Change in Real Hourly Wages for All Female Workers Between Selected Years, by Income Percentile ........................ 113 C.22. Income Type 26: Percentage Change in Real Annual Salary for Females Between Selected Years, by Income Percentile ............................... 114 D.1. Deciles of Nominal Adjusted Household Income, California ............................... 116 D.2. Deciles of Nominal Adjusted Household Income, United States .................................. 117 D.3. Deciles of Nominal Annual Earnings Among Male Workers, California ........................ 118 D.4. Deciles of Nominal Annual Earnings Among Male Workers, United States ...................... 119 D.5. Deciles of Nominal Hourly Wages Among Male Workers, California ........................ 120 D.6. Deciles of Nominal Hourly Wages Among Male Workers, United States ...................... 121 D.7. Deciles of Nominal Annual Earnings Among Female Workers, California ........................ 122 D.8. Deciles of Nominal Annual Earnings Among Female Workers, United States ...................... 123 D.9. Deciles of Nominal Hourly Wages Among Female Workers, California ........................ 124 xxviii D.10. Deciles of Nominal Hourly Wages Among Female Workers, United States ...................... 125 D.11. Regional Trends in the Coefficient of Variation for Real Annual Earnings Among Males, 1969–1994 ......... 126 D.12. Regional Trends in the Coefficient of Variation for Real Annual Earnings Among Females, 1969–1994 ....... 127 D.13. State Rankings for Adjusted Household Income Inequality Based on the Coefficient of Variation: Census ................................. 128 D.14. State Rankings for Male Annual Earnings Inequality Based on the Coefficient of Variation: Census ....... 130 D.15. State Rankings for Female Annual Earnings Inequality Based on the Coefficient of Variation: Census ....... 132 xxix 1. Introduction A fuller understanding of state-level trends in the distribution of income is essential for California. The recent trends not only will determine the need for strengthened state policies to aid low-income families but will affect the likely success of those policies. This report provides a comprehensive description of the widening distribution of income in California. Its findings reveal a general pattern of increasing income inequality in the state, explained by a dramatic decline in income for the poor and near-poor accompanied by income growth for the rich. Subsequent studies by the Public Policy Institute of California will examine the underlying causes of the trends and will explore the relationships between state policy and income inequality. Trends in Income Inequality The upward trend in income inequality in the United States throughout the 1970s and 1980s stands in marked contrast with the 1 trends in the distribution of income from the Great Depression to the late 1960s. Jeffrey Williamson and Peter Lindert (1980) report a dramatic decline in income inequality between the Depression and the end of World War II. From the late 1940s until the late 1970s, inequality fluctuated within a relatively narrow band. This long period of stability in income inequality led to speculation that, with the exception of short-term fluctuations, the distribution of economic wellbeing would remain constant (Blinder, 1980). Tracking changes in inequality, wrote one researcher, was like “watching grass grow” (Aaron, 1978, p. 17). The conventional wisdom was too optimistic. In the early 1980s, Census Bureau reports provided some of the earliest indications of a growing inequality among families. Census Bureau income statistics revealed that family income inequality had reached a postwar low in the late 1960s but had climbed almost constantly from that time. Since the early 1980s, family income inequality has remained higher than in any previous year since the end of the Second World War.1 In recent years, numerous studies have documented the widening distribution of family income and male earnings in the United States. We summarize this work here. Sheldon Danziger and Peter Gottschalk (1995) report that the gap in income between families near the top of the income distribution and those near the bottom has increased, in both recession and recovery, since the recession of the early 1970s. The years 1983 to 1989 stand out as an anomalous period that recorded growth in mean family income along with rising income inequality. ____________ 1U.S. Bureau of the Census, Current Population Reports, P-60 series, various issues. 2 In their comprehensive 1992 review article, Frank Levy and Richard Murnane conclude that the 1970s were a period of either stability or gradual growth in male annual earnings inequality and that the 1980s were a period of rapid increase. Lynn Karoly (1993) shows that this rise in income inequality is explained by a decline in the income of poor families and workers and by growth in the income of the rich. While much attention has been focused on the trends in income inequality at the national level, relatively few studies have investigated income distribution in the state of California. There are many reasons to expect that the trends in the distribution of income in California will differ from those of the nation. Income inequality measures for the country as a whole aggregate regional diversity in economic and demographic trends. California is distinctive in its industrial base, trading partners, racial and ethnic composition, patterns of domestic and international migration, and in the age and education of its workforce. Previous research on the distribution of income in California shows that the state has experienced a rise in income disparity. Jay Chamberlain and Phil Spillberg (1991) report a rising concentration of adjusted gross income in the 1980s: Between 1980 and 1988, the proportion of the total after-tax adjusted income received by the top 20 percent of taxpayers increased from 52 to 57 percent; the proportion received by the top one percent increased from 10 to 16 percent. Karoly (1995) finds that the ratio of the income of wealthy families at the 90th percentile to the income of poor families at the 10th percentile—the 90/10 ratio— 3 increased by 74 percent between 1973 and 1993 in California.2 This rise in inequality was due to growth in the incomes of the rich and a substantial decline in the incomes of the poor. Moreover, the rise in inequality in California was larger than in the nation as a whole, where the 90/10 ratio increased by 54 percent. Research that compares California to other regions of the country is less conclusive. On the one hand, Robert Topel (1994) finds that the western region of the nation, dominated in population by California, experienced the largest increase of any region in male wage inequality between 1972 and 1990. On the other hand, Thomas Husted (1991) shows that between 1981 and 1987, the percentage increase in the Gini coefficient (one index of inequality) was higher in California than in the nation as a whole, but 24 states had larger percentage increases. Nature of the Study This study contributes to the existing research on the distribution of income in California by providing a comprehensive description of state trends and by comparing these to trends of the nation, other regions, and other states. To document the trends in income inequality thoroughly, the study uses five measures of inequality and 26 definitions of income. Data for this analysis come from the annual March file of the Current Population Survey and the decennial Census of Population and ____________ 2The 10th percentile is defined as the level of income that divides the bottom 10 percent from the top 90 percent; similarly, 90 percent of people have incomes below the 90th percentile, whereas only 10 percent have incomes above. 4 Housing.3 The analysis covers the entire period spanned by available public-use files of the Current Population Survey: 1967 to 1994. This study measures the trends for two main types of income: Household income provides a picture of general economic well-being because it includes all sources of money income and it is measured for all people regardless of work status. Labor income, the largest component of household income, measures earnings from work. Labor income reflects changes in the economy and is not directly influenced by changes in household structure. The next two chapters describe results of the study. Chapter 2 describes trends in the distribution of household income and Chapter 3 describes trends in the distribution of male and female labor income. Each chapter analyzes the California experience in relation to that of other regions and states. Chapter 4 presents our conclusions, discusses the relationship between public policy and the distribution of income, and outlines possible explanations for the rise of income inequality in California. Readers interested in greater technical details of the study are directed to the appendices: Appendix A describes the datasets used in the study. Appendix B discusses the representativeness of the California subsample of the Current Population Survey. Appendix C reports on trends in the distribution of alternative measures of income. Appendix D provides supplementary statistics on the distributions of income measures discussed in the text. ____________ 31970 Public Use Sample, 1 percent, and the 1980 and 1990 Public Use Micro Sample, 5 percent. 5 2. Trends in the Distribution of Household Income Household income is a measure of economic well-being that explicitly accounts for income-sharing among members of the same household. It is the most comprehensive measure of income in this study because it is measured for all people regardless of age and work status, and it incorporates income from all reported sources. As this chapter demonstrates, the distribution of household income in California has widened considerably over the past three decades, especially during business cycle recessions. Summary measures of inequality show that the increasing trend in household income inequality was similar for California and the nation until the late 1980s. Since then, the rise in inequality has been much greater in California. Compared to other states, California had one of the highest levels of inequality, even before the recent recession. 6 What Is Household Income? Household income is defined as the sum of income from all sources for all persons living in the same household unit. Because households with many persons require more resources than small households to maintain the same level of consumption, we adjust household income based on the number of household residents.1 We evaluate the distribution of adjusted household income across people, rather than across household units, by assigning to each person the adjusted income of his or her household. This method treats each person equally, rather than implicitly giving less weight in the calculation to people in large households.2 All income statistics reported in this study are adjusted to real 1994 dollars based on the consumer price index computed by the Bureau of Labor Statistics.3 Recent studies suggest that the official consumer price index may exaggerate inflation, thus understating growth and overstating ____________ 1We calculate adjusted household income by dividing total household income by the square root of the number of household residents. Karoly and Burtless (1995) suggest this adjustment factor because it is close to the adjustment for family size implicit in the official poverty thresholds. This adjustment takes into account “economies of scale” made possible through the sharing of common resources in large households. For example, the adjustment implies that a household with four people will require twice, rather than four times, the income of a single person to maintain the same level of consumption. We make the same adjustments to family income based on family size. Median levels of adjusted household and family income reported in the text are multiplied by two to represent income levels for households and families of four persons. For comparison, we also measure changes in the distribution of unadjusted household and family income (see Appendix C). 2Using this method, 50 percent of people live in households with adjusted incomes lower than the median, as opposed to 50 percent of households falling below the median. Similarly, we evaluate the distribution of adjusted family income across people as opposed to family units. For comparison, we measure trends in the distributions of household income across households and family income across families (see Appendix C). 3We use the CPI-U-X1 and allow for differences in the rate of inflation in California and the United States. See Appendix A for details. 7 decline. However, although the consumer price index affects estimated growth trends, the summary measures of inequality used in this report are based on relative income (e.g., the income of the rich relative to the income of the poor) and are not affected by inflation adjustments. The Current Population Survey and the Census report pre-tax money income including wages and salary, farm income, selfemployment income, interest and dividends, welfare receipts, and Social Security and retirement benefits. The income measures are imperfect indices of economic well-being because the data do not include information on tax payments, non-monetary transfers (e.g., housing subsidies, health benefits, food stamps), the return to investments such as owner-occupied housing, or measures of accumulated wealth. However, studies that have used more comprehensive measures of income have found trends in income inequality similar to those for pre-tax money income. (See Appendix C for a review of such studies.) Adjusted Household Income Is Sensitive to the Business Cycle Because the business cycle plays a strong role in the distributional trends we describe, we begin by showing business cycle fluctuations as measured by unemployment and associated fluctuations in household income. Figure 2.1 shows how strongly fluctuations in adjusted household income are related to the business cycle. The upper panel of the figure displays the unemployment rate in California and the United States from 1967 to 1994. Rising rates of unemployment characterize the periods of recession of the early 1970s, mid-1970s, early 1980s, and 8 Median adjusted household income ($) Unemployment rate (%) 10 9 8 7 6 5 4 3 2 1 0 1967 55,000 CA U.S. 1971 1975 1979 1983 1987 1991 1994 50,000 45,000 40,000 35,000 CA U.S. 30,000 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Real median adjusted household income is based on authors’ calculations from the March CPS. Unemployment rates are from the Bureau of Labor Statistics. NOTES: Median incomes have been converted to real 1994 dollars. The U.S. median has been adjusted to reflect the higher cost of living in California. Comparison of median income in California to median income in the United States should be made with caution because of measurement problems in the cost of living index (as described in Appendix A). The trend in median household income is sensitive to the consumer price index. Household income is adjusted for the number of people living in the household. Reported median household income is calibrated to represent a household of four people. Median household income in California in 1988 may not be comparable to other years due to changes in the CPS (as described in Appendix A). Figure 2.1—Trends in the Unemployment Rate and Median Real Adjusted Household Income, 1967–1994 9 early 1990s for both California and the nation. With the exception of the 1980s, unemployment has been higher in California than in the nation, particularly during the recessions of the early 1970s and early 1990s. Median adjusted household income (the lower panel of Figure 2.1) shows a positive growth trend through the mid-1980s for both California and the nation, with higher overall growth in the nation.4 Declines in median household income generally occurred only during periods of recession. However, median household income in California began to stagnate as early as 1987, even before the most recent recession. The greater decline in median household income and the higher unemployment rate in California indicate the stronger effect of the early 1990s recession in the state. (The dip in median household income in 1988 is probably explained by changes in sampling in the Current Population Survey. Results for 1988 are reported in this study but conclusions are not based on statistics specific to 1988.)5 The medians in Figure 2.1 are adjusted both for inflation and for the higher cost of living in California. The position of the Californian median relative to the national median is a function of the adjustments ____________ 4If no adjustments were made to household income for household size and if the distribution were evaluated at the household level and not the person level, the median in the United States would be less than 1 percent higher than the median in California in 1994. The median of unadjusted household income (weighted at the household level) increased in the United States relative to California from 1967 to 1978; after 1978, the relative growth of the U.S. median fluctuated with no clear trend. However, the median of adjusted household income (weighted by persons), the median reported in the text, is a preferred measure of economic well-being because it accounts for the greater resource needs of large households and it applies the same weight to people in large households as to people in small households. 5See Appendix A for further discussion of sampling and other data issues. 10 made for differences in the cost of living.6 For example, in 1994 the cost of living estimate was 9 percent higher in California than in the nation. The cost of living adjustments applied in this figure are calculated from Bureau of Labor Statistics data.7 Because cost of living estimates are imprecise, comparison of the California median to the national median should be made with caution. Although Figure 2.1 shows that the U.S. median adjusted household income was about $3,000 below that of California in 1967 and was about $4,000 above that of California in 1994, this result could be different if a more accurate cost of living index were available. However, the faster rise in the U.S. median does not depend on the cost of living adjustment. The median adjusted household income statistics in Figure 2.1 are the only statistics in this report that are affected by the cost of living adjustment. The Distribution of Household Income Has Widened, Especially During Recessions The most significant widening of the distribution of adjusted household income occurred during periods of recession, particularly in the early 1980s and early 1990s. Overall, the gap between the incomes of people in rich and poor households increased not only because incomes at the top of the distribution rose but also because incomes at the bottom of the distribution fell. ____________ 6For example, if no adjustments were made for cost of living, the median of adjusted household income in the United States would be about half a percent lower than the median in California in 1994. See Appendix D for the nominal value of income at each decile without adjustments for cost of living. 7See Appendix A for the calculation of cost of living adjustment for 1967–1994. 11 A straightforward way to investigate the changing shape of the distribution of income is to examine the relative income positions of low-, middle-, and high-income people. Figure 2.2 illustrates the income trends at the 10th, 20th, 50th (median), 80th, and 90th percentiles of adjusted household income.8 The figure shows the percentage change in income since 1967: For example, the highest point on the graph for California shows that people in the 90th percentile in 1987 had income slightly more than 40 percent higher than people in the 90th percentile in 1967. Although the reported statistics are standardized to the base year of 1967, the figure is not meant to imply that there was no household income inequality in 1967.9 Instead, the figure graphically represents the widening of the distribution and corresponding increases in inequality from its 1967 levels. The absolute decline of income levels for households near the bottom of the distribution in California is a striking feature of the figure. During the 1970s, the income received by households at the 10th and 20th percentiles of the distribution in California fluctuated mildly but showed little overall growth. During the recession of the early 1980s, the ____________ 8People in the 10th percentile have incomes higher than only 10 percent of the population; those in the 90th percentile have incomes higher than 90 percent of the population. People in the 10th and 20th percentiles are in the lower and lower-middle ranks of the income distribution; the median (or 50th percentile) describes the income level of people in the middle of the income distribution; the 80th and 90th percentiles indicate the income levels of people in the upper-middle and upper ranks of the income distribution. 9The information in the figure can be used to calculate the percentage change between any two years by using the following calculation: Add 100 to the values displayed on the figure, take the ratio, subtract 1, and multiply by 100. For example, between the business cycle peaks in 1979 and 1989, adjusted household income at the 10th percentile fell by 8 percent in California (100/108.5 – 1) * 100. 12 Percent change since 1967 Percent change since 1967 60 California 50 10th 40 20th Median 30 80th 90th 20 10 0 –10 –20 –30 1967 1971 1975 1979 1983 1987 1991 1994 60 50 United States 40 30 20 10 0 10th 20th –10 Median 80th –20 90th –30 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Statistics reported in this figure are sensitive to the consumer price index. Adjusted household income in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 2.2—Percentage Change in Real Adjusted Household Income, by Income Percentile, 1967–1994 13 income of households at the 20th percentile fell by 13 percent in California. Before recovering fully from this recession, income at the 20th percentile in California fell again by another 20 percent during the most recent recession. The income decline for households at the 10th percentile was even greater, with income plunging 15 percent and 23 percent during the two recessions. The national distribution of adjusted household income shows the same pattern of sharp decline at the bottom during recessions. However, the decline during recessions was greater in California than in the nation as a whole, especially for low-income households, and the growth during the recovery of the 1980s was weaker in California than in the United States. The trends depicted in Figure 2.2 can easily be misinterpreted. The figure shows that Californians at the 10th percentile in 1994 received 24 percent less income than Californians at the 10th percentile in 1967. The cross-sectional data used in this report do not track the same people over the years. The figure, therefore, does not show that the income of specific people at the 10th percentile declined by 24 percent. The distinction is often subtle. When we say that “the poor got poorer,” we mean that the people who were poor in 1994 were poorer than the people who were poor in 1967, but not that the same people who were poor in 1967 were even poorer in 1994.10 This interpretation issue is ____________ 10As an analogy, imagine a class of nine third-graders lined up in order of height. The height of the fifth child in the line is the median height. Now suppose that four shorter children enter the line, making the total 13. The median child is now the seventh child in the line—the child who was third in the original line. The new median height is lower than the previous median, but no child can be said to have “experienced a decline in height.” 14 particularly important in California where there is a high degree of mobility into and out of the state. Figure 2.2 clearly shows the widening gap between the upper (80th and 90th) and lower (10th and 20th ) percentiles. The income trends displayed in Figure 2.2 demonstrate the strong relationship between business cycle conditions and the widening distribution of household income. For both California and the United States, the recessions in the early 1980s and early 1990s stand out as periods when the distribution of household income widened rapidly, with precipitous drops in income levels at the lower percentiles of the distribution and small shorter-lived declines at the upper percentiles. In California, the widening of the distribution is more substantial than in the nation, showing a larger increase in inequality. Because of this relationship between the business cycle and income inequality, it is important to focus on years in similar points of the business cycle when describing the long-run trends in the distribution of income. Comparing the distributions of adjusted household income in 1967 and 1994, for example, is likely to exaggerate the trends in inequality growth because the economy was strong in 1967 and weak in 1994. To avoid such distortion, Table 2.1 summarizes the trends in Figure 2.2 for selected years at similar points in the business cycle. The first row of the table shows the absolute decline of adjusted household income for the lower-middle of the distribution (the 20th percentile) in California. Between the two major business cycle peaks spanned by our study, 1969 and 1989, the income level of households at the 20th percentile declined by 5 percent. The median of household income grew 16 percent over 15 Table 2.1 Percentage Change in Real Adjusted Household Income Between Selected Years, by Income Percentile: CPS Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th 2 –6 –5 –22 Median 14 2 16 –3 80th 20 5 26 15 Change in 80/20 ratio (%) +18 +12 +32 +48 United States 20th 10 1 11 –1 Median 18 8 27 8 80th 22 14 38 21 Change in 80/20 ratio (%) +10 +13 +24 +22 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. In this and following tables the percentage change between 1969 and 1989 is not equal to the sum of the changes between 1969–1979 and 1979–1989, because the change over the 1980s is calculated from the base year of 1979 and not 1969. For example, if income grew 100 percent between 1969 and 1979 from $10 to $20, and then grew by another 100 percent between 1979 and 1989 to $40, the overall change from 1969 to 1989 would be 400, not 200, percent. the same period. For the upper-middle of the distribution (the 80th percentile), household income grew 26 percent. Even over the 1980s, household income grew more in the United States than in California at each of these percentiles. The widening of the distribution of adjusted household income described in Figure 2.2 can be summarized by the ratio of income at the top of the distribution to income at the bottom. The ratio of the income of the 90th percentile to the income of the 10th percentile, the 90/10 ratio, is often used as a measure of inequality. We use the 80/20 ratio 16 instead, in order to focus on the widening of the middle of the distribution. Table 2.1 illustrates how seriously inequality has grown in California and the nation: The 80/20 ratio increased by 32 percent in California and by 24 percent in the United States between 1969 and 1989. The 80/20 ratios in Table 2.1 suggest that California had much faster growth in inequality than the nation during the 1970s. However, as the next section will show, this finding is not confirmed by other measures of inequality that take into account the entire distribution. This study emphasizes trends in the income distribution up until 1989 because it is impossible to determine whether later changes reflect short-run fluctuations due to the severity of the recent recession or a continuing trend of rapidly rising inequality. Nevertheless, changes between 1976 and 1994 are reported to demonstrate how seriously a deep recession, like the one of the early 1990s, can affect income inequality. As the fourth column in Table 2.1 displays, incorporating the most recent recession reflects the same pattern of income growth as shown up until 1989, but an even bleaker picture emerges, especially in California. Other Summary Measures Also Show Rising Inequality The percentile graphs in the previous section show the widening of the distribution of adjusted household income relative to 1967, but they do not provide an absolute measure of income inequality that would allow us to compare inequality in California and the nation. Since summary measures describe inequality with a single statistic, they make it 17 possible to rank the level of inequality in two different distributions of income (e.g., in the United States and in California). Moreover, the summary measures used in this report are independent of the consumer price index. Even if the consumer price index overstates inflation, the magnitude of the reported summary measures is not affected. The 80/20 ratio reported in the previous section is one summary measure of income inequality, but it suffers from the drawback that it evaluates only two positions in the distribution. There are numerous summary measures of income inequality that evaluate income throughout the distribution, including the extreme top and bottom. This study reports four commonly used and easily calculated measures: the coefficient of variation (CV), Theil’s entropy (ENTROPY), mean log-deviation (MLD), and the variance of the natural logarithm of income (VLN).11 These four were chosen in part to allow for comparability with other studies, particularly Karoly’s (1993) work on income inequality in the United States. There is no a priori best measure of inequality. All four measures agree on what it means to have a perfectly equal society: Each measure is scaled to equal zero when all members of society have the same amount of income. However, the measures do not agree on how to quantify deviations from perfect equality. For example, the VLN is more responsive to reductions in income near the bottom of the distribution: In an economy where nine people have $10 dollars each and one person has $8, the VLN measure will show higher inequality than if the ____________ 11The CV is the standard deviation of income divided by the mean of income. The ENTROPY measure is the mean of [y/mean(y) * ln(y/mean(y))], where y is income. The MLD is the natural logarithm of the mean of income minus the mean of the natural logarithm of income. VLN is the variance of ln(y). 18 anomalous person had $12, even though the deviation is $2 in both cases. This effect reflects the idea that downward deviations from equity have more negative consequences than upward ones. The CV treats upward and downward deviations the same—the CV measure would have the same value if the anomalous person has $8 or $12. If income grows across the distribution, but grows faster for the rich, then the CV will register a greater change in inequality than the VLN will. The MLD and ENTROPY measures emphasize the bottom of the distribution more than the CV but less than the VLN. That is, the VLN is the most responsive to changes at the bottom of the income distribution, followed by the MLD, ENTROPY, and CV, in that order. Because summary measures respond differently to income disparity, they may produce different rankings of income distributions. It is possible to find that income inequality is higher in the United States by some measures and higher in California by others. Similarly, the summary measures may show different time trends for income inequality in California. For this reason, using several summary measures of inequality and comparing the results across measures provide a fuller picture of the trends in income inequality. Figure 2.3 illustrates the inequality of adjusted household income in California and the United States using the four measures of inequality. Overall, the measures show that the level of household income inequality in California was quite similar to that of the nation until the late 1980s. In both California and the United States, the main patterns in income inequality are consistent with a rapid rise in inequality during recession periods. The recession of the early 1980s was a period of dramatic increase in household inequality in both California and the nation: 19 .80 CV .78 .76 CA .74 U.S. .72 .30 ENTROPY .28 CA U.S. .26 .70 .24 .68 .66 .22 .64 .20 .62 .60 .18 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 .49 MLD .44 CA U.S. .39 2.15 1.95 1.75 1.55 VLN CA U.S. .34 1.35 1.15 .29 .95 .24 .75 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Statistics reported in this figure are not sensitive to the consumer price index. Adjusted household income in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 2.3—Summary Measures of Inequality for Real Adjusted Household Income, 1967–1994 20 Between 1979 and 1982, the CV grew by 9 percent in California and by 8 percent in the United States. Over the same period, the VLN grew by 23 percent in California and by 29 percent in the nation, reflecting the greater sensitivity of this measure to the declining income near the bottom of the distribution. Adjusted household income inequality in the United States began another steep increase at the beginning of the most recent recession. In California, the increase began earlier, with small increases as early as 1987 preceding more drastic increases in the early 1990s. The late 1980s and early 1990s stand out as the only period over the last three decades that California has maintained a substantially higher level of adjusted household income inequality than the United States for several consecutive years. This is consistent with the more severe decline in adjusted household income at the median and lower percentiles in California during the most recent recession, as shown in Figure 2.2. Census Data Also Show Rising Household Income Inequality This report focuses primarily on income data from the Current Population Survey (CPS) because its annual data provide a fuller picture of the distribution trends than the decennial Census. Furthermore, the income data in the CPS are likely to be more accurate than the income data in the Census. For example, in 1990, the Census asked respondents about eight specific types of income. In the same year, the CPS asked about more than 20 types of income.12 Results from Census data are ____________ 12In addition, the CPS is conducted by phone by trained survey-takers whereas the Census is taken by mail. For a further discussion of these two datasets, see Appendix A. 21 provided here to address the concern that the CPS may not adequately represent California.13 The Census data do in fact confirm the trends in the distribution of adjusted household income as measured by the CPS.14 Table 2.2, based on Census data, shows that the growth in the upper-middle of the distribution (the 80th percentile) exceeded the growth in the lower-middle of the distribution (the 20th percentile). Table 2.2 Percentage Change in Real Adjusted Household Income Between Selected Years, by Income Percentile: Census Business Cycle Peaks 1969–1979 1979–1989 1969–1989 California 20th Median 80th Change in 80/20 ratio (%) 5 –4 14 0 85 +12 +10 1 14 24 +24 United States 20th Median 80th Change in 80/20 ratio (%) 13 4 18 8 20 14 +6 +9 18 28 36 +15 SOURCE: Based on authors’ calculations from the decennial Census. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. ____________ 13For further discussion of the representativeness of the CPS for California, see Appendix B. 14For a comparison of the income levels at each decile for the Census and CPS, see Appendix D. 22 The Census data also show an absolute decline in adjusted household income at the 20th percentile in California during the 1980s, but the decline is slightly smaller than measured in the CPS, as shown in Table 2.1. Like those calculated from the CPS, the 80/20 ratios show that the widening of the distribution was more pronounced in California than it was in the United States. Table 2.3 shows the levels and trends in the CV using the Census and the CPS. Both datasets show similar levels of adjusted household income inequality in California and the nation and a greater upward trend in inequality in California. The most significant difference between the datasets is that the Census suggests that inequality increased between 1969 and 1979 in California, whereas the CPS data indicate that the increase began after 1979. Table 2.3 Levels and Trends in the Coefficient of Variation for Adjusted Household Income: CPS and Census CV: Level 1969 1979 1989 California CPS Census 0.65 0.66 0.65 0.70 0.74 0.75 United States CPS Census 0.65 0.67 0.64 0.66 0.72 0.73 CV: Percent change 1969–1979 0 5 –1 –1 1979–1989 13 8 12 11 1969–1989 13 13 11 9 SOURCE: Based on authors’ calculations from the March CPS and the decennial census. NOTE: Statistics reported in this table are not sensitive to the consumer price index. 23 Household Income Inequality Rose Faster in California Than in Other Regions and States The large sample size of the Census allows for measurement of income inequality at the state level, even for small states. In 1969, 20 states had higher adjusted household income inequality than California did, as measured by the CV. By 1989, California was the sixth highest state. Over the period 1969 to 1989, only Michigan experienced higher percentage growth than California in adjusted household income inequality (see Appendix D for full state rankings). A limitation of the Census data is that they cannot be used to measure the dramatic increase in California income inequality that occurred during the deep recession of the early 1990s. Fortunately, data for the CPS do cover this period. Because of sample-size limitations, however, the CPS data can only be used to compare California to regions, not to other states. Relative to the other regions of the country, California has experienced higher growth in adjusted household income inequality since 1969. Table 2.4 reports trends in the level and growth of the CV for ten regions of the country: California plus nine geographically defined regions (California is also included as part of the Pacific region).15 The ____________ 15The nine Census regions are New England (Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Connecticut); Middle Atlantic (New York, New Jersey, Pennsylvania); East North Central (Ohio, Indiana, Illinois, Michigan, Wisconsin); West North Central (Minnesota, Iowa, Missouri, North Dakota, South Dakota, Nebraska, Kansas); South Atlantic (Delaware, Maryland, District of Columbia, Virginia, West Virginia, North Carolina, South Carolina, Georgia, Florida); East South Central (Kentucky, Tennessee, Alabama, Mississippi); West South Central (Arkansas, Louisiana, Oklahoma, Texas); Mountain (Montana, Idaho, Wyoming, Colorado, New Mexico, Arizona, Utah, Nevada); and Pacific (Washington, Oregon, California, Alaska, Hawaii). 24 Table 2.4 Regional Trends in the Coefficient of Variation for Real Adjusted Household Income, 1969–1994 Region CV (Rank) Percentage Change in CV (Rank) 1969 1979 1989 1994 1969–1979 1979–1989 1989–1994 California 0.65 0.65 0.74 0.79 0 13 7 (4) (4) (3) (1) (4) (2) (1) New England 0.55 0.60 0.64 0.68 10 6 6 (10) (10) (10) (10) (1) (10) (2) Mid Atlantic 0.64 0.63 0.71 0.74 –2 13 3 (6) (7) (6) (5) (8) (1) (4) E. N. Central 0.59 0.60 0.67 0.69 1 12 3 (9) (9) (9) (8) (3) (5) (5) W. N. Central 0.62 0.61 0.67 0.68 –2 11 0 (8) (8) (8) (9) (9) (7) (9) S. Atlantic 0.67 0.66 0.72 0.73 –1 9 0 (3) (3) (4) (6) (7) (9) (8) E. S. Central 0.72 0.67 0.75 0.74 –6 11 –1 (1) (2) (2) (4) (10) (8) (10) W. S. Central 0.69 0.69 0.78 0.78 0 13 0 (2) (1) (1) (2) (5) (3) (7) Mountain 0.62 0.63 0.71 0.72 2 12 2 (7) (6) (7) (7) (2) (6) (6) Pacific 0.65 0.64 0.72 0.76 –1 12 6 (5) (5) (5) (3) (6) (4) (3) SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are not sensitive to the consumer price index. size of California makes it reasonable to consider the state as its own region: California has more residents than do half of the nine regions. In 1969, California had the fourth-highest level of income inequality of the ten regions. Between 1979 and 1989, California had the secondhighest level of growth in income inequality; between 1989 and 1994, California had the highest. In 1994, California had the highest level of inequality of adjusted household income of the ten regions. 25 Adjusted Family Income Shows Rising Inequality Thus far, we have focused on the distribution trends in adjusted household income among persons. Focusing on household income implicitly assumes income-sharing among household residents regardless of relationship. Many researchers examine the distribution of family income rather than household income. For completeness, we also examined the trends in the distribution of adjusted family income among persons, implicitly assuming that there is no income-sharing among residents of the same household who are not related by blood, marriage, or adoption.16 As shown below, the trend in the distribution of family income exhibits the same widening as the distribution of household income, but the rise in inequality is even more pronounced for adjusted family income. The Census Bureau defines a “family” as the head of household and at least one resident relative: single people living alone and subfamilies unrelated to their household head are not included in the Census Bureau’s sample of families. We use a more comprehensive definition: Single persons living alone are included as their own family; people who do not live alone but are not related to the head of their household are included as separate families.17 This comprehensive definition is preferred to the Census Bureau definition because it includes the entire sample population.18 ____________ 16Karoly and Burtless (1995) suggest an alternative to this assumption: an adjustment for family size that also allows for some sharing among residents of the same household who are not related. 17Separate family-level observations are constructed for each single person and for each secondary family within a household. 18See Appendix C for distribution trends using the Census Bureau definition of primary family. 26 Table 2.5 shows the change in income levels at the 20th, median, and 80th percentiles of the distribution of adjusted family income among persons. Adjusted family income at the 20th percentile fell 11 percent in California between 1969 and 1989. During the same period, income grew by 12 percent at the median and by 24 percent at the 80th percentile. This widening of the distribution led to a 39 percent increase in the 80/20 ratio over the period. In the nation, family income growth was higher at each percentile and the 80/20 ratio increased by 27 percent. Relative to the results for adjusted household income, the growth in the 80/20 ratio for adjusted family income is higher in every period. Figure 2.4 illustrates the rise in inequality for adjusted family income, using the four summary measures discussed above. Adjusted Table 2.5 Percentage Change in Real Adjusted Family Income Between Selected Years, by Income Percentile California 20th Median 80th Change in 80/20 ratio (%) Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 –2 –9 –11 –27 10 1 12 –7 18 5 24 14 +20 +16 +39 +56 United States 20th Median 80th Change in 80/20 ratio (%) 8 17 21 +11 –1 7 –6 7 25 7 13 36 19 +14 +27 +26 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. 27 .83 CV .81 .79 CA .77 U.S. .33 ENTROPY .31 CA .29 U.S. .75 .27 .73 .25 .71 .69 .23 .67 .21 .65 .63 .19 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 .55 MLD .50 CA U.S. .45 .40 .35 2.75 2.55 2.35 2.15 1.95 1.75 1.55 VLN CA U.S. 1.35 .30 1.15 .25 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Statistics reported in this figure are not sensitive to the consumer price index. Adjusted family income in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 2.4—Summary Measures of Inequality for Real Adjusted Family Income, 1967–1994 28 family income inequality exhibits many of the same trends as does adjusted household income inequality. Inequality has increased in both California and the nation since the recession of the early 1980s. Inequality has increased more rapidly in the state than in the nation since the late 1980s, but the faster growth of inequality in California began even earlier for family income than for household income. Levels of adjusted family income inequality were similar to those of adjusted household income inequality in the late 1960s. By the 1990s, measures of inequality were considerably higher for families than for households, especially using those measures that emphasize income at the bottom of the distribution. 29 3. Trends in the Distribution of Labor Income This chapter examines trends in the distribution of labor income, the largest component of household income. There are a number of reasons for looking at labor income inequality as well as household income inequality. Trends in adjusted household income inequality are complicated by societal changes in family size and marriage behavior.1 In contrast, labor income inequality measures the disparity of income of individuals rather than families, and it is not directly affected by changes in household structure. While adjusted household income may be a better indicator of general economic well-being, labor income provides a clearer picture of changes in the economy. ____________ 1The increase in female-headed households has affected the distribution of adjusted household income, as have the falling marriage rates for men. In the past, the wives of low-income men were more likely to have earnings than the wives of high-income men. In recent years, there has been an increase in the number of families with two professional-level salary earners. The increasing correlation of husbands’ and wives’ earnings also has affected the distribution of adjusted household income. 30 As measured by annual earnings and hourly wages, the trends in labor income reveal many of the same patterns found for household income. Since the early 1980s, both California and the nation have experienced growth in labor income inequality. This result holds true for multiple definitions of labor income and measures of inequality. As was true for household income, the rising inequality of male annual earnings began in the early 1970s. In contrast, the inequality of female annual earnings declined substantially during that decade and did not begin to rise until the early 1980s. Since 1975, male hourly wages at the top of the distribution have shown slow growth. For low-wage male workers, hourly wages have declined considerably. For female workers, in contrast, hourly wages have grown near the top of the distribution. Near the bottom of the distribution, female wages declined by a small amount in California and grew by a small amount in the nation. What Is Labor Income? Labor income is income from work. Labor income comprises income from wages, salary, self-employment, and one’s own farm. For people who receive income from their own farm or from selfemployment, however, reports of “earnings” often include income from previous capital investments such as ownership of the farm or business.2 This income from capital is not part of labor income. Therefore, the data sample used to study labor income excludes workers who report that ____________ 2For example, a restaurant owner might show higher net income if she owns, rather than rents, her stoves. 31 their primary occupation was “self-employed”3 and workers who receive a substantial income from self-employment or from their own farm.4 After making these sample exclusions, we compute annual earnings as the sum of earnings from wages and salaries plus income from self-employment and farms. To ensure that the findings do not depend on sample exclusions, we compare these results to distribution trends for total earnings among all adult workers regardless of self-employment or farm owner status. We also measure the trends in the distribution of income from wages and salary for all adults with income from these sources. Finally, we examine trends for a subsample of workers between ages 18 and 55 to remove any effects of early retirement. As is customary, all samples are limited to civilians age 18 and older who are not students and who report some earnings.5 The data on annual earnings include only pre-tax monetary compensation. A brief discussion of the effect of non-monetary ____________ 3People who are self-employed in incorporated businesses are not identified in the CPS before 1975. To maintain the same sample definition throughout all years, these people are not excluded from the sample in any year. 4The sample excludes people who report more income from their farm or business than from wages and salaries and excludes any person reporting an absolute value of more than $2,000 in income in 1994 dollars from their own farm or self-employment. Some wage and salary workers included in the sample receive a small amount of income from farms and self-employment. This income was included in annual earnings to improve estimates of hourly wages, because estimates of annual hours of work include hours worked in the farm or business. The measure of annual earnings used in this study is similar to that of Karoly (1993). Although both studies exclude people who classify themselves as “self-employed,” our study additionally excludes people who receive substantial income from self-employment or farms. Also, Karoly does not include even small amounts of income from farms or self-employment in annual earnings. The results reported here for the nation are similar to those of Karoly. 5The sample also excludes people who report that their primary position was “without pay.” 32 compensation and taxes on the distribution of income can be found in Appendix C. In this chapter, we evaluate trends in both annual earnings and hourly wages. Neither measure by itself allows for a complete understanding of changes in labor income inequality: The distribution of hourly wages gives little indication of total annual earnings; the distribution of annual earnings is confounded by differences in hours of work. Using both of these measures, along with household income as discussed in the previous chapter, provides a more complete picture of income inequality in California. We examine the trends in income inequality for male and female workers separately because of the recent significant changes in the labor force participation of women. Over the years of the study, women’s labor force participation rate has increased from 51 percent to 61 percent in California; between 1975 and 1994, the average hours worked per year among adult women in the labor force increased from 780 to 980.6 Trends in the Distribution of Labor Income Among Males Inequality in male annual earnings and hourly wages is rising. There has been a slow growth in annual earnings and hourly wages near the top of the distribution and a substantial decline near the bottom—and even at the median after 1986 in California. In addition, California had one ____________ 6Labor market participation rates are based on the authors’ calculations from the CPS. The sample includes civilian women age 18 and older. The CPS began including information on hours of work only in 1975. Annual hours are calculated as the product of annual weeks of work and usual hours worked per week of work. 33 of the highest increases of any state in male annual earnings inequality between 1969 and 1989. The Widening Distribution of Male Annual Earnings Figure 3.1 illustrates the trends in annual earnings between 1967 and 1994 for the 10th, 20th, median, 80th, and 90th percentiles of male workers for California and the nation. The figure shows much the same pattern as observed for adjusted household income: The distribution of male annual earnings widened over the past three decades, with the most noticeable increases occurring during periods of recession in the early and mid 1970s, the early 1980s, and the early 1990s. During each recession, male annual earnings fell drastically for the lower and lower-middle positions of the distributions in California and the nation. The decline in male annual earnings was greater in California than in the nation because of slower growth in recovery periods and more rapid decline in the recessions of the early 1970s and early 1990s. For example, in California, men at the 20th percentile in 1971 had 16 percent lower annual earnings than men at the 20th percentile in 1969. For the nation, the decline was 8 percent. The recession of the early 1990s hit California even harder. Between 1989 and 1993, the 20th percentile fell 14 percent in the nation but 27 percent in California. While the trends in the distribution of male earnings are similar to those of adjusted household income in their overall shape, male earnings exhibit much slower growth. Table 3.1 summarizes the trends in male earnings for comparable years in the business cycle. The 80th percentile 34 Percent change since 1967 Percent change since 1967 30 California 20 10 0 –10 –20 10th –30 20th Median 80th –40 90th –50 1967 1971 1975 30 United States 20 1979 1983 1987 1991 1994 10 0 –10 –20 10th 20th –30 Median 80th –40 90th –50 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Sample includes civilians age 18 and older who received wage and salary income. Sample excludes students, those self-employed who are not in incorporated businesses, workers whose primary position is unpaid, workers who receive more farm or self-employment income than wage and salary income, and workers who receive more than $2,000 (real 1994 dollars) from farm or self-employment income. Annual earnings are computed as the sum of earnings from wages, salaries, self-employment and farms. Statistics reported in this figure are sensitive to the consumer price index. Real annual earnings in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.1—Percentage Change in Real Annual Earnings for Males, by Income Percentile, 1967–1994 35 Table 3.1 Percentage Change in Real Annual Earnings for Males Between Selected Years, by Income Percentile: CPS Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –15 –2 11 +30 –21 –12 –5 +21 –33 –27 –13 –20 62 +57 +41 United States 20th Median 80th Change in 80/20 ratio (%) –4 5 11 +16 –14 –7 4 +22 –18 –14 –2 –13 16 4 +42 +21 SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. of male earnings climbed only 6 percent in California between 1969 and 1989, in contrast with a 26 percent growth in adjusted household income. After 1979, the nation displayed a decline in the median of male earnings, particularly during recessions. The median in California fell throughout the period of the study, dropping 13 percent between 1969 and 1989. Surprisingly, the most recent decline in median male earnings began as early as 1987 in California, three years before the most recent recession. Despite the slow growth at the top of the distribution of male annual earnings in California, the 80/20 ratio increased a staggering 57 percent between 1969 and 1989 because of the drastic decline in earnings at the 36 20th percentile. The national 80/20 ratio also shows a remarkable (though smaller) increase of 42 percent. Summary Measures Show Rising Inequality in Male Annual Earnings As Figure 3.2 shows, the summary measures of inequality (introduced in Chapter 2) demonstrate the increasing trend in male annual earnings inequality over the last three decades. Male annual earnings show a clear pattern: Inequality rose sharply during recessions and remained at new, higher levels during recovery periods. In many cases, in fact, inequality continued to increase even during periods of growth.7 Beginning in the 1970s, male earnings inequality was consistently higher in California than in the nation, except for a brief period in the mid 1980s. The measures show that the gap between California and the United States began to widen noticeably as early as 1987. Although inequality did increase in the nation during the most recent recession, levels of inequality in California continue to be considerably higher. Census Data Show Rising Inequality of Male Annual Earnings The Census results, shown in Table 3.2, confirm the large decline in earnings near the bottom of the distribution and the slow growth in ____________ 7All four measures (and especially the VLN) appear to show a decrease in inequality in the early 1980s because of the large spike in inequality between 1979 and 1982. The cause of this spike is clear in Figure 3.1: Earnings fell sharply for the bottom of the distribution between 1979 and 1982 and then showed some compensating recovery in the next few years. If the spike is ignored, the continuing upward pattern of increasing inequality is clear. 37 .80 CV .75 CA U.S. .70 .65 .31 ENTROPY .29 CA .27 U.S. .25 .23 .21 .19 .60 .17 .55 .15 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 .40 MLD .38 .36 CA U.S. .34 .32 1.4 VLN 1.3 CA 1.2 U.S. 1.1 .30 1.0 .28 .9 .26 .24 .8 .22 .7 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are not sensitive to the consumer price index. Real annual earnings in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.2—Summary Measures of Inequality for Male Annual Earnings, 1967–1994 38 Table 3.2 Percentage Change in Real Annual Earnings for Males, by Income Percentile: Census Annual Earnings 1969–1979 1979–1989 1969–1989 California 20th Median 80th Change in 80/20 ratio (%) United States 20th Median 80th Change in 80/20 ratio (%) –18 –3 10 +35 –6 6 11 +18 –18 –33 –12 –14 –7 3 +13 +53 –9 –14 –9 –4 –1 9 +8 +28 SOURCE: Based on authors’ calculations from the decennial Census. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. earnings near the top found by the CPS. The main difference between the Census and CPS results is that over the period 1969 to 1989, the national 80/20 ratio increased by 28 percent according to the Census and by 42 percent according to the CPS. For California, the results are closer: a 53 percent increase according to the Census and a 57 percent increase according to the CPS. The Census data show slower growth but higher levels of inequality. For example, the coefficient of variation for male annual earnings was 0.63 in California in 1969 as calculated from the Census (see Table 3.3). 39 Table 3.3 Levels and Trends in the Coefficient of Variation for Male Annual Earnings: CPS and Census California CPS Census United States CPS Census CV: Level 1969 1979 1989 0.56 0.63 0.56 0.63 0.65 0.72 0.62 0.68 0.75 0.76 0.70 0.71 CV: Percent change 1969–1979 1979–1989 1969–1989 15 14 32 14 11 9 5 13 4 20 26 14 SOURCE: Based on authors’ calculations from the March CPS and the decennial Census. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are not sensitive to the consumer price index. For that same year, the CV calculated from the CPS was 0.56. Despite these differences, both datasets show a substantial rise in inequality in California that exceeded the rise in the United States. The Widening Distribution of Male Hourly Wages We examine trends in hourly wages, in addition to annual earnings, to observe changes in salary separate from changes in hours worked. Hourly wages are calculated by dividing annual earnings by annual hours; annual hours are the product of weeks worked and usual hours worked per week of work.8 Figure 3.3 shows the widening distribution ____________ 8Hourly wages are measured imprecisely because they are calculated from annual data. This imprecision leads to extreme values in some years (e.g., some years have several observations with an hourly wage of less than $1). To avoid fluctuation in the summary measures of inequality due to extreme values, hourly wages were top-coded at 97 percent and bottom-coded at 3 percent in all years. 40 Percent change since 1975 Percent change since 1975 10 California 0 –10 10th 20th –20 Median 80th 90th –30 –40 1975 1978 1981 1984 1987 1990 1994 10 United States 0 –10 10th 20th –20 Median 80th 90th –30 –40 1975 1978 1981 1984 1987 1990 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are sensitive to the consumer price index. Real hourly wages in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.3—Percentage Change in Real Hourly Wages for Males, by Income Percentile, 1975–1994 41 of male hourly wages. (The figure begins in 1975 because information on hours worked per week is not available in earlier years of the CPS.9) The most striking features of Figure 3.3 are the slow growth at the top and the decline at the bottom of the distribution of male hourly wages in both California and the United States. Over most of the period, inequality increased even though wages at the 90th percentile were never more than 10 percent higher than they had been in 1975. Table 3.4 summarizes these trends. Between 1979 and 1989 in California, male hourly wages fell by 21 percent at the 20th percentile and fell by 14 percent at the median. Even at the 80th percentile, wages fell 2 percent. The nation exhibited a similar pattern but with smaller declines. Because of the faster decline in wages at the 20th percentile in California, the 80/20 ratio increased by 24 percent in the state compared to 17 percent in the nation. Summary Measures Show Rising Inequality in Male Wages The summary measures of inequality shown in Figure 3.4 confirm that male wage inequality has risen steadily and significantly in California since 1977. As with male earnings inequality, male hourly wage inequality was higher in California than in the nation for most of the years of the study. Since the late 1980s, however, male wage inequality ____________ 9Before the 1976 survey, the CPS did not ask about hours of work in a usual week in the previous year. This information is needed to calculate hourly wages from annual earnings in the previous year. Hourly wages were not computed from the Census data because the 1970 Census survey did not ask about hours of work per week in the previous year. 42 Table 3.4 Percentage Change in Real Hourly Wages for Males Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1979–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –21 –14 –2 +24 –30 –22 –2 +40 United States 20th Median 80th Change in 80/20 ratio (%) –12 –6 3 +17 –19 –13 1 +25 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. grew more rapidly in California, widening the difference in inequality between California and the nation. Male Labor Income Inequality Rose Faster in California Than in Other Regions and States Relative to the other regions of the country, California experienced high growth in male hourly wage inequality. Table 3.5 shows the CV for 43 .67 CV .65 .63 CA U.S. .61 .59 .57 .55 .53 .51 .49 1975 1979 1983 1987 .21 ENTROPY .20 .19 CA .18 U.S. .17 .16 .15 .14 .13 .12 .11 1991 1994 1975 1979 1983 1987 1991 1994 .22 MLD .21 .20 CA U.S. .19 .18 .46 VLN .44 .42 CA U.S. .40 .38 .17 .36 .16 .34 .15 .32 .14 .30 .13 .28 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are not sensitive to the consumer price index. Real hourly wages in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.4—Summary Measures of Inequality for Male Hourly Wages, 1975–1994 44 Table 3.5 Regional Trends in the Coefficient of Variation for Real Hourly Wages Among Males, 1979–1994 Region CV (Rank) Percentage Change in CV (Rank) 1979 1989 1994 1979–1989 1989–1994 California 0.52 0.61 0.66 (3) (2) (1) New England 0.52 0.54 0.56 (6) (9) (10) Mid Atlantic 0.48 0.57 0.60 (9) (6) (5) E. N. Central 0.45 0.53 0.58 (10) (10) (8) W. N. Central 0.50 0.57 0.57 (8) (7) (9) S. Atlantic 0.54 0.58 0.63 (1) (4) (4) E. S. Central 0.52 0.57 0.60 (4) (5) (6) W. S. Central 0.54 0.63 0.64 (2) (1) (2) Mountain 0.52 0.56 0.59 (5) (8) (7) Pacific 0.51 0.59 0.64 (7) (3) (3) 18 (3) 5 (10) 18 (1) 18 (2) 14 (6) 8 (9) 10 (7) 16 (4) 8 (8) 16 (5) 7 (4) 3 (8) 6 (5) 10 (1) 1 (10) 8 (3) 5 (6) 2 (9) 5 (7) 8 (2) SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are not sensitive to the consumer price index. California and the nine geographic regions, including California’s own Pacific region. In 1979, the CV of male hourly wages in California was the third-highest among the ten regions; in 1989, it was second; by 45 1994, inequality was higher in California than in any other region.10 Appendix D presents the same analysis for male annual earnings. Comparing male annual earnings trends between the states further emphasizes the growth in inequality in California. The Census data show that 20 states had greater inequality than California in 1969, as measured by the CV. Between 1969 and 1989, California was tied with Indiana and Ohio for the fastest percentage growth in male earnings inequality in the country. In 1989, only two states had higher levels of male earnings inequality. Appendix D reports the CV for all 50 states in 1969 and 1989. Other Definitions of Male Labor Income Show Rising Inequality The labor income results reported in the previous sections are based on a data sample that excludes workers who are primarily self-employed or farm owners. This sample definition is preferable because of the difficulty in separating capital income from labor income for people who work in their own business or farm. However, including these workers and looking at the sum of income from wages, salary, self-employment, and farms does not alter the basic trends of decline near the bottom of the distribution, slow growth near the top, and rising inequality. For California between 1969 and 1989, the decline in male annual earnings at the 20th percentile was 30 percent among all workers, compared to 33 percent in the restricted sample of wage and salary workers (shown in Table 3.1). The growth in male earnings at the 80th percentile was about ____________ 10This result is consistent with the regional inequality trends found by Karoly and Klerman (1994). 46 5 percent in both samples. Over the same period, the growth in the CV was 25 percent for all workers and 32 percent for wage and salary workers. The same general trends also hold for the distributions of hourly wages among all workers, of annual earnings and hourly wages among wage and salary workers ages 18 to 55, and of wage and salary income. See Appendix C for further details. Trends in the Distribution of Labor Income Among Females The trends in inequality of female annual earnings are quite different from those of male annual earnings. The distribution of female annual earnings narrowed during the 1970s, when women’s incomes rose substantially near the bottom of the distribution. In the 1980s, the declining inequality of female annual earnings either slowed or reversed itself, depending on which measure of inequality is used. In contrast, all the measures show that the inequality of hourly wages among women increased during the 1980s. The difference between the trends in annual earnings and hourly wages suggests that some of the increase in female earnings, especially in the lower ranks of the distribution, was due to increased hours of work. The levels and trends of female labor income inequality, however, were nearly identical in California and the nation, even over the last decade. The Narrowing, Then Widening, Distribution of Female Annual Earnings In contrast to male annual earnings and adjusted household income, female annual earnings inequality actually declined between 1967 and 47 the early 1980s. As Figure 3.5 shows, growth was fastest among women at the bottom of the distribution during the 1970s in both California and the nation: The upper lines on the figure represent earnings growth at the 10th and 20th percentiles. The relative gains of the lowest-earning women were short-lived, however. In the 1980s, annual earnings began to grow for women in the upper half of the distribution, and, interestingly, did not show the same tendency as male annual earnings to fall during recessions. Earnings at the lower percentiles did continue to grow but fell during recessions, especially in California. Table 3.6 allows us to see these trends clearly. The income at the 20th percentile increased over the 1970s, growing 61 percent in Table 3.6 Percentage Change in Real Annual Earnings for Females Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) 61 22 10 –31 13 82 15 6 29 27 22 35 30 +8 –26 +13 United States 20th Median 80th Change in 80/20 ratio (%) 44 20 16 –20 20 72 11 33 20 39 +0 –20 43 28 33 –7 SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. 48 Percent change since 1967 Percent change since 1967 160 140 California 120 10th 20th 100 Median 80th 80 90th 60 40 20 0 1967 1971 1975 160 140 United States 120 10th 100 20th Median 80 80th 90th 60 40 20 0 1967 1971 1975 1979 1979 1983 1983 1987 1987 1991 1994 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are sensitive to the consumer price index. Real annual earnings in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.5—Percentage Change in Real Annual Earnings for Females, by Income Percentile, 1967–1994 49 California and 44 percent in the nation. Female earnings at the 80th percentile grew at a slower pace, so that the 80/20 ratio fell by 31 percent in California and by 20 percent in the nation over that decade. In California in the 1980s, the 80th percentile grew faster than the 20th, leading to an 8 percent increase in the 80/20 ratio. In the nation, the 80/20 ratio was the same in 1989 as in 1979. Measures of Inequality Show Falling, Then Rising, Inequality in Female Annual Earnings The summary measures in Figure 3.6 all show that female earnings inequality fell during the 1970s and increased beginning in the early 1980s. After 1986, the trends in inequality are dependent on which measure is used. The CV shows a continuing increase throughout the 1980s, whereas the other measures exhibit fluctuations without clear trends.11 Unlike trends in inequality for male earnings and household income, trends in female earnings inequality were virtually identical in California and the nation, even in the late 1980s. Inequality did rise more sharply in California in the early 1990s, but the difference between the state and the nation had narrowed substantially by 1994. Census Data Show a Fall and Then a Rise in Inequality of Female Annual Earnings Table 3.7 shows the trends in the distribution of female annual earnings calculated from the Census data. The picture of growth is ____________ 11The fact that the measures do not agree on whether inequality was higher in the 1990s than in the early 1970s reflects the tremendous growth in the income of the lowest-earning women. The two measures that put more weight on the bottom of the distribution—the MLD and the VLN—do not show as steep an increase in inequality as the other measures do. 50 .84 CV .33 ENTROPY .82 .32 CA CA U.S. .31 U.S. .80 .30 .78 .29 .76 .28 .74 .27 .72 .26 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 .52 1.7 MLD VLN .50 1.6 CA CA .48 U.S. U.S. 1.5 .46 1.4 .44 1.3 .42 1.2 .40 .38 1.1 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are not sensitive to the consumer price index. Real annual earnings in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.6—Summary Measures of Inequality for Female Annual Earnings, 1967–1994 51 Table 3.7 Percentage Change in Real Annual Earnings and Hourly Wages for Females, by Income Percentile: Census California 20th Median 80th Change in 80/20 ratio (%) Annual Earnings 1969–1979 1979–1989 1969–1989 52 13 12 14 9 18 72 27 29 –28 +4 –25 United States 20th Median 80th Change in 80/20 ratio (%) 30 15 11 –15 22 58 12 28 19 33 –2 –16 SOURCE: Based on authors’ calculations from the decennial Census. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. similar to that found using the CPS, shown in Table 3.6. In California, the 80/20 ratio decreased in the 1970s and then increased by a small amount in the 1980s. The overall decline in the 80/20 ratio between 1969 and 1989 was nearly identical in the CPS and the Census (26 percent versus 25 percent for California). The Census data do not show as much growth in female earnings as the CPS data do. Table 3.8 depicts the very similar levels of female annual earnings inequality between the CPS and the Census data, as measured by the CV. In addition, the changes in the CV are also similar—the main 52 Table 3.8 Levels and Trends in the Coefficient of Variation for Female Annual Earnings: CPS and Census California CPS Census United States CPS Census CV: Level 1969 1979 1989 0.78 0.77 0.77 0.77 0.76 0.78 0.74 0.77 0.80 0.84 0.79 0.78 CV: Percent change 1969–1979 1979–1989 1969–1989 –3 4 2 1 –3 –1 8 62 9 31 SOURCE: Based on authors’ calculations from the March CPS and the decennial Census. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are not sensitive to the consumer price index. difference is that the Census shows higher inequality growth in California than the CPS does. The Widening Distribution of Female Hourly Wages The trends in the distribution of female annual earnings reflect the increase in hours worked by women in the labor market. Among women who work, average hours increased 26 percent between 1975 and 1994 in California. Examining the trends in the distribution of hourly wages removes the effect of hours of work. These trends are portrayed in Figure 3.7. Like the distribution of female annual earnings, the distribution of female hourly wages narrowed between 1975 and 1979 as wages for the lowest-paid women rose quickly. During the recession of the early 53 Percent change since 1975 Percent change since 1975 30 California 20 10 0 10th 20th Median 80th 90th –10 1975 1978 1981 1984 30 United States 10th 20 20th Median 80th 90th 10 1987 1990 1994 0 –10 1975 1978 1981 1984 1987 1990 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are sensitive to the consumer price index. Real hourly wage in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.7—Percentage Change in Real Hourly Wages for Females, by Income Percentile, 1975–1994 54 1980s, however, wages fell throughout the distribution. When female wages began to rise again, they grew in a familiar pattern: Wages grew fastest at the upper percentiles. In California, female hourly wages fell at the 10th percentile between 1985 and 1994. In contrast to male wages, female wages near the top of the distribution in California grew over the 1980s, and even in the early 1990s. Moreover, female wages did not show the same strong influence of recessions. In addition, female wages grew only slightly faster in the nation than in California. As Table 3.9 shows, wages at the median increased by 2 percent in California and by 8 percent in the nation between 1979 and 1989. At the 20th percentile, female wages fell 9 Table 3.9 Percentage Change in Real Hourly Wages for Females Between Selected Years, by Income Percentile Business Cycle Peaks 1979–1989 California 20th Median 80th Change in 80/20 ratio (%) –9 2 15 +26 Recessions 1976–1994 –8 8 18 +28 United States 20th Median 80th Change in 80/20 ratio (%) –5 8 16 +22 –2 10 22 +24 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. 55 percent in California compared with 5 percent in the nation over the same period. At the upper end of the distribution, the increase was nearly identical in California and the United States. Measures of Inequality Show Rising Inequality in Female Hourly Wages Female hourly wages show a clear upward trend in inequality beginning in the early 1980s, as depicted in Figure 3.8. This is in contrast to the results for female annual earnings, which depend on the measure of inequality used. Like female annual earnings, hourly wage inequality in California tracks closely with that of the nation. Female Labor Income Inequality Is Similar in California to Other Regions and States Table 3.10 compares female wage inequality in California and in the regions of the country. It confirms the finding, shown in Figure 3.8, that, in contradistinction to household and male labor income, female wages did not show higher levels of inequality in California than in the nation. In the business cycle peak of 1979 and in the business cycle trough of 1994, California’s level of female earnings inequality was firmly in the middle of the regions. Even its apparent high ranking in 1989 is somewhat misleading: Five of the regions had levels of inequality nearly identical to California’s in that year. Appendix D presents the same analysis for female annual earnings. Compared with other states, California had a moderate level of female earnings inequality in 1989 (as measured by the CV)—16 states had higher levels. However, between 1969 and 1989, 39 states experienced larger declines in inequality than California did. 56 .66 CV .64 .62 CA U.S. .60 .58 .56 .54 .52 .19 ENTROPY .18 CA .17 U.S. .16 .15 .14 .13 .50 .12 .48 .11 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 .21 MLD .20 .19 CA .18 U.S. .44 VLN .42 .40 CA .38 U.S. .17 .36 .16 .34 .15 .32 .14 .30 .13 .28 .12 .26 .11 .24 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and calculation of annual earnings. Statistics reported in this figure are not sensitive to the consumer price index. Real hourly wages in 1988 in California may not be comparable to other years due to changes in the CPS. Figure 3.8—Summary Measures of Inequality for Female Hourly Wages, 1975–1994 57 Table 3.10 Regional Trends in the Coefficient of Variation for Real Hourly Wages Among Females, 1979–1994 CV (Rank) Percentage Change in CV (Rank) Region 1979 1989 1994 1979–1989 1989–1994 California New England Mid Atlantic E. N. Central W. N. Central S. Atlantic E. S. Central W. S. Central Mountain Pacific 0.50 (5) 0.48 (9) 0.49 (6) 0.49 (7) 0.48 (10) 0.50 (4) 0.49 (8) 0.51 (2) 0.55 (1) 0.51 (3) 0.57 (2) 0.53 (10) 0.57 (1) 0.55 (8) 0.57 (5) 0.56 (6) 0.55 (7) 0.57 (3) 0.54 (9) 0.57 (4) 0.59 (5) 0.57 (10) 0.60 (2) 0.60 (4) 0.57 (9) 0.61 (1) 0.58 (8) 0.59 (6) 0.60 (3) 0.59 (7) 15 (3) 10 (9) 16 (2) 13 (5) 19 (1) 12 (8) 14 (4) 12 (7) –1 (10) 12 (6) 4 (8) 6 (4) 5 (5) 9 (3) 0 (10) 9 (2) 5 (6) 4 (9) 10 (1) 4 (7) SOURCE: Based on authors’ calculations from the March CPS. NOTES: Hourly wage is calculated as annual earnings divided by the product of annual weeks of work and usual hours worked per week of work. Hourly wage is not available before 1975 in the March CPS. See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are not sensitive to the consumer price index. Other Definitions of Female Labor Income Show Falling, Then Rising, Inequality As was true for male labor income, alternative sample definitions do not alter the basic trends in female labor income reported in the previous sections. Female annual earnings grew rapidly near the bottom of the 58 distribution and inequality declined until the early 1980s. There is one notable difference in results between the sample of all workers and the restricted sample of wage and salary workers: Annual earnings inequality among all female workers followed an increasing trend after 1983 for all the summary measures of inequality. This is in contrast to the trends in inequality after 1983 for female wage and salary workers, which depended on the summary measure used (as shown in Figure 3.6). See Appendix C for further results using the alternative definitions of female annual earnings and hourly wages. 59 4. Conclusions and Implications for Policy and Future Research Our study finds a large increase in income inequality in California over the last three decades for both household income and male earnings. This rise in income inequality is explained by a dramatic decline in income at the lower and lower-middle ranks of the distribution, and a simultaneous growth in income in the upper ranks. The trends in income inequality show a strong relationship to the business cycle: Inequality grew fastest during the recessions of the early 1970s, early 1980s, and early 1990s. Until the late 1980s, the levels and trends in income inequality in California and the nation were similar. Since that time, inequality has grown faster in California than in the United States. Moreover, compared to the nation, California has experienced slower income growth throughout the distribution. 60 Provocative as these findings are, measuring the trends in the distribution of income is only the first step in understanding them and their implications for policy. As California designs programs to promote equity, it will benefit from research on the relationship between existing state policies and income inequality, as well as from a better understanding of the causes of rising inequality in the state. In addition, the income trends measured in this study provide an incomplete picture of the distribution of economic well-being. Distribution trends for morecomprehensive definitions of income (e.g., accounting for taxes and nonmonetary compensation) and the issue of income variability remain to be studied. Public Policy and the Distribution of Income Continued growth in income inequality is not inevitable. As a society, we face a choice as to whether we will act to reverse the trend in growing income inequality. Many policy mechanisms already exist for reducing inequality. Progressive taxes, for example, directly redistribute income. Quality public schools and access to higher education provide an opportunity for people of all income levels to invest in themselves and improve their future incomes. Research on the role of the existing policy mechanisms, as well as identification of new policy options, is essential for understanding how California state policy influences the distribution of income. Some Americans believe that differences in income arise primarily from individual choices, preferences, abilities, investments, and productivity, and that income inequality is a product of an economy that values hard work and talent. Other Americans believe that income 61 differences reflect the unequal distribution of economic opportunity in our society, and that the opportunity to succeed is elusive for those who do not belong to privileged groups. The first viewpoint implies that public policy can affect inequality only by redistributing income; the second implies that policy can reduce inequality by promoting opportunity. Research on the determinants of income distribution and the extent to which policy provides or restricts economic opportunity will suggest avenues for improving opportunities for the less-advantaged. If California seeks to reduce income inequality, the state will benefit from research that identifies policy options that promote equity as well as efficiency in our economy. Labor Market Explanations for Rising Earnings Inequality The similar trends in California and the nation suggest that the same forces that explain the widening of the income distribution in the United States account for the growth in income inequality in California. At the national level, the rise in male earnings inequality has been explained by a combination of factors. Economists agree that changes in the supply and demand of labor have favored skilled workers over less-skilled workers. The underlying forces that have led to these labor market trends include technological change, international competition, immigration, and deunionization.1 However, the effect of each of these forces on the distribution of earnings in California’s distinct economy remains to be studied. ____________ 1Cassidy (1995) provides a straightforward summary of these explanations for rising earnings inequality. 62 Many economists believe that technological change has benefited educated workers who are able to implement new technology and has harmed less-educated workers who may be replaced by mechanized production.2 The effect of technological innovation on California workers may be more pronounced than in the nation. On one hand, the state has a higher percentage of people with at least some college education (23 percent) than the national average (19 percent). On the other hand, the school dropout rate is 14 percent in California, 3 percentage points above the national average of 11 percent.3 There is less agreement on the role of international competition in explaining the rise in income inequality. The cost of a low-skilled workforce is higher in the United States than in other countries, particularly developing countries. Thus, the United States increasingly imports manufactured products and textiles, lowering the labor market demand for low-skilled U.S. workers.4 International competition may have played a different role in the state because California has more Pacific-region trade than the rest of the nation and because a slightly smaller percentage of the state’s workforce is in manufacturing (15 percent compared to the national 16 percent).5 Growth in immigration may have contributed to the rise in income inequality. Immigrants can adversely affect the wage distribution by ____________ 2Krueger (1993) finds evidence that supports this theory. 3Population statistics based on the 1990 Census as reported in U.S. Bureau of the Census (1994), Table 236. 4Borjas, Freeman, and Katz (1992) have found evidence that trade patterns account for a substantial part of the wage losses of high school dropouts. 5Workforce statistics for 1993 reported in U.S. Bureau of the Census (1994), Table 655. 63 raising the number of low-wage workers. Furthermore, by increasing the competition for low-skill employment, immigration can lead to a reduction in the wages offered to natives with low skills.6 The impact of immigration on California is likely to be greater than in the nation, since California has the largest foreign immigration of any state. The decline in the power of unions has reduced the bargaining power of labor with the likely effect of lowering the wages of labor relative to that of management.7 The decline of unions is frequently offered as an explanation for the more rapid growth in earnings inequality in the United States than in other industrialized countries. The sharp rise in income inequality in California beginning in the late 1980s is probably explained, in part, by the same forces that caused the strong recession of the early 1990s. In addition to cuts in defense spending, suggested causes of the severe recession in the state include a decline in residential building, a fall in commercial aircraft orders, and a reduction in spending relative to income.8 The effect of each of these factors on the distribution of income remains to be studied. Demographic Explanations for Rising Family Income Inequality In addition to the economic and labor market forces that explain the increase in earnings inequality in the nation, trends in marriage and ____________ 6Butcher and Card (1991) and Borjas, Freeman, and Katz (1992) find evidence of an effect of immigration on wage inequality. 7Freeman (1993) reports evidence that the decline in unions lowered the wages of blue-collar workers relative to wages of white-collar workers. 8These factors are discussed in a study by the Center for the Continuing Study of the California Economy (1994). 64 female labor force participation may contribute to the rise in household and family income inequality. Declines in the percentage of people who are married may explain a portion of the rise in family income inequality. The growing share of families that rely on the earnings of single mothers has increased the number of low-income families.9 In addition, low-income men are less likely to be married than high-income men and are thus less likely to have a spouse who contributes to family income.10 Trends in marriage behavior may have a larger effect in California than in the nation. Compared to the national average, California had a lower rate of marriage and a higher rate of divorce between 1980 and 1992.11 The growth in the female labor force participation has an undetermined effect on the distribution of family income. As the percentage of women with earnings increased, earnings inequality among women fell. In addition, the rising earnings of married women have increased family income and reduced inequality among married-couple families.12 At the same time, however, the increased contribution of the earnings of wives has further polarized the incomes of single people relative to those of married couples. Furthermore, the correlation of the earnings of husbands and wives has increased: The wives of men with high earnings tend to earn more than the wives of men with low ____________ 9Danziger and Gottschalk (1995) find that the rise in female headship increased the poverty rate by 1.6 percentage points between 1973 and 1991 (Table 5.3, p. 102). 10Burtless (1996) makes this observation. 11Marriage statistics reported in U.S. Bureau of the Census (1994), Table 146. 12Cancian, Danziger, and Gottschalk (1993) find that changes in the earnings of married women reduced income inequality among married-couple families between 1968 and 1988. 65 earnings.13 The effect of female labor force participation may have been different in California because the increase in the average hours worked by women has been smaller than in the nation.14 Additional Measurement Issues There are a number of measurement issues we could not explore with Census Bureau income data that are important for a more complete understanding of the recent trends in income inequality and their implications for public policy in California. The income data used in this study do not account for the effect of taxes and non-monetary compensation (e.g., housing subsidies, health insurance). While national studies show that using more comprehensive measures of income does not substantially change income inequality trends,15 the effect may be different in California. For example, the percentage of people in California without health insurance was 19.3 in 1992, compared to a national average of 14.7 percent.16 The statistics reported in this study describe the distribution of income in each year. Because a person is likely to occupy different places in the distribution of income during his or her lifetime, the distribution ____________ 13Karoly and Burtless (1995) show the rising correlation of earnings between husbands and wives. 14Mean annual hours worked increased from 362 to 576 (59 percent) in California and from 348 to 597 (72 percent) in the United States between 1975 and 1994. These statistics include women who do not work in the labor market (zero hours). Statistics are based on the authors’ calculations from the March CPS. 15See Appendix C for a brief review of the literature on the distribution trends of more comprehensive measures of income. 16Health insurance statistics are reported in U.S. Bureau of the Census (1994), Table 165. 66 of annual income may not accurately reflect the level of inequality in lifetime income. Research at the national level suggests that economic mobility, the changing of positions within the distribution, has remained stable or declined in recent decades.17 However, income variability, the year-to-year fluctuations in income, appears to explain a substantial portion of the increase in male earnings inequality.18 Income mobility and variability remain to be studied in California. The Challenge for the State The combination of the sharp rise in household income inequality in California that began even before the most recent recession, the stagnation and decline of male wages, and the decline of household income for the lower and lower-middle ranks of the distribution pose a challenge to public policy in California. Can state policy help to meet the needs of low-income residents of the state and promote economic equity while not sacrificing economic growth? The answer to this question depends on the causes of recent trends and the policy options for the state. Future reports in this series will address these issues. ____________ 17Hungerford (1993) estimates income mobility in the United States in the 1970s and 1980s. 18Gottschalk and Moffitt (1994) find that one-third to one-half of the increase in the variance of earnings among white males from the 1970s to the 1980s can be explained by increased earnings instability. 67 Appendix A Notes on Data and Methodology This appendix addresses several limitations of the income data and the adjustments for price inflation and cost of living. When applicable, we describe our methodology for reducing the effect of these limitations on the estimated trends in income inequality. Income Data Income data for this study come from two national household surveys collected by the U.S. Bureau of the Census: the decennial Census of Population and Housing (1970, 1980, and 1990)1 and the ____________ 11970 Public Use Sample, 1 percent; 1980 and 1990 Public Use Micro Sample, 5 percent. 69 March Annual Demographic File of the Current Population Survey (public-use files, survey years 1968–1995).2 The Current Population Survey (CPS) and the Census report pre- tax, money income, which includes wages, salary, farm income, self- employment income, Social Security, railroad retirement, Supplemental Social Security, public assistance, welfare, interest, dividends, income from estates and trusts, net rental income, veterans’ payments, unemployment and workers’ compensation, private and government pensions, alimony, child support, regular contributions from persons not living in the same household, and other periodic income. Capital gains are not included. Current Population Survey The March file of the CPS, an annual survey of civilian households, provides detailed demographic information, including income received, for about 5,000 households in California and 50,000 households in the nation.3 The main benefit of using the CPS to study income ____________ 2Each survey has income information from the previous year. This study covers income years 1967–1994. Uniform series data files for CPS survey years 1964–1967, created under the direction of Robert Mare and Christopher Winship, are available from the University of Wisconsin. We chose not to use these files because of possible compatibility problems with the public-use files. We found much smaller average household sizes in the Mare-Winship files relative to the public-use files (e.g., the MareWinship file for 1967 had an average household size of 2.4 persons and the public-use file for 1968 had an average household size of 3.2 persons). The increase in household size leads to a sizable drop in adjusted household income between 1966 and 1967, whereas unadjusted household income shows a slight increase. We interpret the change in household size as evidence of a problem with the data and therefore we report statistics beginning with the 1968 public-use files. 3The March file of the CPS also includes Armed Forces personnel living with civilians. Our measures of household and family income include these households and families. Samples of workers do not include military personnel. About 3,000–4,000 male workers and 2,000–3,000 female workers are in the California sample. The national samples have about ten times as many workers as the California samples. 70 distribution is that it contains annual data, making it possible to observe short-term departures from long-term trends. As this study shows, the business cycle fluctuations observable in those data have strong effects on the distribution of income. The CPS allows us to use comparison years at the same stages of the business cycle when examining inequality changes over time. Over the period of the study, several changes were made in the design of the CPS, which could affect the comparability of the surveys across years. Survey changes that affect the distribution of income will result in one-time jumps in the measures of inequality but not in a pattern of changes across several years. We have confidence in the measured distribution trends discussed in the text because none of these results relies on a change that occurred in a single year. Each decade, the Census Bureau changes the sample design of the survey using population estimates from the most recent Census. The Census Bureau randomly selects a new sample of geographic areas called Primary Sampling Units (PSUs); in California, the PSUs are generally counties. To make the sample representative of all parts of the state, the PSUs are selected from groups of counties with similar population characteristics. Thus, even when the PSUs change, estimates of the income distribution should not be affected because each new PSU should be similar to the one it replaced. All significant sample design changes for this study occurred in 1972–1973 and 1985–1986 (the 1995–1996 redesign was implemented after the March 1995 survey).4 ____________ 4The Census Bureau does rotate “Enumeration Districts” within the Primary Sampling Units. However, substitutions in Enumeration Districts are chosen based on 71 The Census Bureau constructs sample weights such that the CPS sample will represent the national population. The sample weights are based on information from the decennial Census. In survey years 1973, 1982, and 1994, the Census Bureau revised the sample weights to reflect new population estimates from the 1970, 1980, and 1990 Census. Karoly (1993) compares income inequality in the original release of the 1980 CPS and in a reissue of the same survey using the new sample weights. She finds that the change in sampling weights had little effect on the increase in inequality between income years 1978 and 1979. The Census Bureau has changed the survey procedure with respect to Hispanics. In 1976, an additional sample of 2,000 Hispanic households was added to the March CPS to increase Hispanic representation. These households were chosen randomly from Hispanic households interviewed in the November CPS. The addition of these households could affect the measured distribution of income between income years 1974 and 1975. In 1984, the sample weighting procedure was changed to incorporate Hispanics explicitly. This change increased the estimated number of Hispanics and may have affected the distribution of income. In 1994, the Census Bureau automated the CPS survey questionnaire and introduced new sample weights. The Census Bureau (1996) reported that these changes may have increased measured income inequality. Finally, one specific problem with the CPS occurred in a single year of the survey. The median reported income received in 1988 shows an ____________________________________________________ similarity of population characteristics and geographic proximity and should not affect population and income statistics. 72 anomalous decrease in California. Karoly (1995) also reports a decline in income in California in 1988 based on the CPS data. Measures of income in California from other sources do not suggest a dip in 1988. For example, Department of Commerce data show the level of per capita income in 1988 about midway between 1987 and 1989 (California Statistical Abstract, 1995, Table D-7). The probable cause of this reported aberration lies not in California but in Washington. In 1989, funding for the CPS was cut and the sample for California fell to fewer than 3,000 households. The smaller sample size led to a higher sampling error for the 1988 data than in other years and may have affected the representativeness of the sample in that year.5 For this reason, we do not rely heavily on our results for 1988 in the California data. Funding was restored the following year and the California sample size returned to almost 5,000 in 1990. Census of Population and Housing We use the Public Use Sample of the Census to investigate the distribution of income and earnings in 1969, 1979, and 1989. One advantage of using the Census is that its larger sample size leads to more precise statistical estimates. In addition, the Census is designed to survey the entire population and therefore is representative of each state (with the important exception of undercount problems). The main limitation of the Census is that with only three years of data, we cannot distinguish long-run trends from short-run business cycle effects. However, because ____________ 5The sample weights were adjusted to reflect the smaller sample size, but the representativeness of the sample may still have been affected because the sample was not cut randomly, but only in Los Angeles. 73 the Census years are all business cycle peaks, changes in the distribution of income as reported in the Census are likely to represent trends rather than cyclical fluctuations. Although we expect to observe similar trends in the distribution of income using the CPS and the Census, income data from the two sources are not identical. For example, in 1990, the Census asked respondents about eight specific types of income. In the same year, the CPS asked about more than 20 types of income, making it less likely that a respondent will omit a source of income than when answering the Census questions. Also, the CPS is collected over the phone by trained survey-takers, who help improve the survey accuracy relative to the Census, which is done by mail. For this reason, we expect the CPS to reflect more accurately the sum of income from all sources as well as earnings and wages. The changes in the Census survey procedures were not as significant as changes in the CPS for measuring the trends in the distribution of income. It is worth noting that the Public Use Sample in 1970 (1 percent of the population) was much smaller than the Public Use Microdata Samples in 1980 and 1990 (5 percent of the population). Also, income in 1970 was reported as a range (e.g., $100–$199); we used the median of the ranges in our calculations. We did not calculate hourly wages with the Census data because the 1970 Census asks about hours of work in the previous week, as opposed to in a usual week in the previous year,6 and these hours are reported as a range (e.g., 1 to 14 hours). ____________ 6For this same reason, we do not calculate hourly wages in the CPS before 1975. 74 Top-Codes Both the CPS and the decennial Census restrict responses to income questions to a certain range. Responses outside of the range are “topcoded”: reported at the range cutoff points. For instance, from 1967 to 1975, sampled households with income above $50,000 were reported as income at $50,000 in the CPS. The range for reporting incomes changed over time. Increasing the magnitude of the top-code can increase measured income inequality even when the true underlying distribution of income does not change. To limit biases in our measures of inequality due to changing top-codes, we standardized the percentage top-coded across every year for each type of income for both surveys. Similarly, we recoded the same percentage in California and the United States. For example, the highest percentage of persons affected by the top-code of household income in the CPS was 98.8 percent (in 1975 in California). We recoded household income in every year of the CPS so that 98.8 percent of people were top-coded in both the state and the nation. Despite this recode, the top-coding can still affect estimates of trends in income distribution. The recode consistently top-codes total household income but not its component parts. In some cases, a person will have one component of income top-coded so that the sum of household income is affected by this top-code even when his or her household income is below the top of the range for household income. For example, a person with a salary of over $50,000 in 1980 will have that component of his income top-coded. This top-coding will affect the sum of income in his household, even though household income was not top-coded in that year. The measure of annual earnings used in the text 75 is not affected by this problem because we only need to recode based on a single component: income from wages and salary. The household size adjustment results in an additional problem with top-coding. A household with income top-coded at $50,000 in 1970 may not be in the top of the distribution of adjusted household income in that year if several people share that income. Top-coding will also dampen the magnitude of levels of inequality by masking the distribution of income above the cutoff points. As a result, an increasing concentration of income among the super-rich (the top 1 percent of income recipients) will not register in our measures of income inequality. Similarly, if the spread of income above the top-code is greater in some areas of the country than in others, the top-code will affect our comparisons of California to other states and regions. Thus, although we have recoded the data for consistent top-codes, the trends in adjusted household income are still affected by top-coding. Top-coding changed in each Census and in the CPS in years 1976, 1981, 1982, 1985, and 1989. Imputation Procedures In both the CPS and the Census, some respondents do not answer some of the income questions or answer inconsistently. When this happens, the Census Bureau uses a “hot deck” procedure to impute the missing income information from another person or household with similar characteristics. The hot deck procedure has changed over time in the CPS and in every year of the Census. In the CPS in 1976, education was added to the list of items used to define a hot deck match and the procedure was changed so that earnings, 76 weeks of work, and hours per week are supplied by the same matched observation. Juhn, Murphy, and Pierce (1993) show that these changes lowered estimates of hourly wage inequality. The trends in the distribution of hourly wages in this study begin with the 1976 survey and therefore are not affected by this change. In 1989, the CPS hot deck procedure was changed so that all income items are supplied by the same matched observation. In addition, the processing system was updated and more sources of income were added to the questionnaire. These changes led to an increase in aggregate income. To allow for comparisons with earlier survey years, the 1988 survey was reissued using the new 1989 processing system. Although we report income statistics for income year 1987 only from the reissue of the March 1988 CPS, all statistics in this study were also calculated with the original 1988 survey. A comparison of the results based on the two surveys shows that the new processing system reduces income inequality for all income metrics and all inequality measures, but the change is small. For example, the coefficient of variation of adjusted household income based on the original 1988 survey was 110.9; it was 110.7 based on the reissue with the new processing system. (See Appendix D for the decile levels of income in both issues of the 1988 survey. The original survey is labeled 1987a; the reissue, which we used, is labeled 1987.) Consumer Price Index and Cost of Living Adjustment All income statistics reported in this study have been adjusted to 1994 dollars based on the consumer price index computed by the Bureau of Labor Statistics (BLS). The consumer price index for California is 77 calculated by the California Department of Finance based on the population-weighted sum of the consumer price indices for San Francisco and Los Angeles (and San Diego between 1965 and 1986). The consumer price index used in this report is based on all urban consumers (CPI-U). In 1983, the method for calculating the CPI-U was changed to include a rental equivalence measure for owner-occupied housing. At the national level, the consumer price index was reissued for the years 1967 to 1982 to reflect this change (CPI-U-X1). The CPI-U-X1 series is the preferred price index because the CPI-U overstated inflation during the 1970s due to housing cost estimation procedures; after 1982, the CPI-U is the same as the CPI-U-X1. Because the CPI-U-X1 series is not available at the level of metropolitan areas before 1983, however, the California price index is based on the CPI-U. To construct a CPI-U-X1 series for California, we assumed that the ratio of (CPI-U)/(CPI-U-X1) in the national statistics is the same for the California statistics. Using this assumption and the CPI-U and CPI-U-X1 series for the nation and the CPI-U series for California, we computed an estimate of the CPI-U-X1 for California. The consumer price index provided by the BLS does not adjust for cost of living differences among regions. If the cost of living is higher in California than the national average, a higher income in California will have less purchasing power than a lower income elsewhere in the nation. Because of the difficulty in measuring the regional cost of living, the BLS stopped reporting this statistic in 1981. It is possible to create a cost of living series using the 1981 estimate of a 8.4 percent7 higher cost of ____________ 7The BLS reported a cost of living index for 24 standard metropolitan statistical areas (SMSAs) in 1981. Using the BLS index, McMahon (1991) calculated an index of 78 living in California and adjusting by the California consumer price index to create a yearly cost of living estimate. However, this estimate may not be accurate enough to allow reliable income comparisons between California and the nation. Median household income, reported in Figure 2.1, has been adjusted in this manner. All other statistics and figures reported in the text are insensitive to the cost of living adjustments. The price and cost of living adjustments for conversion to 1994 California dollars are summarized in Table A.1. The first column shows the CPI-U for California as reported by the California Department of Finance. The second column shows the CPI-U and the fourth column shows the CPI-U-X1 for the nation, as calculated by the BLS. The third column shows our calculation of a CPI-U-X1 for California using the assumptions described above. The fifth column converts Column 3 so that the 1994 value is equal to 1. The sixth column converts Column 4 to reflect the higher cost of living in California. As described above, the series was calculated by making the cost of living 8.41 percent higher in California than in the nation in 1981. To convert income data to 1994 California dollars, we multiply California data by Column 5 and national data by Column 6. The “Ideal” Data Several improvements in the quality of the data and the accuracy of the analysis could be made if California were to collect state-level data for ____________________________________________________ 108.41 in California (where the population-weighted average for the United States is 100). 79 Table A.1 Price and Cost of Living Adjustments, California and United States, 1967–1994 Year 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 12 CPI-U CA U.S. 33.0 34.4 36.1 37.9 39.3 40.6 43.0 47.4 52.3 55.6 59.5 64.4 71.3 82.4 91.4 97.3 98.9 103.8 108.6 112.0 116.6 121.9 128.0 135.0 140.6 145.6 149.4 151.5 33.4 34.8 36.7 38.8 40.5 41.8 44.4 49.3 53.8 56.9 60.6 65.2 72.6 82.4 90.9 96.5 99.6 103.9 107.6 109.6 113.6 118.3 124.0 130.7 136.2 140.3 144.5 148.2 34 CPI-U-X1 CA U.S. 35.9 37.3 38.8 40.3 41.8 43.1 45.7 49.9 54.6 58.0 62.1 66.7 72.7 82.3 90.6 96.4 98.9 103.8 108.6 112.0 116.6 121.9 128.0 135.0 140.6 145.6 149.4 151.5 36.3 37.7 39.4 41.3 43.1 44.4 47.2 51.9 56.2 59.4 63.2 67.5 74.0 82.3 90.1 95.6 99.6 103.9 107.6 109.6 113.6 118.3 124.0 130.7 136.2 140.3 144.5 148.2 56 COLA CA U.S. 4.22 4.50 4.07 4.33 3.91 4.15 3.76 3.96 3.62 3.79 3.51 3.68 3.31 3.46 3.04 3.15 2.77 2.91 2.61 2.75 2.44 2.58 2.27 2.42 2.08 2.21 1.84 1.98 1.67 1.81 1.57 1.71 1.53 1.64 1.46 1.57 1.40 1.52 1.35 1.49 1.30 1.44 1.24 1.38 1.18 1.32 1.12 1.25 1.08 1.20 1.04 1.16 1.01 1.13 1.00 1.10 SOURCES: Column 1, California Department of Finance; Columns 2 and 4, U.S. Bureau of Labor Statistics; Columns 3, 5, and 6, authors’ calculations. the study of the economy, including income inequality. Ideally, the state dataset would use a sample representative of the population of California that would be big enough to look at subregions and groups within the population. The dataset would be consistently collected over several 80 years and would include a panel component (i.e., would interview the same people over time). It would add to the value of the survey to have accurate inflation and cost of living estimates for the state. 81 Appendix B Using the Current Population Survey to Represent California1 The weighting procedure in the Current Population Survey (CPS) makes the survey representative of the nation as a whole. In calculating the March file weights, the Census Bureau does not attempt to correct for population distributions within states (e.g., California’s distinct racial distribution). Therefore, the California subsample of the March CPS may not accurately represent the population of the state. Before beginning our analysis of the distribution of income in California based ____________ 1The information on Census Bureau weighting procedures comes from the U.S. Department of Commerce and the Bureau of the Census (1978) and subsequent publications regarding redesign and revision of the CPS. 82 on the CPS data, we first evaluated the ability of the CPS to represent California.2 After conducting the March survey, the Census Bureau calculates a weight for each observation in the sample. The weight is based on a combination of factors, including adjustments to make the survey population match the national population’s distributions of age, sex, and race (with full interactions: age within sex within race). In survey years before 1978, the national sample weights did not specifically take into account the total number of people within each state. Since that time, several changes have been implemented in the calculation of the weights so that estimates of state populations based on the CPS sample are consistent with estimates of state total populations from other sources. However, the estimates of state populations within sex, age, and racial groups are not adjusted by the state-specific weights. Compared to the nation, California has had very different distributions of these characteristics, especially race. We suspected that adjusting the sample to match the U.S. distributions might severely affect the sample distributions for California. Furthermore, the Census Bureau bases the national weights on the decennial Census. Between Census updates, the weights are based on calculations of the population change. California’s distinct population trends provide another reason to suspect that the sample distributions would not accurately reflect the changes in California’s population. ____________ 2In addition to weighting procedures, sample design issues (e.g., changes in Primary Sampling Units) can also affect the representativeness of the CPS at the state level. The CPS sample design is discussed in Appendix A. 83 The Census Bureau does calculate labor force estimates at the state level from the CPS. For some states, this requires a state supplemental sample. However, the Census Bureau deemed the California subsample in the CPS large enough to make a state supplement for California unnecessary. Thus, we anticipated that the potential problem for our study was not sample size but rather whether the population distributions would be representative of the state. Although the Census Bureau believes that the California sample is large enough, it does construct state-specific weights (based on each state’s population distribution of race by residence) when it calculates statespecific labor force statistics from the CPS. The Census Bureau’s statespecific sample weights are not calculated for the March demographic survey used in this study (and are not provided in the public-use data). However, if the California subsample is found to be not representative of the state, state-specific weights for California could be constructed for the March survey using estimates of the California population by age, sex, and race (available from the Department of Finance). To determine whether a reweight of the California data was necessary, we evaluated the California subsample of the CPS on the distributions of residence, sex, age, and race—the same characteristics that the Census Bureau uses to weight the national sample. The CPS distributions were compared to the Census distributions for the years 1970, 1980, and 1990.3 Although the Census is also a national survey, it ____________ 3We expected that the representativeness of the CPS would be particularly poor in Census years. The Census Bureau recalculates the national weights based on each Census and applies updates of them in subsequent years—new weights were introduced in 1973, 1982, and 1994. Therefore, the weights in Census years are based on updates of tenyear-old population estimates. 84 is designed to survey the entire population in each state. The 5 percent sample of the Census used in this study is randomly chosen from the national population and is therefore representative at the state level (except for undercount problems). Table B.1 reports the distributions of farm households, sex, age, race, and ethnicity for California from the Census and the CPS.4 Judging by the similarity of the distributions, we concluded that the California subsample of the national CPS appears to represent the California population accurately with respect to these characteristics. Thus, we used the national sample weights in our calculations and did not reweight at the California level. The race and ethnicity distributions do show an interesting difference between the CPS and the Census. In 1980 and 1990, the distributions of race in the CPS do not match well with the distributions of race in the Census. However, when race and ethnicity are combined, the distributions from the two surveys match closely. This pattern suggests that people respond differently to race questions in the CPS and in the Census. Hispanic respondents in the Census are much more likely to record race as “other” than Hispanic respondents in the CPS. Although the statistics reported in Table B.1 certainly suggest that the California subsample of the CPS can be used to represent the state, further research is required to verify that the interacted distributions (sex within age within race) and the intercensal distributions are representative. Such verification is beyond the scope of this report. ____________ 4Deciles of the income distributions in the CPS and the Census also match fairly closely. See Appendix D. 85 Table B.1 Percentage of Population in Each Category: Census and CPS Characteristic 1970 1980 Census CPS Census CPS 1990 Census CPS Farm % non-farm household 90 n/a 99 99 Sex % male 48 48 49 49 Age in years 0–4 8 9 8 8 5–9 10 11 7 7 10–14 10 10 8 7 15–19 9 999 20–24 8 8 9 10 25–29 7 799 30–34 6 699 35–39 6 676 40–44 6 655 45–49 6 755 50–54 5 555 55–59 5 555 60–64 4 344 65–69 3 344 70–74 2 233 75–79 2 222 80+ 2 2 2 2 Hispanic Mexican n/a n/a 15 15 Other Hispanic n/a n/a 4 3 Not Hispanic n/a n/a 80 82 Race White 90 90 77 86 Black 7 688 Native American n/a n/a n/a n/a Asian n/a n/a n/a n/a Other 3 4 16 6 Race and ethnicity White, non-Hispanic n/a n/a 67 68 Black, non-Hispanic n/a n/a 7 8 Asian, non-Hispanic n/a n/a 5 n/a Hispanic n/a n/a 19 18 Other, non-Hispanic n/a n/a 1 6 99 49 8 8 7 7 8 9 10 8 7 6 5 4 4 4 3 2 2 21 5 74 69 7 1 10 13 57 7 9 26 1 99 49 9 8 7 7 8 9 9 8 7 5 4 4 4 4 3 2 2 21 4 75 83 7 1 9 1 58 7 9 25 1 SOURCES: Authors’ calculations from the March CPS and the Census. NOTE: Percentages may not add to 100 due to rounding. 86 Appendix C Trends in the Distributions of Alternative Measures of Income The first section of this appendix reviews the literature on the trends in the distributions of income for measures that account for taxes and non-monetary compensation and transfers. The second section presents results for alternative measures of money income not discussed fully in the text. Income Other Than Money Income The Current Population Survey (CPS) and the Census measure only pre-tax money income. When taxes and non-monetary transfers (e.g., health insurance, housing subsidies) are incorporated in the income measure, the decline in annual earnings is often diminished and the level of income inequality is generally lower. However, national research 87 shows that the growth in income inequality remains at levels similar to pre-tax money income. Pre-tax money income measures are imperfect indices of economic well-being. Money income is not reduced for payments such as personal taxes, Social Security, and union dues. Money income does not include non-monetary compensation such as health insurance, employer contributions to retirement programs, and room and board. Money income also does not include non-monetary transfers such as Medi-Cal, housing subsidies, food stamps, and energy assistance. Money income does not include the return to non-financial investments, such as owneroccupied housing. Studies that adjust money income for tax payments have reported rising inequality trends similar to those found for pre-tax income. Chamberlain and Spillberg (1991) report that the share of pre-tax adjusted gross income going to the top 20 percent of the distribution increased 9.5 percent between 1980 and 1988; the share of after-tax income increased 9.3 percent. Moreover, Gramlich, Kasten, and Sammartino (1993) and Pechman (1990) find that for the nation during the 1980s, inequality in after-tax income increased even more than inequality in pre-tax income. National studies that attempt to account for non-monetary benefits and taxes find that the trends in money income inequality are confirmed. The U.S. House of Representatives Committee on Ways and Means (1989) reports similar trends in the quintile shares of money income and more comprehensive income (after-tax income, including food and housing benefits). For example, the share of money income received by the poorest 20 percent of families fell by 9.8 percent between 1979 and 88 1987; their share of comprehensive income fell by 9.2 percent. Levy (1987) finds that the level of income inequality is lower when taxes, Medicare, Medicaid, food stamps, and fringe benefits are included in income, but that the trends in income inequality in Census and CPS data are essentially the same. This result may be different in California, which has lower health insurance rates than the rest of the country. Consumption data provide an alternative measure of economic wellbeing. Cutler and Katz (1991, 1992) find that changes in the distribution of expenditures parallel changes in the distribution of money income during the 1980s. Alternative Measures of Money Income Chapter 2 describes trends in the distribution of adjusted household income among persons. This measure of income was chosen because it allows for income-sharing among members of the same household, accounts for the greater income needs of large households, and counts each person equally regardless of household size. With CPS data it is possible to create alternative measures of income that vary the incomepooling unit (e.g., income-sharing within the family versus within the household), the size adjustment, and the unit of analysis (e.g., each person counts as a unit versus each household counts as a unit). Table C.1 lists the 12 types of household and family income examined in this study. All 12 measures use the sum of income received from all reported sources. Chapter 3 describes the trends in the distributions of annual earnings and hourly wages among people who are primarily employees (i.e., people who receive most of their earnings from wages and salary as 89 Table C.1 Alternative Measures of Household and Family Income Income Measure Income Pooling Unit of Size Location of Analysis Adjustment Results 1. Adjusted household income Household among persons residents Person n Chapter 2 2. Unadjusted household income among persons Household residents Person None Appendix C 3. Adjusted household income Household among households residents Household n Appendix C 4. Unadjusted household income among households Household residents Household None Appendix C 5. Adjusted family income among persons Family members Person n Chapter 2 6. Unadjusted family income among persons Family members Person None Appendix C 7. Adjusted family income among families Family members Family n Appendix C 8. Unadjusted family income among families Family members Family None Appendix C 9. Adjusted primary family income among persons Primary family members Person n Appendix C 10. Unadjusted primary family Primary family income among persons members Person None Appendix C 11. Adjusted primary family income among families Primary family members Family n Appendix C 12. Unadjusted primary family Primary family income among families members Family None Appendix C NOTES: A “family” includes all people living in the same household who are related by blood, marriage, or adoption. Separate family observations are created for single people and secondary families (families not related to their head of household). A “primary family” includes the head of household and relatives. Single people and secondary families are excluded. There is only one primary family per household. opposed to farm ownership and self-employment). The sample was restricted to employees because income from self-employment and one’s own farm often includes not only income from labor but also income from capital investments. To ensure that the measured trends were not a 90 result of limiting the sample to employees, the study also examined the distributions of annual earnings and hourly wages without this sample restriction. In addition, we examined the distributions limited to workers ages 18 to 55 (to remove any effects of early retirement) and the distributions of income from wages and salary only (for comparison to earlier national studies). Table C.2 summarizes the seven measures of labor income examined in this study. The trends in each income measure were estimated for males and females separately. The alternative measures generally display similar results to those discussed in the text. For family income and male earnings and hourly wages, we find the same five results: Inequality has increased in California since the early 1970s, the level and trends in inequality were similar in California and the nation until the late 1980s when inequality in California grew more rapidly, inequality increased most rapidly during recessions, income in the lower percentiles declined, and income growth was slower in California than in the nation. There are few exceptions to these trends. When family income is weighted at the family level, the VLN measure of inequality shows that California had higher inequality as early as 1979. When male annual earnings includes all workers, the VLN measure shows higher inequality in the United States than in California before 1975 and no substantial difference between the United States and California in the 1990s. For male income from wages and salary, the level of the VLN measure of inequality essentially recovers to pre-recession levels after the recession of the early 1980s. Despite these differences, the basic trends remain fairly consistent and the measures of income discussed in the text are preferred (as discussed above). 91 Table C.2 Alternative Measures of Labor Income Income Measure 13. Male annual earnings among workers 20. Female annual earnings among workers Sample Includes Source of Location of Anyone Who: Earnings Unit Ages Results Receives earnings primarily from wages and salary All earnings Annual 18 and Chapter 3 over 14. Male hourly wages among workers 21 Female hourly wages among workers Receives earnings primarily from wages and salary All earnings Hourly 18 and Chapter 3 over 15. Male annual earnings, workers ages 18 to 55 22. Female annual earnings, workers ages 18 to 55 Receives earnings primarily from wages and salary All earnings Annual 18 to Appendix C 55 16. Male hourly wages, workers ages 18 to 55 23. Female hourly wages, workers ages 18 to 55 Receives earnings primarily from wages and salary All earnings Hourly 18 to Appendix C 55 17. Male annual Receives All earnings Annual 18 and Appendix C earnings among all income over workers from wages, 24. Female annual salary, self- earnings among all employment, workers or own farm 92 Table C.2—continued Income Measure Sample Includes Source of Location of Anyone Who: Earnings Unit Ages Results 18. Male hourly wages Receives All earnings Hourly 18 and Appendix C among all workers income over 25. Female hourly from wages, wages among all salary, self- workers employment, or own farm 19. Male annual wages Receives any Earnings Annual 18 and Appendix C and salary income from from wages over 26. Female annual wages and and salary wages and salary salary NOTES: The income category “Receives earnings primarily from wages and salary” excludes people who report more income from their farm or business than from wages and salaries and excludes any person reporting an absolute value of more than $2,000 in 1994 dollars in income from their own farm or self-employment. Some wage and salary workers included in the sample receive a small amount of income from farms and self-employment. This income was included in annual earnings to improve estimates of hourly wages because estimates of annual hours of work include hours worked in the farm or business. The income category “All earnings” includes earnings from wages, salary, self-employment, or own farm. Hourly wages are not calculated for income types 19 and 26 because hours of work includes hours worked in selfemployment or own farm. All samples exclude military personnel, students, people with earnings less than or equal to zero, people under age 18, and workers whose primary occupation is “without pay.” The trends in the distribution of household income show more sensitivity to the adjustments for household size and weighting by persons. When each household is counted as a single unit (as opposed to each person) or no adjustments are made for household size, the upward trend in the VLN is less clear. Household income growth is generally, but not always, higher in the United States than in California. When household income is weighted at the household level, income at the 20th percentile does not decline between 1969 and 1989, but it does decline 93 between 1976 and 1994. Although these differences are notable, unadjusted and unweighted household income does not reflect economic well-being as accurately as adjusted household income (weighted at the person level) because it does not adjust for the greater needs of large households and it gives less weight to people in large households. For all the measures of female annual earnings, inequality declined until the early 1980s, the level and trends in inequality were similar in California and the nation, and income in the lower percentiles increased. When female annual earnings are measured for all female workers regardless of self-employment or own farm status, the decline in inequality does not begin until after 1975 and all measures of inequality show a rising trend after 1983. This is in contrast to the trends discussed in the text for female annual earnings among workers who receive earnings primarily from wages and salary: The trends in inequality after the early 1980s depended on which measure of inequality was used. For the measures of female hourly wages, inequality has increased since the early 1980s, the level and trends in inequality were similar in California and the nation, and wages in the lower percentiles fell. There were no substantial exceptions to this pattern. For annual earnings and hourly wages among both males and females, the trends in the distributions remain nearly identical when the age range is restricted to ages 18 to 55. Tables C.3 through C.22 and their associated figures provide summary statistics for the trends in the distributions of each alternative measure of income that is not described fully in the text. 94 Income Type 2: Unadjusted Household Income, Weighted by Persons Table C.3 Percentage Change in Real Unadjusted Household Income Among Persons Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –7 4 13 +21 –2 –8 –20 2 7 –2 7 21 14 +9 +32 +42 United States 20th Median 80th Change in 80/20 ratio (%) –1 9 14 +15 –1 3 11 +13 –2 –7 12 2 27 15 +29 +24 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .77 CV .75 CA .73 U.S. 2.35 VLN 2.15 CA 1.95 U.S. .71 1.75 .69 1.55 .67 1.35 .65 1.15 .63 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.1—Summary Measures of Inequality for Unadjusted Household Income Among Persons, 1967–1994 95 Income Type 3: Adjusted Household Income, Weighted at the Household Level Table C.4 Percentage Change in Real Adjusted Household Income Among Households Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) 0 11 18 +18 88 8 19 10 30 +2 +20 –11 3 17 +31 United States 20th Median 80th Change in 80/20 ratio (%) 13 15 18 +5 5 18 10 26 14 36 +9 +15 5 9 20 +15 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .83 CV .81 CA .79 U.S. .77 2.15 1.95 1.75 VLN CA U.S. .75 1.55 .73 1.35 .71 1.15 .69 .67 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.2—Summary Measures of Inequality for Adjusted Household Income Among Households, 1967–1994 96 Income Type 4: Unadjusted Household Income, Weighted at the Household Level Table C.5 Percentage Change in Real Unadjusted Household Income Among Households Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –3 2 10 +13 14 10 –9 684 10 21 18 –3 +10 +29 United States 20th Median 80th Change in 80/20 ratio (%) 3 5 11 +8 470 5 11 2 11 23 14 +7 +15 +15 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .81 CV .79 CA U.S. .77 2.35 2.15 1.95 VLN CA U.S. .75 1.75 .73 1.55 .71 1.35 .69 1.15 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.3—Summary Measures of Inequality for Unadjusted Household Income Among Households, 1967–1994 97 Income Type 6: Unadjusted Family Income, Weighted by Persons Table C.6 Percentage Change in Real Unadjusted Family Income Among Persons Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –15 1 10 +29 –4 –2 6 +10 –18 –25 0 –8 16 12 +42 +50 United States 20th Median 80th Change in 80/20 ratio (%) –4 7 12 +17 –5 1 10 +15 –9 –14 8 –2 23 13 +35 +31 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .81 CV .79 2.95 VLN 2.75 .77 CA U.S. .75 .73 .71 .69 .67 2.55 2.35 2.15 1.95 1.75 1.55 1.35 CA U.S. .65 1.15 .63 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.4—Summary Measures of Inequality for Unadjusted Family Income Among Persons, 1967–1994 98 Income Type 7: Adjusted Family Income, Weighted at the Family Level Table C.7 Percentage Change in Real Adjusted Family Income Among Families Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) 0 6 15 +15 –1 –1 6 12 7 23 +9 +25 –16 –2 13 +34 United States 20th 12 2 14 Median 13 8 21 80th 17 13 33 Change in 80/20 ratio (%) +5 +11 +16 0 6 17 +17 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .86 CV .84 3.3 VLN 3.1 .82 CA U.S. .80 .78 .76 .74 .72 2.9 CA 2.7 U.S. 2.5 2.3 2.1 1.9 1.7 .70 1.5 .68 1.3 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.5—Summary Measures of Inequality for Adjusted Family Income Among Families, 1967–1994 99 Income Type 8: Unadjusted Family Income, Weighted at the Family Level Table C.8 Percentage Change in Real Unadjusted Family Income Among Families Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –12 –8 5 +19 4 –9 3 –5 5 11 +2 +22 –15 –5 12 +32 United States 20th Median 80th Change in 80/20 ratio (%) 2 1 10 +7 02 23 8 19 +8 +17 –8 –5 10 +20 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .90 CV .88 .86 CA .84 U.S. .82 3.95 3.45 2.95 VLN CA U.S. .80 2.45 .78 .76 1.95 .74 1.45 .72 .70 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.6—Summary Measures of Inequality for Unadjusted Family Income Among Families, 1967–1994 100 Income Type 9: Adjusted Primary Family Income, Weighted by Persons Table C.9 Percentage Change in Real Adjusted Primary Family Income Among Persons Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th 2 –13 –11 –28 Median 14 0 14 –5 80th 20 5 25 15 Change in 80/20 ratio (%) +18 +20 +42 +59 United States 20th 10 –2 9 –4 Median 18 7 27 9 80th 22 13 38 21 Change in 80/20 ratio (%) +10 +15 +27 +26 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .80 CV .78 .76 CA .74 U.S. .72 2.4 VLN 2.2 CA 2.0 U.S. 1.8 .70 1.6 .68 1.4 .66 1.2 .64 .62 1.0 .60 .8 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.7—Summary Measures of Inequality for Adjusted Primary Family Income Among Persons, 1967–1994 101 Income Type 10: Unadjusted Primary Family Income, Weighted by Persons Table C.10 Percentage Change in Real Unadjusted Primary Family Income Among Persons Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th –15 –4 –18 –25 Median 1 –2 0 –8 80th 10 6 16 12 Change in 80/20 ratio (%) +29 +10 +42 +50 United States 20th –4 –5 –9 –14 Median 7 1 8 –2 80th 12 10 23 13 Change in 80/20 ratio (%) +17 +15 +35 +31 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .85 CV 3.3 VLN .80 CA U.S. 2.8 CA U.S. .75 2.3 .70 1.8 .65 1.3 .60 .8 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.8—Summary Measures of Inequality for Unadjusted Primary Family Income Among Persons, 1967–1994 102 Income Type 11: Adjusted Primary Family Income, Weighted at the Family Level Table C.11 Percentage Change in Real Adjusted Primary Family Income Among Families Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th 1 –8 –8 –23 Median 11 2 14 1 80th 19 6 27 16 Change in 80/20 ratio (%) +19 +16 +37 +51 United States 20th 10 –1 10 –1 Median 16 8 25 8 80th 19 15 36 21 Change in 80/20 ratio (%) +8 +15 +24 +23 SOURCE; Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .80 CV .78 .76 CA .74 U.S. 2.2 VLN 2.0 CA 1.8 U.S. .72 1.6 .70 .68 1.4 .66 1.2 .64 1.0 .62 .60 .8 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.9—Summary Measures of Inequality for Adjusted Primary Family Income Among Families, 1967–1994 103 Income Type 12: Unadjusted Primary Family Income, Weighted at the Family Level Table C.12 Percentage Change in Real Unadjusted Primary Family Income Among Families Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –3 7 14 +18 –5 –8 19 8 23 +13 +33 –20 0 16 +44 United States 20th Median 80th Change in 80/20 ratio (%) 5 11 14 +9 –3 2 4 16 13 29 +16 +26 –4 4 17 +22 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .78 .76 .74 CA U.S. .72 .70 .68 .66 .64 .62 2.4 VLN 2.2 CA 2.0 U.S. 1.8 1.6 1.4 1.2 1.0 .60 .8 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.10—Summary Measures of Inequality for Unadjusted Primary Family Income Among Families, 1967–1994 104 Income Type 15: Annual Earnings Among Male Workers Ages 18 to 55 Table C.13 Percentage Change in Real Annual Earnings for Males Ages 18 to 55 Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –22 –6 10 +40 –22 –39 –12 –17 –5 4 +22 +71 –30 –23 2 +46 United States 20th Median 80th Change in 80/20 ratio (%) –7 3 11 +19 –17 –22 –8 –5 4 16 +25 +49 –20 –13 2 +27 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .80 CV .75 CA U.S. .70 .65 .60 1.25 1.15 1.05 VLN CA U.S. .95 .85 .75 .55 .65 .50 .55 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.11—Summary Measures of Inequality for Annual Earnings for Males Ages 18 to 55, 1967–1994 105 Income Type 16: Hourly Wages Among Male Workers Ages 18 to 55 Table C.14 Percentage Change in Real Hourly Wages for Males Ages 18 to 55 Between Selected Years, by Income Percentile Business Cycle Peaks 1979–1989 Recessions 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –21 –14 –5 +20 –31 –23 –6 +36 United States 20th Median 80th Change in 80/20 ratio (%) –13 –8 1 +16 –20 –13 –2 +22 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .65 CV .63 CA .61 U.S. .59 .45 VLN .43 CA .41 U.S. .39 .57 .37 .55 .35 .53 .33 .51 .31 .49 .29 .47 .27 .45 .25 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.12—Summary Measures of Inequality for Hourly Wages for Males Ages 18 to 55, 1975–1994 106 Income Type 17: Annual Earnings Among All Male Workers Table C.15 Percentage Change in Real Annual Earnings for All Male Workers Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th –13 –19 –30 –31 median –5 –10 –14 –20 80th 6 –2 4 –2 Change in 80/20 ratio (%) +22 +20 +47 +41 United States 20th median 80th Change in 80/20 ratio (%) –5 3 10 +15 –14 –5 4 +21 –18 –15 –2 –12 15 3 +40 +21 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .85 CV .80 CA U.S. .75 .70 .65 1.5 VLN 1.4 CA U.S. 1.3 1.2 1.1 1.0 .60 .9 .55 .8 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.13—Summary Measures of Inequality for Annual Earnings for All Male Workers, 1967–1994 107 Income Type 18: Hourly Wages Among All Male Workers Table C.16 Percentage Change in Real Hourly Wages for All Male Workers Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1979–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) United States 20th Median 80th Change in 80/20 ratio (%) –19 –12 –4 +19 –10 –5 2 +13 –30 –21 –4 +37 –16 –13 0 +20 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .69 CV .67 CA .65 U.S. .63 .53 VLN CA .48 U.S. .61 .43 .59 .57 .38 .55 .53 .33 .51 .49 .28 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.14—Summary Measures of Inequality for Hourly Wages for All Male Workers, 1975–1994 108 Income Type 19: Annual Income from Wages and Salary Among Male Workers Table C.17 Percentage Change in Real Annual Salary for Males Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –3 –4 6 +9 –14 –16 –12 –15 –1 6 +15 +26 –14 –23 4 +21 United States 20th Median 80th Change in 80/20 ratio (%) –5 5 8 +14 –5 –10 –6 –1 6 15 +12 +28 –3 –8 6 +10 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .85 CV .80 CA U.S. 1.65 VLN 1.55 CA 1.45 U.S. .75 1.35 1.25 .70 `` 1.15 .65 1.05 .60 .95 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.15—Summary Measures of Inequality for Annual Salary for Males, 1967–1994 109 Income Type 22: Annual Earnings Among Female Workers Ages 18 to 55 Table C.18 Percentage Change in Real Annual Earnings for Females Ages 18 to 55 Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) 62 22 11 –31 13 83 6 29 22 36 +8 –26 25 29 30 +5 United States 20th Median 80th Change in 80/20 ratio (%) 48 19 16 –22 24 85 19 42 24 44 0 –22 56 25 34 –14 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .84 CV .82 .80 CA U.S. 1.7 VLN 1.6 1.5 CA U.S. .78 1.4 .78 1.3 .74 1.2 .72 1.1 .70 1.0 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.16—Summary Measures of Inequality for Annual Earnings for Females Ages 18 to 55, 1967–1994 110 Income Type 23: Hourly Wages Among Female Workers Ages 18 to 55 Table C.19 Percentage Change in Real Hourly Wages for Females Ages 18 to 55 Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1979–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –9 2 15 +27 –7 10 21 +31 United States 20th Median 80th Change in 80/20 ratio (%) –4 8 16 +21 –1 11 22 +24 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .64 CV .62 CA .60 U.S. .58 .43 VLN .41 CA .39 U.S. .37 .35 .56 .33 .31 .54 .29 .52 .27 .50 .25 .48 .23 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.17—Summary Measures of Inequality for Hourly Wages for Females Ages 18 to 55, 1975–1994 111 Income Type 24: Annual Earnings Among All Female Workers Table C.20 Percentage Change in Real Annual Earnings for All Female Workers Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) 77 22 10 –38 12 7 22 +10 98 30 35 –32 28 27 28 0 United States 20th Median 80th Change in 80/20 ratio (%) 42 21 16 –18 30 85 51 14 38 24 20 39 34 –8 –25 –11 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .84 CV .72 .80 .78 2.0 VLN 1.9 CA CA U.S. 1.8 U.S. 1.7 1.6 .76 1.5 1.4 .74 1.3 .72 1.2 .70 1.1 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.18—Summary Measures of Inequality for Annual Earnings for All Female Workers, 1967–1994 112 Income Type 25: Hourly Wages Among All Female Workers Table C.21 Percentage Change in Real Hourly Wages for All Female Workers Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1979–1989 1976–1994 California 20th Median 80th Change in 80/20 ratio (%) –9 2 15 +26 –8 7 18 +28 United States 20th Median 80th Change in 80/20 ratio (%) –5 10 16 +23 –1 11 22 +23 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .66 CV .64 CA .62 U.S. .60 .54 VLN .49 CA U.S. .44 .58 .39 .56 .54 .34 .52 .29 .50 .48 .24 1975 1979 1983 1987 1991 1994 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.19—Summary Measures of Inequality for Hourly Wages for All Female Workers, 1975–1994 113 Income Type 26: Annual Income from Wages and Salary Among Female Workers Table C.22 Percentage Change in Real Annual Salary for Females Between Selected Years, by Income Percentile Business Cycle Peaks Recessions 1969–1979 1979–1989 1969–1989 1976–1994 California 20th 53 23 89 Median 22 10 34 80th 12 18 33 Change in 80/20 ratio (%) –26 –4 –30 28 27 25 –2 United States 20th 36 20 64 50 Median 13 15 30 27 80th 12 24 39 30 Change in 80/20 ratio (%) –18 +4 –15 –14 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this table are sensitive to the consumer price index, except for the 80/20 ratio. .88 CV .86 1.80 VLN 1.75 1.70 1.65 .84 1.60 .82 1.55 1.50 .80 1.45 CA .78 U.S. 1.40 1.35 CA U.S. .76 1.30 1967 1971 1975 1979 1983 1987 1991 1994 1967 1971 1975 1979 1983 1987 1991 1994 SOURCE: Based on authors’ calculations from the March CPS. NOTE: Statistics reported in this figure are not sensitive to the consumer price index. Figure C.20—Summary Measures of Inequality for Annual Salary for Females, 1967–1994 114 Appendix D Supplementary Statistics This appendix provides additional statistics on the trends in the distributions of income for those measures of income discussed in the main text. The appendix contains several tables. Tables D.1 through D.10 show deciles of the distributions of adjusted household income, male annual earnings, male hourly wages, female annual earnings, and female hourly wages. Reported decile levels are in nominal terms. The price index from Table A.1 is provided for cost of living and inflation adjustments (multiply the income level by the price index). Then, Tables D.11 and D.12 show regional comparisons of the coefficient of variation for male and female annual earnings. Finally, Tables D.13 through D.15 give state comparisons of the coefficient of variation for adjusted household income, male annual earnings, and female annual earnings. 115 Table D.1 Deciles of Nominal Adjusted Household Income, California Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 3499 5281 1968 3914 5676 1969 4061 5997 1969 (C) 3667 5831 1970 4175 6203 1971 4159 6179 1972 4213 6197 1973 4654 7008 1974 5074 7419 1975 5266 8088 1976 6174 8894 1977 6692 9442 1978 7188 10347 1979 7690 11439 1979 (C) 7446 11553 1980 8653 12732 1981 8516 12874 1982 8697 13187 1983 8490 13594 1984 9903 15561 1985 10141 15970 1986 10543 16270 1987a 11132 17711 1987 11320 17792 1988 10825 17669 1989 12462 18878 1989 (C) 13101 20500 1990 12132 19401 1991 11790 18288 1992 11787 18771 1993 11257 17615 1994 11205 18002 6720 7989 9236 10651 12408 14901 19249 7246 8665 10003 11621 13583 16151 21194 7627 9090 10695 12465 14578 17315 22437 7647 9263 10854 12656 14779 17718 22900 7800 9491 11291 13145 15309 18185 23079 8030 9785 11485 13266 15448 18976 24496 8195 10213 12310 14533 17184 20417 26103 9248 11489 13781 16075 18567 21705 28150 9855 12262 14633 17017 19610 23145 29210 10512 13343 15792 18473 21717 26122 33852 11525 14195 17070 20046 23365 27636 35836 12278 15500 18712 22007 25788 30645 39637 13881 17124 20743 24223 28765 34064 44608 15561 19078 22841 26709 32373 38951 49068 15530 19420 23210 27492 32436 39332 50926 16946 21016 25495 30023 35586 43226 54066 17139 21451 26840 31797 38266 46669 60000 17588 22643 27853 33964 41121 50260 64044 18224 24066 29087 35233 42658 51945 67108 21220 26308 31441 37827 46015 55472 72588 22018 27314 33606 40559 48734 59935 78713 22731 29082 35929 43431 51897 63873 82836 24027 30758 36958 44456 53882 67134 88947 24013 30781 37034 44434 53788 66402 87723 24144 31056 36850 45152 55029 69489 91351 25999 33511 40876 49150 58213 72184 96805 27944 35330 43070 51962 62367 77000 102615 26180 34007 41117 49901 61079 75996 100742 25297 33529 41709 50295 61040 76628 102252 25718 34583 41924 51974 63210 78740 103648 24265 31757 39969 50999 64284 81672 107405 25963 33305 43330 55040 66503 82618 110106 4.22 4.07 3.91 3.91 3.76 3.62 3.51 3.31 3.04 2.77 2.61 2.44 2.27 2.08 2.08 1.84 1.67 1.57 1.53 1.46 1.40 1.35 1.30 1.30 1.24 1.18 1.18 1.12 1.08 1.04 1.01 1.00 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTE: Household income is adjusted for household size: Reported deciles are calibrated to represent a household of four. 116 Table D.2 Deciles of Nominal Adjusted Household Income, United States Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 2773 1968 3089 1969 3380 1969 (C) 3060 1970 3511 1971 3695 1972 3929 1973 4319 1974 4683 1975 4935 1976 5373 1977 5784 1978 6306 1979 6938 1979 (C) 6710 1980 7320 1981 7734 1982 7651 1983 7953 1984 8616 1985 9084 1986 9375 1987a 9728 1987 9865 1988 10395 1989 11385 1989(C) 11314 1990 11507 1991 11535 1992 11641 1993 11629 1994 12343 4294 4761 5220 5050 5465 5658 6060 6713 7202 7539 8246 8903 9814 10807 10760 11559 12185 12550 12837 14013 14787 15503 16109 16269 17046 18245 18779 18796 18931 19142 19316 20332 5602 6209 6784 6736 7072 7323 8010 8851 9476 9970 10783 11749 13119 14475 14432 15422 16366 17069 17657 19254 20139 21222 22170 22382 23349 25060 25470 25512 26021 26256 26383 27900 6767 7442 8169 8198 8546 8940 9811 10740 11556 12225 13362 14505 16033 17882 17947 19255 20516 21458 22321 24125 25481 26813 28001 28300 29565 31519 32000 32193 32972 33499 33802 35094 7965 8703 9579 9650 10019 10512 11540 12595 13604 14583 15916 17285 19025 21220 21311 22909 24601 25900 27043 29194 30740 32372 34173 34471 35843 38355 38682 39019 40070 41121 41413 43070 9166 10028 11077 11201 11664 12242 13452 14710 15821 16977 18566 20261 22341 24935 25020 26892 29095 30701 32260 34857 36731 38620 40629 40860 42955 45643 46052 46446 48084 49384 50242 52192 10679 11688 12888 13106 13625 14271 15718 17232 18455 19843 21721 23778 26206 29240 29456 31756 34428 36466 38378 41505 43735 46086 48432 48852 51327 54520 55007 55629 57254 58902 60558 62897 12761 14005 15337 15704 16260 17124 18734 20530 22037 23760 25729 28336 31209 35033 35348 38110 41359 44202 46739 50590 53358 56253 59224 59469 62401 66719 67199 68278 70090 72078 75345 77440 16395 4.50 17802 4.33 19639 4.15 20350 4.15 20791 3.96 21954 3.79 24186 3.68 26047 3.46 28093 3.15 30262 2.91 32682 2.75 35940 2.58 39817 2.42 44182 2.21 45368 2.21 48264 1.98 52786 1.81 57358 1.71 60170 1.64 65506 1.57 69485 1.52 73222 1.49 76889 1.44 77557 1.44 82020 1.38 87567 1.32 88912 1.32 89170 1.25 91503 1.20 94658 1.16 99102 1.13 101849 1.10 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTE: Household income is adjusted for household size: Reported deciles are calibrated to represent a household of four. 117 Table D.3 Deciles of Nominal Annual Earnings Among Male Workers, California Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 2200 4099 1968 2434 4496 1969 2342 4499 1969 (C) 2150 4550 1970 2502 4503 1971 1997 4095 1972 2231 4498 1973 2484 4796 1974 2554 5006 1975 2797 5106 1976 2716 5490 1977 3004 6009 1978 3506 6793 1979 3991 7184 1979 (C) 3805 7005 1980 4003 7406 1981 3886 7511 1982 4008 7514 1983 4856 8989 1984 4501 8403 1985 5018 9032 1986 5190 9222 1987a 5411 9953 1987 5603 9705 1988 5488 9977 1989 5822 9970 1989(C) 6000 10136 1990 5988 9980 1991 6014 10023 1992 5997 9995 1993 4980 9960 1994 6000 10400 5499 6499 5994 6994 6006 7346 6050 7350 6081 7505 6081 7590 6447 7997 6702 8392 6862 8510 7358 9411 7986 9982 8253 10516 9490 11987 9977 12555 10005 12505 10408 13530 11016 14266 10520 14327 12465 15981 12034 16925 12545 17061 12974 17964 13739 18020 13307 18010 14168 17959 13957 18418 15000 19142 14272 18962 14032 19316 14992 19990 13546 17928 15000 19200 7399 7993 8369 8450 8665 8988 9496 9991 10012 11012 11979 12981 14385 15433 15435 16713 17847 18034 19976 20807 21813 22954 23512 23012 22948 23927 24000 24042 25057 24987 23904 25000 8130 8992 9508 9650 9890 9987 10995 11689 12014 13015 14175 15022 16622 17959 18205 20015 21031 22042 23971 25008 26093 27537 28515 28015 27936 29410 29971 29941 30069 30984 29880 30772 9098 10498 9991 11689 10608 12010 10750 12050 11007 12808 11485 13183 12095 14294 12988 14986 14017 16019 15017 17520 16091 18717 17526 20030 19127 22234 20952 24943 21005 25005 23018 26020 24869 29043 26049 31059 27367 32960 29009 34953 30107 36128 31937 37925 33991 40021 33418 40021 32925 39909 34894 41872 35000 41000 34931 44100 36083 45103 38481 45977 36852 45816 40000 50000 12997 14303 15012 15050 15510 16279 17893 18483 20024 21424 22989 25037 27471 30930 30505 34026 36053 40076 41949 43014 46164 48903 50026 50026 50884 54833 52032 58883 59136 60969 59760 65000 4.22 4.07 3.91 3.91 3.76 3.62 3.51 3.31 3.04 2.77 2.61 2.44 2.27 2.08 2.08 1.84 1.67 1.57 1.53 1.46 1.40 1.35 1.30 1.30 1.24 1.18 1.18 1.12 1.08 1.04 1.01 1.00 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. 118 Table D.4 Deciles of Nominal Annual Earnings Among Male Workers, United States Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 2001 1968 2101 1969 2275 1969 (C) 2150 1970 2210 1971 2100 1972 2403 1973 2597 1974 2623 1975 2745 1976 2895 1977 3206 1978 3661 1979 3997 1979 (C) 3905 1980 4003 1981 3999 1982 3911 1983 4006 1984 4211 1985 4809 1986 4992 1987a 4982 1987 4982 1988 5388 1989 6024 1989 (C) 6000 1990 5979 1991 5810 1992 5780 1993 6026 1994 6580 3609 4891 5604 4001 5002 6002 4091 5503 6603 4050 5450 6550 4160 5601 6803 4131 5694 7001 4596 6072 7510 5003 6664 8005 5025 7009 8511 5133 7218 8985 5590 7794 9750 6011 8279 10287 6691 8988 11083 7352 9994 11992 7125 10005 12005 7506 10186 13011 7838 10998 13997 7521 10730 14039 7956 11115 14921 8341 12031 15506 9017 12524 16031 9225 12979 16972 9868 13741 17618 9888 13552 17538 9978 14392 17961 10543 15061 19077 10906 15000 19000 10603 14948 18935 10017 15026 19334 10436 15052 20070 11047 15064 20086 12000 16000 20800 6504 7204 7002 7803 7503 8465 7550 8550 7806 8982 8002 9203 8911 10013 9561 11007 10013 11515 10482 11980 11478 13027 12022 14426 13476 15579 14862 16989 15005 17025 15513 18215 16854 19996 17047 20055 18025 21029 19048 23059 20039 24047 20149 24960 21245 25411 20926 25012 21952 26641 23093 28113 22802 27000 23419 27904 24042 29050 25087 30105 25107 30129 25000 30772 8005 8960 9675 9850 10001 10402 11415 12108 13017 13976 14974 16430 17976 19987 20005 21017 22995 24066 25034 27069 28054 29951 29895 29895 31132 33134 31500 33385 35061 35122 35652 37964 9406 10003 11005 11050 11701 12002 13017 14409 15020 15973 17649 19091 20772 22985 23005 25020 26994 28077 30041 32081 34066 34943 35874 35874 37917 40162 38000 39863 41071 43150 44189 46000 11507 12004 13806 14050 14502 15003 16521 18012 19025 19966 21961 24044 25267 28082 29005 30024 33983 36100 38052 40102 42082 44927 46317 45839 48893 50202 48098 51323 53092 55318 58249 60000 4.50 4.33 4.15 4.15 3.96 3.79 3.68 3.46 3.15 2.91 2.75 2.58 2.42 2.21 2.21 1.98 1.81 1.71 1.64 1.57 1.52 1.49 1.44 1.44 1.38 1.32 1.32 1.25 1.20 1.16 1.13 1.10 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. 119 Table D.5 Deciles of Nominal Hourly Wages Among Male Workers, California Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1975 2.41 3.24 1976 2.56 3.52 1977 2.70 3.76 1978 2.88 3.92 1979 3.17 4.32 1980 3.46 4.81 1981 3.49 4.81 1982 3.76 5.01 1983 3.84 5.27 1984 3.85 5.37 1985 3.94 5.44 1986 4.08 5.76 1987a 4.15 5.77 1987 4.17 5.77 1988 4.32 5.82 1989 4.44 5.98 1990 4.57 6.11 1991 4.73 6.26 1992 4.81 6.53 1993 4.64 6.17 1994 4.81 6.41 4.09 4.88 5.73 6.35 7.22 8.42 4.32 5.23 6.08 6.91 7.86 9.05 4.67 5.54 6.46 7.31 8.35 9.63 4.99 5.95 6.99 7.99 9.12 10.57 5.37 6.57 7.67 8.76 10.00 11.51 5.77 7.11 8.39 9.62 11.07 12.83 6.08 7.50 8.81 10.30 11.94 13.70 6.42 7.86 9.63 11.13 12.81 15.07 6.81 8.34 9.70 11.52 13.28 15.61 6.93 8.66 10.13 12.02 13.77 16.22 7.06 9.12 10.72 12.54 14.62 17.37 7.20 9.12 11.20 13.21 15.35 17.75 7.42 9.38 11.42 13.68 15.87 18.76 7.50 9.38 11.17 13.47 15.49 18.28 7.48 9.40 11.17 13.43 15.71 19.19 7.67 9.59 11.56 14.24 16.49 19.82 7.70 9.60 12.00 14.39 16.79 19.96 8.19 10.12 12.30 14.55 17.35 21.03 8.24 10.25 12.88 15.38 18.26 21.97 7.76 9.85 11.97 14.59 17.62 21.55 8.33 10.10 12.39 15.38 18.75 23.08 10.38 11.04 11.80 13.09 14.39 15.93 17.33 19.27 19.63 20.20 21.44 22.36 24.05 23.09 24.55 24.92 26.32 27.34 28.83 28.73 29.62 2.77 2.61 2.44 2.27 2.08 1.84 1.67 1.57 1.53 1.46 1.40 1.35 1.30 1.30 1.24 1.18 1.12 1.08 1.04 1.01 1.00 SOURCE: Authors’ calculations from the March CPS. 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTES: Hourly wages not calculated before 1975 because earlier CPSs did not ask respondents about their hours of work in a usual week in the previous year (annual earnings refers to earnings in the previous year). Hourly wages not calculated for the Census because the 1970 Census also did not ask about weekly hours of work in the previous year. 120 Table D.6 Deciles of Nominal Hourly Wages Among Male Workers, United States Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987a 1987 1988 1989 1990 1991 1992 1993 1994 2.36 3.17 3.85 4.57 5.27 5.93 6.72 7.68 9.60 2.91 2.46 3.35 4.08 4.80 5.60 6.37 7.20 8.35 10.24 2.75 2.61 3.56 4.38 5.17 5.99 6.86 7.78 9.02 11.13 2.58 2.84 3.78 4.69 5.52 6.40 7.36 8.45 9.60 12.00 2.42 3.07 4.16 5.09 6.03 7.08 8.17 9.37 10.68 13.15 2.21 3.30 4.47 5.50 6.54 7.70 8.85 10.10 11.84 14.43 1.98 3.47 4.81 5.77 7.02 8.19 9.61 11.06 12.83 15.86 1.81 3.55 4.82 6.05 7.26 8.68 10.03 11.75 13.76 17.08 1.71 3.57 4.90 6.26 7.51 8.83 10.35 12.04 14.35 17.81 1.64 3.67 5.06 6.42 7.71 9.40 11.03 12.66 14.94 18.80 1.57 3.85 5.30 6.74 8.19 9.63 11.48 13.21 15.56 19.27 1.52 3.99 5.46 6.91 8.40 9.98 11.76 13.82 16.22 20.45 1.49 3.85 5.44 6.87 8.32 9.76 11.56 13.82 16.18 20.21 1.44 4.00 5.70 7.13 8.62 10.12 11.98 14.37 16.77 20.91 1.44 4.23 5.79 7.42 8.97 10.55 12.47 14.47 17.27 21.93 1.38 4.46 6.11 7.72 9.41 11.10 13.03 15.45 18.34 23.17 1.32 4.57 6.23 7.73 9.58 11.24 13.18 15.33 18.69 23.96 1.25 4.67 6.26 8.01 9.63 11.56 13.58 16.11 19.26 24.56 1.20 4.82 6.51 8.20 9.91 12.01 14.23 16.84 19.68 25.57 1.16 4.83 6.53 8.21 9.95 12.05 14.28 16.90 20.21 26.34 1.13 5.00 6.77 8.55 10.22 12.16 14.42 17.31 20.98 27.47 1.10 SOURCE: Authors’ calculations from the March CPS. 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTES: Hourly wages not calculated before 1975 because earlier CPSs did not ask respondents about their hours of work in a usual week in the previous year (annual earnings refers to earnings in the previous year). Hourly wages not calculated for the Census because the 1970 Census also did not ask about weekly hours of work in the previous year. 121 Table D.7 Deciles of Nominal Annual Earnings Among Female Workers, California Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 366 1968 400 1969 375 1969 (C) 450 1970 450 1971 399 1972 520 1973 599 1974 601 1975 701 1976 799 1977 929 1978 999 1979 1297 1979 (C) 1305 1980 1601 1981 1602 1982 2004 1983 1998 1984 2001 1985 2007 1986 1996 1987a 2401 1987 2501 1988 2993 1989 3274 1989 (C) 3000 1990 2994 1991 3157 1992 2998 1993 2789 1994 3000 900 1700 2400 3068 3999 4799 5799 6999 4.22 999 1698 2498 3397 4196 5015 6011 7493 4.07 993 1699 2502 3503 4333 5304 6265 7771 3.91 1050 1950 2950 3750 4650 5550 6550 8050 3.91 1006 1871 3002 3903 4815 5904 7005 8223 3.76 1011 1804 2796 3941 4994 5992 7040 8988 3.62 1299 2234 3287 4498 5498 6697 7797 9796 3.51 1399 2318 3312 4396 5695 6994 8117 10091 3.31 1464 2467 3504 4601 5857 7008 8589 11013 3.04 1511 2729 3954 5050 6507 8009 9570 12013 2.77 1996 2995 4193 5450 6988 8485 9982 12877 2.61 2003 3294 4807 6009 7511 9013 11016 14021 2.44 2399 3896 5095 6793 7992 9989 11987 14984 2.27 2993 4590 5986 7981 9379 10975 12970 17161 2.08 3005 4505 6005 7875 9505 11005 13465 17365 2.08 3493 5004 7005 9007 10508 12510 15011 19815 1.84 3505 5308 7395 9546 11577 13812 16149 21031 1.67 3866 6011 8015 10019 12263 15029 18034 23044 1.57 4195 6193 8984 10987 12984 15781 19177 24970 1.53 4258 6442 9137 11604 14004 16983 20006 26008 1.46 4215 7025 9706 12143 15053 18064 22078 28602 1.40 4890 7171 9980 12974 15968 18962 22954 29941 1.35 5003 7804 10506 14005 17009 20011 24513 31016 1.30 5003 7671 10405 13209 16753 20011 24776 31016 1.30 5522 8082 11195 14467 17959 21566 25941 33923 1.24 5982 8773 11964 14954 17945 22930 27915 35891 1.18 6000 9393 12000 15800 19500 23421 28050 36000 1.18 5988 8982 11976 14970 19960 23952 28943 36927 1.12 6014 9622 12829 16574 20046 25057 30069 38438 1.08 6497 9995 12993 16991 20989 25487 30984 39980 1.04 5976 9362 12948 16932 20916 24900 31872 42828 1.01 6000 10000 13000 18000 23000 28000 34000 43000 1.00 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. 122 Table D.8 Deciles of Nominal Annual Earnings Among Female Workers, United States Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1967 300 1968 350 1969 390 1969 (C) 450 1970 400 1971 450 1972 501 1973 500 1974 575 1975 653 1976 735 1977 801 1978 939 1979 1039 1979 (C) 1195 1980 1154 1981 1300 1982 1474 1983 1502 1984 1540 1985 1703 1986 1797 1987a 1993 1987 1993 1988 2130 1989 2465 1989 (C) 2500 1990 2491 1991 2719 1992 2950 1993 3013 1994 3000 750 850 925 1050 1000 1040 1172 1201 1352 1547 1757 2004 2001 2498 2565 2856 2999 3169 3455 3609 3908 3994 4504 4599 4989 5020 5219 5406 5810 6021 6026 6288 1301 1500 1578 1750 1700 1800 2003 2001 2303 2496 2822 3005 3396 3997 4005 4323 4879 5014 5464 5922 6012 6330 6975 6975 7284 8032 8000 8013 8834 9031 9039 9625 2001 2748 2156 3001 2371 3107 2550 3250 2500 3345 2712 3554 3004 3845 3002 4003 3259 4296 3604 4776 3993 5021 4276 5510 4894 5992 5496 6995 5505 7005 6005 7599 6499 8098 7019 9025 7510 9613 8020 10025 8260 10520 8892 10982 9487 11958 9565 11958 9978 12473 10442 13053 10500 13110 10962 13952 11565 14585 12042 15052 12051 15566 12500 16000 3302 3581 3912 4050 4100 4371 4706 5003 5207 5890 6189 6913 7490 8195 8165 9101 9998 10930 11716 12031 13025 13897 14768 14748 14967 16065 16000 16942 18031 19066 19164 20000 4003 4201 4647 4850 5001 5201 5607 6004 6408 6988 7587 8015 8988 9994 10005 11009 11997 13036 14019 15038 15630 16673 17680 17738 18459 20081 19917 20330 21638 23080 23099 24000 4953 6004 5002 6452 5527 7003 5750 7150 6001 7601 6284 8002 6904 8511 7205 9006 7910 9895 8486 10694 9014 11480 10018 12190 10631 13482 11992 14990 12005 15005 13011 16613 14497 17996 15765 20055 17023 21433 18046 23059 19338 25049 19968 25958 20926 27520 21126 27404 22950 29517 24097 31126 24000 30130 24914 31890 26045 34560 28098 35508 29024 37661 30000 39200 4.50 4.33 4.15 4.15 3.96 3.79 3.68 3.46 3.15 2.91 2.75 2.58 2.42 2.21 2.21 1.98 1.81 1.71 1.64 1.57 1.52 1.49 1.44 1.44 1.38 1.32 1.32 1.25 1.20 1.16 1.13 1.10 SOURCE: Authors’ calculations from the CPS and Census (C). 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. 123 Table D.9 Deciles of Nominal Hourly Wages Among Female Workers, California Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1975 1.64 2.19 2.59 3.03 3.49 3.99 4.55 5.29 6.67 2.77 1976 1.86 2.40 2.75 3.24 3.74 4.22 4.80 5.61 6.93 2.61 1977 1.96 2.50 3.00 3.51 3.97 4.51 5.14 6.05 7.54 2.44 1978 2.12 2.79 3.29 3.75 4.32 4.80 5.40 6.38 7.97 2.27 1979 2.49 3.12 3.63 4.21 4.80 5.40 6.21 7.13 9.11 2.08 1980 2.69 3.37 4.07 4.74 5.29 6.00 6.90 8.18 10.13 1.84 1981 2.89 3.61 4.38 5.06 5.78 6.48 7.45 8.82 11.07 1.67 1982 3.01 3.85 4.67 5.32 6.25 7.16 8.19 9.63 12.04 1.57 1983 3.08 4.00 4.82 5.71 6.45 7.37 8.61 10.08 12.64 1.53 1984 3.13 4.15 5.00 5.96 6.84 7.85 9.14 11.04 13.55 1.46 1985 3.35 4.34 5.31 6.27 7.24 8.44 9.76 11.67 14.47 1.40 1986 3.20 4.24 5.33 6.40 7.49 8.73 10.00 12.00 15.21 1.35 1987a 3.46 4.69 5.77 6.83 8.00 9.31 10.63 12.83 16.35 1.30 1987 3.50 4.69 5.77 6.73 7.87 9.24 10.63 12.71 16.03 1.30 1988 3.74 4.88 5.99 7.14 8.39 9.59 11.51 13.43 16.79 1.24 1989 3.83 4.98 6.23 7.48 8.63 10.20 11.98 14.38 18.33 1.18 1990 3.91 5.23 6.39 7.68 9.21 10.59 12.48 14.97 19.19 1.12 1991 4.25 5.57 6.79 8.25 9.64 11.38 13.23 15.69 20.05 1.08 1992 4.32 5.77 7.08 8.65 10.09 11.66 13.45 16.34 20.82 1.04 1993 4.15 5.58 7.11 8.62 10.02 11.92 14.11 17.24 22.03 1.01 1994 4.25 5.77 7.21 8.88 10.58 12.31 14.42 17.31 22.70 1.00 SOURCE: Authors’ calculations from the March CPS. 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTES: Hourly wages not calculated before 1975 because earlier CPSs did not ask respondents about their hours of work in a usual week in the previous year (annual earnings refers to earnings in the previous year). Hourly wages not calculated for the Census because the 1970 Census also did not ask about weekly hours of work in the previous year. 124 Table D.10 Deciles of Nominal Hourly Wages Among Female Workers, United States Price Year 10% 20% 30% 40% 50% 60% 70% 80% 90% Index 1975 1.54 2.00 2.38 2.70 3.08 3.50 4.04 4.80 5.87 2.91 1976 1.68 2.20 2.50 2.89 3.33 3.83 4.32 5.05 6.30 2.75 1977 1.78 2.34 2.71 3.13 3.58 4.04 4.68 5.48 6.84 2.58 1978 1.92 2.50 2.92 3.36 3.84 4.32 4.95 5.76 7.24 2.42 1979 2.22 2.88 3.30 3.75 4.26 4.80 5.50 6.44 8.17 2.21 1980 2.41 3.10 3.59 4.09 4.70 5.29 6.01 7.15 8.95 1.98 1981 2.56 3.35 3.85 4.44 5.01 5.77 6.67 7.74 9.77 1.81 1982 2.75 3.50 4.11 4.82 5.41 6.27 7.23 8.56 10.63 1.71 1983 2.86 3.61 4.33 5.01 5.78 6.59 7.70 9.06 11.53 1.64 1984 2.89 3.76 4.47 5.19 6.01 6.91 8.09 9.64 12.05 1.57 1985 3.01 3.85 4.70 5.45 6.26 7.23 8.56 10.12 12.77 1.52 1986 3.00 3.90 4.80 5.67 6.54 7.68 8.89 10.56 13.44 1.49 1987a 3.06 4.01 4.81 5.71 6.68 7.72 9.13 10.72 13.78 1.44 1987 3.19 4.15 4.98 5.92 6.95 8.00 9.49 11.21 14.37 1.44 1988 3.30 4.32 5.28 6.24 7.20 8.50 9.98 11.99 14.97 1.38 1989 3.40 4.59 5.58 6.61 7.72 8.93 10.55 12.55 16.16 1.32 1990 3.65 4.79 5.75 6.79 7.97 9.34 10.95 13.08 16.77 1.25 1991 3.85 5.01 6.01 7.16 8.35 9.63 11.56 13.76 17.53 1.20 1992 3.99 5.14 6.27 7.37 8.68 10.13 12.06 14.47 18.53 1.16 1993 4.02 5.24 6.40 7.73 9.01 10.51 12.31 15.02 19.31 1.13 1994 4.17 5.40 6.50 7.69 9.13 10.72 12.60 15.38 20.13 1.10 SOURCE: Authors’ calculations from the March CPS. 1987a is based on the original release of the March 1988 CPS, which used the processing system of previous years. NOTES: Hourly wages not calculated before 1975 because earlier CPSs did not ask respondents about their hours of work in a usual week in the previous year (annual earnings refers to earnings in the previous year). Hourly wages not calculated for the Census because the 1970 Census also did not ask about weekly hours of work in the previous year. 125 Table D.11 Regional Trends in the Coefficient of Variation for Real Annual Earnings Among Males, 1969–1994 CV (Rank) Percentage Change in CV (Rank) Region 1969 1979 1989 1994 1969–1979 1979–1989 1989–1994 California 0.56 0.65 0.75 0.78 15 14 4 (4) (2) (2) (1) (3) (6) (8) New England 0.51 0.63 0.65 0.69 25 2 7 (9) (4) (10) (10) (1) (10) (7) Mid Atlantic 0.53 0.58 0.67 0.72 9 16 8 (7) (9) (8) (7) (6) (3) (4) E. N. Central 0.50 0.54 0.65 0.70 8 19 8 (10) (10) (9) (8) (8) (1) (3) W. N. Central 0.52 0.60 0.70 0.69 16 15 0 (8) (8) (5) (9) (2) (4) (10) S. Atlantic 0.62 0.66 0.70 0.76 6 6 9 (1) (1) (4) (4) (9) (9) (1) E. S. Central 0.62 0.62 0.69 0.74 0 11 7 (2) (6) (6) (5) (10) (8) (5) W. S. Central 0.60 0.65 0.77 0.77 9 19 0 (3) (3) (1) (2) (7) (2) (9) Mountain 0.54 0.61 0.68 0.73 13 11 8 (6) (7) (7) (6) (4) (7) (2) Pacific 0.56 0.62 0.71 0.76 11 15 7 (5) (5) (3) (3) (5) (5) (6) SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this table are not sensitive to the consumer price index. 126 Table D.12 Regional Trends in the Coefficient of Variation for Real Annual Earnings Among Females, 1969–1994 Region California New England Mid Atlantic E. N. Central W. N. Central S. Atlantic E. S. Central W. S. Central Mountain Pacific 1969 0.78 (3) 0.71 (10) 0.72 (9) 0.76 (5) 0.76 (6) 0.77 (4) 0.75 (8) 0.81 (1) 0.76 (7) 0.80 (2) CV (Rank) 1979 1989 0.76 0.80 (3) (2) 0.73 0.76 (5) (7) 0.72 0.78 (8) (6) 0.72 0.78 (7) (5) 0.71 0.80 (10) (3) 0.72 0.76 (9) (8) 0.73 0.73 (6) (10) 0.75 0.81 (4) (1) 0.79 0.74 (1) (9) 0.76 0.79 (2) (4) 1994 0.81 (7) 0.76 (10) 0.80 (8) 0.84 (1) 0.80 (9) 0.81 (5) 0.81 (3) 0.81 (4) 0.84 (2) 0.81 (6) Percentage Change in CV (Rank) 1969–1979 1979–1989 1989–1994 –3 4 1 (4) (6) (7) 240 (2) (7) (10) 083 (3) (2) (5) –5 8 8 (7) (3) (3) –7 12 0 (10) (1) (9) –6 6 7 (8) (5) (4) –3 1 11 (5) (9) (2) –6 7 1 (9) (4) (8) 4 –6 12 (1) (10) (1) –4 4 2 (6) (8) (6) SOURCE: Based on authors’ calculations from the March CPS. NOTES: See the notes to Figure 3.1 for sample criteria and the calculation of annual earnings. Statistics reported in this figure are not sensitive to the consumer price index. 127 Table D.13 State Rankings for Adjusted Household Income Inequality Based on the Coefficient of Variation: Census State Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma 1969 CV Rank 0.71 4 0.65 15 0.65 15 0.70 5 0.63 21 0.62 23 0.57 42 0.60 28 0.69 6 0.69 6 0.59 34 0.60 28 0.60 28 0.56 47 0.60 28 0.62 23 0.68 9 0.73 2 0.57 42 0.61 26 0.57 42 0.57 42 0.59 34 0.78 1 0.65 15 0.61 26 0.63 21 0.59 34 0.55 50 0.58 38 0.72 3 0.65 15 0.64 20 0.62 23 0.57 42 0.65 15 1989 CV Rank 0.71 6 0.62 37 0.70 8 0.70 8 0.71 6 0.66 20 0.62 37 0.61 43 0.69 13 0.70 8 0.61 43 0.64 27 0.67 16 0.61 43 0.62 37 0.65 22 0.70 8 0.76 1 0.60 46 0.63 32 0.63 32 0.65 22 0.62 37 0.76 1 0.67 16 0.65 22 0.62 37 0.64 27 0.57 50 0.63 32 0.74 4 0.73 5 0.66 20 0.64 27 0.64 27 0.70 8 Change Percent Rank 0 43 –4 49 8 12 1 42 13 2 6 21 7 15 3 32 0 43 2 38 4 28 7 15 13 2 99 3 32 5 25 3 32 5 25 6 21 3 32 10 7 15 1 6 21 –3 48 3 32 7 15 –1 46 8 12 4 28 7 15 2 38 12 4 2 38 2 38 11 6 7 15 128 Table D.13—continued 1969 1989 Change State CV Rank CV Rank Percent Rank Oregon Pennsylvania Rhode Island South Carolina South Dakota Tennessee Texas Utah Vermont Virginia Washington West Virginia Wisconsin Wyoming 0.60 28 0.58 38 0.60 28 0.69 6 0.68 9 0.67 12 0.68 9 0.56 47 0.58 38 0.67 12 0.58 38 0.66 14 0.56 47 0.59 34 0.65 22 0.65 22 0.62 37 0.67 16 0.64 27 0.69 13 0.75 3 0.60 46 0.60 46 0.67 16 0.63 32 0.69 13 0.60 46 0.63 32 9 12 3 –2 –6 4 10 8 4 0 9 5 7 6 19 4 32 46 50 28 7 12 28 43 9 25 15 21 SOURCE: Authors’ calculations from the 1970 and 1990 Census. NOTES: Ties in rank are reported with the highest common rank. For example, if two states are tied for first, both states are reported with rank 1 and the next highest state is reported with rank 3. The CV values for California reported in this table are lower than reported in the text due to top-coding differences. A greater amount of topcoding was required to achieve consistency across all states than was required for consistency between California and the nation as a whole. For consistent comparison across states, adjusted household income was top-coded at 4 percent in each state. The top-code used in the text figures was 2.42 percent. The CV values reported in the text are more accurate for California; the values above are more accurate for comparison with other states. 129 Table D.14 State Rankings for Male Annual Earnings Inequality Based on the Coefficient of Variation: Census State Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma 1969 CV Rank 0.60 7 0.57 15 0.56 18 0.61 6 0.55 21 0.54 26 0.53 30 0.55 21 0.63 2 0.62 3 0.53 30 0.53 30 0.51 40 0.48 49 0.52 35 0.55 21 0.57 15 0.60 7 0.52 35 0.56 18 0.52 35 0.50 43 0.52 35 0.67 1 0.56 18 0.53 30 0.55 21 0.54 26 0.49 46 0.54 26 0.62 3 0.55 21 0.59 10 0.58 13 0.48 49 0.57 15 1989 CV Rank 0.63 15 0.63 15 0.67 3 0.64 8 0.67 3 0.64 8 0.59 35 0.58 42 0.66 5 0.63 15 0.60 30 0.61 25 0.61 25 0.58 42 0.58 42 0.60 30 0.64 8 0.66 5 0.56 47 0.59 35 0.58 42 0.60 30 0.59 35 0.66 5 0.63 15 0.64 8 0.60 30 0.62 22 0.56 47 0.61 25 0.69 1 0.63 15 0.61 25 0.64 8 0.59 35 0.64 8 Change Percent Rank 5 40 10 31 20 6 4 43 22 1 18 12 11 30 5 40 4 43 2 48 13 22 16 14 19 9 22 1 12 24 9 33 12 24 10 31 8 36 5 40 12 24 21 4 15 16 –3 50 12 24 20 6 8 36 16 14 14 14 13 22 12 24 15 16 3 47 9 33 22 1 14 14 130 Table D.14—continued 1969 1989 Change State CV Rank CV Rank Percent Rank Oregon 0.51 40 Pennsylvania 0.50 43 Rhode Island 0.52 35 South Carolina 0.58 13 South Dakota 0.59 10 Tennessee 0.59 10 Texas 0.60 7 Utah 0.50 43 Vermont 0.51 40 Virginia 0.62 3 Washington 0.49 46 West Virginia 0.54 26 Wisconsin 0.49 46 Wyoming 0.53 30 0.62 22 0.59 35 0.57 46 0.61 25 0.62 22 0.63 15 0.68 2 0.60 30 0.56 47 0.63 15 0.59 35 0.64 8 0.56 47 0.59 35 20 19 9 4 4 8 14 19 8 2 21 18 15 12 6 9 33 43 43 36 19 9 36 48 4 12 16 24 SOURCE: Authors’ calculations from the 1970 and 1990 Census. NOTES: Ties in rank are reported with the highest common rank. For example, if two states are tied for first, both states are reported with rank 1 and the next highest state is reported with rank 3. The CV values for California reported in this table are lower than reported in the text due to top-coding differences. A greater amount of top-coding was required to achieve consistency across all states than was required for consistency between California and the nation as a whole. For consistent comparison across states, adjusted household income was top-coded at 4 percent in each state. The top-code used in the text figures was 0.93 percent. The CV values reported in the text are more accurate for California; the values above are more accurate for comparison with other states. 131 Table D.15 State Rankings for Female Annual Earnings Inequality Based on the Coefficient of Variation: Census State Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma 1969 CV Rank 0.73 18 0.79 5 0.74 15 0.69 32 0.70 29 0.73 18 0.65 46 0.69 32 0.72 22 0.70 29 0.67 39 0.79 5 0.65 46 0.68 36 0.75 10 0.73 18 0.69 32 0.77 7 0.71 25 0.69 32 0.67 39 0.71 25 0.72 22 0.76 9 0.68 36 0.77 7 0.75 10 0.66 42 0.66 42 0.66 42 0.80 2 0.65 46 0.65 46 0.81 1 0.70 29 0.75 10 1989 CV Rank 0.70 13 0.69 17 0.70 13 0.68 24 0.69 17 0.69 17 0.63 45 0.65 39 0.67 31 0.68 24 0.62 49 0.74 2 0.67 31 0.69 17 0.67 31 0.68 24 0.71 9 0.72 5 0.65 39 0.63 45 0.63 45 0.73 4 0.66 37 0.70 13 0.69 17 0.72 5 0.68 24 0.66 37 0.62 49 0.65 39 0.74 2 0.68 24 0.63 45 0.70 13 0.69 17 0.71 9 Change Percent Rank –5 23 –13 48 –7 37 –2 12 –1 9 –6 29 –2 12 –6 29 –7 37 –3 15 –6 29 –6 29 41 16 –10 46 –7 37 33 –6 29 –8 42 –9 44 –5 23 33 –8 42 –7 37 16 –6 29 –9 44 –1 9 –5 23 –1 9 –6 29 41 –3 15 –14 49 –2 12 –4 21 132 Table D.15—continued 1969 1989 Change State CV Rank CV Rank Percent Rank Oregon 0.75 10 Pennsylvania 0.66 42 Rhode Island 0.67 39 South Carolina 0.65 46 South Dakota 0.80 2 Tennessee 0.68 36 Texas 0.73 18 Utah 0.74 15 Vermont 0.72 22 Virginia 0.71 25 Washington 0.74 15 West Virginia 0.75 10 Wisconsin 0.71 25 Wyoming 0.80 2 0.71 9 0.68 24 0.64 43 0.65 39 0.67 31 0.67 31 0.71 9 0.72 5 0.64 43 0.68 24 0.69 17 0.72 5 0.67 31 0.75 1 –5 2 –5 0 –17 –3 –3 –3 –11 –3 –7 –4 –5 –6 23 5 23 8 50 15 15 15 47 15 37 21 23 29 SOURCE: Authors’ calculations from the 1970 and 1990 Census. NOTES: Ties in rank are reported with the highest common rank. For example, if two states are tied for first, both states are reported with rank 1 and the next highest state is reported with rank 3. The CV values for California reported in this table are lower than reported in the text due to top-coding differences. A greater amount of top-coding was required to achieve consistency across all states than was required for consistency between California and the nation as a whole. For consistent comparison across states, adjusted household income was top-coded at 4 percent in each state. The top-code used in the text figures was 0.13 percent. The CV values reported in the text are more accurate for California; the values above are more accurate for comparison with other states. 133 Bibliography Aaron, Henry (1978), Politics and the Professors: The Great Society in Perspective, Brookings Institution, Washington, D.C. Blinder, Alan (1980), “The Levels and Distribution of Economic WellBeing: Introduction and Preview,” in Martin Feldstein (ed.), The American Economy in Transition, University of Chicago Press, Chicago. Borjas, George, Richard Freeman, and Larry Katz (1992), “On the Labor Market Effects of Immigration and Trade,” in George Borjas and Richard Freeman (eds.), Immigration and the Work Force: Economic Consequences for the United States and Source Areas, University of Chicago Press, Chicago. Burtless, Gary (1996), “Trends in the Level and Distribution of U.S. Living Standards, 1973–1993,” unpublished manuscript, Brookings Institution, Washington, D.C. Butcher, Kristine, and David Card (1991), “Immigration and Wages: Evidence from the 1980s,” American Economic Review, Vol. 81, No. 2, May 1991, pp. 292–296. California Department of Finance (1995), Statistical Abstract 1995, Sacramento, California. 135 Cancian, Maria, Sheldon Danziger, and Peter Gottschalk (1993), “Working Wives and Family Income Inequality Among Married Couples,” in Sheldon Danziger and Peter Gottschalk (eds.), Uneven Tides: The Rising Inequality in America, Russell Sage Foundation, New York, pp. 195–221. Cassidy, J. (1995), “Who Killed the Middle Class?” The New Yorker, October 16, pp. 113–124. Center for the Continuing Study of the California Economy (1994), California Economic Growth—1994 Edition, Palo Alto, California. Chamberlain, Jay D., and Phil Spillberg (1991), “Trends in the Distribution of Income and Personal Income Tax Burden for California Taxpayers,” Occasional Papers Series, California Franchise Tax Board, FTB/OPS 91-01, Sacramento, California. Cutler, David M., and Larry F. Katz (1991), “Macroeconomic Performance and the Disadvantaged,” Brookings Papers on Economic Activity, Vol. 2, pp. 1–61. Cutler, David M., and Larry F. Katz (1992), “Rising Inequality? Changes in the Distribution of Income and Consumption in the 1980s,” American Economic Review, Vol. 82, No. 2, pp. 546–551. Danziger, Sheldon, and Peter Gottschalk (1995), America Unequal, Russell Sage Foundation, New York. Freeman, Richard (1993), “How Much Has Deunionization Contributed to the Rise in Male Earnings Inequality?” in Sheldon Danziger and Peter Gottschalk (eds.), Uneven Tides: The Rising Inequality in America, Russell Sage Foundation, New York, pp. 133– 163. Gottschalk, Peter, and Robert Moffitt (1994), “The Growth of Earnings Instability in the U.S. Labor Market,” Brookings Papers on Economic Activity, Vol. 2, pp. 217–272. Gramlich, Edward M., Robert Kasten, and Frank Sammartino (1993), “Growing Inequality in the 1980s: The Role of Federal Taxes and Cash Transfers,” in Sheldon Danziger and Peter Gottschalk (eds.), 136 Uneven Tides: Rising Inequality in America, Russell Sage Foundation, New York, pp. 225–249. Hungerford, Thomas (1993), “U.S. Income Mobility in the Seventies and Eighties,” Review of Income and Wealth, Vol. 39, No. 4, pp. 403– 417. Husted, Thomas (1991), “Changes in State Income Inequality from 1981 to 1987,” The Review of Regional Studies, Vol. 21, No. 3, pp. 249–260. Juhn, Chinhui, Kevin M. Murphy, and Brooks Pierce (1993), “Wage Inequality and the Rise in the Return to Skill,” Journal of Political Economy, Vol. 101, June, pp. 410–442. Karoly, Lynn (1993), “The Trend in Inequality Among Families, Individuals, and Workers in the United States: A Twenty-Five Year Perspective,” in Sheldon Danziger and Peter Gottschalk (eds.), Uneven Tides: Rising Inequality in America, Russell Sage Foundation, New York, pp. 19–97. Karoly, Lynn (1995), unpublished testimony prepared for the California Legislature Assembly Committee on Labor and Employment, October 23. Karoly, Lynn, and Jacob Alex Klerman (1994), “Using Regional Data to Reexamine the Contribution of Demographic and Sectoral Changes to Increasing U.S. Wage Inequality,” in J. H. Bergstrand, T. F. Cosimano, J. W. Houck, and R. G. Sheehan (eds.), The Changing Distribution of Income in an Open U.S. Economy, North-Holland, Amsterdam, pp. 183–216. Karoly, Lynn, and Gary Burtless (1995), “Demographic Change, Rising Earnings Inequality, and the Distribution of Personal Well-Being, 1959–1989,” Demography, Vol. 32, No. 3, pp. 379–405. Krueger, Alan (1993), “How Computers Have Changed the Wage Structure: Evidence from Microdata, 1984–89,” Quarterly Journal of Economics, Vol. 108, February, pp. 33–60. Levy, Frank (1987), Dollars and Dreams: The Changing American Income Distribution, Russell Sage Foundation, New York. 137 Levy, Frank, and Richard J. Murnane (1992), “U.S. Earnings Levels and Earnings Inequality: A Review of Recent Trends and Proposed Explanations,” Journal of Economic Literature, Vol. 30, September, pp. 1333–1381. McMahon, Walter (1991), “Geographical Cost of Living Differences: An Update,” AREUEA Journal, Vol. 19, No. 3, pp. 426–450. Pechman, Joseph A. (1990), “The Future of the Income Tax,” American Economic Review, Vol. 80, No. 1, pp. 1–20. Topel, Robert (1994), “Regional Labor Markets and the Determinants of Wage Inequality,” American Economic Review, Vol. 84, No. 2, pp. 17– 22. U.S. Bureau of the Census (1990), Current Population Reports, Series P-60, Washington, D.C. U.S. Bureau of the Census (1994), Statistical Abstract of the United States, Washington, D.C. U.S. Bureau of the Census (1996), Current Population Reports, P-60-191, Washington, D.C. U.S. Bureau of Labor Statistics (1982), “Family Budgets for 1981: Final Report,” Monthly Labor Review, U.S. Department of Labor, Washington, D.C. U.S. Department of Commerce and the U.S. Bureau of the Census (1978), The Current Population Survey: Design and Methodology, January, Washington, D.C. U.S. House of Representatives, Committee on Ways and Means (1989), “Trends in Family Income and Income Inequality,” Background Material and Data on Programs Within the Jurisdiction of the Committee on Ways and Means, U.S. Government Printing Office, Washington, D.C. Williamson, Jeffrey, and Peter Lindert (1980), American Inequality: A Macroeconomic History, Academic Press, New York. 138 About the Authors DEBORAH S. REED Deborah Reed, a research fellow at the Public Policy Institute of California, is a specialist in labor economics and development resources. Her research interests include labor markets, public policy, and poverty in the United States and Brazil. Reed recently completed a post-doctoral fellowship at the Population Studies Center at the University of Michigan. A recipient of fellowships from the Mellon Foundation and Yale University, she has served as a consultant for the World Bank in addition to her teaching and research activities. Reed received an A.B. (1989) in economics from the University of California, Berkeley, and an M.A. (1990) in economics from Yale University. She will receive her Ph.D. in economics from Yale University in 1996. MELISSA GLENN HABER Melissa Glenn Haber is a research assistant at the Public Policy Institute of California. Before joining PPIC, Haber was a research associate at the Family Welfare Research Group where she developed a cost-benefit model to evaluate a state-wide teen pregnancy prevention program funded by the California Office of Family Planning. While at the Research Group, she also produced a report summarizing the cost, causes, and incidence of adolescent childbearing in the United States and in California. Haber received a B.A. (1991) in comparative religion from Harvard University and a Master’s degree in Public Policy (1995) from the University of California, Berkeley. LAURA A. MAMEESH Laura Mameesh is a research assistant at the Public Policy Institute of California. She has worked for Los Angeles Mayor Richard Riordan on a business tax relief policy aimed at retaining targeted industries in Los Angeles. Her analysis included calculating the cost to the city of providing tax relief as well as researching the cost of doing business in Los Angeles compared with competitor cities. Prior to her work in Los Angeles, Mameesh worked at the Law & Economics Consulting Group on anti-trust cases. Mameesh received her B.A. (1990) in economics from Mills College and is near completion of a Master’s degree in Public Policy from the University of Southern California." ["post_date_gmt"]=> string(19) "2017-05-20 09:35:02" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(8) "r_796drr" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2017-05-20 02:35:02" ["post_modified_gmt"]=> string(19) "2017-05-20 09:35:02" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(50) "http://148.62.4.17/wp-content/uploads/R_796DRR.pdf" ["menu_order"]=> int(0) ["post_mime_type"]=> string(15) "application/pdf" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" ["status"]=> string(7) "inherit" ["attachment_authors"]=> bool(false) }