Donate
Independent, objective, nonpartisan research

Higher Education as a Driver of Economic Mobility, Technical Appendix

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(20) "1218hjr-appendix.pdf" ["wpmf_size"]=> string(6) "318636" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(12157) "Higher Education as a Driver of Economic Mobility Technical Appendix Hans Johnson, Marisol Cuellar Mejia, and Sarah Bohn with research support from Sergio Sanchez Supported with funding from the College Futures Foundation and the Sutton Family Fund Wage Premiums Methods For our analysis of wage premiums we rely on Mincer’s human-capital wage equations. Specifically, wage premiums were estimated using regressions of log of real annual wages on education categorical variables (less than high school, some college, associate degree, bachelor’s degree, master’s degree, professional degree, and doctorate, with high school omitted), age, age squared, categorical variables for race/ethnicity (Hispanic, African American, Asian, and other, with white omitted), and dummy variables for marital status and citizenship status. We ran separate regressions by year and gender, limiting our sample to full-time and year-round workers 25-to64-years-old employed in the public or private sector. Workers in the military and institutionalized or unincorporated self-employed workers are excluded. Annual wages below $1,700 are also dropped. For comparative purposes, we ran all our specifications for both California and the rest of the nation. Data to estimate these equations are from the public use files of the American Community Survey and the decennial censuses. Regression-adjusted wage premiums take account of the differing age and racial/ethnic distribution of each educational group; consequently, it is a better measure than a “wage premium” computed by simply dividing college-graduate wage by the high-school-only wage, for example. Although we controlled for personal characteristics to make comparisons between individuals who are as similar as can be observed, we do not have quasi-experimental variation concerning who goes to college. Thus, caution is necessary in making causal interpretations of the estimated wage premiums, since the potential problem of selection bias from nonrandom sorting on unobservables remains. A critical question is whether the wage gains enjoyed by college graduates would have occurred for those individuals even if they did not attend college. One argument in the debate over the causality between schooling and earnings is that colleges select individuals who would have succeeded in the labor market even if they did not attend college (known as the selection effect). The other argument is that the skills and knowledge acquired in college lead to better labor market outcomes, including higher wages. The best research suggests that the college wage premium, as estimated in our standard wage models, is an accurate measure of the causal effect of college. In a thorough review and analysis of the extensive literature on wages and education, David Card (1999) concluded that the selection effect does not exceed 10 percent of the estimated schooling coefficient. That estimate is derived from studies of twins with different educational attainment. Other approaches, including instrumental variable (IV) estimates, are often higher than classic ordinary least squares (OLS) estimates from standard human capital earnings functions. Although it is unclear to what extend this is due to measurement error or inadequate instrumentation, Card notes that one possibility is that OLS approaches actually understate the causal value of a degree (see Trostel et al. 2002). The Role of Colleges in Intergenerational Mobility Research from the Equality of Opportunity Project has found that some colleges are more successful than others in producing high earners from economically diverse student bodies. Specifically, they find that a group of midtier non-selective institutions are important contributors to economic mobility, both because their graduates earn incomes similar to those who emerge from more selective universities and because these mid-tier institutions enroll many students from low-income families. The good news is that many California institutions top the list of colleges that have the highest upward mobility rates: California State University at Los Angeles (CSULA), Cal Poly-Pomona, and Glendale Community College are in the top 10 out of over 2,400 institutions nationwide. Also, five UC campuses are ranked in the top 50 colleges with the highest economic mobility: Irvine (#12), Riverside (#19), Los Angeles (#24), Berkeley (#39) and San Diego (#41). For almost all UC campuses, more than half of alumni whose parents come from the bottom income quintile move to the top income quintile in adulthood. On average, private nonprofit institutions and community colleges have lower mobility rates relative to CSU and UC PPIC.ORG Technical Appendix Higher Education as a Driver of Economic Mobility 2 but for very different reasons. In the case of the private nonprofit colleges this result is driven by low access rates (measured as the share of students with parents in the bottom quintile of the income distribution) while in the case of the community colleges it is driven by low success rates (measured as the share of those students who reach to the top quintile of the income distribution by age 34). Additional Tables and Charts The tables and charts below provide additional, detailed information about the distribution of wages by educational level and major, wage premiums over time, employment and other outcomes by educational level, and shares of students enrolled in two-year colleges by income level. TABLE A1 Median wages by educational level 1990 2000 2010 2016 Growth between 1990 and 2016 Less than high school 28,916 27,933 25,358 26,000 -10 High school 42,355 41,899 38,589 36,000 -15 Some college 48,214 48,883 48,512 45,000 -7 Associate degree 51,786 54,469 55,127 50,000 -3 Bachelor's degree 62,500 69,832 71,665 71,000 14 Graduate degree 80,357 87,989 93,716 96,000 19 SOURCE: Authors’ calculations based on the 1990 and 2000 Decennial Census and 2010 and 2016 American Community Survey one-year estimates. NOTE: Full-time, year-round workers ages 25 to 64. TABLE A2 Wage premium relative to workers with a high school degree (percent) 1990 2000 2010 2016 Change 2000–2016 (percentage points) California Less than high school -19.3 -22.3 -25.1 -22.4 0 Some college 13.0 16.5 19.2 20.6 4 Associate degree 18.3 23.5 29.4 28.7 5 Bachelor's degree 38.4 50.1 57.7 62.1 12 Graduate degree 58.5 72.2 84.7 90.7 19 BA plus 45.3 57.9 68.2 73.1 15 Rest of US Less than high school -19.8 -17.7 -19.6 -16.4 Some college 15.2 16.2 17.4 17.0 1 Associate degree 21.7 23.5 27.3 24.6 1 Bachelor's degree 44.1 49.8 56.2 57.4 8 Graduate degree 62.1 70.2 83.1 83.3 13 BA plus 50.5 57.1 66.6 67.6 11 SOURCE: Authors’ calculations based on the 1990 and 2000 Decennial Census and 2010 and 2016 American Community Survey one-year estimates. NOTE: Full-time, year-round workers ages 25 to 64. These estimates are regression adjusted for age, race/ethnicity, gender and citizenship. PPIC.ORG Technical Appendix Higher Education as a Driver of Economic Mobility 3 TABLE A3 Wage distribution by educational level, 2016 Mean 10th percentile 25th percentile Median Less than high school 32,503 14,400 19,600 26,000 High school 44,148 18,000 25,000 36,000 some college 55,982 20,000 30,000 45,000 Associate degree 60,870 23,000 34,000 50,000 Bachelor's degree 90,526 30,000 45,000 71,000 Graduate degree 126,535 42,000 65,000 96,000 SOURCE: Authors’ calculation based on 2016 American Community Survey one-year estimates. NOTE: Full-time year-round workers ages 25 to 64. 75th percentile 39,000 52,000 70,000 77,000 107,000 149,000 90th percentile 55,000 78,000 100,000 110,000 155,000 225,000 TABLE A4 75th/25th wage differential 1990 2016 25th percentile 75th percentile 75th/25th gap 25th percentile 75th percentile 75th/25th gap Less than high school 19,643 44,643 25,000 19,600 39,000 19,400 High school some college 28,571 34,059 57,143 66,607 28,571 32,548 25,000 30,000 52,000 70,000 27,000 40,000 Associate degree Bachelor's degree 35,714 44,643 71,429 86,071 35,714 41,429 34,000 45,000 77,000 107,000 43,000 62,000 Graduate degree 57,143 114,286 57,143 65,000 149,000 84,000 SOURCE: Authors’ calculations based on the 1990 Decennial Census and 2016 American Community Survey one-year estimates. NOTE: Full-time year-round workers ages 25 to 64. Increase in the 75th/ 25th gap -22 -6 23 20 50 47 PPIC.ORG Technical Appendix Higher Education as a Driver of Economic Mobility 4 FIGURE A1 Wage distribution within majors 10th 25th 50th 75th 90th Engineering Computers Business Science, medicine Social Science, Law Liberal Arts Other Education High school $0 $20,000 $40,000 $60,000 $80,000 $100,000 $120,000 $140,000 $160,000 Annual wages SOURCE: Authors’ calculations based on 2016 American Community Survey one-year estimates. NOTE: Full-time, year-round workers ages 25 to 64.Sample restricted to workers with a bachelor’s degree or less. Predicted wages from OLS regressions. TABLE A5 Variable definitions and sources for Figures 3 and 4 in the report Poverty rate Social safety net recipients Wage premium Unemployment rate Labor force participation Full-time employment Health Insurance through Employment Measure The California Poverty Measure (CPM) accounts for the cost of living and a range of family needs and resources, including social safety net benefits. Refers only to CalWORKs, General Assistance, CalFresh, Supplemental Security Income, and federal housing subsidies. Regression-adjusted wage premium relative to high school graduates Adults age 25–64, not enrolled in college, excludes armed forces Adults age 25-64 Workers age 25-64 working 35 hours or more Full-time year-round workers age 25-64 Retirement Plan through Employment Homeownership rate Marital status Full-time year-round workers age 25-64 Educational attainment of the head/householder, age 25 and older Adults age 25-64 Source Using the CPM, which is based on ACS data from IPUMS, but available only at PPIC Using the CPM, which is based on ACS data from IPUMS, but available only at PPIC Authors’ calculations based on 2016 ACS 1-year estimates Authors’ calculations based on 2016 ACS 1-year estimates Authors’ calculations based on 2016 ACS 1-year estimates Authors’ calculations based on 2016 ACS 1-year estimates Authors’ calculations based on 2016 ACS 1-year estimates Authors' calculations using 2017 CPS Authors’ calculations based on 2016 ACS 1-year estimates Authors’ calculations based on 2016 ACS 1-year estimates PPIC.ORG Technical Appendix Higher Education as a Driver of Economic Mobility 5 TABLE A6 Measures of well-being by educational attainment. Less than high school Poverty 35 Welfare recipients 44 Unemployment rate 8 Labor force participation 66 Health insurance through employment 44 Retirement plan through employment 17 Full-time employment 52 Homeownership rate 39 Share married SOURCES: See Table A5. NOTE: All values are expressed in percentage terms. 55 High school 21 33 7 73 66 32 58 49 50 Some college 16 24 6 78 76 40 63 53 51 Associate degree 14 19 4 79 77 44 64 60 55 Bachelor's degree 9 10 4 85 81 46 73 61 58 Advanced degree 7 7 3 88 87 53 77 68 66 FIGURE A2 Students from low-income families are more likely to attend a community college Share of HS graduates attending two-year colleges 70 60 57 50 44 40 30 20 10 0 Under $30,000 62 43 63 39 57 33 44 29 California Rest of US 40 18 $30,000-49,999 $50,000-74,999 $75,000 - 99,999 $100,000 - 149,999 $150,000 and over Family income SOURCE: Author’s calculations based on October Current Population Survey 2007–2016. NOTE: Restricted to recent high school graduates. PPIC.ORG Technical Appendix Higher Education as a Driver of Economic Mobility 6 The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonpartisan research. Public Policy Institute of California 500 Washington Street, Suite 600 San Francisco, CA 94111 T: 415.291.4400 F: 415.291.4401 PPIC.ORG PPIC Sacramento Center Senator Office Building 1121 L Street, Suite 801 Sacramento, CA 95814 T: 916.440.1120 F: 916.440.1121" } ["___content":protected]=> string(171) "

Higher Education as a Driver of Economic Mobility, Technical Appendix

" ["_permalink":protected]=> string(100) "https://www.ppic.org/publication/higher-education-as-a-driver-of-economic-mobility/1218hjr-appendix/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(17576) ["ID"]=> int(17576) ["post_author"]=> string(1) "4" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2018-12-10 20:44:56" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(17389) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(69) "Higher Education as a Driver of Economic Mobility, Technical Appendix" ["post_type"]=> string(10) "attachment" ["slug"]=> string(16) "1218hjr-appendix" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(20) "1218hjr-appendix.pdf" ["wpmf_size"]=> string(6) "318636" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(12157) "Higher Education as a Driver of Economic Mobility Technical Appendix Hans Johnson, Marisol Cuellar Mejia, and Sarah Bohn with research support from Sergio Sanchez Supported with funding from the College Futures Foundation and the Sutton Family Fund Wage Premiums Methods For our analysis of wage premiums we rely on Mincer’s human-capital wage equations. Specifically, wage premiums were estimated using regressions of log of real annual wages on education categorical variables (less than high school, some college, associate degree, bachelor’s degree, master’s degree, professional degree, and doctorate, with high school omitted), age, age squared, categorical variables for race/ethnicity (Hispanic, African American, Asian, and other, with white omitted), and dummy variables for marital status and citizenship status. We ran separate regressions by year and gender, limiting our sample to full-time and year-round workers 25-to64-years-old employed in the public or private sector. Workers in the military and institutionalized or unincorporated self-employed workers are excluded. Annual wages below $1,700 are also dropped. For comparative purposes, we ran all our specifications for both California and the rest of the nation. Data to estimate these equations are from the public use files of the American Community Survey and the decennial censuses. Regression-adjusted wage premiums take account of the differing age and racial/ethnic distribution of each educational group; consequently, it is a better measure than a “wage premium” computed by simply dividing college-graduate wage by the high-school-only wage, for example. Although we controlled for personal characteristics to make comparisons between individuals who are as similar as can be observed, we do not have quasi-experimental variation concerning who goes to college. Thus, caution is necessary in making causal interpretations of the estimated wage premiums, since the potential problem of selection bias from nonrandom sorting on unobservables remains. A critical question is whether the wage gains enjoyed by college graduates would have occurred for those individuals even if they did not attend college. One argument in the debate over the causality between schooling and earnings is that colleges select individuals who would have succeeded in the labor market even if they did not attend college (known as the selection effect). The other argument is that the skills and knowledge acquired in college lead to better labor market outcomes, including higher wages. The best research suggests that the college wage premium, as estimated in our standard wage models, is an accurate measure of the causal effect of college. In a thorough review and analysis of the extensive literature on wages and education, David Card (1999) concluded that the selection effect does not exceed 10 percent of the estimated schooling coefficient. That estimate is derived from studies of twins with different educational attainment. Other approaches, including instrumental variable (IV) estimates, are often higher than classic ordinary least squares (OLS) estimates from standard human capital earnings functions. Although it is unclear to what extend this is due to measurement error or inadequate instrumentation, Card notes that one possibility is that OLS approaches actually understate the causal value of a degree (see Trostel et al. 2002). The Role of Colleges in Intergenerational Mobility Research from the Equality of Opportunity Project has found that some colleges are more successful than others in producing high earners from economically diverse student bodies. Specifically, they find that a group of midtier non-selective institutions are important contributors to economic mobility, both because their graduates earn incomes similar to those who emerge from more selective universities and because these mid-tier institutions enroll many students from low-income families. The good news is that many California institutions top the list of colleges that have the highest upward mobility rates: California State University at Los Angeles (CSULA), Cal Poly-Pomona, and Glendale Community College are in the top 10 out of over 2,400 institutions nationwide. Also, five UC campuses are ranked in the top 50 colleges with the highest economic mobility: Irvine (#12), Riverside (#19), Los Angeles (#24), Berkeley (#39) and San Diego (#41). For almost all UC campuses, more than half of alumni whose parents come from the bottom income quintile move to the top income quintile in adulthood. On average, private nonprofit institutions and community colleges have lower mobility rates relative to CSU and UC PPIC.ORG Technical Appendix Higher Education as a Driver of Economic Mobility 2 but for very different reasons. In the case of the private nonprofit colleges this result is driven by low access rates (measured as the share of students with parents in the bottom quintile of the income distribution) while in the case of the community colleges it is driven by low success rates (measured as the share of those students who reach to the top quintile of the income distribution by age 34). Additional Tables and Charts The tables and charts below provide additional, detailed information about the distribution of wages by educational level and major, wage premiums over time, employment and other outcomes by educational level, and shares of students enrolled in two-year colleges by income level. TABLE A1 Median wages by educational level 1990 2000 2010 2016 Growth between 1990 and 2016 Less than high school 28,916 27,933 25,358 26,000 -10 High school 42,355 41,899 38,589 36,000 -15 Some college 48,214 48,883 48,512 45,000 -7 Associate degree 51,786 54,469 55,127 50,000 -3 Bachelor's degree 62,500 69,832 71,665 71,000 14 Graduate degree 80,357 87,989 93,716 96,000 19 SOURCE: Authors’ calculations based on the 1990 and 2000 Decennial Census and 2010 and 2016 American Community Survey one-year estimates. NOTE: Full-time, year-round workers ages 25 to 64. TABLE A2 Wage premium relative to workers with a high school degree (percent) 1990 2000 2010 2016 Change 2000–2016 (percentage points) California Less than high school -19.3 -22.3 -25.1 -22.4 0 Some college 13.0 16.5 19.2 20.6 4 Associate degree 18.3 23.5 29.4 28.7 5 Bachelor's degree 38.4 50.1 57.7 62.1 12 Graduate degree 58.5 72.2 84.7 90.7 19 BA plus 45.3 57.9 68.2 73.1 15 Rest of US Less than high school -19.8 -17.7 -19.6 -16.4 Some college 15.2 16.2 17.4 17.0 1 Associate degree 21.7 23.5 27.3 24.6 1 Bachelor's degree 44.1 49.8 56.2 57.4 8 Graduate degree 62.1 70.2 83.1 83.3 13 BA plus 50.5 57.1 66.6 67.6 11 SOURCE: Authors’ calculations based on the 1990 and 2000 Decennial Census and 2010 and 2016 American Community Survey one-year estimates. NOTE: Full-time, year-round workers ages 25 to 64. These estimates are regression adjusted for age, race/ethnicity, gender and citizenship. PPIC.ORG Technical Appendix Higher Education as a Driver of Economic Mobility 3 TABLE A3 Wage distribution by educational level, 2016 Mean 10th percentile 25th percentile Median Less than high school 32,503 14,400 19,600 26,000 High school 44,148 18,000 25,000 36,000 some college 55,982 20,000 30,000 45,000 Associate degree 60,870 23,000 34,000 50,000 Bachelor's degree 90,526 30,000 45,000 71,000 Graduate degree 126,535 42,000 65,000 96,000 SOURCE: Authors’ calculation based on 2016 American Community Survey one-year estimates. NOTE: Full-time year-round workers ages 25 to 64. 75th percentile 39,000 52,000 70,000 77,000 107,000 149,000 90th percentile 55,000 78,000 100,000 110,000 155,000 225,000 TABLE A4 75th/25th wage differential 1990 2016 25th percentile 75th percentile 75th/25th gap 25th percentile 75th percentile 75th/25th gap Less than high school 19,643 44,643 25,000 19,600 39,000 19,400 High school some college 28,571 34,059 57,143 66,607 28,571 32,548 25,000 30,000 52,000 70,000 27,000 40,000 Associate degree Bachelor's degree 35,714 44,643 71,429 86,071 35,714 41,429 34,000 45,000 77,000 107,000 43,000 62,000 Graduate degree 57,143 114,286 57,143 65,000 149,000 84,000 SOURCE: Authors’ calculations based on the 1990 Decennial Census and 2016 American Community Survey one-year estimates. NOTE: Full-time year-round workers ages 25 to 64. Increase in the 75th/ 25th gap -22 -6 23 20 50 47 PPIC.ORG Technical Appendix Higher Education as a Driver of Economic Mobility 4 FIGURE A1 Wage distribution within majors 10th 25th 50th 75th 90th Engineering Computers Business Science, medicine Social Science, Law Liberal Arts Other Education High school $0 $20,000 $40,000 $60,000 $80,000 $100,000 $120,000 $140,000 $160,000 Annual wages SOURCE: Authors’ calculations based on 2016 American Community Survey one-year estimates. NOTE: Full-time, year-round workers ages 25 to 64.Sample restricted to workers with a bachelor’s degree or less. Predicted wages from OLS regressions. TABLE A5 Variable definitions and sources for Figures 3 and 4 in the report Poverty rate Social safety net recipients Wage premium Unemployment rate Labor force participation Full-time employment Health Insurance through Employment Measure The California Poverty Measure (CPM) accounts for the cost of living and a range of family needs and resources, including social safety net benefits. Refers only to CalWORKs, General Assistance, CalFresh, Supplemental Security Income, and federal housing subsidies. Regression-adjusted wage premium relative to high school graduates Adults age 25–64, not enrolled in college, excludes armed forces Adults age 25-64 Workers age 25-64 working 35 hours or more Full-time year-round workers age 25-64 Retirement Plan through Employment Homeownership rate Marital status Full-time year-round workers age 25-64 Educational attainment of the head/householder, age 25 and older Adults age 25-64 Source Using the CPM, which is based on ACS data from IPUMS, but available only at PPIC Using the CPM, which is based on ACS data from IPUMS, but available only at PPIC Authors’ calculations based on 2016 ACS 1-year estimates Authors’ calculations based on 2016 ACS 1-year estimates Authors’ calculations based on 2016 ACS 1-year estimates Authors’ calculations based on 2016 ACS 1-year estimates Authors’ calculations based on 2016 ACS 1-year estimates Authors' calculations using 2017 CPS Authors’ calculations based on 2016 ACS 1-year estimates Authors’ calculations based on 2016 ACS 1-year estimates PPIC.ORG Technical Appendix Higher Education as a Driver of Economic Mobility 5 TABLE A6 Measures of well-being by educational attainment. Less than high school Poverty 35 Welfare recipients 44 Unemployment rate 8 Labor force participation 66 Health insurance through employment 44 Retirement plan through employment 17 Full-time employment 52 Homeownership rate 39 Share married SOURCES: See Table A5. NOTE: All values are expressed in percentage terms. 55 High school 21 33 7 73 66 32 58 49 50 Some college 16 24 6 78 76 40 63 53 51 Associate degree 14 19 4 79 77 44 64 60 55 Bachelor's degree 9 10 4 85 81 46 73 61 58 Advanced degree 7 7 3 88 87 53 77 68 66 FIGURE A2 Students from low-income families are more likely to attend a community college Share of HS graduates attending two-year colleges 70 60 57 50 44 40 30 20 10 0 Under $30,000 62 43 63 39 57 33 44 29 California Rest of US 40 18 $30,000-49,999 $50,000-74,999 $75,000 - 99,999 $100,000 - 149,999 $150,000 and over Family income SOURCE: Author’s calculations based on October Current Population Survey 2007–2016. NOTE: Restricted to recent high school graduates. PPIC.ORG Technical Appendix Higher Education as a Driver of Economic Mobility 6 The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonpartisan research. Public Policy Institute of California 500 Washington Street, Suite 600 San Francisco, CA 94111 T: 415.291.4400 F: 415.291.4401 PPIC.ORG PPIC Sacramento Center Senator Office Building 1121 L Street, Suite 801 Sacramento, CA 95814 T: 916.440.1120 F: 916.440.1121" ["post_date_gmt"]=> string(19) "2018-12-11 04:44:56" ["comment_status"]=> string(6) "closed" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(16) "1218hjr-appendix" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2018-12-10 20:45:25" ["post_modified_gmt"]=> string(19) "2018-12-11 04:45:25" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(60) "https://www.ppic.org/wp-content/uploads/1218hjr-appendix.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) }