Donate
Independent, objective, nonpartisan research

R 905DRR

Authors

R 905DRR

Tagged with:

Publication PDFs

Database

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

View live version

object(Timber\Post)#3711 (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_905DRR.pdf" ["wpmf_size"]=> string(6) "284747" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(180714) "Educational Progress Across Immigrant Generations in California ••• Deborah Reed Laura E. Hill Christopher Jepsen Hans P. Johnson 2005 Library of Congress Cataloging-in-Publication Data Educational progress across immigrant generations in California / Deborah Reed ... [et al.]. p. cm. Includes bibliographical references. ISBN: 1-58213-091-4 1. Immigrants—Education—California. 2. Children of immigrants— Education—California. 3. Educational attainment—California. I. Reed, Deborah, 1967LC3732.C2E38 2005 371.826’912’09794—dc22 2005021485 Copyright © 2005 by Public Policy Institute of California All rights reserved San Francisco, CA Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source and the above copyright notice is included. PPIC does not take or support positions on any ballot measure or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Foreword Of all the political and economic uncertainties that beset California these days, perhaps none is more worrisome than the capabilities of our future labor force. The fiscal and governance challenges confronting the state will become even more complicated if, 30 years from now, a substantial part of the state’s working age population is mired in marginal jobs. In the United States, as well as California, the key to avoiding that scenario has always been the process of intergenerational improvement in education and skills. Californians have typically translated the American Dream into a belief that their children will have better jobs and be better off than they were at the same age. One interpretation of “better off” is that California’s youth today—many of whom are the children of immigrants—should be graduating from high school and college at greater rates than their parents did. The question is, are they? The authors of Educational Progress Across Immigrant Generations in California set out to answer that question with financial support from the William and Flora Hewlett Foundation. They conclude that there is good news. When the educational attainment of second and third generations is compared directly with that of their parents or their parents’ generation, the authors find strong intergenerational progress for all major immigrant groups. However, even by the third generation, Mexican Americans in California have not attained the educational levels that whites have. In other words, there is some progress but even by the third generation only 11 percent of Mexican American adults have earned a bachelor’s degree. In contrast, among third-and-latergeneration whites, more than a third have a bachelor’s degree. Even with this note of caution it is heartening to learn that the traditional vision of intergenerational improvement is still working in California. The authors’ findings have important implications for the state’s education policy. About 30 percent of California’s children are growing iii up in families where neither parent has completed high school. One consequence of this low educational attainment is that as many as 95 percent of these children might not earn a bachelor’s degree; the low educational attainment of parents makes it less likely that their children will attain high levels of education. Among these children at risk of low educational achievement, Mexican Americans make up a large percentage. Policymakers could target at-risk children for extra attention. In addition, community colleges, a resource important to the Latino population, could provide more targeted English language, remedial, and vocational resources. Such courses will become increasingly important for workforce training, especially for those who do not go on to complete a bachelor’s degree, at a time when the value of education and skills in the California economy will continue to grow. More broadly, better educational attainment by today’s children should mean better attainment in later generations. Looking into California’s future 20 years out, the authors of a recently released PPIC study, entitled California 2025: Taking on the Future, concluded that “If California’s youth do not get a college education, they face the prospect of low or no employment, lack of opportunities for high-paying jobs, and greater likelihood of depending on public health and social services. They will also generate lower tax revenues for supporting the state’s infrastructure and other service needs.” Even with all of the adaptive capabilities of California’s huge and complex economy, it does not and will not have the capacity to absorb a lowskilled, poorly educated population indefinitely. The policy issue is right before us—elected officials, the business community, the nonprofit sector, and the millions of state residents: Education has always been and will continue to be the key to success for this and future generations. It is time to address this critical component of the education issue— educating and training our immigrant youth to be highly productive members of the labor force in future years and decades. David W. Lyon President and CEO Public Policy Institute of California iv Summary More than half of all California youth ages 13 to 24 have a foreignborn parent. Because a large number of these immigrant parents have a limited education, lack of improvement in educational attainment from one generation to the next would have serious implications for the state economically as well as socially. Education is an important determinant of social and economic well-being, such as income, health, home ownership, and civic participation. The value of education in the California labor market has increased substantially in recent decades and projections suggest that workers without a college education will continue to see their earnings erode. In this context, it is particularly disconcerting that some recent studies have suggested that intergenerational progress may stall between the second and third generations. The concern for educational progress is particularly acute for Mexican Americans who, even by the third generation, have very low levels of educational attainment. This study examines educational attainment among California’s youth by race, ethnicity, and immigrant generation. It finds that intergenerational progress has not stalled but rather that second- and third-generation immigrants have made substantial educational progress when compared with their parents. Nevertheless, the low educational attainment of Mexican Americans remains a concern and the study examines the factors that influence educational attainment and the policy implications of the findings. Key Facts and Findings Latinos are the largest racial and ethnic group among youth and more than half have a foreign-born parent. California’s youth (ages 13 to 24) are tremendously diverse. However, Latinos constitute a plurality, at 41 percent of the youth population, followed by whites, at 37 percent, Asians, at 11 percent, and African Americans, at 7 percent. Most of the v state’s Latino youth are of Mexican ancestry (84%) and over 60 percent of them were born in the United States. Overall, one in four youth is a first-generation immigrant (i.e., born in a foreign country). About the same share are second-generation immigrants (i.e., born in the United States with at least one foreign-born parent). Racial and ethnic differences in educational attainment are strongly influenced by immigration. Of the major racial and ethnic groups in California, young adults of Mexican descent have the lowest levels of education. Of those ages 25 to 29, only 51 percent have earned a high school diploma, compared to 93 percent of non-Hispanic whites. However, the rate for Mexican American youth born in the United States is substantially higher—76 percent. Although Asians typically have higher levels of education than whites have, less than 80 percent of foreign-born Asians from Cambodia and Laos have completed high school. Mexican youth who come to the United States as teens often do not attend high school here. The older their age at arrival, the less likely Mexican youth are to attend school in California. Among those ages 16 to 18 and who recently arrived in the United States, less than half are enrolled in school. Among men, many are working; among women, substantial numbers are working, married, or raising children. In contrast, over 90 percent of Asian immigrant youth, including those from Southeast Asia, are enrolled in school at ages 16 to 18, regardless of their age at arrival. Among youth in immigrant families, there is tremendous variation in family income and parental education. Among young immigrants ages 13 to 17, about one-third of those from Mexico are living in poor families and only 17 percent have a mother who finished high school (maternal education is measured only for those living with their mothers). Among youth from Cambodia and Laos, over half are living in poor families and only 19 percent have a mother who finished high school. In comparison, among Filipino immigrants, less than 10 percent are living in poor families and over 90 percent have a mother who finished high school. These differences in family characteristics contribute to racial and ethnic differences in educational attainment for vi immigrant youth, which, in turn, contribute to education differences for their second-generation children. Substantial educational progress is made between the first, second, and third generations of immigrants. When the educational attainment of second- and third-generation immigrants is compared directly with that of their parents or with that of their parents’ generation, we find strong intergenerational progress for all major immigrant groups. In particular, although some research has suggested that educational progress stalls between the second and third generations for Mexican Americans, we find that college graduation rates of third-generation immigrants are more than twice those of their parents. Further, although over half of their parents did not graduate from high school, about eight in 10 third-generation Mexican Americans have graduated from high school. Even by the third generation, however, Mexican Americans in California have lower educational attainment than whites have. Despite strong intergenerational progress, less than 85 percent of thirdand-later-generation Mexican American adults, ages 25 to 34, have finished high school and only 11 percent have completed a bachelor’s degree. (“Third-and-later” generation includes youth with both parents born in the United States but the data do not identify whether their grandparents or great-grandparents were born in the United States.) In comparison, among third-and-later-generation whites, 95 percent earned a high school diploma and over a third have a bachelor’s degree. Differences in family characteristics explain most of the lower educational attainment of Mexican Americans. Among Mexican American youth, parental education, parental English language ability, and family income are substantially lower than among white youth. Using national data, we find that these differences in family economic resources and family characteristics generally account for the lower educational attainment of Mexican American youth relative to white youth. Policy Considerations The findings of this study have important implications for education-related policy in California. Mexican immigrant youth who vii arrive at age 15 or older are among the least educated Californians. Improving their educational attainment is particularly challenging because many do not enroll in California schools but are working and raising families. For these youth, adult education programs in school districts and community colleges may provide better schedules for parttime, evening, and weekend coursework. Another approach would be to target the workplace, offering programs in collaboration with employers to help workers develop English language and literacy skills. In addition, as these youth become parents, programs that work with young children can assist parents with parental support and literacy improvement. For second and third generations, and for immigrants who do enter California schools, the quality of the K–12 public education system is clearly a key factor in success. Several recent and continuing reforms have sought to improve California schools, particularly in the areas of student achievement, teacher quality, and quality of facilities. In addition, English language learning is of concern for the children of immigrants. For students whose own parents have limited educational experience, programs of educational counseling and tutoring may be particularly helpful. Our analysis suggests that about 30 percent of California’s children are growing up in families where neither parent has completed high school and that as many as 95 percent of these children might not achieve a bachelor’s degree. Among these children at risk of low educational achievement, Mexican Americans make up a large share (68%). The success of students in California’s community colleges is of particular importance for improving Latino postsecondary education because almost 80 percent of Latinos who enroll in public higher education enter through community colleges. Of great concern, however, is the low transfer rate to four-year institutions, and transfers are especially low among Latino students. In addition to preparing students for transfers, community colleges provide English language, remedial, and vocational courses. As the value of education and skills in the California economy continues to grow, these courses will become increasingly important to workforce training, especially for those who do not go on to complete a bachelor’s degree. viii Investing in an education system that works for all of California’s children will pay off for generations to come. Better educational attainment for today’s children will translate to better family economic resources for the next generation of parents. These higher resources will help the next generation of California’s children to attain even higher levels of education. ix Contents Foreword......................................... iii Summary......................................... v Figures .......................................... xiii Tables ........................................... xv Acknowledgments................................... xvii 1. INTRODUCTION .............................. Conclusions and Organization of the Report ............. A Note on Data ................................. 2. STATISTICAL PORTRAIT OF CALIFORNIA YOUTH... Demographic Characteristics ........................ Educational Attainment ........................... 1 2 3 5 5 9 3. INTERGENERATIONAL PROGRESS AND EDUCATIONAL ATTAINMENT ................... Measuring Intergenerational Progress .................. Intergenerational Progress Has Been Impressive ........... First Generation to Second Generation ............... Second Generation to Third Generation .............. Improvement Is Still Needed for Some Groups ........... Implications for the Future ......................... 15 15 17 18 22 26 28 4. DETERMINANTS OF THE EDUCATIONAL ATTAINMENT OF FOREIGN-BORN YOUTH ........ Recent Historical Context .......................... Family Characteristics ............................. Age at Arrival ................................... 5. DETERMINANTS OF THE EDUCATIONAL ATTAINMENT OF U.S.-BORN YOUTH ............. Family Characteristics ............................. 31 31 32 34 41 41 xi Factors Contributing to Racial and Ethnic Differences in Educational Attainment ........................ California Youth At Risk of Low Educational Attainment ................................. 42 50 6. CONCLUSIONS ............................... 53 Policy Directions ................................ 54 Directions for Future Research....................... 56 Appendix A. Notes on Data and Methods ........................ B. Detailed Nativity and Ancestry....................... C. Detailed Model Results ............................ 57 71 73 References ........................................ 77 About the Authors .................................. 87 Related PPIC Publications............................. 89 xii Figures 2.1. Percentage Distribution of the Educational Attainment of Young California Adults by Race, Ethnicity, and Immigrant Generation, Ages 25 to 34............... 12 4.1. Age at Arrival of Foreign-Born Mexicans and Asians..... 35 4.2. Activities of Foreign-Born Mexican and Asian Men by Age at Arrival, Ages 16 to 18 ..................... 36 4.3. Activities of Foreign-Born Mexican and Asian Women by Age at Arrival, Ages 16 to 18 ................... 38 5.1. High School Completion Rates Relative to Third-and- Later-Generation Whites ........................ 44 5.2. Grade Level Completion Rates of U.S.-Born Young California Adults with No High School Diploma, Ages 19 to 24, 2000 ............................... 46 5.3. Postsecondary School Attendance Relative to Whites .... 47 5.4. College Completion Rates Relative to Whites ......... 48 xiii Tables 2.1. Top Ten Countries of Birth and Ancestral Backgrounds of California Youth, Ages 13 to 24, 2000 ............ 6 2.2. Percentage Distribution of California Youth by Race, Ethnicity, and Immigrant Generation, Ages 13 to 24, 2004 ...................................... 7 2.3. Percentage Distribution of Youth in California and the Rest of the Nation by Race, Ethnicity, and Immigrant Generation, Ages 13 to 24, 1970 and 2000 ........... 9 2.4. Percentage Distribution of the Educational Attainment of Young California Adults by Race, Ethnicity, and Immigrant Generation, Ages 25 to 29............... 10 3.1. Percentage Distribution of the Educational Attainment of Groups in the First and Second Generation ......... 19 3.2. Percentage Distribution of the Educational Attainment of First-Generation Parents and Their SecondGeneration Adult Children ...................... 20 3.3. Percentage Distribution of the Educational Attainment of Groups in the Second and Third-and-Later Generations ................................. 24 3.4. Percentage Distribution of the Educational Attainment of Second-Generation Parents and Their ThirdGeneration Adult Children ...................... 25 3.5. Percentage Distribution of the Projected Educational Attainment of Second- and Third-Generation Descendants of Mexican Immigrants ............... 29 4.1. Percentage Distribution of the Family Characteristics of Foreign-Born California Youth by Race and Ethnicity, Ages 13 to 17, 2000 ........................... 34 5.1. Percentage Distribution of the Family Characteristics of U.S.-Born California Youth by Race and Ethnicity, Ages 13 to 17, 2000 ............................... 42 xv 5.2. Percentage Distribution of College Attendance and Completion Rates of U.S.-Born Youth by Race and Ethnicity, Ages 25 to 29, 2000.................... 49 5.3. Percentage Distribution of Students in Higher Education Institutions in California by Race and Ethnicity, 2002 .............................. 49 5.4. Percentage Distribution of Educational Attainment by Family Characteristics, Ages 0 to 9, 2000 ............ 51 A.1. Sample Sizes for Data Used in Table 3.1 ............. 62 A.2. Sample Sizes for Data Used in Tables 3.2 and 3.4 ...... 62 A.3. Sample Sizes for Data Used in Table 3.3 ............. 63 A.4. Sample Sizes and Percentage Distribution of College Completion Rates in the NELS (2000) and CPS (1996–2000) ................................ 67 A.5. Percentage Distribution of NELS Model Projections Versus Actual Bachelor’s Degree Attainment for Californians by Race and Ethnicity................. 68 A.6. Percentage Distribution of Educational Attainment by Generation: Mother-to-Child .................... 69 A.7. Percentage Distribution of Educational Attainment by Generation: Father-to-Child ..................... 70 B.1. Number of Youth and Educational Attainment for the Top 26 Countries of Birth Reported by Foreign-Born Californians, Ages 13 to 24, 2000.................. 71 B.2. Number of Youth and Educational Attainment for the Top 25 Ancestries Reported by U.S.-Born Californians, Ages 13 to 24, 2000 ........................... 72 C.1. Logistic Probability Model for High School Completion ................................. 74 C.2. Logistic Probability Model for Postsecondary School Enrollment ................................. 75 C.3. Logistic Probability Model for a Bachelor’s Degree...... 76 xvi Acknowledgments The authors wish to thank the William and Flora Hewlett Foundation for its generous support for this research project. We appreciate the continuing interest of foundation President Paul Brest in the topic of this report, and we thank Michael Wald and Tia Martinez for many helpful suggestions in the development of this project. Frank Furstenberg, Arturo Gonzalez, Donald Hernandez, Joyce Peterson, and Rubén G. Rumbaut provided valuable comments on an earlier draft. Amanda Bailey, Pedro Cerdán, and Shelly de Alth assisted in the research analysis. The methods and content of this report were determined solely by the authors. xvii 1. Introduction Nearly half of all Californians today are first- or second-generation immigrants. As that share of the California population continues to grow, it is increasingly important to understand the nature of intergenerational progress for immigrant groups.1 Recent research has called into question the intergenerational progress of immigrants, particularly educational progress between the second generation and the third generation. Because California has such large numbers of immigrants with limited education, a lack of improvement in educational attainment from one generation to the next would have serious implications for the state economically as well as socially. Educational progress is particularly important because education plays a role in determining racial and ethnic differences in other areas of social and economic well-being, such as poverty, health status, employment, home ownership, and civic participation (Reyes, 2001; Reed, 2003a). Furthermore, the economic value of education has increased substantially in the California labor market in recent decades (Betts, 2000). Over the past three decades, workers with a high school diploma or less have seen their average earnings erode whereas average earnings for those with a college degree have grown (Reed, 2003b). Neumark (2005) suggests that over the next two decades, the value of a college education in the California labor market will continue to grow. There are a number of state and local programs designed to improve the lives of youth as well as to steer them in the direction of positive future outcomes. Youth ages 13 to 24 are of critical concern because during these ages youth are preparing for the transition to adulthood _____________ 1Myers, Pitkin, and Park (2005) describe the growing presence of long-term and second-generation immigrants in the future California population and discuss implications for social and economic achievement. 1 with its increased economic challenges and responsibilities and often with new marriage and parenting relationships.2 During these ages, many potentially life-changing decisions are often made, including the decisions to finish high school, to go to college, and perhaps to start a family.3 Conclusions and Organization of the Report This report provides evidence that the intergenerational progress of immigrants has not stalled. Among all groups studied, including Mexican Americans, significant educational progress has occurred from the first to the second generation and from the second to the third generation. Nevertheless, even by the third generation, Mexican Americans have lower educational attainment than do whites and Asian Americans, the other large immigrant group. The report analyzes the factors that influence educational attainment, particularly that of Mexican Americans, and considers the implications for policy. Chapter 2 provides a statistical portrait of California youth, focusing on race, ethnicity, and immigrant generation and differences in educational attainment among groups. Chapter 3 compares the educational attainment of young adults to that of their parents (or their parents’ generation) and shows that all immigrant groups have made progress, including the first three generations of Mexican Americans. However, even accounting for this progress, the currently low educational attainment of Mexican Americans is likely to remain low in the coming decades. The findings in this chapter point to the need to understand the factors that contribute to this low educational _____________ 2See Setterson, Furstenberg, and Rumbaut (2005) for studies of the transition to adulthood in the United States and Western Europe. 3This report is part of a larger study that includes analysis of teen fertility and the transition from school to work. We chose to limit the report to education-related topics to present a more cohesive and focused discussion. Readers interested in our analysis of teen fertility and early work success are referred to Reed, Jepsen, and Hill (2005). For a more detailed description of teen fertility, see Johnson (2003). See Neumark (2004) for a discussion of school-to-career programs in California. The larger study also included an investigation of factors that contribute to positive educational and socioeconomic outcomes for youth who become teen parents and those who drop out of high school (see Hill and Jepsen, 2005). 2 attainment. Chapter 4 examines the determinants of educational attainment for foreign-born youth, finding that relative to Asian immigrants, Mexican immigrants are more likely to arrive after age 14 and to enter the workforce or family life and less likely to enroll in school. Chapter 5 examines the determinants of educational attainment among U.S.-born youth, finding that low parental education and low family income explain most of the lower educational attainment of Mexican American youth relative to white youth. Chapter 6 draws on all chapter findings to discuss policy directions for improving educational attainment. A Note on Data The Current Population Survey (CPS) and the decennial Censuses, used to measure youth educational attainment for this study, were chosen because they are representative of California.4 However, the data have two important limitations. First, they do not include information on involvement with the criminal justice system.5 Previous research suggests the importance of racial, ethnic, and immigrant differences in this area (Blumstein, 2001; Rumbaut, 2004). Second, they do not include detailed measures of youth background, such as attitudes about education, school conditions, and the quality of family relationships. Rumbaut (1999) provides a broad discussion of these issues for immigrant children and youth in San Diego.6 _____________ 4We focus on youth educational attainment and do not address the important but broader issues of health status or farmworker conditions. See Hernandez (1998, 1999) for national research on these topics. Diversity of health status across racial, ethnic, and immigrant groups in California is a topic of ongoing research at PPIC. 5The CPS represents the noninstitutional, civilian population. The Census represents the entire population, including those who are incarcerated, but it does not include information on past incarcerations or broader measures of interactions with the criminal justice system. See Appendix A for further descriptions of these data. 6Rumbaut (1995) describes four case studies of immigrant groups in California high schools. See also Portes and Rumbaut (2001) and Rumbaut and Portes (2001). The National Educational Longitudinal Study (NELS), used in this study to estimate models of youth educational attainment, includes information on some of these topics, but the subsample of youth in California is small and not necessarily representative. See Kao and Tienda (1995) and Kao (1999, 2004) for national measures of these issues based on the NELS. 3 The tremendous diversity of the California population makes it difficult to characterize the population and socioeconomic conditions by racial, ethnic, or ancestry group.7 The main analyses in this study focus on four major groups: Mexican Americans, Asian Americans, nonHispanic whites, and African Americans. We distinguish Mexican Americans from other Latinos because of their unique historical context and related socioeconomic conditions.8 Wherever possible we describe conditions for Central Americans, but their numbers are not sufficient to study intergenerational progress from the second generation to the third generation. Latino backgrounds also include Puerto Ricans, Cubans, Peruvians, and others. Their socioeconomic conditions are too diverse to make aggregation meaningful and their overall numbers in most of our data samples are too small for separate analysis. California’s Asian groups are particularly diverse. Where data permit, we distinguish Southeast Asians from refugee-sending countries (Laos, Cambodia, and Vietnam) from other Asians because of their substantially different historical context and socioeconomic conditions. We also consider the two Asian backgrounds that represent the largest number of Asian youth: Vietnamese and Filipinos. _____________ 7California is also regionally diverse and the nature of educational progress differs across regions. See Johnson and Hayes (2004) for an analysis of related issues for the Central Valley. 8See Chapter 4 for a discussion of the historical context of immigration for major immigrant groups. 4 2. Statistical Portrait of California Youth This chapter documents the tremendous diversity of California youth in terms of race, ethnicity, and immigrant generation. The analysis demonstrates the large and growing share of first- and secondgeneration immigrants among California youth and also describes substantial differences in educational attainment across racial, ethnic, and immigrant groups. The portrait presented here provides a context for the following chapters, which focus on the largest immigrant groups (Mexicans and Asians) and on those with particularly low educational attainment (Mexicans and Southeast Asians).1 Demographic Characteristics The youth population ages 13 to 24 in California numbers almost 6.6 million. By 2010, it is expected to increase to almost 7 million and to continue to make up close to 18 percent of the state population.2 The tremendous diversity of this population makes it difficult to characterize the backgrounds of its members. About one in four California youth was born outside the United States. California’s foreign-born youth are from almost 150 countries, mainly Mexico, the Philippines, El Salvador, Vietnam, and Guatemala (Table 2.1). California’s U.S.-born youth are from nearly 200 different ancestral _____________ 1Hill (2004) presents a more complete description of California’s youth, including regional information as well as statistics on youth activities, parental employment, health status, teen parenting, and citizenship. See also Ramakrishnan and Johnson (2005) for a description of second-generation immigrants in California, including statistics by California region. 2Population estimates are based on California Department of Finance population projections for 2005 and 2010. 5 Table 2.1 Top Ten Countries of Birth and Ancestral Backgrounds of California Youth, Ages 13 to 24, 2000 Country of Birth Number Foreign-Born 1. Mexico 783,124 2. Philippines 76,753 3. El Salvador 59,612 4. Vietnam 58,701 5. Guatemala 42,795 6. Korea 28,228 7. Taiwan 25,859 8. India 23,576 9. Thailand 22,822 10. China 22,337 Ancestry U.S.-Born 1. Mexican 2. African American 3. German 4. Irish 5. English 6. Italian 7. American 8. Filipino 9. White 10. Chinese Number 1,228,338 310,810 279,195 210,186 178,050 161,383 158,956 107,742 94,380 82,943 SOURCE: Authors’ calculations from the 2000 Census. NOTES: See Appendix Tables B.1 and B.2 for statistics on the top 25 countries of birth and ancestral backgrounds. Ancestry is self-identified and no specific categories are supplied (see Appendix A). backgrounds, mainly Mexican, African American, German, Irish, and English (Table 2.1).3 To further describe the youth population, we focus on the four largest racial and ethnic groups: Latinos, whites, Asians, and African Americans. Among California youth, Latinos are the largest ethnic group, comprising about 41 percent of all youth (Table 2.2). Thirteen percent of California youth were born in Mexico. The majority of Latino youth are of Mexican ancestry and were born in the United States. Fourteen percent of youth are second-generation Mexican Americans (born in the United States with at least one parent born _____________ 3The identification of racial, ethnic, and immigrant generation is complex. See Appendix A for a description of our analysis. See Rumbaut (2004) for a detailed analysis of these complexities. 6 Table 2.2 Percentage Distribution of California Youth by Race, Ethnicity, and Immigrant Generation, Ages 13 to 24, 2004 White Mexican Central American Other Latino Southeast Asian Other Asian African American Native American Multiracial and all others Total of column ForeignBorn 2 13 2 1 1 4 <1 <1 Second Generation 4 14 1 2 1 3 1 <1 Third-andLater Generation 31 7 <1 1 <1 1 6 1 Total of Row 37 34 3 4 3 8 7 1 1 12 4 25 27 49 100 SOURCES: Authors’ calculations from the 2000 Census and the Current Population Survey (March 2003, 2004). NOTES: Some rows and columns do not sum to totals because of rounding. See Appendix A for details on data and measurement. outside the United States). Seven percent are third-and-later-generation Mexican Americans. Throughout the report, “third-and-later generation” refers to people who were born in the United States and whose parents were both born in the United States. Most of the analysis relies on data that do not identify later generations.4 The next largest racial and ethnic group is non-Hispanic whites. Among white youth, the vast majority are third-and-later-generation Americans. Similarly, African Americans (7% of youth) and Native _____________ 4Data from the Current Population Survey do not distinguish between the third generation and those in later generations. The data report the birthplace of parents but not of grandparents. Rumbaut (2004) demonstrates that further categorizations of immigrant generations into the “1.5 generation” (immigrants who arrive by age 10) and the “2.5 generation” (those with one foreign-born parent and one U.S.-born parent) are useful in understanding diversity and educational outcomes. See also Ramakrishnan (2003). 7 Americans (about 1% of youth) are also primarily third-and-latergeneration Americans.5 The third largest racial and ethnic group is Asians, about 11 percent of the youth population. This population is very diverse with about half born in foreign countries, particularly the Philippines, Vietnam, Korea, Taiwan, India, Thailand, and China (Table 2.1). For U.S.-born Asian youth, the most common backgrounds are Filipino and Chinese. Among Asians, those from Southeast Asian refugee-sending countries (Cambodia, Laos, and Vietnam) tend to live in markedly worse socioeconomic conditions including higher poverty and lower educational attainment (see Chapter 4).6 For this reason, where possible, our analysis separates youth from these Southeast Asian backgrounds, about 3 percent of all youth. Overall, just under half of California’s youth are third-and-latergeneration immigrants. About one-fourth are second-generation immigrants and about one-fourth are first-generation immigrants. Because of the influx of immigrants in their late teens and early twenties, the share of youth who are foreign-born tends to increase with age. Among youth age 13, only 14 percent are foreign-born. At age 18, 22 percent of youth are foreign-born, and at age 24, the foreign-born share is 38 percent. Just over half of California’s youth—52 percent—are male. The primary factor increasing the share of males relative to females is immigration from Mexico and Central America—56 percent of these youth are male. Youth in California today are notably more diverse than California youth of past decades and also much more diverse than youth in the rest of the nation (Table 2.3). In 1970, the share of California youth who were white was 73 percent. In 2000, the share who were white was lower, at 66 percent for the rest of the nation, but in California the share was much lower, at only 37 percent. In the rest of the nation in 2000, as _____________ 5Some Native Americans are foreign-born, mainly from Canada. 6See Reyes (2001) for an analysis of socioeconomic indicators across Asian ethnic groups. 8 Table 2.3 Percentage Distribution of Youth in California and the Rest of the Nation by Race, Ethnicity, and Immigrant Generation, Ages 13 to 24, 1970 and 2000 White Latino Asian African American Multiracial and all others California, 1970 73 17 2 7 1 California, 2000 37 41 11 7 5 Rest of U.S., 2000 66 13 3 15 3 Third-and-later generation 83 49 81 Second generation 10 27 10 First generation 7 25 9 SOURCES: Authors’ calculations from decennial Censuses (1970, 2000). NOTES: Some columns do not sum to 100 percent because of rounding. See Appendix A for details on data and measurement. in California in 1970, over 80 percent of youth were third-and-latergeneration Americans. Educational Attainment California’s demographic diversity may be a source of pride and economic strength, but this diversity can create challenges when there are large and continuing disparities in socioeconomic conditions between groups. As discussed in the introduction, education is a key factor in determining economic success, health status, and other socioeconomic conditions.7 Among first-generation immigrant youth, there is tremendous diversity in educational attainment. Immigrants from Mexico and, to a lesser extent, Central America have particularly low levels of completing _____________ 7The association between education and earnings is not entirely causal (Griliches, 1977). National research that adjusts for noncausal factors finds strong labor market returns to education (Jaeger and Page, 1996; Arias and McMahon, 2001). See also Kane and Rouse (1999) for national estimates of the economic return to attending community college. 9 high school, attending college, and getting a bachelor’s degree (Table 2.4).8 Asian and white immigrants have relatively high educational attainment, although bachelor’s degree completion is lower among immigrants from Southeast Asian refugee-sending countries and Filipino Table 2.4 Percentage Distribution of the Educational Attainment of Young California Adults by Race, Ethnicity, and Immigrant Generation, Ages 25 to 29 Men White Mexican Central American Vietnamese Other Southeast Asian Filipino Other Asian Black or African American High School Diploma First Second+ 89 92 34 73 45 84 82 79 92 94 95 97 87 81 Some College Bachelor’s Degree First Second+ First Second+ 72 71 14 42 26 64 68 54 75 84 86 89 62 52 48 35 3 11 5 20 30 17 29 37 65 65 24 13 Women White Mexican Central American Vietnamese Other Southeast Asian Filipino Other Asian Black or African American 91 38 50 83 72 96 95 85 94 79 83 97 98 85 76 78 18 53 27 65 67 49 83 86 86 93 68 61 46 40 5 14 7 24 34 14 38 48 63 71 29 19 SOURCE: Authors’ calculations from the 2000 Census. NOTES: See Appendix A for measurement details. High school diploma includes the General Equivalency Diploma (GED). Sample sizes are not sufficient for secondand-later-generation Southeast Asians. The smallest sample used was 196 (for firstgeneration black women). _____________ 8We use ages 25 to 29 to measure education among young adults because by age 25 many have completed their education. However, particularly among Latinos, educational attainment does continue to improve at older ages. See Reed (2005). 10 immigrants. Black immigrant youth also have lower levels of college attainment than do other immigrants.9 Throughout this report, we use aggregated groups to describe socioeconomic conditions. However, within each aggregated racial and ethnic category in Table 2.4, there is a fair degree of diversity. For example, among first-generation immigrants from Central America, those from Guatemala tend to have lower educational attainment and those from Nicaragua have higher educational attainment (see Appendix Table B.1 for educational attainment among foreign-born youth by country of birth). Among Asian immigrants, those from India, Taiwan, and Hong Kong have particularly high rates of college completion. Among white immigrants, those from Armenia tend to have lower college completion. Even among young adults born in the United States, the aggregated groups mask tremendous diversity (see Appendix Table B.2 for educational attainment among U.S.-born youth by detailed ancestry group).10 Among U.S.-born (second-and-later-generation) Mexican Americans, educational attainment is much higher than for those born in Mexico. Indeed, if we were to calculate the high school completion rate among all young adults of Mexican descent, the rate would be only 51 percent. Because many Mexican immigrants do not enroll in school in the United States, this rate does not reflect the experience of Mexican Americans in California schools (see Chapter 4 and Fry, 2003).11 However, even among U.S.-born Mexican Americans, educational _____________ 9The rows of Table 2.4 are distinguished by reported race, ethnicity, and ancestry and not by country of origin. Many Southeast Asian immigrant youth were born outside Cambodia, Laos, and Vietnam, especially in Thailand and Malaysia. African American or black immigrants do not necessarily come from Africa. 10Rumbaut and Portes (2001), Portes and Rumbaut (2001), and Rumbaut (2004) provide detailed descriptions of the lives of the children of immigrants by ethnicity for San Diego and South Florida, including information on parental relationships, attitudes, and activities not available in the CPS or Census data. Waldinger and Bozorgmehr (1996) describe detailed conditions by ethnicity for Los Angeles. 11Educational statistics in Table 2.4 are for young adults living in California, but even for U.S.-born adults, these statistics may not reflect the educational attainment of young adults who grew up in California (because of domestic migration). See Reed (2005) for a discussion of education by race and ethnicity for Californians. For measures of high school dropout rates in California, see The Civil Rights Project (2005). 11 attainment is substantially below that of U.S.-born whites and somewhat below that of African Americans. For Mexican Americans in California, the education of the third-and-later generation is not substantially different from that of the second generation (Figure 2.1). 12 Among Asians, other than those from Southeast Asia, and among whites, educational attainment across all three generations is substantially higher than for Mexican Americans. Within each of these groups, educational attainment is fairly similar across the generations.13 Percentage 100 80 60 40 20 0 First Second Third+ First Second Third+ First Second Third+ Generation Generation Generation White Mexican American Asian American No high school diploma High school diploma Bachelor’s degree SOURCES: Authors’ calculations from the Current Population Survey (1996–2004). NOTES: The age range was increased from 25–29 to 25–34 to increase sample sizes. The smallest sample has 129 observations (Third+ –generation Asian Americans). Sample sizes were not sufficient to investigate other groups shown in Table 2.4. Figure 2.1—Percentage Distribution of the Educational Attainment of Young California Adults by Race, Ethnicity, and Immigrant Generation, Ages 25 to 34 _____________ 12Educational attainment by immigrant generation is measured using the 1996–2004 Current Population Survey. The CPS tends to measure higher high school completion than does the Census, in part because the Current Population Survey does not include institutionalized populations. 13Several studies at the national level have found higher educational achievements among the second generation relative to the first or third generations (Smith, 2003; Farley and Alba, 2002; Chiswick and DebBurman, 2004; Kao, 2004; Kao and Tienda, 1995). 12 Table 2.4 and Figure 2.1 describe educational attainment for young adults in California by immigrant generation, but these statistics should not be used to assess intergenerational progress. The intergenerational progress of these young adults should be measured against the achievement of their own parents or against people of their parents’ generation (as opposed to other young adults). We turn to the analysis of intergenerational progress in the next chapter. 13 3. Intergenerational Progress and Educational Attainment The previous chapter described the low educational attainment of immigrants from Mexico, Central America, and Southeast Asia. Because immigrants from these regions and their children constitute about onethird of all California youth, it is important for all Californians that these children acquire more education than their parents to gain the skills required by the state’s economy and the income necessary to provide for their families. In this chapter, we examine the recent history of intergenerational progress in the United States. We demonstrate substantial educational progress between second-generation immigrants and their foreign-born parents from all major immigrant-sending regions—progress that continues between the second and third generations for Mexicans, the state’s largest immigrant group. Because of the recent record, we project a high degree of educational progress for the third generation (grandchildren) of today’s Mexican immigrants. Despite this progress, Latino generations in years to come will still fall far short of the educational levels currently attained by U.S.-born whites. The analysis of intergenerational progress in this chapter is based on immigrants and their descendents nationwide. Use of a national sample is appropriate and avoids potentially biasing measures of intergenerational progress. For example, if immigrant families who have been successful have been more likely to remain in California, then limiting the analysis to those in California would overestimate the intergenerational progress of immigrants. Measuring Intergenerational Progress Improvements in educational attainment from the first generation to the second generation are well documented (Chavez, 1991; Card, 15 DiNardo, and Estes, 1990; Perlmann and Waldinger, 1997; and Farley and Alba, 2002), but some researchers have found little if any progress from the second to the third generations, particularly for Latinos or Mexicans (Borjas, 1994; Grogger and Trejo, 2002; Trejo, 1997, 2003; Chapa, 1990; Camarota, 2001; Kao and Tienda, 1995; and Kao, 2004). Measuring progress from the second to the third generation is difficult because of a lack of data on grandparents’ place of birth, and the research evidence that calls into question the intergenerational progress of recent immigrants is not definitive.1 Findings that point to weak progress between second and third generations often fail to measure the appropriate type of progress—progress in relation to one’s own parents (or progress from one generation of adults to an older previous generation)—or fail to consider nationally representative data. One study that compares current young adults with those of their parents’ generation finds strong progress from the first to the second generation and continuing progress to the third generation (Smith, 2003).2 In this chapter, we measure intergenerational progress in education using two methods.3 In both methods, we use nationally representative data from the Current Population Survey to estimate changes in educational attainment from one generation to the next, comparing attainment for adults with that of their parents. The CPS includes questions on the place of birth of all household members as well as that of the parents of those household members. We measure intergenerational change first by comparing the educational attainment of one entire generation to that of a previous generation and, second, by directly comparing an adult’s educational attainment with that of his/her parents. The first approach, a comparison of specific age groups from entire generations, uses a large sample but suffers from not directly linking parents with their descendants. In this sample, we cannot identify the third generation apart from fourth and later generations. We are careful to identify the age groups and immigrant generations that allow the most appropriate _____________ 1See Johnson (2005) for a more complete discussion. 2Smith and Edmonston (1997) also find progress across the first three generations. 3See Appendix A for details of our methodology. 16 comparison of one generation with a previous one. In particular, we compare educational attainment for young adults (ages 30 to 39) from one generation with the attainment of older adults (ages 57 to 66) from a previous generation. Young adults in their thirties have largely completed their formal education. On average, their parents are 27 years older. Thus, older adults ages 57 to 66 from a previous generation will include many of the parents of these young adults, and this group therefore serves as a benchmark for measuring intergenerational progress. The second approach has the advantage of explicitly linking parents with their adult descendants by examining households in which adult children and their parents live together. This allows us to precisely measure true (parent-to-child) generational change. Moreover, unlike the first approach, we are able to specifically identify the third generation. The disadvantage of this approach is that the sample is not large and is subject to selection bias because it requires that adult children be living with their parents.4 Because of sample size limitations, we use this multiple-generation household approach only for the state’s largest immigrant groups. Intergenerational Progress Has Been Impressive We find strong intergenerational progress in education. Across all generations for all groups, we consistently find that children acquire more education than their parents. This is true for U.S.-born children of immigrants (the second generation compared to the first generation), as well as for third-generation descendants of immigrants. The results are consistent for both approaches. That is, we find similar levels of progress in both the smaller sample that explicitly links parents with their adult children and the larger sample that compares one entire generation to another entire generation. _____________ 4See Appendix A for a more complete discussion of the sample. NELS includes information on parental education and young adult educational attainment (to age 26) regardless of living arrangements. Findings on intergenerational educational progress based on the NELS are consistent with the results shown in this chapter (see Appendix Tables A.6 and A.7). 17 First Generation to Second Generation There are dramatic increases in educational attainment from the first generation to the second generation for all ethnic groups. These increases are evident in comparisons of the entire second generation (ages 30 to 39) with the entire first generation (ages 57 to 66) (Table 3.1) as well as in comparisons of individuals specifically with their own parents (Table 3.2).5 The increases are evident for highly educated immigrant groups as well as for less educated immigrant groups. Overall, for immigrants from all regions, second-generation descendants born in the United States are much better educated than their parents’ generation. Whereas over one-third of the first generation have not graduated from high school, more than nine of 10 secondgeneration descendants have graduated from high school (Table 3.1). College graduation rates are much greater for young second-generation adults than for their parents’ generation, with over one-third of the second generation graduating from college compared to fewer than one in four in the first generation. These improvements from the first generation to the second generation are much larger than for other generations of U.S. natives. Among U.S.-born individuals with both parents born in the United States, 22 percent of the older generation has completed college compared to 28 percent of the younger generation (Table 3.1 bottom panel, third+ generation). Moreover, the young second generation in the United States is more likely than other U.S. natives to have graduated from college. Results for the intergenerational household sample of parents and adult children are similar (Table 3.2). In this sample, second-generation adult children are twice as likely as their first-generation parents to complete college. Although patterns of progression vary with ethnicity and region of origin, intergenerational progress is observed for all groups. The largest _____________ 5We focus on changes from the first generation to the second generation, but results are similar for changes from the first generation to the 1.5 generation (individuals born abroad but who came to the United States at age 10 or younger and therefore primarily if not wholly educated in the United States). Results for the 1.5 generation are available from the authors. 18 Table 3.1 Percentage Distribution of the Educational Attainment of Groups in the First and Second Generations Less Than High School Ethnic Group Diploma All ethnic groups First generation, ages 57 to 66 37 Second generation, ages 30 to 39 9 Mexican First generation, ages 57 to 66 78 Second generation, ages 30 to 39 18 Central Americana First generation, ages 57 to 66 54 Second generation, ages 30 to 39 16 Southeast Asianb First generation, ages 57 to 66 38 Second generation, ages 30 to 39 13 Other Asian First generation, ages 57 to 66 21 Second generation, ages 30 to 39 3 White (non-Hispanic) First generation, ages 57 to 66 20 Second generation, ages 30 to 39 4 Black or African American First generation, ages 57 to 66 35 Second generation, ages 30 to 39 11 All ethnic groups Third+ generation, ages 57 to 66 18 Third+ generation, ages 30 to 39 8 High School Diploma 26 26 12 33 26 26 31 22 22 14 32 24 34 23 38 35 Some Bachelor’s College Degree 13 23 30 35 64 34 15 9 11 32 26 12 19 24 41 12 44 25 58 18 29 29 43 14 17 33 33 22 22 29 28 SOURCES: Authors’ analysis of the Current Population Survey (1996–2004). NOTE: See Appendix Table A.1 for sample sizes. aCentral Americans are identified as Latinos born in Central America (Belize, Guatemala, Honduras, Costa Rica, El Salvador, or Panama) or with at least one parent born in Central America. The second generation includes the 1.5 generation. Our sample of Central Americans for the 1.5 and second generation ages 30 to 39 is 310. bSoutheast Asians are identified as Asians born in Southeast Asia (Vietnam, Cambodia, or Laos) or with at least one parent born in Southeast Asia. The second generation includes the 1.5 generation. Our sample of Southeast Asians for the 1.5 and second generation ages 30 to 39 is 169. 19 Table 3.2 Percentage Distribution of the Educational Attainment of First-Generation Parents and Their Second-Generation Adult Children Ethnic Group All ethnic groups First-generation parents Second-generation descendants Mexican origin First-generation parents Second-generation descendants Central Americana First-generation parents Second-generation descendants Southeast Asiana First-generation parents Second-generation descendants Other Asian First-generation parents Second-generation descendants White First-generation parents Second-generation descendants Less Than High School Diploma 38 10 75 14 45 10 29 2 16 1 25 8 High School Diploma 30 27 15 36 33 32 22 25 26 10 38 26 Some College 16 33 7 38 11 43 11 21 18 23 20 33 College Graduate 15 30 3 12 11 15 37 52 41 66 17 34 SOURCES: Authors’ analysis of the Current Population Survey (1996–2004). NOTES: Data in the table include only households with parents and adult children in which the adult children are ages 25 to 49. Results are the average of mother-to-child and father-to-child comparisons. See Appendix Table A.2 for sample sizes. aFor Central Americans and Southeast Asians, we combine the 1.5 and second generations to increase sample sizes. group of immigrants in California is from Mexico (see Chapter 2).6 Mexican immigrants are among the least educated populations in the United States. In the sample of older first-generation Mexican immigrants, almost four in five have not graduated from high school. Although first-generation immigrants from Mexico tend to have very _____________ 6Mexicans are also the largest group of immigrants to the United States (U.S. Census Bureau, 2002). 20 low levels of education, their children show remarkable gains. Among young adults in the second generation, over four in five have graduated from high school (Table 3.1, Mexican panel). Progress is also evident at the other end of the educational spectrum. The share of college graduates is about four times greater among the second generation than the first. Results for the multiple-generation household sample are nearly identical, also showing strong gains from the first generation of Mexican parents to their adult children (Table 3.2). Even with this progress, however, educational attainment levels for second-generation Mexican Americans are lower than those of other U.S. natives. Although Central American immigrants show higher levels of educational attainment than Mexican immigrants, a majority of older Central American immigrants have not graduated from high school and only one in nine is a college graduate (Table 3.1, Central American panel). By the second generation, the proportion of Central Americans who are college graduates is near the average for all U.S.-born, and the share who do not have a high school diploma drops tremendously (from 54% for the first generation to 16% for the second generation). The sample of multiple-generation households for Central Americans also shows that adult second- and 1.5-generation children are much better educated than their parents, but progress does not appear to be as dramatic for college graduation.7 In contrast to immigrants from Mexico and Central America, Asian immigrants (not including those from Southeast Asia) tend to be better educated than other immigrants; indeed, they are more likely than U.S. natives to graduate from college. This is particularly the case for immigrants from East Asia, the Philippines, and India. It is not surprising then that their second-generation descendants have especially high levels of education, although the increase across generations is remarkable (Table 3.1, Other Asian panel). Over half of secondgeneration descendants of Asian immigrants have graduated from college and not completing high school is extremely uncommon. _____________ 7The 1.5 generation is made up of foreign-born individuals who came to the United States by age 10. The 1.5 generation in this sample is less likely than the second generation to graduate from college. We combine these groups for Central Americans to increase our sample sizes. 21 Even adult children of Southeast Asian immigrants, the least educated Asian immigrant group, show tremendous increases in educational attainment compared to their parents. Still, their high school dropout rates are higher than those of whites and other Asians. College graduation rates are remarkably high for second-generation Southeast Asians.8 The results for Southeast Asian immigrants and their descendants outlined here are dominated by Vietnamese, the largest and best educated of the Southeast Asian groups. Results for Hmongs, Laotians, and Cambodians could be very different. Results from the multigenerational household sample confirm these general patterns but should be interpreted with caution because of the small sample size. Non-Hispanic white immigrants in California and the United States come from many different countries and regions, including Canada, Europe, and the Middle East. The older first generation has an educational profile that is similar to that of third-and-later-generation older adults in the United States, with the share not completing high school almost as large as the share completing college. The second generation that includes their children is far better educated, with 43 percent graduating from college and only 4 percent not completing high school. Finally, blacks also show strong intergenerational progress from the first generation to the second, with the proportion graduating from college almost twice as high in the second generation as in the first. However, California is home to relatively few black immigrants. Second Generation to Third Generation Measuring educational attainment is difficult for the third generation. Because the CPS does not include information on grandparents’ place of birth, we are able to identify only third-and-later generations in our large sample. In the smaller multigenerational households sample, we are able to identify those of the third generation precisely, because they live with their second-generation parents. Having _____________ 8Note that the strong educational progress of Southeast Asians is consistent with their relatively high level of school attendance shown in Chapter 4. Because of the small sample size, we combined the 1.5 and second generations for Southeast Asians. 22 arrived primarily in the late 1970s and the 1980s, the immigrant population from Southeast Asian refugee-sending countries has not been in the United States long enough to develop a sizable third generation.9 Similarly, third-generation Central American immigrants are too few to report in our multigenerational household sample and cannot be identified in other households of the CPS.10 Black immigrants are an even more recently arrived group, on average, and have not yet developed a sizable second generation of older adults. Intergenerational progress continues from the second generation to the third generation, although progress is not as dramatic. For every group, third-generation adults are better educated than their secondgeneration parents (Tables 3.3 and 3.4). For less educated groups, these improvements are similar in scale, although generally not quite as large, as between first-generation immigrants and their second-generation children. Among all third-and-later-generation young adults, only 8 percent were high school dropouts, compared to 14 percent for the second generation of older adults (Table 3.3). Among adult children living with their second-generation parents, nine of 10 had graduated from high school and more than one in four had graduated from college, compared to only 75 percent and 16 percent (respectively) of their parents (Table 3.4). Third-generation Asians show strong intergenerational progress in the multigenerational household sample that allows comparisons of adult children with their parents (Table 3.4) but only modest progress in the larger sample that compares the entire third-and-later generation with the older second generation (Table 3.3). However, both samples show that third- or third-and-later-generation Asians are much better educated than their peers of other ethnicities, with high college graduation rates and low levels of high school dropouts relative to other third-generation groups. _____________ 9There are too few third-generation Southeast Asians to measure progress from the second to the third generations. In our multigenerational household sample, we observed no third-generation Southeast Asians. 10The CPS combines Central Americans with South Americans in its ethnic categories. 23 Table 3.3 Percentage Distribution of the Educational Attainment of Groups in the Second and Third-and-Later Generations Ethnic Group Mexican Second generation, ages 57–66 Third+ generation, ages 30–39 Asian Second generation, ages 57–66 Third+ generation, ages 30–39 Black or African American Second generation, ages 57–66 Third+ generation, ages 30–39 White (non-Hispanic) Second generation, ages 57–66 Third+ generation, ages 30–39 All ethnic groups Second generation, ages 57–66 Third+ generation, ages 30–39 Third+ generation, ages 57–66 Less Than High School High School Diploma Diploma 47 28 20 38 7 27 4 26 n/a 12 40 10 36 7 33 14 35 8 34 18 38 Some College 18 29 29 26 32 23 29 23 29 22 Bachelor’s Degree 7 12 37 44 16 30 31 28 28 22 SOURCES: Authors’ analysis of the Current Population Survey (1996–2004). NOTES: Asian includes Pacific Islanders. See Appendix Table A.3 for sample sizes. Similarly for whites, the comparison of third-and-later-generation young adults with the entire second-generation of older adults shows only slight improvements in educational attainment (Table 3.3), whereas the comparison of third-generation adults with their second-generation parents who live in the same household shows much more substantial improvements (Table 3.4). In either case, the third (or third-and-later) generation is well educated, with high school dropout rates quite low and college completion quite high. Third-generation Mexicans also show substantial improvements in education over the second generation. Among third-generation Mexicans who live with their parents, college graduation rates are twice those of their second-generation parents (although still relatively low at only 11 percent); and although over half of their U.S.-born parents did 24 Table 3.4 Percentage Distribution of the Educational Attainment of Second-Generation Parents and Their Third-Generation Adult Children Ethnic Group All ethnic groups Second-generation parents Third-generation descendants Mexican origin Second-generation parents Third-generation descendants Asian Second-generation parents Third-generation descendants White (not Hispanic) Second-generation parents Third-generation descendants Less Than High School Diploma 25 9 54 18 13 2 21 8 High School Diploma 39 36 28 39 42 21 42 36 Some College 19 29 13 33 30 41 20 27 Bachelor’s Degree 16 26 5 11 15 36 18 29 SOURCES: Authors’ analysis of the Current Population Survey (1996–2004). NOTES: Data in the table include only households with parents and adult children in which the adult children are ages 25 to 49. Results are the average of mother-to-child and father-to-child comparisons. See Appendix Table A.2 for sample sizes. not graduate from high school, more than eight in 10 third-generation descendants of Mexicans did graduate from high school (Table 3.4). Results from the larger sample (which compares entire generations but does not compare parents directly with their children) are consistent with the smaller parent-to-child sample findings presented above, showing strong although not quite as dramatic intergenerational progress from the second to the third generations (Table 3.3). This is an important finding: Increases in educational attainment do not cease between the second and third generations among Mexican Americans. Nonetheless, sizable proportions of the third generation have not completed high school and the proportion of college graduates is quite low. Although third-generation Mexicans are better educated than their parents, they are not better educated than second-generation Mexicans of similar age. This conundrum is important to understand and decipher, as it has led to confusion even among scholars who study 25 intergenerational change. Second-generation parents (Table 3.4) are much less educated than second-generation adult children (Table 3.2). Indeed, second-generation parents are only slightly better educated than first-generation parents, with very few college graduates among either group. This is because today’s second-generation parents were educated (or not) at a very different time in the United States. Not only were educational attainment levels much lower overall many decades ago for the entire U.S. population, certain groups—including Mexican Americans—had less access to public education. For example, in the 1930s the California legislature withheld state funds to local school districts that did not segregate Mexican children (Surace, 1982, citing Heizer and Almquist, 1977). Finally, the parents of today’s secondgeneration parents were less educated than today’s first-generation immigrants. In many of the primary sending countries of immigrants to the United States, including Mexico, educational attainment levels are on the rise. For example, recent immigrants from Mexico are much better educated than Mexican immigrants who came to the United States 30 or 40 years ago, a reflection of rising education levels in Mexico. Because parental educational attainment is such an important determinant of educational attainment for their children, these low levels of educational attainment of second-generation parents have important implications for third-generation educational attainment. In sum, cross-sectional comparisons of all third-generation adults with the entire second generation show only marginal differences for those of Mexican descent. However, when we compare the third generation with their parents (who form the older, least-educated part of the second generation), we see dramatic gains. Thus, intergenerational progress for Mexican Americans from the second to the third generations is strong when it is properly measured by comparing attainment for individuals with attainment of their own parents. Improvement Is Still Needed for Some Groups Increases in educational attainment from one generation to the next, although impressive, have not been sufficient for all groups to reach parity with other U.S. natives. This disparity can be partly explained by the increasing educational attainment among U.S. natives, as well. For 26 example, younger U.S.-born adults are better educated than the older cohort that includes their parents: Among third-and-later-generation young adults, the proportion who did not complete high school is less than half the rate of third-and-later-generation older adults (8% for those ages 30 to 39 versus 18% for those ages 57 to 66). Of particular concern is the state’s largest immigrant group, Mexicans. Second- and thirdgeneration Mexicans have not reached parity with other U.S.-born groups. Rates of college graduation are low, and rates of high school dropout are higher for Mexican-origin descendants than for other U.S. natives. The pattern for Asians is very different, with remarkably high levels of educational attainment—much higher than for U.S.-born whites. Even many first-generation Asian immigrants are well educated: 50 percent of the youngest adult Asian immigrants are college graduates. Second-generation Asians have even higher levels of education. Thirdand-later generations still exceed the educational levels of their parents and other U.S.-born whites. Levels of college education for U.S.-born Southeast Asians are also higher than for U.S.-born whites, even though their parents are the least educated Asian group. One cautionary note is evident, however, in the relatively high rate of high school dropouts among 1.5-generation Southeast Asians—a rate that is twice that of U.S.born whites. It is possible that Vietnamese immigrants and their descendants have met with much success educationally, whereas Laotians and Cambodians—among the least educated of any immigrant groups in the United States—have not.11 In other words, the diversity of educational attainment might be a reflection of the diversity of the immigrant groups included in the category, with the relatively numerous Vietnamese composing the best educated Southeast Asians and the less numerous Hmong, Laotians, and Cambodians composing the least educated Southeast Asians. _____________ 11Vietnamese immigrants tend to have higher levels of education than other Southeast Asians. Chapter 4 shows the higher family economic resources of Vietnamese compared to other Southeast Asians. 27 Implications for the Future Using historical educational progress across generations as a basis, we develop projections of educational attainment for the descendants of the most recent immigrants to the United States. These projections allow us to consider the future path of intergenerational progress. To develop the projections, we used transition probabilities based on an individual’s generation, maternal education, paternal education, and region of origin. Details of the method are included in Appendix A. The premise of the projections is that within an immigrant group, the historical relationship between an individual’s educational attainment and that of his/her parents will remain static over time. This relationship is, of course, a strong one—for example, the children of parents who are college graduates are much more likely than children of parents who did not complete high school to graduate from college. Nonetheless, other factors also play a role in the determination of educational attainment, and this set of projections implicitly assumes that those other factors are both embodied in the historical patterns and will remain unchanged in the future. The purpose of these projections is not to forecast what will happen, rather it is to suggest what would happen if historical patterns persist. Because Mexicans are the single largest group of immigrants and have the lowest levels of education, we focus on that group exclusively in this discussion. The large majority of Mexican immigrants have not completed high school. Indeed, well over half have not graduated from high school. Nonetheless, the vast majority of their children and grandchildren will graduate from high school. We project that only 13 percent of the children of today’s immigrants will not graduate from high school (Table 3.5), and only 11 percent of their grandchildren (the third generation) will not complete high school. A small but not insignificant proportion of the second generation will even graduate from college. That proportion is projected to increase from the second generation to the third generation but, at 17 percent, will remain far lower than that of other U.S. natives. 28 Table 3.5 Percentage Distribution of the Projected Educational Attainment of Second- and Third-Generation Descendants of Mexican Immigrants Less than high school diploma High school diploma Some college Bachelor’s degree FirstGeneration Mothers 66 21 8 5 SecondGeneration Children 13 34 39 13 ThirdGeneration Grandchildren 11 30 42 17 SOURCE: Authors’ projections. NOTES: Data in the table are based on the parent-to-child education transition matrix (see Appendix A). First-generation numbers are based on the Current Population Survey (2004) for Mexican immigrants ages 30 to 39. These projections suggest strong intergenerational gains in educational attainment from the first generation to the second generation and continuing if less dramatic gains from the second generation to the third generation. However, they also suggest that the gap between U.S.-born whites and the third-generation descendants of Mexican immigrants will continue into the future. Of particular concern are the relatively low rates of completing college. These patterns persist into the third generation because the tremendous gains into “some college” by the second generation do not translate into college graduation for the third generation. Of course, these projections are based on historical patterns and future patterns could be different. Indeed, it is the goal of public policy in general and education policy in particular to improve these outcomes. 29 4. Determinants of the Educational Attainment of Foreign-Born Youth The education and socioeconomic conditions of the first generation set the stage for the subsequent progress of the second generation. In this chapter, we explore the determinants of differences in educational attainment for those of the first generation, beginning with the recent historical context of their migration. We then turn to today’s foreignborn youth, suggesting that the low educational attainment of firstgeneration Mexican youth can be understood as a continuation of the historical pattern whereby youth from families with low socioeconomic resources come to the United States as young adults primarily for work opportunities. This analysis provides a background for discussing policy challenges and possible directions for improving educational attainment among first-generation Mexican youth—topics we address in the concluding chapter. Recent Historical Context During the 20th century, Mexican immigration to the United States was primarily made up of low-skilled workers seeking positions in agriculture, production, and service jobs.1 Even during the 1940s and 1950s, when foreign immigration to the United States was severely restricted, many Mexicans came to California as agricultural workers _____________ 1See the introduction to Waldinger and Bozorgmehr (1996) for a brief history of migration to California. The volume also provides an account of the ethnic transition of Los Angeles since the 1970s, including socioeconomic conditions of specific ethnic groups. See López and Stanton-Salazar (2001) for a brief discussion of the history of Mexican immigration to California and an analysis of second-generation Mexican Americans in San Diego. See Massey, Durand, and Malone (2002) for a history of Mexican immigration and related U.S. policies. 31 under the Bracero program (1942–1964). Still others came as unauthorized workers in production and service sectors. The easing of immigration restrictions by the Hart-Cellar Act (1965) brought additional flows of Mexican workers and their family members. Flows increased after the passage of the Immigration Reform and Control Act of 1986 (IRCA), which legalized the status of those who had been U.S. residents since 1982 and “special agricultural workers.” These residents and workers were then joined by family members under family reunification provisions, which were further expanded in 1990. Immigration from Central America has primarily been more recent, beginning in earnest during the late 1970s with Salvadorans and Guatemalans fleeing political unrest at home. Thus the historical context for immigration from Central America is not as closely tied to work opportunities in low-wage sectors as is Mexican immigration. The recent history of Asian migration has been profoundly different from that of Mexican migration. The Immigration Restriction Acts (1921, 1924) severely limited Asian immigration through very low country-of-origin quotas. The Hart-Cellar Act lifted these quotas and allowed immigration to reunite families and for those with scarce and needed skills. In the years that followed, many new Asian immigrants, notably Koreans and Asian Indians, came as skilled workers and paved the way for later immigrants to come as kin. The relatively high educational attainment of recent Asian immigrants can be understood partially as a consequence of this “skill-based” reason for entry into the United States. During the late 1970s and the 1980s, a different wave of Asian immigrants arrived: refugees fleeing communist regimes in Vietnam, Cambodia, and Laos. The first Southeast Asian immigrants were typically elites but were followed by refugees with relatively low educational attainment. Family Characteristics Among today’s immigrant youth, family characteristics follow a pattern that might be expected given the recent context of immigration: Immigrants from Mexico, Central America, and Southeast Asia tend to 32 have low resources, whereas those from other parts of Asia tend to have higher economic resources. These family resources are important because they help determine the educational attainment of firstgeneration youth. Parental education, particularly maternal education, is strongly linked to a child’s cognitive development and school success (Haveman and Wolfe, 1995). Research has shown that higher educational attainment is also associated with having a mother who speaks English, living with both parents, and having family income above the poverty level.2 To examine family resources among immigrant youth, we focus on those ages 13 to 17 who live with their families (Table 4.1).3 Immigrant youth from Mexico have particularly low family resources. Among those who live with their mother, only 17 percent have a mother with a high school diploma. Among Central American immigrant youth, family resources are also low, but not as low as for Mexican youth. For example, 31 percent have a mother who has completed high school. Among Southeast Asian immigrants, especially those from Cambodia and Laos, maternal education is low and the share living above poverty is also particularly low (only 46%). In contrast, Filipino youth are in families with high resources: 91 percent have a mother who completed high school, and 92 percent live above the poverty line. Among Asians from other backgrounds, there is a fair amount of diversity, but family resources tend to be relatively high. _____________ 2See Haveman and Wolfe (1995) for a review of the literature on the determinants of youth outcomes. See also Pirog and Magee (1997), Denton, West, and Walston (2003), Mare and Winship (1988), Grogger and Trejo (2002), and Chuang (1997). Mayer (1997) finds that family income has only a small effect on educational attainment. She calculates that if family income were doubled for poor families, the overall high school dropout rate would fall from 17.3 to 16.1 percent. She argues that the effect is small because many basic necessities are met through government programs, including Food Stamps, Medicaid, and housing subsidies. 3See Beavers and D’Amico (2005) for a study of the conditions of children in immigrant families nationally with state comparisons. 33 Table 4.1 Percentage Distribution of the Family Resources of Foreign-Born California Youth by Race and Ethnicity, Ages 13 to 17, 2000 Mother Mother Youth Youth Family Family Has High Has Mother Lives Lives Income Receives Share School Bachelor’s Speaks with Both with Any Above Public of Total Diploma Degree English Parents Relative Poverty Assistance White Mexican Central American Vietnamese Other Southeast Asian Filipino Other Asian 9 56 7 4 4 6 11 83 17 31 38 19 91 79 35 72 2 27 5 42 2 25 2 24 55 92 34 50 81 64 59 79 77 77 81 97 76 93 64 93 69 98 65 97 46 97 92 96 81 21 9 7 47 66 5 5 SOURCE: Authors’ calculations from the 2000 Census. NOTES: Maternal education and English ability include only youth who are co-resident with their biological, adoptive, or stepmother. “Mother speaks English” is the percentage who speak only English at home plus those whose mother speaks English “well” or “very well.” The share living with a relative includes only those who live in a household headed by a relative. The 2000 Census does not describe family relationships for people who are unrelated to the household head. See Appendix A for further details. Age at Arrival A growing body of research has noted that age at arrival is an important determinant of socioeconomic outcomes for immigrants. Hill (2004) finds that among immigrants in California, those who arrived as young children have similar educational outcomes to native-born youth of the same race and ethnicity.4 Among young adults from Mexico, a large share arrive at ages 15 and older. Of those who arrived by age 18, about half arrived after age 14 (Figure 4.1).5 Of those who arrived by age 24, 72 percent arrived after _____________ 4See Chiswick and DebBurman (2004) for a review of studies of age at arrival and post-immigration schooling. See also Rumbaut (2004) and Gonzalez (2003). 5Figure 4.1 reports the age of migration for two age cohorts to show that the major patterns are consistent. For a single age cohort, the age pattern could be dominated by an event such as IRCA. This act allowed legalization of immigrants who were working in the United States in 1982 and earlier. Following IRCA, there was a spike in immigration from Mexico in 1988–1990 as people joined newly legalized family and friends in the United States. For the older cohort, this spike begins about age 14. For the younger cohort, the spike begins at about age 8. However, for both cohorts, the highest levels of immigration are after age 14. 34 20 Mexicans 15 Ages 25 to 30 in 2000 Ages 19 to 24 in 2000 Percentage 10 5 0 20 Southeast Asians 15 Ages 25 to 30 in 2000 Ages 19 to 24 in 2000 Percentage 10 5 0 20 Other Asians 15 Ages 25 to 30 in 2000 Ages 19 to 24 in 2000 Percentage 10 5 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Age at arrival SOURCE: Authors’ calculations from the 2000 Census. NOTES: Data in the figure include foreign-born youth who arrived in the United States before age 19. Age at arrival does not differ substantially between men and women. Figure 4.1—Age at Arrival of Foreign-Born Mexicans and Asians 35 age 14 (not shown). In contrast, Southeast Asian immigrants were more likely than Mexican youth to arrive at early ages. Among Southeast Asians who arrived in the United States by age 24, about half arrived after age 14.6 Among other Asians, immigration is fairly evenly spread across all ages but tends to be slightly higher at ages 17 and 18. Among other Asian immigrants who arrived by age 24, just over half arrived after age 14. The high share of Mexican immigrant youth who arrive in their late teens is particularly important because these youth are much less likely than other youth to enroll in school. Among Mexican immigrant men, almost 90 percent of those who came to the United States as young children were enrolled in school at ages 16 to 18, compared to only about a third of those who arrived at age 17 (Figure 4.2, first panel). Enrolled in school 100 80 60 Working 100 Mexican 80 Southeast Asian Other Asian 60 Percentage 40 Other Asian 40 Southeast Asian 20 Mexican 20 00 2 5 8 11 14 17 2 5 8 11 14 17 Age at arrival Age at arrival SOURCE: Authors’ calculations from the 2000 Census. NOTES: Data in the table include foreign-born youth who arrived in the United States before age 19 and who have not completed high school. Reported statistics are three-year moving averages. Figure 4.2—Activities of Foreign-Born Mexican and Asian Men by Age at Arrival, Ages 16 to 18 _____________ 6Arrival by age 24 is measured for immigrants ages 25 to 30. Immigration from the Southeast Asian refugee-sending nations peaked around 1980. For the older cohort, this peak occurred when they were ages 6 to 11. For the younger cohort, this peak occurred when they were age 5 and under. 36 Among Asian and Southeast Asian immigrant men ages 16 to 18, school enrollment was over 90 percent even for those arriving at age 17. Mexican men arriving in their late teens are much more likely than other youth to be working.7 About half of those who arrived after age 14 were working at ages 16 to 18 (Figure 4.2, second panel). Among Mexican men who arrived at younger ages and among Asians, work participation at ages 16 to 18 was only around 20 percent (with the exception of Southeast Asian men who arrived in their late teens, for whom work participation was about 33 percent). College-age youth, those ages 19 to 21, had a similar pattern (not shown in the figure). Among Mexican men who arrived at age 19, only 11 percent were enrolled in school and 70 percent were working. Among Southeast Asian men who arrived in their late teens, 64 percent were enrolled in school and 44 percent were working.8 Among other Asian men who arrived at age 19, 74 percent were enrolled in school and 39 percent were working. Mexican women who arrived in their late teens were also less likely than other youth to be enrolled in school at ages 16 to 18. Among those who arrived at age 17, only 43 percent of Mexicans were enrolled in school compared to 87 percent of Southeast Asians and 94 percent of other Asians (Figure 4.3). Work participation was fairly similar between the groups at about 20 percent, but Mexican women were much more likely to be married (21% were married compared to less than 2% of Southeast Asian and other Asian women) and much more likely to be raising young children (7% compared to less than 2% of Southeast Asian and other Asian women). Among immigrant women of college age, ages 19 to 21, Mexicans were also far less likely to be in school (not shown in the figure). Only 14 percent of Mexican women who arrived at age 19 were enrolled in school, 36 percent were working, 41 percent were married, and 22 percent were raising a young child. Southeast Asian women who arrived _____________ 7In Figures 4.2 and 4.3, the share working includes those who are working while in school. 8Because of small sample sizes, Southeast Asian statistics are averaged over those who arrived between ages 16 and 21 (but, similar to other groups, are limited to men ages 19 to 21 at the time of the 2000 Census). 37 Enrolled in school 100 80 60 Working 100 Mexican 80 Other Asian Southeast Asian 60 Percentage 40 Other Asian 40 Southeast Asian 20 Mexican 20 00 2 5 8 11 14 17 2 5 8 11 14 17 Age at arrival Age at arrival SOURCE: Authors’ calculations from the 2000 Census. NOTES: Data in the table include foreign-born youth who arrived in the United States before age 19 and who have not completed high school. Reported statistics are three-year moving averages. Figure 4.3—Activities of Foreign-Born Mexican and Asian Women by Age at Arrival, Ages 16 to 18 in their late teens were much more likely to be in school (70%) but were also fairly likely to be working (40%), married (29%), and raising young children (17%).9 In contrast, other Asian women who arrived at age 19 were very likely to be in school (81%), fairly likely to be working (30%), and very unlikely to be married (10%) or raising a child (3%). The analysis of age at arrival shows that Mexican immigrant youth are more likely than Southeast Asian youth or other Asian youth to have come to the United States in their late teens. Mexican youth who arrive in their late teens have low levels of school enrollment, high levels of work participation among men, and high levels of marriage and childrearing among women. These findings suggest that many Mexican immigrant men come to California for work and many women come to work and to join their spouses or start a family. _____________ 9Because of small sample sizes, Southeast Asian statistics are averaged over those who arrived between ages 16 and 21 (but, similar to other groups, are limited to women ages 19 to 21 at the time of the 2000 Census). 38 In contrast, Asian immigrants are not particularly likely to arrive in their late teens and even among those who do, school enrollment is high, work participation among men is substantially lower than that of Mexican men, and family formation among women is lower than among Mexican women. These findings suggest that many Asian youth come to California to attend school. Many Southeast Asians came to California as refugees to escape oppressive conditions in their native countries. School enrollment is fairly high among Southeast Asians, even among those who arrived in their late teens. This chapter has identified two factors that contribute to the lower educational attainment of Mexican immigrant youth relative to Asian immigrant youth. First, family characteristics associated with youth success, such as parental education, are much lower among Mexican immigrant youth than among Asian youth (other than those from Southeast Asia). Second, for immigrant youth from Mexico, many arrive after age 14 and a large share do not appear to enroll in school in the United States. Their educational attainment is therefore quite low. Among Southeast Asian youth, family resources are also very low. Southeast Asian immigrant youth are more likely than Mexican immigrant youth to be living in poverty. However, Southeast Asian immigrants are much more likely than Mexican immigrants to enroll in school. One partial explanation for this difference may be the refugee status of many Southeast Asian immigrants. As refugees, Southeast Asian youth tend to come to the United States with their families and their families are eligible for public assistance. More than half of Southeast Asian youth receive some form of public assistance income, whereas among Mexican youth the share is less than 10 percent (Table 4.1). Public assistance income likely makes it possible for Southeast Asian youth to choose school because their family income is less dependent on their work earnings. 39 5. Determinants of the Educational Attainment of U.S.-Born Youth Low educational attainment among Mexican immigrants helps explain the low educational attainment of second-and-later-generation Mexican Americans because parental education is a key determinant of attainment (see Chapter 4). In this chapter, we explore the relationships between parental education and other family characteristics and educational attainment among U.S.-born youth. In particular, we examine the extent to which differences in family characteristics can explain racial and ethnic differences in educational attainment. In addition to providing a better understanding of the factors that explain racial and ethnic differences, the analysis is used to estimate the share of California children at risk of not finishing high school or college and to suggest mechanisms for targeting educational support programs. Family Characteristics Perhaps not surprisingly, the family characteristics of U.S.-born (second-and-later-generation) youth follow similar patterns to those shown for immigrant youth in the previous chapter. Mexican American youth tend to have low maternal education: Fewer than half have a mother with a high school diploma and only 5 percent have a mother with a bachelor’s degree (Table 5.1). Maternal education is also low among Southeast Asians. Southeast Asian and Latino youth are less likely than Asian and white youth to be growing up in a family with income above the poverty level.1 Family income is particularly low among Southeast Asians with ancestry other than Vietnamese. _____________ 1The 2000 Census measures poverty in 1999. Since 1999, a recession and unstable economic conditions have led to mild growth in poverty (Reed, 2004). 41 Table 5.1 Percentage Distribution of the Family Characteristics of U.S.-Born California Youth by Race and Ethnicity, Ages 13 to 17, 2000 Mother Mother Youth Family Has High Has Mother Lives with Income School Bachelor’s Speaks Both Above Share Diploma Degree English Parents Poverty White 44 93 29 99 71 92 Mexican American 29 46 5 70 66 78 Central American 2 44 7 66 63 79 Vietnamese 1 55 10 55 78 76 Other Southeast Asian 1 21 2 35 77 52 Filipino 2 95 53 99 81 95 Other Asian 4 84 44 77 87 92 African American 8 84 14 99 37 73 Family Income Above Low Income 81 49 47 54 22 86 80 51 SOURCE: Authors’ calculations from the 2000 Census. NOTES: See the notes to Table 4.1. The low-income threshold is twice the federal poverty threshold for each family. Factors Contributing to Racial and Ethnic Differences in Educational Attainment To explore the extent to which these differences in family characteristics explain racial and ethnic differences in educational attainment, we use longitudinal data that measure family resources in the 8th grade and follow youth until roughly age 26. We use a national sample because such longitudinal data are not available for California. However, because of similarities in family resources and youth outcomes by race and ethnicity, the national results appear to be applicable to California (see Appendix A). We focus our discussion on findings for Mexican American youth because of their relatively low educational attainment. Findings for African Americans and Asian Americans are 42 reported in the figures. The survey has too few Southeast Asians for analysis.2 Compared to third-and-later-generation white men (hereafter referred to as “white men” for ease of exposition), third-and-latergeneration Mexican American men were 7 percentage points less likely to complete high school (Figure 5.1).3 Using a statistical model, we adjust for parental education to compare Mexican Americans with whites who have similar parental education backgrounds. After this adjustment, third-and-later-generation Mexican American men have high school completion rates that are not statistically distinguishable from those of white men.4 In other words, parental education and any factors associated with parental education (such as family income, family size, or even neighborhood characteristics) explain the lower high school completion rates of third-and-later-generation Mexican American men. For third-and-later-generation Mexican American women, the gap with white women is substantially reduced when we adjust for parental education and the gap is eliminated when we further adjust for family income, single-parent families, speaking English in the home, family size, and region of residence. Among second-generation Mexican American men, their lower high school completion rates are within the sampling error of the survey. For second-generation Mexican American women, low parental education explains their gap with white women. Overall, then, the lower high school completion rates of Mexican Americans relative to whites is explained by their lower family resources, particularly parental education and related factors. _____________ 2We do not report results for second-generation whites because the NELS sample appears to be not representative. For third-and-later-generation Asians, the NELS sample is too small and also appears to be not representative. See Appendix Table A.4. 3Figure 5.1 and related figures show marginal effects from logistic regressions (measured at the means of other variables). Statistical significance is based on the coefficient estimate. See Appendix C for model results. 4We adjust for parental education by including dummy variables for mother having a high school diploma, attending college, and having a bachelor’s degree plus a dummy for both parents not completing high school and a dummy for both parents having bachelor’s degrees. 43 Percentage point difference with whites 30 Baseline Adjust for parental education Add adjustments for additional family factors 20 Add adjustments for school and neighborhood poverty 10 0 –10 –20 Men –30 Women ssttttseeshhhiihcceieirrMMMMrrcoocAddAeeeedfdfonnor++ArxxxxA+i+iiiiinddnssciccccciggddggggaaaaaaaaeeeegeegnnnnnnnnnnnnenneeeeeAAAAAAAAnrreenrrrraaeeammammmmmttmraartiitttiieeeeeeeeiiaaoorrrrrrtrrtooiiiiiiiiooiinnnccncccccconnoaaaaaaaannnnnnnnnn,,,,,,,, SOURCE: Authors’ calculations from NELS. NOTES: High school completion defined as diploma by about age 20 (six years after 8th grade). Figure shows percentage point differential with third-and-higher-generation whites based on a logistic regression. White-filled bars signify difference with whites is not statistically significant at the 10 percent level. Figure 5.1—High School Completion Rates Relative to Third-and-LaterGeneration Whites In contrast, second-generation Asian men are more likely than white men to complete high school and less than half of the difference can be explained by parental education and other family factors.5 When we add adjustments for whether they attend a high-poverty school and whether they live in a low-income neighborhood, about half of the gap can be explained.6 The high school completion rate among second-generation _____________ 5Several studies have shown the importance of “attitudes” and parenting styles for educational outcomes, especially among Asians (Fuligni, 1997, 1998; Kao and Tienda, 1995; and Kao, 2004). 6We use the share of students who are eligible for a free or reduced price meal as a measure of school poverty. We use poverty rate in the neighborhood of the school. For 44 Asian American women is not statistically different from that of white women. Among young Californians who leave school before finishing high school, the age at leaving school is not substantially different across racial and ethnic groups. For most groups, the share of youth who are not enrolled in school at ages 13, 14, 15, 16, and 17 is 3 percent or less.7 For Mexican Americans, the share out of school is 3 percent or less through age 15, 4 percent for age 16, and 5 percent for age 17. Among those who do not complete high school, the majority have completed 11th or 12th grade (Figure 5.2).8 The share is slightly lower for Mexican Americans (62%) than for whites (70%). Similar to high school completion, enrollment in postsecondary education tends to be lower for Mexican Americans than for whites but the gaps disappear when we adjust for parental education (Figure 5.3).9 Indeed, second-generation Mexican Americans have higher postsecondary school attendance than whites with similar parental education backgrounds. Mexican Americans are substantially less likely than whites to complete a bachelor’s degree (Figure 5.4). The college completion gap disappears after adjusting for family characteristics, except in the case of third-and-later-generation Mexican American women for whom adjustments cut the gap by only about half. For Mexican Americans, gaps with whites in bachelor’s degree completion are much larger than gaps in postsecondary school enrollment (comparing Figures 5.3 and 5.4). Findings from the 2000 ______________________________________________________________ research on neighborhood and school effects, see Duncan and Hoffman (1990), Brewster (1994), Patterson (2000), Wolfe, Wilson, and Haveman (2001), and Light and Strayer (2000, 2002). 7We rely on the 2000 Census for this analysis because the NELS sample for youth not completing high school is too small. 8The 2000 Census allowed respondents to indicate that they completed the 12th grade but did not achieve a high school diploma. For example, a person may have attended an entire year of schooling after the 11th grade but not completed all requirements for a high school diploma. 9Postsecondary school attendance is measured as enrollment in any postsecondary institution. Postsecondary school attendance and achieving a bachelor’s degree are measured in 2000, when most NELS respondents were age 26. 45 50 White 40 Mexican American 30 African American Percentage 20 10 0 8th grade or less 9th grade 10th grade 11th grade 12th grade SOURCE: Authors’ calculations from the 2000 Census. NOTE: The sample is too small to calculate the grade level distribution for Asians and Central Americans. Figure 5.2—Grade Level Completion Rates of U.S.-Born Young California Adults with No High School Diploma, Ages 19 to 24, 2000 Census confirm that not only do Mexican Americans have relatively low college enrollment, but of those who have attended college, the share that graduated with a bachelor’s degree is particularly low for these groups in California and the rest of the nation (Table 5.2).10 For most major racial and ethnic groups, young adults from California are slightly more likely than those from other states to have attended college, although they are not more likely to have completed a bachelor’s degree (Table 5.2).11 Perhaps the large size of the community college system helps explain the higher college enrollment of Californians. In 2002, the California Community College (CCC) system enrolled over 146,000 first-time freshmen compared to less than _____________ 10Measures of bachelor’s degree attainment in the Census (Table 5.2) exceed those of the CPS (Figure 2.1) for each major racial and ethnic group in California and nationally. 11Migration to and from other states to attend college or shortly after finishing college tends to increase college attendance and completion rates for California. To estimate college rates for those who were educated in California, Table 5.2 reports statistics by state of birth regardless of state of current residence (see Reed, 2005). 46 ssttttseeshhhiihcceieirrMMMMrrcoocAddAeeeedfdfonnor++ArxxxxA+i+iiiiinddnssciccccciggddggggaaaaaaaaeeeegeegnnnnnnnnnnnnenneeeeeAAAAAAAAnrreenrrrraaeeammammmmmttmraartiitttiieeeeeeeeiiaaoorrrrrrrtrtooiiiiiiiiooiinnnccncccccconnoaaaaaaaannnnnnnnnn,,,,,,,, Percentage point difference with whites Baseline 40 Adjust for parental education Add adjustments for additional family factors 30 Add adjustments for school and neighborhood poverty 20 10 0 –10 –20 Men Women –30 SOURCE: Authors’ calculations from NELS. NOTES: Postsecondary school attendance is defined as enrollment in any postsecondary educational institution as of about age 26. See the notes to Figure 5.1. Figure 5.3—Postsecondary School Attendance Relative to Whites 30,000 in the University of California (UC) system and less than 40,000 in the California State University (CSU) system (Table 5.3).12 The representation of Latino students in the CCC system matches that of public high school graduates in California (Table 5.3), whereas Latinos are underrepresented in the CSU and UC systems. Owing to the large size of the CCC system and the relatively large share of Latinos entering CCCs, almost 80 percent of all Latino freshman in public _____________ 12Enrollment data include first-time freshman enrollees from private high schools and from out of state. If limited to graduates from California public high schools, the share of whites in UC would be slightly lower (34%) and the share would be higher for Latinos (15%) and Asians (35%). 47 Percentage point difference with whites Baseline Adjust for parental education Add adjustments for additional family factors 40 Add adjustments for school and neighborhood poverty 30 20 10 0 –10 –20 Men –30 Women ssttttseeshhhiihcceieirrMMMMrrcoocAddAeeeedfdfonnor++ArxxxxA+i+iiiiinddnssciccccciggddggggaaaaaaaaeeeegeegnnnnnnnnnnnnenneeeeeAAAAAAAAnrreenrrrraaeeammammmmmttmraartiitttiieeeeeeeeiiaaoorrrrrrtrrtooiiiiiiiiooiinnnccncccccconnoaaaaaaaannnnnnnnnn,,,,,,,, SOURCE: Authors’ calculations from NELS. NOTES: College completion is defined as attaining a bachelor’s degree as of about age 26. See the notes to Figure 5.1. Figure 5.4—College Completion Ratio Relative to Whites higher education in California are in the CCC system.13 Thus, from a policy perspective, the success of students in the CCC system is an important determinant of the success of Latino youth, an issue we return to in the concluding chapter. Transfer rates from CCC are fairly low, with about 48,000 students per year transferring to CSU, about 11,000 students per year transferring to UC, and about 8,000 students per year transferring to private institutions (California Postsecondary Education Commission, 2002). _____________ 13See also Lopez, Puddefoot, and Gándara (2000). Fry (2002) documents the disproportional enrollment of Hispanics in community colleges at the national level. 48 Table 5.2 Percentage Distribution of College Attendance and Completion Rates of U.S.-Born Youth by Race and Ethnicity, Ages 25 to 29, 2000 Born in California Born Elsewhere in the U.S. Some Bachelor’s Graduation Some Bachelor’s Graduation College Degree Rate College Degree Rate White 69 31 44 65 32 49 Mexican American 48 12 26 45 13 28 Central American 65 23 35 65 27 41 Vietnamese 63 31 50 Other Southeast Asian Filipino 82 39 48 84 48 57 Other Asian 91 66 72 88 70 79 African American 54 15 28 48 15 31 SOURCE: Authors’ calculations from the 2000 Census. NOTES: The column labeled “Graduation Rate” shows the percentage with a bachelor’s degree of those who attended college. The sample size is too small to calculate statistics for Vietnamese in California and for other Southeast Asians nationally. Table 5.3 Percentage Distribution of Students in Higher Education Institutions in California by Race and Ethnicity, 2002 Public High School CCC CCC CSU CSU UC UC Graduates Freshmen Transfers Freshmen Graduates Freshmen Graduates White 43 39 42 39 43 37 40 Latino 33 33 20 23 20 14 12 Filipino 3 4 36 4 5 4 Other Asian 12 10 14 14 14 33 30 African American 7 7 57 5 3 3 Number 323,573 146,240 49,029 38,710 59,025 29,664 34,716 SOURCE: Reed (2005), based on California Postsecondary Education Commission Data. NOTES: CCC transfers include transfers to private institutions. Columns do not sum to 100 percent because some students do not identify their race or ethnicity. 49 The share transferring to four-year institutions is particularly low for Latinos (Table 5.3). One explanation for low transfer rates is that students attend CCCs for other programs such as workforce training, vocational or occupational education, and remedial education. Thus, the CCC system provides important opportunities for educational attainment beyond simply the transfer function. Research on the efficacy of community college provides mixed results. In an analysis of research based on the NELS, Fry (2004) finds that Latino students are more likely than similarly prepared white students to choose less-selective and “open” admissions colleges such as community colleges. Fry argues that because selective colleges tend to have higher graduation rates, more Latino students would earn bachelor’s degrees if they were to choose the same colleges.14 In contrast, Sandy, Gonzalez, and Hilmer (forthcoming) find that differences in student quality (and not institutional quality) explain most of the lower bachelor’s degree completion rates for students who attend community college. Furthermore, Gonzalez and Hilmer (forthcoming) find that access to a community college increases the chances that Latino students finish a bachelor’s degree. Apparently, although some Latino students and others may have been hindered in their progress at community colleges (e.g., by lack of counseling, inability to transfer units, or limited availability of courses), many others have found that community colleges promote their progress by providing greater access through low fees, flexible schedules, and local availability. California Youth At Risk of Low Educational Attainment Estimates of the relationship between family resources and young adult educational attainment can be used to describe the size and nature of the California population at risk for low attainment in the future. If the relationships measured in the NELS were to continue to hold, we would _____________ 14Fry (2004) also finds that Latinos have lower graduation rates than whites when comparing those in the same colleges. He argues that lower completion rates may be related to Latinos’ delayed enrollment in college, greater financial responsibility for family members, and living at home rather than in campus housing. 50 expect about 17 percent of California’s children ages 0 to 9 to become high school dropouts (Table 5.4). Of course, California has had many educational reforms over the last decade and further reforms are likely.15 However, conditions for higher education may erode, particularly in light of current fiscal constraints and the growing number of college-age youth (Hanak and Barbour, 2005). Other changes, such as income growth or large waves of new immigrants, may also affect educational and other outcomes. Therefore, the statistics in Table 5.4 do not tell us what will happen but do provide a sense of the size and nature of the child population most at risk for low attainment. The analysis suggests that the children most at risk of low attainment are those whose parents have not completed high school. Roughly 30 percent of California children have parents with no high school diploma. Table 5.4 Percentage Distribution of Educational Attainment by Family Characteristics, Ages 0 to 9, 2000 Family Characteristic All Neither parent has a high school diploma Mother has a high school diploma Mother attended college Mother has a bachelor’s degree Both parents have a bachelor’s degree Parents do not speak English Family income < $25,000 Raised by single parent Share of Children 30 17 26 7 11 12 29 23 No High School Diploma 17 36 15 7 2 <1 29 31 29 No Bachelor’s Degree 68 95 82 61 15 4 91 88 83 SOURCE: Authors’ calculations from the NELS (1994, 2000) and the 2000 Census. NOTES: Children living with a single parent who has not completed high school are included in “Neither parent has a high school diploma.” “No high school diploma” is estimated at age 20 and “No bachelor’s degree” is estimated at age 26. See Appendix A for a discussion of measurement issues. _____________ 15See Reed (2005) for a review of recent educational reforms in California. 51 If the relationships measured during the 1990s were to hold, 36 percent of these children would drop out of high school and 95 percent would not receive a bachelor’s degree (by age 26). The risk of low attainment is also very high among children whose parents do not speak English, families with low income, and children being raised by a single parent. Among California children at risk of low attainment, Mexican Americans make up a disproportionate share. For example, among children with parents who do not have a high school diploma, Mexican Americans make up 68 percent, whereas among all children that number is 40 percent. Thus, this analysis provides further evidence that we should expect the relatively low educational attainment of Mexican Americans to continue in the next generation (see also Chapter 3). 52 6. Conclusion Our results have much to say about what we can expect for the future of California. We find evidence of substantial educational progress between second-generation immigrants and their foreign-born parents—progress that continues between the second and third generations. These findings are particularly important because prior research comparing same-age cohorts of second- and third-generation immigrants has found that second-generation immigrants tend to have higher educational attainment, suggesting the possibility that educational progress stalls between the second and third generations. Using the recent record as a basis, we project a high degree of educational progress for the third-generation grandchildren of today’s Mexican immigrants. Despite continued intergenerational progress, educational attainment for third-and-later-generation Mexican Americans still lags that of U.S.-born whites. Without future improvements, Mexican American generations in years to come will fall far short of the educational levels currently attained by U.S.-raised whites.1 The stakes are high—13 percent of California’s youth are immigrants from Mexico and another 21 percent are U.S.-born Mexican Americans. In addition, Central American immigrants, whose educational attainment is somewhat similar to that of Mexican immigrants, are growing in number (currently representing 2% of the youth population). Low educational attainment is of particular concern in light of the growing importance of education in the California labor market. Today, more so than in past decades, economic success is strongly related to educational attainment. Education can also be linked to health outcomes and civic participation. Thus, education is arguably one of the most important factors for the integration of immigrants into California. _____________ 1See López and Stanton-Salazar (2001) for a stark perspective on the future for Mexican Americans. 53 Policy Directions The findings of this study have important implications for education-related policy in California. Other areas of public policy, such as federal immigration policy and state health care policy, are relevant for immigrant integration, but these broader policy issues were not the focus of this study and we do not discuss them here.2 Improving educational attainment and other conditions among Mexican immigrants who arrive as youth is particularly challenging. Among young adult Mexican immigrants in California, almost three of every four arrived after age 14. Many of these youth are not enrolled in California schools. Thus, youth programs that operate in and through schools will not reach these Mexican youth, who have some of the lowest educational attainment in California. One approach is to encourage them to enroll in high schools. For those who do, English language training and educational counseling may be particularly important. However, a substantial share of those who are not in school are working and raising families. For these youth, adult education programs in school districts and community colleges may provide better schedules for part-time, evening, and weekend coursework in vocational or academic subjects.3 Another approach would be to target the workplace, potentially working with employers who may see advantages to workers developing English language or literacy skills. In addition, as these youth become parents, policies could be devised to engage them through their young children, by offering parental support and literacy courses.4 For second and third generations, and for immigrants who do enter California schools, the quality of the K–12 public education system is clearly a key factor in success. Recent research suggests that the quality of public schools in California is relatively low (Carroll et al., 2005; The Civil Rights Project, 2005) and that Latino and African American students are more likely than other groups to be in low-performing _____________ 2The Little Hoover Commission (2002) describes a broad range of state and local policy approaches for integrating immigrants in California. 3See de Cos (2004) for a study of adult education in California. 4For example, the First Five commissions in California support a variety of parenting-support services, as do the federal Head Start programs. 54 schools (Reed, 2005). Several recent reforms have sought to improve California schools, particularly in the areas of student achievement, teacher quality, and adequate facilities (Reed, 2005). In addition to these basic elements of school quality, English language learning is of particular concern for the children of immigrants.5 One additional approach would be to target students from poor families for educational counseling, tutoring, and other mentoring.6 For example, such programs could target children whose parents have not completed high school. We find that about 30 percent of California’s children are growing up in families where neither parent has completed a high school diploma and that as many as 95 percent of these children might not achieve a bachelor’s degree.7 Mexican American children make up 68 percent of these at-risk children but they constitute only 40 percent of all children. Almost 80 percent of Latinos who enroll in public higher education enter through the CCC system. Through open admissions policies, low fees, flexible schedules, and local availability, community colleges provide greater access to a college education. Of great concern, however, is the low transfer rate to four-year institutions; transfers are especially low among Latino students.8 In addition to preparing students for transfers, community colleges provide English language, remedial, and vocational courses. As the value of education and skills in the California economy continues to grow, these courses will become increasingly important to _____________ 5See Jepsen and de Alth (2005) for analysis of English language learners in California schools. See also Rumbaut and Cornelius (1995). 6The Advancement Via Individual Determination (AVID) program is an example of a program that targets middle and high school students for additional mentoring and support. See Torres and Marquez (2005) for a discussion of parental outreach programs for postsecondary education. 7Head Start is an example of a program targeting poor children early in their educational careers. Wald and Martinez (2003) suggest an alternative targeting approach focused on youth who have poor outcomes during their teens, including those not completing high school, those deeply involved in the criminal justice system, young unmarried mothers, and adolescents in the foster care system. 8See California Postsecondary Education Commission (2002) for a summary of recent policies to address student transfer from CCC. 55 workforce training, especially for those who do not go on to complete a bachelor’s degree. Directions for Future Research The analysis of this report suggests important areas for future research. Although education policy is clearly a target area for promoting the integration of immigrants in California, research has not provided clear direction on specific policies that will improve school quality and further educational attainment, particularly among disadvantaged youth. Ongoing and future research at PPIC is working to address specific policy options in the areas of K–12 financing and reform, community college education, and adult English language training. Policy research on out-of-school immigrant youth who arrive in their teenage years is important as they represent a substantial share of California’s youth and their children will be a substantial share of the next generation of Californians. Their activities, attainment, and resources are different from those of youth who were born in the United States. Future PPIC research will examine resources, needs, and policies for these immigrant youth. Finally, California is undergoing major demographic transitions as the second-generation population continues to grow. PPIC will be developing projections of the demographic and socioeconomic conditions of California’s future population. These projections have important implications for California policy in the areas of education and workforce training, health and human services, and civic participation. 56 Appendix A Notes on Data and Methods Throughout this report we have relied on several sources of data as well as a variety of methods. In this appendix, we provide further details on the data and methods. The appendix is arranged by chapter and, within chapters, generally follows the order that data and methods are discussed in the main text. Chapter 2 Information on the number of youth by country of birth, ancestry, race, and ethnicity comes from the 2000 Census (5-percent Public Use Microdata Sample).1 We include as “U.S.-born” those who were born in one of the 50 United States and those who were born abroad of American parents. Youth born in Puerto Rico or other outlying areas of the United States are included as foreign-born. For detailed ancestral backgrounds (Table 2.1 and Appendix Table B.2), we used the first of two ancestries recorded. In some cases, we aggregated similar responses (for example, combining “Mexican” with “Mexican American”). In cases where youth identified a single Hispanic or Asian identity in response to “Hispanic origin” and “Race” questions, we used these to determine detailed ancestry. Table 2.2, which shows aggregated race and ethnicity, is based mainly on responses to “Hispanic origin” and “Race” questions with two exceptions. First, because a substantial number of Hispanics did not identify a specific Hispanic background, we used detailed ancestry to allocate these youth first to Mexican American, then to Central American, and finally to “all other Latinos.” Second, for Southeast Asians, we used ancestry to include _____________ 1Although the Census Bureau attempts to include undocumented immigrants in the Census and Current Population Survey, the number of immigrant youth is likely to be underreported. 57 Cambodian, Khmer, Hmong, Laotian, Meo, Western Lao, Vietnamese, Katu, Ma, and Mnong. The 2000 Census does not have information that allows us to separately identify second-generation U.S.-born youth from later generations. We use the CPS (March Annual Social and Economic Supplement, 1996–2004 combined)2 to estimate these shares among U.S.-born youth, separately for each row of Table 2.2.3 We identify the second generation as those with either parent born outside the 50 United States. We identify “third-and-later generations” as those with both parents born in one of the 50 United States. We then applied these shares to the count of U.S.-born youth in the Census by race and ethnicity. All analysis with the Census and CPS uses the sample weights (individual-level weights). Rumbaut (2004) provides a detailed discussion of the complexity of using the CPS to identify and describe the first and second generations. A serious limitation of the CPS is that it includes only the noninstitutionalized, civilian adult population. As noted by Rumbaut (2004) and Blumstein (2001), experiences with the criminal justice system differ substantially across racial, ethnic, and immigrant groups. Another issue that affects our analysis is that later generations and youth of mixed parentage are less likely to identify a specific ancestry. Chapter 3 Developing estimates of immigrant intergenerational progress is made difficult by the lack of large datasets that record an individual’s nativity, his/her parents’ nativity, and his/her grandparents’ nativity. Immigration researchers have been frustrated by this lack of data on the _____________ 2We do not use the CPS for 1994 and 1995 because of concerns raised by Jeffrey Passel at the Urban Institute regarding the representativeness of race, ethnicity, and immigrant generation in these years. 3The CPS does not identify detailed ancestry. We use “Hispanic origin” to identify subgroups of U.S.-born Hispanics. We separate Central American immigrants from South American immigrants using country of birth. For Southeast Asian immigrants, we use country of birth of youth and parents (Vietnam, Laos, and Cambodia). Many Southeast Asian youth whose parents were refugees from these countries were themselves born in Thailand and Malaysia. We identify these youth from the birthplace of their parents. 58 place of birth of a U.S. native’s parents and grandparents (Hirschman, 1994). The primary sources of information on the population of the United States, the Decennial Censuses and the CPS (March Annual Social and Economic Supplement), do not include information on place of birth of an individual’s grandparents. The last three decennial Censuses have not even collected information on the place of birth of an individual’s parents. Beginning in 1994, the CPS has included information on the place of birth of an individual’s parents but still does not collect information on grandparents’ nativity. Such information is necessary to identify the third-generation descendants of immigrants. The General Social Survey (GSS), a much smaller survey than the CPS and conducted only in English, does collect information on foreign birth of an individual’s parents and grandparents and also collects information on ancestry but not specific country of birth. We used the GSS to estimate intergenerational change for Mexican Americans and found that the results are consistent with those presented in this report (see Johnson, 2005). In this chapter, we use recent CPS data to estimate changes in educational attainment across three generations of immigrants and their descendants. We use the following definitions of generations: The first generation consists of anyone born outside the United States (not of American parents) and who came to the United States at age 11 or older; the 1.5 generation consists of those foreign-born individuals who came to the United States as young children; the second generation consists of U.S.-born individuals with at least one foreign-born parent; and the third generation consists of individuals with at least one foreign-born grandparent. We develop two samples from the CPS data to estimate intergenerational changes. In one sample, we consider only those adults who live with their parents. This multigenerational household sample links parents with their children and provides for direct identification of first, second, and third generations. Specific countries of origin can be identified in the public use data for an individual and his/her parents, allowing for the identification of the first- and second-generation descendants of immigrants by country of origin. Third-generation descendants of immigrants can be directly identified by examining the place of birth of the parents of the adult child’s parents, something that 59 can be done only if the adult child lives with his/her parents. In addition, because the CPS includes data on the educational attainment of all household members, information on the educational attainment of the adult child and his/her parents is available in the multigenerational samples of the CPS. This permits an analysis of generational changes that compares an individual with his/her parents. Moreover, aggregation bias in ethnic categories is not a problem with this sample: Because we compare adult children with their parents, ethnic groups are consistently represented in the different generations. The tables show the average of educational attainment for mothers and fathers of a previous generation in comparison to that of their adult children. The primary concern with the multigenerational sample is whether residents of such households are representative of all immigrants and their descendants. On the one hand, we might think that living in an extended family is an indication that an individual has not adopted U.S. norms, as such households are uncommon among U.S. natives. In addition, adult children might live with their parents because they lack the resources to establish their own households. Both of these factors suggest that such individuals would have lower educational attainment. On the other hand, living in an extended family might be a marker of a strong and supportive family environment or an indication that an adult child has the resources to care for his/her parents. If this is the case, then we might expect such individuals to have higher levels of education. Results from the data suggest that adult children who live with their parents have similar educational attainment profiles as adults who do not live with their parents. In the other sample, we compare all individuals of one generation with all individuals from another generation. In this much larger sample, parents and children are not linked, and the third generation must be identified indirectly. Indeed, third-and-later-generation descendants of immigrants must be grouped together. Any adult born in the United States to parents born in the United States is assigned to the third-and-later generation. Many, if not most, of the third-and-later generation are fourth-and-later generations. Third-and-later-generation descendants of Mexican and Central/South American immigrants are identified on the basis of ancestry; specifically, those U.S. natives who 60 report one of five Mexican or Central/South American ethnicities and whose parents were born in the United States constitute the third-andlater generation.4 Third-and-later-generation descendants of Asian immigrants are identified on the basis of race; specifically, those U.S. natives who report their race as Asian or Pacific Islander and whose parents were born in the United States constitute the third-and-later generation. Aggregation bias could be a problem with the Asian category, with Asian subgroups possibly disproportionately represented in the different generations. This approach produces a much larger sample than the multigenerational sample but suffers from not directly linking parents with their descendants. In addition, third-and-latergeneration assignment is dependent on responses to race or ancestry questions, and some descendants of first-generation immigrants might not identify with that immigrant’s race or ancestry. In other work on descendants of Mexicans, we show that this most likely causes a downward bias of third-generation educational attainment. We are careful to identify the age groups in combination with the immigrant generation that provide for the most appropriate comparison of one generation with a previous generation. We do so by comparing older members of earlier generations with younger members of subsequent generations—that is, by comparing a generation that includes parents with a subsequent generation that includes their children. In both of our samples, we consider the largest immigrant groups to California and the United States. Data are from the 1996–2004 CPS (combined surveys). Sample sizes are shown below (Tables A.1 to A.3). Samples for some groups are so small (less than 100) that we did not publish results. For Central Americans and Southeast Asians, we combine the 1.5 generation with the second generation to evaluate changes that occur between immigrants and their children. This is necessary because of the small samples for these groups. _____________ 4The CPS groups Central Americans with South Americans in its reporting of ancestry. 61 Table A.1 Sample Sizes for Data Used in Table 3.1 All ethnic groups First generation, ages 57–66 Second generation, ages 30–39 Mexican First generation, ages 57–66 Second generation, ages 30–39 Central American First generation, ages 57–66 Second generation, ages 30–39 Southeast Asian First generation, ages 57–66 Second generation, ages 30–39 Other Asian First generation, ages 57–66 Second generation, ages 30–39 White (non-Hispanic) First generation, ages 57–66 Second generation, ages 30–39 Black or African American First generation, ages 57–66 Second generation, ages 30–39 All ethnic groups Third+ generation, ages 57–66 Third+ generation, ages 30–39 15,004 14,191 2,708 2,825 503 310 455 169 2,635 809 4,035 7,076 875 384 86,973 176,757 NOTE: Central American and Southeast Asian second-generation samples include the 1.5 generation. Table A.2 Sample Sizes for Data Used in Tables 3.2 and 3.4 First:second generation Second:third generation Total 3,577 8,645 Mexican 815 890 Central American 176 Insuf. Southeast Asian 113 Insuf. Other Asian 352 430 White 1,303 4,984 NOTES: The Central American and Southeast Asian second-generation samples include the 1.5 generation. “Insuf.” indicates that the sample size was less than 100, and no results are provided. 62 Table A.3 Sample Sizes for Data Used in Table 3.3 Mexican Second generation, ages 57–66 Third+ generation, ages 30–39 Asian Second generation, ages 57–66 Third+ generation, ages 30–39 Black or African American Second generation, ages 57–66 Third+ generation, ages 30–39 White (non-Hispanic) Second generation, ages 57–66 Third+ generation, ages 30–39 All ethnic groups Second generation, ages 57–66 Third+ generation, ages 30–39 Third+ generation, ages 57–66 1,228 6,030 384 1,287 Insuf. 20,230 7,723 143,591 9,795 176,757 86,973 NOTE: “Insuf.” indicates that the sample size was less than 100, and no results are provided. Finally, using the historical progress across generations with respect to educational attainment as a basis, we develop projections of educational attainment for the descendants of the most recent immigrants to the United States. To develop the projections, we use transition probabilities based on an individual’s generation, maternal education, and region of origin. These transition probabilities are derived from the CPS multigenerational household sample that links parents with their children. Thus, we have education levels for both the adult child and his/her parents. The transition probabilities are developed separately for first- to second-generation changes in educational attainment, and second- to third-generation changes. We report the results only for descendants of Mexican immigrants. Five educational attainments are measured—less than 9th grade, 9th to 12th grade but no high school diploma, high school diploma, some college, and bachelor’s degree—but only four are reported. The accuracy of the projections depends on the stability of the transition probabilities over 63 time. In addition, the method assumes that the transition probabilities we observe in multigenerational households are an accurate representation for all second- and third-generation individuals, regardless of whether they live in a multigenerational household. Chapter 4 Analysis in this chapter, focused on foreign-born youth, is based on the 2000 Census. We chose the 2000 Census over the CPS because of the larger sample size. Identification of race and ethnicity follows the same strategy as in Chapter 2. However, among these immigrant youth, Hispanic subgroups, Vietnamese, and Filipinos are identified by country of birth. Southeast Asians are identified by ancestry and country of birth (including, for example, immigrants of Vietnamese descent who were born in Thailand). Information on maternal education and English language ability is available only for those youth living with their mother. For this reason, we report family characteristics for youth ages 13 to 17—ages when most immigrant youth (79%) are living with their mother. The share living with their mother is lower among Mexican and Central American immigrants (about 75%) and higher among white immigrants (90%). Chapter 5 Analysis of family characteristics and other topics, focused on U.S.born youth but not by immigrant generation, is based on the 2000 Census. We chose the 2000 Census over the CPS because of the larger sample size. Identification of race and ethnicity follows the strategy described for Chapter 2. Vietnamese and Filipinos are separately identified based on “Race” (using any combination that includes Vietnamese or Filipino, respectively) and ancestry. Information on maternal education and English language ability, available only for youth living with their mother, is available for 86 percent of U.S.-born youth ages 13 to 17. The share living with their mother is lower among African Americans (77%). 64 To examine the role of family characteristics in explaining differences in youth attainment, we use the 1988 NELS.5 The NELS is a nationally representative sample of 8th-graders who were first surveyed in the spring of 1988. Follow-up surveys were conducted in 1990, 1992, 1994, and 2000. The NELS is a valuable data source for this study because it is longitudinal—we observe characteristics (e.g., parental education) at early ages that can then be related to attainment of the same youth at later ages. In addition, the NELS covers enough years to allow observation of young adult attainment, such as completion of a bachelor’s degree by age 26. For this reason, we chose not to use more recent longitudinal surveys such as the National Longitudinal Survey of Youth from 1997 and the Adolescent Health Survey. However, in follow-up work on this issue, we do expect to use these newer surveys. Measures from the NELS are not defined exactly the same as those from the Census or CPS. In general, the NELS has more detailed information for the specific cohort. For example, to measure high school completion in NELS, we use high school diploma received by 1994 when most respondents are approximately age 20.6 In comparison, using the Census we measured high school diploma or GED received for those ages 25 to 29. Relative to the 2000 Census and the CPS, the NELS measures unusually high numbers of students who report having enrolled in postsecondary education. The difference appears to be due to a difference in the surveys. In the Census and CPS, the question on educational attainment refers specifically to college and to college degrees. In the NELS, the question refers to “college, university, or vocational, technical, or trade school for academic credit.” The NELS _____________ 5For other studies of determinants of youth outcomes, see Haveman and Wolfe (1995) and Pirog and Magee (1997). For studies of differences between the foreign-born and the native-born, see Mare and Winship (1988), White and Kaufman (1997), and Fligstein and Fernandez (1985). For studies of the relationship between outcomes and behaviors such as drug and alcohol use, crime, gang activity, and youth employment, see Chuang (1997), Mensch and Kandel (1988), Eckstein and Wolpin (1999), Schoenhals, Tienda, and Schneider (1998), and Moore and Zaff (2002). 6This is the same definition of high school graduation used in Grogger and Trejo (2002). Graduation is defined as of 1994 because most respondents were still in high school when the previous survey was administered in the spring of 1992. Graduation does not include those with a General Equivalency Diploma. 65 also has several follow-up questions and thus is more likely to capture enrollments that respondents may fail to mention (e.g., because they were of short durations) in the Census and CPS. Therefore, the magnitudes of group differences as measured in NELS are not directly comparable to those shown in Chapters 2 and 5. Following Grogger and Neal (2000), we use NELS sample weights in our estimates (sample weights are specific to each year of the survey). Grogger and Trejo (2002) compare the NELS with the CPS and conclude that the NELS is representative of the U.S.-born population and immigrants arriving by age 5 for whites, African Americans, and Mexican Americans. They do not consider Asian Americans or separately consider second versus third-and-later generations. We compared the 2000 NELS and the CPS and found the NELS to be fairly representative of third-and-later-generation whites, second- and thirdand-later-generation Mexican Americans, second-generation Asian Americans, and African Americans (Table A.4).7 Among Asians, the NELS does not have adequate samples of any subgroup. We therefore model an aggregated second-generation Asian category, including Southeast Asians.8 For third-and-later-generation Asian Americans, the sample size was quite small and appeared to be not representative. For second-generation whites, the sample was also apparently not representative. We did not analyze “Other Hispanics” because the sample size was relatively small and the group of the greatest relevance for California, Central Americans, was not separately identified. Kao (2004) and Kao and Tienda (1995) use the NELS to study the academic achievement of high school students by immigrant generation, including first-generation immigrants. However, the NELS is not representative of most first-generation immigrant youth (who never attended the 8th grade in the United States, with many arriving after age 14). We include immigrants arriving before age 5 with the second generation based on their similar characteristics (Hill, 2004) and to _____________ 7We do not include second-generation blacks or Native Americans because of their small sample sizes in the NELS. 8The sample of Southeast Asians is small enough that dropping them from the analysis does not change the results. 66 Table A.4 Sample Sizes and Percentage Distribution of College Completion Rates in the NELS (2000) and CPS (1996–2000) Men White, second generation White, third+ generation Mexican American, second generation Mexican American, third+ generation Other Hispanic, second generation Other Hispanic, third+ generation Asian American, second generation Asian American, third+ generation African American, third+ generation Women White, second generation White, third+ generation Mexican American, second generation Mexican American, third+ generation Other Hispanic, second generation Other Hispanic, third+ generation Asian American, second generation Asian American, third+ generation African American, third+ generation NELS Sample Bachelor’s Size Degree 203 5,368 211 235 128 91 240 69 855 215 4,710 261 253 132 96 274 69 731 35 28 7 13 37 22 62 18 11 61 36 14 14 27 13 59 28 23 NOTE: CPS results are for those ages 25 to 29. Difference CPS in Sample Bachelor’s Bachelor’s Size Degree Degree 1,498 38 3 25,836 31 3 1,355 12 5 1,488 12 –1 966 20 –16 506 18 –5 614 55 –7 259 44 26 3,585 14 2 1,615 44 –17 28,335 35 –1 1,530 12 –2 1,653 12 –2 1,139 21 –6 603 18 4 570 57 –2 274 46 18 5,243 16 –7 increase the sample sizes for the second generation. We remove immigrants arriving at age 5 or older. To examine high school completion, we use the 1994 survey to measure “high school diploma” at a time when most respondents were about age 20. For postsecondary schooling and bachelor’s degree completion, we use the 2000 survey, when most respondents were about age 26. For detailed model results from the NELS logistic regression analysis, see Appendix C. We use the NELS at the national level because the California subsample is small and not necessarily representative. We include a California indicator (dummy variable) in the models along with other 67 regional indicators. We argue that the model results are likely to be relevant for California because racial and ethnic differences in education are similar for California youth and youth from the rest of the nation (see Table 5.2 and Reed, 2005).9 To verify the applicability of NELS model estimates to California youth, we use model coefficients from the NELS to project the bachelor’s degree completion rates of Californiaborn children ages 6 to 9 in the 1980 Census.10 We then compare these projections to bachelor’s degree attainment for California-born young adults ages 26 to 29 in the 2000 Census. There are many reasons why the projections would not match the 2000 Census, including that measures of education and income differ in the two surveys. However, the results confirm that the NELS models are applicable to California (Table A.5). For the estimations in Table 5.4, we followed the same approach, using California children under age 10 in the 2000 Census. The NELS allows us to assess parent-to-child educational progress without limiting the analysis to young adults who live with their parents (as was necessary for the analysis in Chapter 3). The analysis in Chapter Table A.5 Percentage Distribution of NELS Model Projections Versus Actual Bachelor’s Degree Attainment for Californians by Race and Ethnicity White Mexican American Asian African American NELS Projection 35 11 53 19 2000 Census 31 13 58 16 SOURCES: Authors’ calculations from NELS (2000) and the 1980 and 2000 Censuses. _____________ 9Asians, especially Filipinos, from California are somewhat less likely to achieve a bachelor’s degree than are their counterparts nationally. 10For these NELS models, we combined immigrant generations because the Census does not distinguish between second and third-and-later generations. Instead, we add English language proficiency of the parents to the NELS models. 68 3 relies on the CPS to study more racial and ethnic groups than are represented in the NELS. In particular, as shown in Table A.4, the NELS is not representative of third-and-later-generation Asian Americans. However, as a check that our findings on intergenerational progress are robust, we show here that all of the groups represented in the NELS demonstrate substantial educational progress from parent to child (Tables A.6 and A.7). Table A.6 Percentage Distribution of Educational Attainment by Generation: Mother-to-Child Ethnic Group All ethnic groups Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Mexican Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Asian Foreign-born parents Second-generation youth White (non-Hispanic) Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Black or African American Second+-generation parents Third+-generation youth Less Than High School Diploma High School Diploma Some College Bachelor’s Degree 31 19 31 19 6 8 48 39 11 29 42 18 5 16 46 33 65 15 18 2 14 14 61 11 25 31 40 4 10 15 59 16 16 15 28 41 1 2 33 64 11 21 44 24 2 6 41 51 9 30 41 20 5 15 43 36 15 24 52 10 6 18 57 20 SOURCES: Authors’ calculations from the NELS (2000). NOTE: In our analysis of the NELS, second-generation youth also include foreign-born youth who arrived in the United States by age 5. 69 Table A.7 Percentage Distribution of Educational Attainment by Generation: Father-to-Child Ethnic Group All ethnic groups Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Mexican Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Asian Foreign-born parents Second-generation youth White (non-Hispanic) Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Black or African American Second+-generation parents Third+-generation youth Less Than High School Diploma High School Diploma Some College Bachelor’s Degree 30 15 27 28 6 8 46 40 12 22 37 30 4 15 44 38 72 6 20 2 13 19 57 12 28 22 39 12 8 11 63 18 4 17 29 50 5 2 31 63 10 20 32 39 1 6 39 53 10 22 36 32 4 15 41 40 25 21 41 14 4 16 56 25 SOURCE: Authors’ calculations from the NELS (2000). NOTE: In our analysis of the NELS, second-generation youth also include foreign-born youth who arrived in the United States by age 5. 70 Appendix B Detailed Nativity and Ancestry Table B.1 Number of Youth and Educational Attainment for the Top 26 Countries of Birth Reported by Foreign-Born Californians, Ages 13 to 24, 2000 Mexico Philippines El Salvador Vietnam Guatemala Korea Taiwan India Thailand China Iran Hong Kong Armenia Japan Nicaragua Honduras Laos Canada Russia Peru Indonesia Cambodia Ukraine South Korea England Pakistan Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Population 783,124 76,753 59,612 58,701 42,795 28,228 25,859 23,576 22,822 22,337 16,536 14,819 13,588 12,731 11,479 11,370 11,283 9,250 7,733 7,614 7,114 7,096 6,769 5,982 5,941 5,011 High School Diploma (%) 36 94 46 83 38 97 99 96 90 90 95 98 86 96 70 40 71 97 95 87 94 81 91 97 97 85 Bachelor’s Degree (%) 4 34 5 33 5 56 73 81 52 59 48 70 24 45 12 4 13 62 59 20 60 23 37 59 55 44 SOURCE: Authors’ calculation from the 2000 Census. NOTES: Includes all countries of birth with California populations over 5,000. Education is reported for adults ages 25 to 29. 71 Table B.2 Number of Youth and Educational Attainment for the Top 25 Ancestries Reported by U.S.-Born Californians, Ages 13 to 24, 2000 Mexican African American German Irish English Italian American Filipino White Chinese Hispanic Scotch Irish French European American Indian Norwegian Portuguese Japanese Vietnamese Polish Salvadoran Swedish Korean Dutch Puerto Rican Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Population 1,228,338 310,810 279,195 210,186 178,050 161,383 158,956 107,742 94,380 82,943 72,515 66,807 46,820 46,422 38,530 37,770 37,754 34,913 34,878 34,659 32,360 31,934 31,781 27,848 26,601 High School Diploma (%) 76 86 95 94 96 94 88 95 85 97 83 95 93 97 84 98 94 97 78 96 81 98 96 94 79 Bachelor’s Degree (%) 12 17 40 39 45 41 25 39 20 70 19 43 30 54 13 43 26 55 34 54 15 48 62 42 18 SOURCE: Authors’ calculation from the 2000 Census. NOTES: Includes all ancestries with California populations over 25,000. Specific race or Hispanic identifiers were used in place of “ancestry” when available. For example, a U.S.-born Californian identifying Hispanic background as “Mexican” is included as Mexican regardless of ancestry identification. Ancestry was used when the race or Hispanic identifier was not specific (e.g., “Other Hispanic”). Education is reported for adults ages 25 to 29. 72 Appendix C Detailed Model Results The tables in this appendix provide coefficient estimates from logistic regressions for high school completion by age 20, postsecondary school attendance by age 26, and bachelor’s degree completion by age 26. For a description of the methods, see Appendix A. 73 Table C.1 Logistic Probability Model for High School Completion Men Women Mexican American, second generation Mexican American, third+ generation Black, third+ generation Asian American, second generation Mother has a high school diploma Mother has some college Mother has a bachelor’s degree Neither parent has a high school diploma Both parents have a bachelor’s degree Family income Family income squared Raised in single– parent home Number of siblings Family size School offers vocational education % receiving free lunch at school Neighborhood poverty Constant Observations Pseudo R-squared (1) (2) (3) (1) (2) 0.358+ 0.663** 1.358** 0.129 0.389+ (1.65) (3.00) (5.42) (–0.68) (1.90) –0.205 0.041 0.294 –0.513** –0.141 (–0.97) (–0.19) (–1.16) (–2.68) (–0.71) –0.509** –0.382** –0.144 –0.069 0.17 (6.44) (4.59) (–1.48) (–0.71) (–1.63) 0.87+ 0.924+ 0.935+ 0.065 0.104 (1.85) (1.93) (1.81) (–0.2) (–0.31) 0.452** 0.264* 0.171 0.794** 0.622** (3.55) (2.02) (–1.13) (6.31) (4.83) 1.039** 0.834** 0.83** 0.693** 0.48** (8.14) (6.34) (5.46) (5.96) (3.96) 1.94** 1.342** 1.442** 2.033** 1.471** (10.69) (6.94) (6.75) (9.77) (6.76) –0.945** –0.747** –1.046** –0.741** –0.651** (5.72) (4.42) (5.32) (4.77) (4.11) 0.644** 0.405* 0.442* 1.729** 1.509** (3.78) (2.29) (2.29) (6.02) (5.18) 0.023** 0.016** 0.028** (6.78) (4.20) (8.42) –0.005** –0.002 –0.008** (2.62) (–0.89) (5.15) –0.241* –0.359** 0.026 (2.30) (3.04) (–0.26) –0.079** –0.079** –0.05* (2.99) (2.67) (2.03) –0.026 0.006 –0.015 (–0.89) (–0.19) (–0.52) 0.107 (–0.49) –0.716** (3.36) –0.766+ (1.88) 1.123** 0.495* 2.134** 1.419** 0.331 (9.29) (2.30) (6.56) (12.23) (–1.58) 7,331 7,331 7,331 6,671 6,671 0.0886 0.1145 0.2963 0.1067 0.1423 (3) 0.608* (2.51) –0.113 (–0.48) 0.373** (3.01) 0.123 (–0.33) 0.659** (4.38) 0.559** (3.92) 1.495** (6.27) –0.527** (2.84) 1.594** (5.26) 0.025** (6.46) –0.006** (3.78) 0.066 (–0.59) –0.041 (–1.46) –0.002 (–0.07) –0.319 (–0.78) –0.461+ (1.92) –0.478 (–1.07) 2.526** (5.37) 6,671 0.3415 SOURCE: Authors’ calculations from the NELS (1994). NOTES: T-ratio is given in parentheses. Models 2 and 3 also include regional and California indicators. +Significant at 10 percent. *Significant at 5 percent. **Significant at 1 percent. 74 Table C.2 Logistic Probability Model for Postsecondary School Enrollment Men Women (1) (2) (3) (1) (2) (3) Mexican American, second generation Mexican American, third+ generation Black, third+ generation Asian American, second generation Mother has a high school diploma Mother has some college Mother has a bachelor’s degree Neither parent has a high school diploma Both parents have a bachelor’s degree Family income Family income squared Raised in singleparent home Number of siblings Family size School offers vocational education % receiving free lunch at school Neighborhood poverty Constant Observations Pseudo R-squared 0.531* 0.697** 1.058** 0.411* 0.821** (2.24) (2.89) (4.20) (2.23) (4.00) 0.404+ 0.443+ 0.664* –0.13 0.314 (1.65) (1.76) (2.53) (–0.65) (–1.48) –0.161+ –0.05 0.101 0.388** 0.77** (1.78) (–0.52) (–1) (3.28) (6.03) 2.709** 2.582** 2.73** 0.218 0.342 (3.02) (2.86) (3.00) (–0.66) (–0.99) 0.45** 0.279+ 0.258+ 0.756** 0.53** (3.09) (1.85) (1.67) (6.28) (4.18) 1.263** 0.996** 0.964** 1.432** 1.12** (8.63) (6.57) (6.21) (11.86) (8.73) 3.503** 2.726** 2.703** 3.334** 2.566** (13.08) (9.89) (9.71) (12.63) (9.41) –0.652** –0.444* –0.433* –0.476** –0.323* (3.39) (2.22) (2.12) (3.06) (1.98) 1.964** 1.631** 1.701** 1.607** 1.324** (8.59) (6.98) (7.19) (6.70) (5.36) 0.035** 0.03** 0.041** (9.53) (7.74) (11.63) –0.009** –0.007** –0.012** (4.87) (3.61) (7.94) 0.199 0.224+ –0.055 (–1.57) (1.74) (–0.48) –0.111** –0.107** –0.082** (3.60) (3.45) (3.15) 0.019 0.04 –0.112** (–0.56) (–1.18) (3.74) –0.594** (3.45) –0.953** (4.63) –0.383 (–0.93) 0.108 –0.594* 0.491 0.442** –0.127 (–0.78) (2.43) (–1.6) (3.88) (–0.56) 5,626 5,626 5,626 5,613 5,613 0.1254 0.1528 0.1790 0.1286 0.1834 0.7** (3.29) 0.298 (–1.37) 0.743** (5.54) 0.348 (–1) 0.509** (3.91) 1.118** (8.50) 2.493** (9.06) –0.226 (–1.36) 1.34** (5.34) 0.04** (10.95) –0.011** (7.54) 0.009 (–0.08) –0.084** (3.16) –0.104** (3.39) –0.386+ (1.68) –0.513* (2.23) 1.327** (2.79) 0.572+ (1.73) 5,613 0.2100 SOURCE: Authors’ calculations from the NELS (2000). NOTES: T-ratio is given in parentheses. Models 2 and 3 also include regional and California indicators. +Significant at 10 percent. *Significant at 5 percent. **Significant at 1 percent. 75 Table C.3 Logistic Probability Model for a Bachelor’s Degree Men Women (1) (2) (3) (1) (2) (3) Mexican American, second generation Mexican American, third+ generation Black, third+ generation Asian American, second generation Mother has a high school diploma Mother has some college Mother has a bachelor’s degree Neither parent has a high school diploma Both parents have a bachelor’s degree Family income Family income squared Raised in singleparent home Number of siblings Family size School offers vocational education % receiving free Lunch at school Neighborhood poverty Constant Observations Pseudo R-squared –0.845* –0.303 0.097 –0.325 0.105 (2.08) (–0.73) (–0.23) (–1.38) (–0.42) –0.643* –0.247 –0.093 –0.899** –0.605* (2.02) (–0.75) (–0.27) (3.60) (2.34) –0.809** –0.666** –0.444** –0.209+ –0.156 (6.18) (4.96) (3.20) (1.87) (–1.31) 0.883** 0.872** 0.888** 0.608** 0.677** (3.07) (2.92) (2.87) (2.63) (2.73) 0.57* 0.341 0.283 0.723** 0.67** (2.50) (–1.44) (–1.17) (4.48) (4.01) 1.055** 0.798** 0.706** 1.22** 1.161** (4.77) (3.44) (3.00) (7.83) (7.11) 2.85** 2.267** 2.224** 2.727** 2.353** (12.52) (9.33) (9.01) (16.10) (13.05) –1.11** –0.866* –0.871* –0.954** –0.734** (2.85) (2.14) (2.13) (3.78) (2.82) 0.965** 0.836** 0.838** 1.083** 0.865** (8.68) (6.88) (6.77) (10.62) (7.99) 0.021** 0.014** 0.017** (7.12) (4.69) (6.87) –0.005** –0.003* –0.003** (4.20) (2.30) (3.15) –0.362** –0.357* 0.302** (2.38) (2.30) (2.59) –0.131** –0.112** –0.131** (3.51) (2.94) (4.72) –0.067+ –0.057 –0.02 (1.66) (–1.38) (–0.64) –0.544** (4.40) –1.303** (4.82) –1.33* (2.52) –1.951** –2.128** –0.829* –1.717** –2.326** (8.98) (6.86) (2.38) (11.17) (9.61) 5,626 5,626 5,626 5,613 5,613 0.1748 0.2070 0.2291 0.1614 0.1949 0.18 (–0.7) –0.541* (2.04) –0.014 (–0.11) 0.66** (2.64) 0.635** (3.74) 1.126** (6.78) 2.267** (12.39) –0.677* (2.57) 0.794** (7.21) 0.015** (5.88) –0.002* (2.43) 0.338** (2.87) –0.128** (4.51) –0.014 (–0.45) –0.228+ (1.80) –1.051** (4.75) 0.454 (–1.05) –1.687** (5.90) 5,613 0.2082 SOURCE: Authors’ calculations from the NELS (2000). NOTES: T-ratio is given in parentheses. Models 2 and 3 also include regional and California indicators. +Significant at 10 percent. *Significant at 5 percent. **Significant at 1 percent. 76 References Arias, Omar, and Walter W. McMahon, “Dynamic Rates of Return to Education in the U.S.,” Economics of Education Review, Vol. 20, No. 2, 2001, pp. 121–138. Beavers, Laura, and Jean D’Amico, Children in Immigrant Families: U.S. and State-Level Findings from the 2000 Census, The Annie E. Casey Foundation and the Population Reference Bureau, Washington, D.C., January 2005. Betts, Julian R., The Changing Role of Education in the California Labor Market, Public Policy Institute of California, San Francisco, California, 2000. Blumstein, Alfred, “Race and Criminal Justice,” America Becoming: Racial Trends and Their Consequences, Vol. 2, 2001, pp. 21–31. Borjas, George J., “Long-Run Convergence of Ethnic Skill Differentials: The Children and Grandchildren of the Great Migration,” Industrial and Labor Relations Review, Vol. 47, No. 4, 1994, pp. 553–573. Brewster, K. L., “Race Differences in Sexual Activity among Adolescent Women: The Role of Neighborhood Characteristics,” American Sociological Review, Vol. 59, 1994, pp. 408–424. California Postsecondary Education Commission, Student Transfer in California Postsecondary Education, Sacramento, California, February 2002. Camarota, Steven A., “Immigration from Mexico: Assessing the Impact on the United States,” Center for Immigration Studies, Washington, D.C., 2001. Card, David, John DiNardo, and Eugena Estes, “The More Things Change: Immigrants and the Children of Immigrants in the 1940s, 77 the 1970s, and the 1990s,” National Bureau of Economic Research Working Paper No. 6519, Cambridge, Massachusetts, 1990. Carroll, Stephen J., Cathy Krop, Jeremy Arkes, Peter A. Morrison, and Ann Flanagan, California’s K–12 Public Schools: How Are They Doing? The RAND Corporation, Santa Monica, California, 2005. Chapa, Jorge, “The Myth of Hispanic Progress: Trends in the Educational and Economic Attainment of Mexican Americans,” Journal of Hispanic Policy, Vol. 4, 1990, pp. 3–18. Chavez, Linda, Out of the Barrio: Toward a New Politics of Hispanic Assimilation, Basic Books, New York, New York, 1991. Chiswick, Barry R., and Noyna DebBurman, “Educational Attainment: Analysis by Immigrant Generation,” Economics of Education Review, Vol. 23, No. 4, 2004, pp. 361–379. Chuang, Hwei-Lin, “High School Youths’ Dropout and Re-Enrollment Behavior,” Economics of Education Review, Vol. 16, No. 2, 1997, pp. 171–186. The Civil Rights Project, Harvard University, “Confronting the Graduation Rate Crisis in California,” 2005, available at www. civilrightsproject.harvard.edu/research/dropouts/dropouts05.pdf. de Cos, Partricia, Educational Opportunities for Adults in California, California Research Bureau, Sacramento, California, February 2004. Denton, Kristin, Jerry West, and Jill Walston, Reading—Young Children’s Achievement and Classroom Experiences, U.S. Department of Education, National Center for Education Statistics, NCES 2003070, Washington, D.C., 2003. Duncan, Greg J., and Saul Hoffman, “Welfare Benefits, Economic Opportunities, and Out-of-Wedlock Births among Black Teenage Girls,” Demography, Vol. 27, No. 4, 1990, pp. 519–535. Eckstein, Zvi, and Kenneth I. Wolpin, “Why Youths Drop Out of High School: The Impact of Preferences, Opportunities, and Abilities,” Econometrica, Vol. 67, No. 6, 1999, pp. 1295–1339. 78 Farley, Reynold, and Richard Alba, “The New Second Generation in the United States,” International Migration Review, Vol. 36, No. 2, 2002, pp. 669–701. Fligstein, Neil, and Roberto M. Fernandez, “Educational Transitions of Whites and Mexican-Americans,” in George J. Borjas and Marta Tienda, eds., Hispanics in the U.S. Economy, Academic Press, Orlando, Florida, 1985. Fry, Richard, “Latinos in Higher Education: Many Enroll, Too Few Graduate,” Pew Hispanic Center, Washington, D.C., 2002. Fry, Richard, “Hispanic Youth Dropping Out of U.S. Schools: Measuring the Challenge,” Pew Hispanic Center, Washington, D.C., June 2003. Fry, Richard, “Latino Youth Finishing College: The Role of Selective Pathways,” Pew Hispanic Center, Washington, D.C., June 2004. Fuligni, Andrew J., “The Academic Achievement of Adolescents from Immigrant Families: The Roles of Family Background, Attitudes, and Behavior,” Child Development, Vol. 68, No. 2, April 1997, pp. 261–273. Fuligni, Andrew J., “The Adjustment of Children from Immigrant Families,” Current Directions in Psychological Science, Vol. 7, No. 4, August 1998, pp. 159–183. Gonzalez, Arturo, “The Education and Wages of Immigrant Children: The Impact of Age at Arrival,” Economics of Education Review, Vol. 22, No. 2, 2003. Gonzalez, Arturo, and Michael J. Hilmer, “The Role of Two-Year Colleges in the Improving Situation of Hispanic Post-Secondary Education,” Economics of Education Review (forthcoming). Griliches, Zvi, “Estimating the Returns to Schooling—Some Econometric Problems,” Econometrica, Vol. 45, No. 1, 1977, pp. 1–22. 79 Grogger, Jeffrey, and Derek Neal, “Further Evidence on the Effects of Catholic Secondary Schooling,” Brookings-Wharton Papers on Urban Affairs, Washington, D.C., 2000. Grogger, Jeffrey, and Stephen J. Trejo, Falling Behind or Moving Up? The Intergenerational Progress of Mexican Americans, Public Policy Institute of California, San Francisco, California, 2002. Hanak, Ellen, and Elisa Barbour, Sizing Up the Challenge: California’s Infrastructure Needs and Trade-offs, Occasional Paper, Public Policy Institute of California, San Francisco, California, 2005. Haveman, Robert, and Barbara Wolfe, “The Determinants of Children’s Attainment: A Review of Methods and Findings,” Journal of Economic Literature, Vol. 33, No. 4, 1995, pp. 1829–1878. Heizer, Robert F., and Alan J. Almquist, The Other Californians: Prejudice and Discrimination under Spain, Mexico, and the United States to 1920, University of California Press, Berkeley, California, 1977. Hernandez, Donald, ed., From Generation to Generation: The Health and Well-Being of Children in Immigrant Families, National Research Council and Institute of Medicine, National Academy Press, Washington, D.C., 1998. Hernandez, Donald, ed., Children of Immigrants: Health, Adjustment, and Public Assistance, National Research Council and Institute of Medicine, National Academy Press, Washington, D.C., 1999. Hill, Laura E., The Socioeconomic Well-Being of California’s Immigrant Youth, Public Policy Institute of California, San Francisco, California, 2004. Hill, Laura E., and Christopher Jepsen, “Positive Outcomes from Poor Starts: Predictors of Dropping Back In,” Working Paper, Public Policy Institute of California, San Francisco, California, 2005. Hirschman, Charles, “Problems and Prospects of Studying Immigrant Adaptation from the 1990 Population Census: From Generational Comparisons to the Process of ‘Becoming American,’” International Migration Review, Vol. 28, No. 4, 1994, pp. 690–713. 80 Jaeger, David A., and Marianne E. Page, “Degrees Matter: New Evidence on the Sheepskin Effects in the Returns to Education,” The Review of Economics and Statistics, Vol. 78, No. 4, 1996, pp. 63–84. Jepsen, Christopher, and Shelley de Alth, English Learners in California Schools, Public Policy Institute of California, San Francisco, California, 2005. Johnson, Hans P., “Maternity Before Maturity: Teen Birth Rates in California,” California Counts, Vol. 4, No. 3, Public Policy Institute of California, San Francisco, California, February 2003. Johnson, Hans P., “Increases in Educational Attainment: Measuring the Progress of the Third Generation,” Working Paper, Public Policy Institute of California, San Francisco, California, 2005. Johnson, Hans P., and Joseph Hayes, The Central Valley at a Crossroads: Migration and Its Implications, Public Policy Institute of California, San Francisco, California, 2004. Kane, Thomas, and Cecilia Rouse, “The Community College: Educating Students at the Margin Between College and Work,” Journal of Economic Perspectives, Vol. 13, No. 1, 1999, pp. 63–84. Kao, Grace, “Psychological Well-Being and Educational Attainment Among Immigrant Youth,” in Donald Hernandez, ed., Children of Immigrants: Health, Adjustment, and Public Assistance, National Research Council and Institute of Medicine, National Academy Press, Washington, D.C., 1999. Kao, Grace, “Parental Influences on the Educational Outcomes of Immigrant Youth,” International Migration Review, Vol. 38, No. 2, Summer 2004, pp. 427–449. Kao, Grace, and Marta Tienda, “Optimism and Achievement: The Educational Performance of Immigrant Youth,” Social Science Quarterly, Vol. 76, No. 1, March 1995, pp. 1–19. Light, Audrey, and Wayne Strayer, “The Determinants of College Completion: School Quality or Student Ability?” Journal of Human Resources, Vol. 35, No. 2, 2000, pp. 299–332. 81 Light, Audrey, and Wayne Strayer, “From Bakke to Hopwood: Does Race Affect College Attendance and Completion?” The Review of Economics and Statistics, Vol. 84, No. 1, 2002, pp. 34–44. Little Hoover Commission, We the People: Helping Newcomers Become Californians, Sacramento, California, 2002. López, David E., and Ricardo D. Stanton-Salazar, “Mexican Americans: A Second Generation at Risk,” in Rubén G. Rumbaut and Alejandro Portes, eds., Ethnicities, University of California Press, Berkeley and Los Angeles, California, 2001. Lopez, Elias, Ginny Puddefoot, and Patricia Gándara, eds., A Coordinated Approach to Raising the Socio-Economic Status of Latinos in California, California Research Bureau, Sacramento, California, 2000. Mare, Robert D., and Christopher Winship, “Ethnic and Racial Patterns of Educational Attainment and School Enrollment,” in Gary D. Sandefur and Marta Tienda, eds., Divided Opportunities: Minorities, Poverty, and Social Policy, Plenum Press, New York, New York, 1988. Massey, Douglas S., Jorge Durand, and Nolan J. Malone, Beyond Smoke and Mirrors: Mexican Immigration in an Era of Economic Integration, Russell Sage Foundation, New York, New York, 2002. Mayer, Susan E., What Money Can’t Buy: Family Income and Children’s Life Chances, Harvard University Press, Cambridge, Massachusetts, 1997. Mensch, Barbara S., and Denise B. Kandel, “Dropping Out of High School and Drug Involvement,” Sociology of Education, Vol. 61, No. 2, 1988, pp. 95–113. Moore, Kristin Anderson, and Jonathan F. Zaff, “Building a Better Teenager: A Summary of ‘What Works’ in Adolescent Development,” Child Trends Research Brief, Publication #2002-57, Washington, D.C., 2002. Myers, Dowell, John Pitkin, and Julie Park, California Demographic Futures: Projections to 2030, by Immigrant Generations, Nativity, and Time of Arrival in U.S., School of Policy, Planning, and 82 Development, University of Southern California, Los Angeles, California, 2005. Neumark, David, The Effects of School-to-Career Programs on Postsecondary Enrollment and Employment, Public Policy Institute of California, San Francisco, California, 2004. Neumark, David, California’s Economic Future and Infrastructure Challenges, Occasional Paper, Public Policy Institute of California, San Francisco, California, 2005. Patterson, Rhiannon, “Neighborhood Effects on High-School Drop-Out Rates and Teenage Childbearing: Tests for Non-Linearities, RaceSpecific Effects, Interactions with Family Characteristics, and Endogenous Causation Using Geocoded California Census Microdata,” unpublished paper, January 2000. Perlmann, Joel, and Roger Waldinger, “Second Generation Decline? Children of Immigrants, Past and Present—A Reconsideration,” International Migration Review, Vol. 31, No. 4, 1997, pp. 893–922. Pirog, Maureen, and Chris Magee, “High School Completion: The Influence of Schools, Families, and Adolescent Parenting,” Social Sciences Quarterly, Vol. 78, No. 1, 1997, pp. 710–724. Portes, Alejandro, and Rubén G. Rumbaut, Legacies: The Story of the Immigrant Second Generation, University of California Press, Berkeley and Los Angeles, California, 2001. Ramakrishnan, Karthick, “Children of Mixed Nativity: Accounting for the 2.5 Generation,” Working Paper, Public Policy Institute of California, San Francisco, California, 2003. Ramakrishnan, S. Karthick, and Hans P. Johnson, “Second Generation Immigrants in California,” California Counts, Vol. 6, No. 4, Public Policy Institute of California, San Francisco, California, May 2005. Reed, Deborah, “The Growing Importance of Education in California,” Occasional Paper, Public Policy Institute of California, San Francisco, California, 2003a. 83 Reed, Deborah, Racial and Ethnic Wage Gaps in the California Labor Market, Public Policy Institute of California, San Francisco, California, 2003b. Reed, Deborah, “Recent Trends in Income and Poverty,” California Counts, Vol. 5, No. 3, Public Policy Institute of California, San Francisco, California, February 2004. Reed, Deborah, “Educational Resources and Outcomes in California, by Race and Ethnicity,” California Counts, Vol. 6, No. 3, Public Policy Institute of California, San Francisco, California, February 2005. Reed, Deborah, Christopher Jepsen, and Laura E. Hill, “Transitions to Work for Racial, Ethnic, and Immigrant Groups,” Working Paper, Public Policy Institute of California, San Francisco, California, 2005. Reyes, Belinda I., ed., A Portrait of Race and Ethnicity in California: An Assessment of Social and Economic Well-Being, Public Policy Institute of California, San Francisco, California, 2001. Rumbaut, Rubén G., “The New Californians: Comparative Research Findings on the Educational Progress of Immigrant Children,” in Rubén G. Rumbaut and Wayne A. Cornelius, eds., California’s Immigrant Children: Theory, Research, and Implications for Educational Policy, Center for U.S.-Mexican Studies, University of California, San Diego, La Jolla, California, 1995. Rumbaut, Rubén G., “Passages to Adulthood: The Adaptation of Children of Immigrants in Southern California,” in Donald Hernandez, ed., Children of Immigrants: Health, Adjustment, and Public Assistance, National Research Council and Institute of Medicine, National Academy Press, Washington, D.C., 1999. Rumbaut, Rubén G., “Ages, Life Stages, and Generational Cohorts: Decomposing the Immigrant First and Second Generations in the United States,” International Migration Review, Vol. 38, No. 3, Fall 2004, pp. 1160–1205. Rumbaut, Rubén G., and Wayne A. Cornelius, eds., California’s Immigrant Children: Theory, Research, and Implications for 84 Educational Policy, Center for U.S.-Mexican Studies, University of California, San Diego, La Jolla, California, 1995. Rumbaut, Rubén G., and Alejandro Portes, eds., Ethnicities: Children of Immigrants in America, University of California Press, Berkeley and Los Angeles, California, 2001. Sandy, Jonathan, Arturo Gonzalez, and Michael J. Hilmer, “Alternative Paths to College Completion: The Effect of Attending a Two Year School on the Probability of Completing a Four Year Degree,” Economics of Education Review (forthcoming). Schoenhals, Mark, Marta Tienda, and Barbara Schneider, “The Educational and Personal Consequences of Adolescent Employment,” Social Forces, Vol. 77, No. 2, 1998, pp. 723–761. Setterson, Richard A., Jr., Frank Furstenberg, Jr., and Rubén G. Rumbaut, eds., On the Frontier of Adulthood: Theory, Research, and Public Policy, University of Chicago Press, Chicago, Illinois, 2005. Smith, James P., “Assimilation across the Latino Generations,” American Economic Review, May 2003, pp. 315–319. Smith, James P., and Barry Edmonston, eds., The New Americans, National Research Council, National Academy Press, Washington, D.C., 1997. Surace, Samuel J., “Achievement, Discrimination, and Mexican Americans,” Comparative Studies in Society and History, Vol. 24, No. 2, 1982, pp. 315–339. Torres, Celina, and Amalia Marquez, Reaching Higher Ground: Parental Outreach Programs at the Post-Secondary Level, Tomás Rivera Policy Institute, University of Southern California, Los Angeles, California, 2005. Trejo, Stephen J., “Why Do Mexican Americans Earn Low Wages?” Journal of Political Economy, Vol. 105, No. 6, December 1997, pp. 1235–1268. 85 Trejo, Stephen J., “Intergenerational Progress of Mexican-Origin Workers in the U.S. Labor Market,” Journal of Human Resources, Vol. 38, No. 3, Summer 2003, pp. 467–489. U.S. Census Bureau, Foreign-Born Population of the United States, Current Population Survey—March 2002, Detailed Tables (PPL-162), 2002, available at www.census.gov/population/www/socdemo/ foreign/cps2002.html. Wald, Michael, and Tia Martinez, “Connected by 25: Improving the Life Chances of the Country’s Most Vulnerable 14–24 Year Olds,” Working Paper, William and Flora Hewlett Foundation, Menlo Park, California, November 2003. Waldinger, Roger, and Mehdi Bozorgmehr, eds., Ethnic Los Angeles, Russell Sage Foundation, New York, New York, 1996. White, Michael J., and Gayle Kaufman, “Language Usage, Social Capital, and School Completion among Immigrants and Native-Born Ethnic Groups,” Social Science Quarterly, Vol. 78, No. 2, 1997, pp. 385–398. Wolfe, Barbara, Kathryn Wilson, and Robert Haveman, “The Role of Economic Incentives in Teenage Nonmarital Childbearing Choices,” Journal of Public Economics, Vol. 81, 2001, pp. 473–511. 86 About the Authors DEBORAH REED Deborah Reed is a research fellow and director of the Population Program at the Public Policy Institute of California. She is a specialist in labor economics with research interests in labor markets, income distribution, public policy, and poverty. A recipient of fellowships from the Mellon Foundation and Yale University, she has served as a consultant to the World Bank in addition to her teaching and research activities. She holds a Ph.D. in economics from Yale. LAURA E. HILL Laura E. Hill is a research fellow at the Public Policy Institute of California, where she researches immigrants, international migration, race and ethnicity, and youth. Before joining PPIC as a research fellow, she was a research associate at The SPHERE Institute and a National Institute of Aging postdoctoral fellow. She has a Ph.D. in demography from the University of California, Berkeley. CHRISTOPHER JEPSEN Christopher Jepsen is a research fellow at the Public Policy Institute of California. His primary research area is the economics of education, with a particular focus on student achievement. Specific research topics include English learners, class size reduction, immigrant intergenerational progress in education, and student demographics in the Central Valley. He holds a Ph.D. in economics from Northwestern University. HANS P. JOHNSON Hans P. Johnson is a demographer whose research interests include international and domestic migration, population estimates and projections, and state and local demography. Before joining PPIC, he was senior demographer at the California Research Bureau, where he conducted research for the state legislature and the governor’s office on population issues. He has also worked as a demographer at the California Department of Finance, specializing in population projections. He holds a Ph.D. in demography from the University of California, Berkeley. 87 Related PPIC Publications Falling Behind or Moving Up? The Intergenerational Progress of Mexican Americans Jeffrey Grogger and Stephen J. Trejo The Socioeconomic Well-Being of California’s Immigrant Youth Laura E. Hill English Learners in California Schools Christopher Jepsen and Shelley de Alth “Second-Generation Immigrants in California” California Counts: Population Trends and Profiles Volume 6, Number 4, May 2005 S. Karthick Ramakrishnan and Hans P. Johnson Racial and Ethnic Wage Gaps in the California Labor Market Deborah Reed and Jennifer Cheng “Educational Resources and Outcomes in California, by Race and Ethnicity” California Counts: Population Trends and Profiles Volume 6, Number 3, February 2005 Deborah Reed PPIC publications may be ordered by phone or from our website (800) 232-5343 [mainland U.S.] (415) 291-4400 [outside mainland U.S.] www.ppic.org 89" } ["___content":protected]=> string(102) "

R 905DRR

" ["_permalink":protected]=> string(106) "https://www.ppic.org/publication/educational-progress-across-immigrant-generations-in-california/r_905drr/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(8317) ["ID"]=> int(8317) ["post_author"]=> string(1) "1" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-05-20 02:36:34" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(3496) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(8) "R 905DRR" ["post_type"]=> string(10) "attachment" ["slug"]=> string(8) "r_905drr" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(12) "R_905DRR.pdf" ["wpmf_size"]=> string(6) "284747" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(180714) "Educational Progress Across Immigrant Generations in California ••• Deborah Reed Laura E. Hill Christopher Jepsen Hans P. Johnson 2005 Library of Congress Cataloging-in-Publication Data Educational progress across immigrant generations in California / Deborah Reed ... [et al.]. p. cm. Includes bibliographical references. ISBN: 1-58213-091-4 1. Immigrants—Education—California. 2. Children of immigrants— Education—California. 3. Educational attainment—California. I. Reed, Deborah, 1967LC3732.C2E38 2005 371.826’912’09794—dc22 2005021485 Copyright © 2005 by Public Policy Institute of California All rights reserved San Francisco, CA Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source and the above copyright notice is included. PPIC does not take or support positions on any ballot measure or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Foreword Of all the political and economic uncertainties that beset California these days, perhaps none is more worrisome than the capabilities of our future labor force. The fiscal and governance challenges confronting the state will become even more complicated if, 30 years from now, a substantial part of the state’s working age population is mired in marginal jobs. In the United States, as well as California, the key to avoiding that scenario has always been the process of intergenerational improvement in education and skills. Californians have typically translated the American Dream into a belief that their children will have better jobs and be better off than they were at the same age. One interpretation of “better off” is that California’s youth today—many of whom are the children of immigrants—should be graduating from high school and college at greater rates than their parents did. The question is, are they? The authors of Educational Progress Across Immigrant Generations in California set out to answer that question with financial support from the William and Flora Hewlett Foundation. They conclude that there is good news. When the educational attainment of second and third generations is compared directly with that of their parents or their parents’ generation, the authors find strong intergenerational progress for all major immigrant groups. However, even by the third generation, Mexican Americans in California have not attained the educational levels that whites have. In other words, there is some progress but even by the third generation only 11 percent of Mexican American adults have earned a bachelor’s degree. In contrast, among third-and-latergeneration whites, more than a third have a bachelor’s degree. Even with this note of caution it is heartening to learn that the traditional vision of intergenerational improvement is still working in California. The authors’ findings have important implications for the state’s education policy. About 30 percent of California’s children are growing iii up in families where neither parent has completed high school. One consequence of this low educational attainment is that as many as 95 percent of these children might not earn a bachelor’s degree; the low educational attainment of parents makes it less likely that their children will attain high levels of education. Among these children at risk of low educational achievement, Mexican Americans make up a large percentage. Policymakers could target at-risk children for extra attention. In addition, community colleges, a resource important to the Latino population, could provide more targeted English language, remedial, and vocational resources. Such courses will become increasingly important for workforce training, especially for those who do not go on to complete a bachelor’s degree, at a time when the value of education and skills in the California economy will continue to grow. More broadly, better educational attainment by today’s children should mean better attainment in later generations. Looking into California’s future 20 years out, the authors of a recently released PPIC study, entitled California 2025: Taking on the Future, concluded that “If California’s youth do not get a college education, they face the prospect of low or no employment, lack of opportunities for high-paying jobs, and greater likelihood of depending on public health and social services. They will also generate lower tax revenues for supporting the state’s infrastructure and other service needs.” Even with all of the adaptive capabilities of California’s huge and complex economy, it does not and will not have the capacity to absorb a lowskilled, poorly educated population indefinitely. The policy issue is right before us—elected officials, the business community, the nonprofit sector, and the millions of state residents: Education has always been and will continue to be the key to success for this and future generations. It is time to address this critical component of the education issue— educating and training our immigrant youth to be highly productive members of the labor force in future years and decades. David W. Lyon President and CEO Public Policy Institute of California iv Summary More than half of all California youth ages 13 to 24 have a foreignborn parent. Because a large number of these immigrant parents have a limited education, lack of improvement in educational attainment from one generation to the next would have serious implications for the state economically as well as socially. Education is an important determinant of social and economic well-being, such as income, health, home ownership, and civic participation. The value of education in the California labor market has increased substantially in recent decades and projections suggest that workers without a college education will continue to see their earnings erode. In this context, it is particularly disconcerting that some recent studies have suggested that intergenerational progress may stall between the second and third generations. The concern for educational progress is particularly acute for Mexican Americans who, even by the third generation, have very low levels of educational attainment. This study examines educational attainment among California’s youth by race, ethnicity, and immigrant generation. It finds that intergenerational progress has not stalled but rather that second- and third-generation immigrants have made substantial educational progress when compared with their parents. Nevertheless, the low educational attainment of Mexican Americans remains a concern and the study examines the factors that influence educational attainment and the policy implications of the findings. Key Facts and Findings Latinos are the largest racial and ethnic group among youth and more than half have a foreign-born parent. California’s youth (ages 13 to 24) are tremendously diverse. However, Latinos constitute a plurality, at 41 percent of the youth population, followed by whites, at 37 percent, Asians, at 11 percent, and African Americans, at 7 percent. Most of the v state’s Latino youth are of Mexican ancestry (84%) and over 60 percent of them were born in the United States. Overall, one in four youth is a first-generation immigrant (i.e., born in a foreign country). About the same share are second-generation immigrants (i.e., born in the United States with at least one foreign-born parent). Racial and ethnic differences in educational attainment are strongly influenced by immigration. Of the major racial and ethnic groups in California, young adults of Mexican descent have the lowest levels of education. Of those ages 25 to 29, only 51 percent have earned a high school diploma, compared to 93 percent of non-Hispanic whites. However, the rate for Mexican American youth born in the United States is substantially higher—76 percent. Although Asians typically have higher levels of education than whites have, less than 80 percent of foreign-born Asians from Cambodia and Laos have completed high school. Mexican youth who come to the United States as teens often do not attend high school here. The older their age at arrival, the less likely Mexican youth are to attend school in California. Among those ages 16 to 18 and who recently arrived in the United States, less than half are enrolled in school. Among men, many are working; among women, substantial numbers are working, married, or raising children. In contrast, over 90 percent of Asian immigrant youth, including those from Southeast Asia, are enrolled in school at ages 16 to 18, regardless of their age at arrival. Among youth in immigrant families, there is tremendous variation in family income and parental education. Among young immigrants ages 13 to 17, about one-third of those from Mexico are living in poor families and only 17 percent have a mother who finished high school (maternal education is measured only for those living with their mothers). Among youth from Cambodia and Laos, over half are living in poor families and only 19 percent have a mother who finished high school. In comparison, among Filipino immigrants, less than 10 percent are living in poor families and over 90 percent have a mother who finished high school. These differences in family characteristics contribute to racial and ethnic differences in educational attainment for vi immigrant youth, which, in turn, contribute to education differences for their second-generation children. Substantial educational progress is made between the first, second, and third generations of immigrants. When the educational attainment of second- and third-generation immigrants is compared directly with that of their parents or with that of their parents’ generation, we find strong intergenerational progress for all major immigrant groups. In particular, although some research has suggested that educational progress stalls between the second and third generations for Mexican Americans, we find that college graduation rates of third-generation immigrants are more than twice those of their parents. Further, although over half of their parents did not graduate from high school, about eight in 10 third-generation Mexican Americans have graduated from high school. Even by the third generation, however, Mexican Americans in California have lower educational attainment than whites have. Despite strong intergenerational progress, less than 85 percent of thirdand-later-generation Mexican American adults, ages 25 to 34, have finished high school and only 11 percent have completed a bachelor’s degree. (“Third-and-later” generation includes youth with both parents born in the United States but the data do not identify whether their grandparents or great-grandparents were born in the United States.) In comparison, among third-and-later-generation whites, 95 percent earned a high school diploma and over a third have a bachelor’s degree. Differences in family characteristics explain most of the lower educational attainment of Mexican Americans. Among Mexican American youth, parental education, parental English language ability, and family income are substantially lower than among white youth. Using national data, we find that these differences in family economic resources and family characteristics generally account for the lower educational attainment of Mexican American youth relative to white youth. Policy Considerations The findings of this study have important implications for education-related policy in California. Mexican immigrant youth who vii arrive at age 15 or older are among the least educated Californians. Improving their educational attainment is particularly challenging because many do not enroll in California schools but are working and raising families. For these youth, adult education programs in school districts and community colleges may provide better schedules for parttime, evening, and weekend coursework. Another approach would be to target the workplace, offering programs in collaboration with employers to help workers develop English language and literacy skills. In addition, as these youth become parents, programs that work with young children can assist parents with parental support and literacy improvement. For second and third generations, and for immigrants who do enter California schools, the quality of the K–12 public education system is clearly a key factor in success. Several recent and continuing reforms have sought to improve California schools, particularly in the areas of student achievement, teacher quality, and quality of facilities. In addition, English language learning is of concern for the children of immigrants. For students whose own parents have limited educational experience, programs of educational counseling and tutoring may be particularly helpful. Our analysis suggests that about 30 percent of California’s children are growing up in families where neither parent has completed high school and that as many as 95 percent of these children might not achieve a bachelor’s degree. Among these children at risk of low educational achievement, Mexican Americans make up a large share (68%). The success of students in California’s community colleges is of particular importance for improving Latino postsecondary education because almost 80 percent of Latinos who enroll in public higher education enter through community colleges. Of great concern, however, is the low transfer rate to four-year institutions, and transfers are especially low among Latino students. In addition to preparing students for transfers, community colleges provide English language, remedial, and vocational courses. As the value of education and skills in the California economy continues to grow, these courses will become increasingly important to workforce training, especially for those who do not go on to complete a bachelor’s degree. viii Investing in an education system that works for all of California’s children will pay off for generations to come. Better educational attainment for today’s children will translate to better family economic resources for the next generation of parents. These higher resources will help the next generation of California’s children to attain even higher levels of education. ix Contents Foreword......................................... iii Summary......................................... v Figures .......................................... xiii Tables ........................................... xv Acknowledgments................................... xvii 1. INTRODUCTION .............................. Conclusions and Organization of the Report ............. A Note on Data ................................. 2. STATISTICAL PORTRAIT OF CALIFORNIA YOUTH... Demographic Characteristics ........................ Educational Attainment ........................... 1 2 3 5 5 9 3. INTERGENERATIONAL PROGRESS AND EDUCATIONAL ATTAINMENT ................... Measuring Intergenerational Progress .................. Intergenerational Progress Has Been Impressive ........... First Generation to Second Generation ............... Second Generation to Third Generation .............. Improvement Is Still Needed for Some Groups ........... Implications for the Future ......................... 15 15 17 18 22 26 28 4. DETERMINANTS OF THE EDUCATIONAL ATTAINMENT OF FOREIGN-BORN YOUTH ........ Recent Historical Context .......................... Family Characteristics ............................. Age at Arrival ................................... 5. DETERMINANTS OF THE EDUCATIONAL ATTAINMENT OF U.S.-BORN YOUTH ............. Family Characteristics ............................. 31 31 32 34 41 41 xi Factors Contributing to Racial and Ethnic Differences in Educational Attainment ........................ California Youth At Risk of Low Educational Attainment ................................. 42 50 6. CONCLUSIONS ............................... 53 Policy Directions ................................ 54 Directions for Future Research....................... 56 Appendix A. Notes on Data and Methods ........................ B. Detailed Nativity and Ancestry....................... C. Detailed Model Results ............................ 57 71 73 References ........................................ 77 About the Authors .................................. 87 Related PPIC Publications............................. 89 xii Figures 2.1. Percentage Distribution of the Educational Attainment of Young California Adults by Race, Ethnicity, and Immigrant Generation, Ages 25 to 34............... 12 4.1. Age at Arrival of Foreign-Born Mexicans and Asians..... 35 4.2. Activities of Foreign-Born Mexican and Asian Men by Age at Arrival, Ages 16 to 18 ..................... 36 4.3. Activities of Foreign-Born Mexican and Asian Women by Age at Arrival, Ages 16 to 18 ................... 38 5.1. High School Completion Rates Relative to Third-and- Later-Generation Whites ........................ 44 5.2. Grade Level Completion Rates of U.S.-Born Young California Adults with No High School Diploma, Ages 19 to 24, 2000 ............................... 46 5.3. Postsecondary School Attendance Relative to Whites .... 47 5.4. College Completion Rates Relative to Whites ......... 48 xiii Tables 2.1. Top Ten Countries of Birth and Ancestral Backgrounds of California Youth, Ages 13 to 24, 2000 ............ 6 2.2. Percentage Distribution of California Youth by Race, Ethnicity, and Immigrant Generation, Ages 13 to 24, 2004 ...................................... 7 2.3. Percentage Distribution of Youth in California and the Rest of the Nation by Race, Ethnicity, and Immigrant Generation, Ages 13 to 24, 1970 and 2000 ........... 9 2.4. Percentage Distribution of the Educational Attainment of Young California Adults by Race, Ethnicity, and Immigrant Generation, Ages 25 to 29............... 10 3.1. Percentage Distribution of the Educational Attainment of Groups in the First and Second Generation ......... 19 3.2. Percentage Distribution of the Educational Attainment of First-Generation Parents and Their SecondGeneration Adult Children ...................... 20 3.3. Percentage Distribution of the Educational Attainment of Groups in the Second and Third-and-Later Generations ................................. 24 3.4. Percentage Distribution of the Educational Attainment of Second-Generation Parents and Their ThirdGeneration Adult Children ...................... 25 3.5. Percentage Distribution of the Projected Educational Attainment of Second- and Third-Generation Descendants of Mexican Immigrants ............... 29 4.1. Percentage Distribution of the Family Characteristics of Foreign-Born California Youth by Race and Ethnicity, Ages 13 to 17, 2000 ........................... 34 5.1. Percentage Distribution of the Family Characteristics of U.S.-Born California Youth by Race and Ethnicity, Ages 13 to 17, 2000 ............................... 42 xv 5.2. Percentage Distribution of College Attendance and Completion Rates of U.S.-Born Youth by Race and Ethnicity, Ages 25 to 29, 2000.................... 49 5.3. Percentage Distribution of Students in Higher Education Institutions in California by Race and Ethnicity, 2002 .............................. 49 5.4. Percentage Distribution of Educational Attainment by Family Characteristics, Ages 0 to 9, 2000 ............ 51 A.1. Sample Sizes for Data Used in Table 3.1 ............. 62 A.2. Sample Sizes for Data Used in Tables 3.2 and 3.4 ...... 62 A.3. Sample Sizes for Data Used in Table 3.3 ............. 63 A.4. Sample Sizes and Percentage Distribution of College Completion Rates in the NELS (2000) and CPS (1996–2000) ................................ 67 A.5. Percentage Distribution of NELS Model Projections Versus Actual Bachelor’s Degree Attainment for Californians by Race and Ethnicity................. 68 A.6. Percentage Distribution of Educational Attainment by Generation: Mother-to-Child .................... 69 A.7. Percentage Distribution of Educational Attainment by Generation: Father-to-Child ..................... 70 B.1. Number of Youth and Educational Attainment for the Top 26 Countries of Birth Reported by Foreign-Born Californians, Ages 13 to 24, 2000.................. 71 B.2. Number of Youth and Educational Attainment for the Top 25 Ancestries Reported by U.S.-Born Californians, Ages 13 to 24, 2000 ........................... 72 C.1. Logistic Probability Model for High School Completion ................................. 74 C.2. Logistic Probability Model for Postsecondary School Enrollment ................................. 75 C.3. Logistic Probability Model for a Bachelor’s Degree...... 76 xvi Acknowledgments The authors wish to thank the William and Flora Hewlett Foundation for its generous support for this research project. We appreciate the continuing interest of foundation President Paul Brest in the topic of this report, and we thank Michael Wald and Tia Martinez for many helpful suggestions in the development of this project. Frank Furstenberg, Arturo Gonzalez, Donald Hernandez, Joyce Peterson, and Rubén G. Rumbaut provided valuable comments on an earlier draft. Amanda Bailey, Pedro Cerdán, and Shelly de Alth assisted in the research analysis. The methods and content of this report were determined solely by the authors. xvii 1. Introduction Nearly half of all Californians today are first- or second-generation immigrants. As that share of the California population continues to grow, it is increasingly important to understand the nature of intergenerational progress for immigrant groups.1 Recent research has called into question the intergenerational progress of immigrants, particularly educational progress between the second generation and the third generation. Because California has such large numbers of immigrants with limited education, a lack of improvement in educational attainment from one generation to the next would have serious implications for the state economically as well as socially. Educational progress is particularly important because education plays a role in determining racial and ethnic differences in other areas of social and economic well-being, such as poverty, health status, employment, home ownership, and civic participation (Reyes, 2001; Reed, 2003a). Furthermore, the economic value of education has increased substantially in the California labor market in recent decades (Betts, 2000). Over the past three decades, workers with a high school diploma or less have seen their average earnings erode whereas average earnings for those with a college degree have grown (Reed, 2003b). Neumark (2005) suggests that over the next two decades, the value of a college education in the California labor market will continue to grow. There are a number of state and local programs designed to improve the lives of youth as well as to steer them in the direction of positive future outcomes. Youth ages 13 to 24 are of critical concern because during these ages youth are preparing for the transition to adulthood _____________ 1Myers, Pitkin, and Park (2005) describe the growing presence of long-term and second-generation immigrants in the future California population and discuss implications for social and economic achievement. 1 with its increased economic challenges and responsibilities and often with new marriage and parenting relationships.2 During these ages, many potentially life-changing decisions are often made, including the decisions to finish high school, to go to college, and perhaps to start a family.3 Conclusions and Organization of the Report This report provides evidence that the intergenerational progress of immigrants has not stalled. Among all groups studied, including Mexican Americans, significant educational progress has occurred from the first to the second generation and from the second to the third generation. Nevertheless, even by the third generation, Mexican Americans have lower educational attainment than do whites and Asian Americans, the other large immigrant group. The report analyzes the factors that influence educational attainment, particularly that of Mexican Americans, and considers the implications for policy. Chapter 2 provides a statistical portrait of California youth, focusing on race, ethnicity, and immigrant generation and differences in educational attainment among groups. Chapter 3 compares the educational attainment of young adults to that of their parents (or their parents’ generation) and shows that all immigrant groups have made progress, including the first three generations of Mexican Americans. However, even accounting for this progress, the currently low educational attainment of Mexican Americans is likely to remain low in the coming decades. The findings in this chapter point to the need to understand the factors that contribute to this low educational _____________ 2See Setterson, Furstenberg, and Rumbaut (2005) for studies of the transition to adulthood in the United States and Western Europe. 3This report is part of a larger study that includes analysis of teen fertility and the transition from school to work. We chose to limit the report to education-related topics to present a more cohesive and focused discussion. Readers interested in our analysis of teen fertility and early work success are referred to Reed, Jepsen, and Hill (2005). For a more detailed description of teen fertility, see Johnson (2003). See Neumark (2004) for a discussion of school-to-career programs in California. The larger study also included an investigation of factors that contribute to positive educational and socioeconomic outcomes for youth who become teen parents and those who drop out of high school (see Hill and Jepsen, 2005). 2 attainment. Chapter 4 examines the determinants of educational attainment for foreign-born youth, finding that relative to Asian immigrants, Mexican immigrants are more likely to arrive after age 14 and to enter the workforce or family life and less likely to enroll in school. Chapter 5 examines the determinants of educational attainment among U.S.-born youth, finding that low parental education and low family income explain most of the lower educational attainment of Mexican American youth relative to white youth. Chapter 6 draws on all chapter findings to discuss policy directions for improving educational attainment. A Note on Data The Current Population Survey (CPS) and the decennial Censuses, used to measure youth educational attainment for this study, were chosen because they are representative of California.4 However, the data have two important limitations. First, they do not include information on involvement with the criminal justice system.5 Previous research suggests the importance of racial, ethnic, and immigrant differences in this area (Blumstein, 2001; Rumbaut, 2004). Second, they do not include detailed measures of youth background, such as attitudes about education, school conditions, and the quality of family relationships. Rumbaut (1999) provides a broad discussion of these issues for immigrant children and youth in San Diego.6 _____________ 4We focus on youth educational attainment and do not address the important but broader issues of health status or farmworker conditions. See Hernandez (1998, 1999) for national research on these topics. Diversity of health status across racial, ethnic, and immigrant groups in California is a topic of ongoing research at PPIC. 5The CPS represents the noninstitutional, civilian population. The Census represents the entire population, including those who are incarcerated, but it does not include information on past incarcerations or broader measures of interactions with the criminal justice system. See Appendix A for further descriptions of these data. 6Rumbaut (1995) describes four case studies of immigrant groups in California high schools. See also Portes and Rumbaut (2001) and Rumbaut and Portes (2001). The National Educational Longitudinal Study (NELS), used in this study to estimate models of youth educational attainment, includes information on some of these topics, but the subsample of youth in California is small and not necessarily representative. See Kao and Tienda (1995) and Kao (1999, 2004) for national measures of these issues based on the NELS. 3 The tremendous diversity of the California population makes it difficult to characterize the population and socioeconomic conditions by racial, ethnic, or ancestry group.7 The main analyses in this study focus on four major groups: Mexican Americans, Asian Americans, nonHispanic whites, and African Americans. We distinguish Mexican Americans from other Latinos because of their unique historical context and related socioeconomic conditions.8 Wherever possible we describe conditions for Central Americans, but their numbers are not sufficient to study intergenerational progress from the second generation to the third generation. Latino backgrounds also include Puerto Ricans, Cubans, Peruvians, and others. Their socioeconomic conditions are too diverse to make aggregation meaningful and their overall numbers in most of our data samples are too small for separate analysis. California’s Asian groups are particularly diverse. Where data permit, we distinguish Southeast Asians from refugee-sending countries (Laos, Cambodia, and Vietnam) from other Asians because of their substantially different historical context and socioeconomic conditions. We also consider the two Asian backgrounds that represent the largest number of Asian youth: Vietnamese and Filipinos. _____________ 7California is also regionally diverse and the nature of educational progress differs across regions. See Johnson and Hayes (2004) for an analysis of related issues for the Central Valley. 8See Chapter 4 for a discussion of the historical context of immigration for major immigrant groups. 4 2. Statistical Portrait of California Youth This chapter documents the tremendous diversity of California youth in terms of race, ethnicity, and immigrant generation. The analysis demonstrates the large and growing share of first- and secondgeneration immigrants among California youth and also describes substantial differences in educational attainment across racial, ethnic, and immigrant groups. The portrait presented here provides a context for the following chapters, which focus on the largest immigrant groups (Mexicans and Asians) and on those with particularly low educational attainment (Mexicans and Southeast Asians).1 Demographic Characteristics The youth population ages 13 to 24 in California numbers almost 6.6 million. By 2010, it is expected to increase to almost 7 million and to continue to make up close to 18 percent of the state population.2 The tremendous diversity of this population makes it difficult to characterize the backgrounds of its members. About one in four California youth was born outside the United States. California’s foreign-born youth are from almost 150 countries, mainly Mexico, the Philippines, El Salvador, Vietnam, and Guatemala (Table 2.1). California’s U.S.-born youth are from nearly 200 different ancestral _____________ 1Hill (2004) presents a more complete description of California’s youth, including regional information as well as statistics on youth activities, parental employment, health status, teen parenting, and citizenship. See also Ramakrishnan and Johnson (2005) for a description of second-generation immigrants in California, including statistics by California region. 2Population estimates are based on California Department of Finance population projections for 2005 and 2010. 5 Table 2.1 Top Ten Countries of Birth and Ancestral Backgrounds of California Youth, Ages 13 to 24, 2000 Country of Birth Number Foreign-Born 1. Mexico 783,124 2. Philippines 76,753 3. El Salvador 59,612 4. Vietnam 58,701 5. Guatemala 42,795 6. Korea 28,228 7. Taiwan 25,859 8. India 23,576 9. Thailand 22,822 10. China 22,337 Ancestry U.S.-Born 1. Mexican 2. African American 3. German 4. Irish 5. English 6. Italian 7. American 8. Filipino 9. White 10. Chinese Number 1,228,338 310,810 279,195 210,186 178,050 161,383 158,956 107,742 94,380 82,943 SOURCE: Authors’ calculations from the 2000 Census. NOTES: See Appendix Tables B.1 and B.2 for statistics on the top 25 countries of birth and ancestral backgrounds. Ancestry is self-identified and no specific categories are supplied (see Appendix A). backgrounds, mainly Mexican, African American, German, Irish, and English (Table 2.1).3 To further describe the youth population, we focus on the four largest racial and ethnic groups: Latinos, whites, Asians, and African Americans. Among California youth, Latinos are the largest ethnic group, comprising about 41 percent of all youth (Table 2.2). Thirteen percent of California youth were born in Mexico. The majority of Latino youth are of Mexican ancestry and were born in the United States. Fourteen percent of youth are second-generation Mexican Americans (born in the United States with at least one parent born _____________ 3The identification of racial, ethnic, and immigrant generation is complex. See Appendix A for a description of our analysis. See Rumbaut (2004) for a detailed analysis of these complexities. 6 Table 2.2 Percentage Distribution of California Youth by Race, Ethnicity, and Immigrant Generation, Ages 13 to 24, 2004 White Mexican Central American Other Latino Southeast Asian Other Asian African American Native American Multiracial and all others Total of column ForeignBorn 2 13 2 1 1 4 <1 <1 Second Generation 4 14 1 2 1 3 1 <1 Third-andLater Generation 31 7 <1 1 <1 1 6 1 Total of Row 37 34 3 4 3 8 7 1 1 12 4 25 27 49 100 SOURCES: Authors’ calculations from the 2000 Census and the Current Population Survey (March 2003, 2004). NOTES: Some rows and columns do not sum to totals because of rounding. See Appendix A for details on data and measurement. outside the United States). Seven percent are third-and-later-generation Mexican Americans. Throughout the report, “third-and-later generation” refers to people who were born in the United States and whose parents were both born in the United States. Most of the analysis relies on data that do not identify later generations.4 The next largest racial and ethnic group is non-Hispanic whites. Among white youth, the vast majority are third-and-later-generation Americans. Similarly, African Americans (7% of youth) and Native _____________ 4Data from the Current Population Survey do not distinguish between the third generation and those in later generations. The data report the birthplace of parents but not of grandparents. Rumbaut (2004) demonstrates that further categorizations of immigrant generations into the “1.5 generation” (immigrants who arrive by age 10) and the “2.5 generation” (those with one foreign-born parent and one U.S.-born parent) are useful in understanding diversity and educational outcomes. See also Ramakrishnan (2003). 7 Americans (about 1% of youth) are also primarily third-and-latergeneration Americans.5 The third largest racial and ethnic group is Asians, about 11 percent of the youth population. This population is very diverse with about half born in foreign countries, particularly the Philippines, Vietnam, Korea, Taiwan, India, Thailand, and China (Table 2.1). For U.S.-born Asian youth, the most common backgrounds are Filipino and Chinese. Among Asians, those from Southeast Asian refugee-sending countries (Cambodia, Laos, and Vietnam) tend to live in markedly worse socioeconomic conditions including higher poverty and lower educational attainment (see Chapter 4).6 For this reason, where possible, our analysis separates youth from these Southeast Asian backgrounds, about 3 percent of all youth. Overall, just under half of California’s youth are third-and-latergeneration immigrants. About one-fourth are second-generation immigrants and about one-fourth are first-generation immigrants. Because of the influx of immigrants in their late teens and early twenties, the share of youth who are foreign-born tends to increase with age. Among youth age 13, only 14 percent are foreign-born. At age 18, 22 percent of youth are foreign-born, and at age 24, the foreign-born share is 38 percent. Just over half of California’s youth—52 percent—are male. The primary factor increasing the share of males relative to females is immigration from Mexico and Central America—56 percent of these youth are male. Youth in California today are notably more diverse than California youth of past decades and also much more diverse than youth in the rest of the nation (Table 2.3). In 1970, the share of California youth who were white was 73 percent. In 2000, the share who were white was lower, at 66 percent for the rest of the nation, but in California the share was much lower, at only 37 percent. In the rest of the nation in 2000, as _____________ 5Some Native Americans are foreign-born, mainly from Canada. 6See Reyes (2001) for an analysis of socioeconomic indicators across Asian ethnic groups. 8 Table 2.3 Percentage Distribution of Youth in California and the Rest of the Nation by Race, Ethnicity, and Immigrant Generation, Ages 13 to 24, 1970 and 2000 White Latino Asian African American Multiracial and all others California, 1970 73 17 2 7 1 California, 2000 37 41 11 7 5 Rest of U.S., 2000 66 13 3 15 3 Third-and-later generation 83 49 81 Second generation 10 27 10 First generation 7 25 9 SOURCES: Authors’ calculations from decennial Censuses (1970, 2000). NOTES: Some columns do not sum to 100 percent because of rounding. See Appendix A for details on data and measurement. in California in 1970, over 80 percent of youth were third-and-latergeneration Americans. Educational Attainment California’s demographic diversity may be a source of pride and economic strength, but this diversity can create challenges when there are large and continuing disparities in socioeconomic conditions between groups. As discussed in the introduction, education is a key factor in determining economic success, health status, and other socioeconomic conditions.7 Among first-generation immigrant youth, there is tremendous diversity in educational attainment. Immigrants from Mexico and, to a lesser extent, Central America have particularly low levels of completing _____________ 7The association between education and earnings is not entirely causal (Griliches, 1977). National research that adjusts for noncausal factors finds strong labor market returns to education (Jaeger and Page, 1996; Arias and McMahon, 2001). See also Kane and Rouse (1999) for national estimates of the economic return to attending community college. 9 high school, attending college, and getting a bachelor’s degree (Table 2.4).8 Asian and white immigrants have relatively high educational attainment, although bachelor’s degree completion is lower among immigrants from Southeast Asian refugee-sending countries and Filipino Table 2.4 Percentage Distribution of the Educational Attainment of Young California Adults by Race, Ethnicity, and Immigrant Generation, Ages 25 to 29 Men White Mexican Central American Vietnamese Other Southeast Asian Filipino Other Asian Black or African American High School Diploma First Second+ 89 92 34 73 45 84 82 79 92 94 95 97 87 81 Some College Bachelor’s Degree First Second+ First Second+ 72 71 14 42 26 64 68 54 75 84 86 89 62 52 48 35 3 11 5 20 30 17 29 37 65 65 24 13 Women White Mexican Central American Vietnamese Other Southeast Asian Filipino Other Asian Black or African American 91 38 50 83 72 96 95 85 94 79 83 97 98 85 76 78 18 53 27 65 67 49 83 86 86 93 68 61 46 40 5 14 7 24 34 14 38 48 63 71 29 19 SOURCE: Authors’ calculations from the 2000 Census. NOTES: See Appendix A for measurement details. High school diploma includes the General Equivalency Diploma (GED). Sample sizes are not sufficient for secondand-later-generation Southeast Asians. The smallest sample used was 196 (for firstgeneration black women). _____________ 8We use ages 25 to 29 to measure education among young adults because by age 25 many have completed their education. However, particularly among Latinos, educational attainment does continue to improve at older ages. See Reed (2005). 10 immigrants. Black immigrant youth also have lower levels of college attainment than do other immigrants.9 Throughout this report, we use aggregated groups to describe socioeconomic conditions. However, within each aggregated racial and ethnic category in Table 2.4, there is a fair degree of diversity. For example, among first-generation immigrants from Central America, those from Guatemala tend to have lower educational attainment and those from Nicaragua have higher educational attainment (see Appendix Table B.1 for educational attainment among foreign-born youth by country of birth). Among Asian immigrants, those from India, Taiwan, and Hong Kong have particularly high rates of college completion. Among white immigrants, those from Armenia tend to have lower college completion. Even among young adults born in the United States, the aggregated groups mask tremendous diversity (see Appendix Table B.2 for educational attainment among U.S.-born youth by detailed ancestry group).10 Among U.S.-born (second-and-later-generation) Mexican Americans, educational attainment is much higher than for those born in Mexico. Indeed, if we were to calculate the high school completion rate among all young adults of Mexican descent, the rate would be only 51 percent. Because many Mexican immigrants do not enroll in school in the United States, this rate does not reflect the experience of Mexican Americans in California schools (see Chapter 4 and Fry, 2003).11 However, even among U.S.-born Mexican Americans, educational _____________ 9The rows of Table 2.4 are distinguished by reported race, ethnicity, and ancestry and not by country of origin. Many Southeast Asian immigrant youth were born outside Cambodia, Laos, and Vietnam, especially in Thailand and Malaysia. African American or black immigrants do not necessarily come from Africa. 10Rumbaut and Portes (2001), Portes and Rumbaut (2001), and Rumbaut (2004) provide detailed descriptions of the lives of the children of immigrants by ethnicity for San Diego and South Florida, including information on parental relationships, attitudes, and activities not available in the CPS or Census data. Waldinger and Bozorgmehr (1996) describe detailed conditions by ethnicity for Los Angeles. 11Educational statistics in Table 2.4 are for young adults living in California, but even for U.S.-born adults, these statistics may not reflect the educational attainment of young adults who grew up in California (because of domestic migration). See Reed (2005) for a discussion of education by race and ethnicity for Californians. For measures of high school dropout rates in California, see The Civil Rights Project (2005). 11 attainment is substantially below that of U.S.-born whites and somewhat below that of African Americans. For Mexican Americans in California, the education of the third-and-later generation is not substantially different from that of the second generation (Figure 2.1). 12 Among Asians, other than those from Southeast Asia, and among whites, educational attainment across all three generations is substantially higher than for Mexican Americans. Within each of these groups, educational attainment is fairly similar across the generations.13 Percentage 100 80 60 40 20 0 First Second Third+ First Second Third+ First Second Third+ Generation Generation Generation White Mexican American Asian American No high school diploma High school diploma Bachelor’s degree SOURCES: Authors’ calculations from the Current Population Survey (1996–2004). NOTES: The age range was increased from 25–29 to 25–34 to increase sample sizes. The smallest sample has 129 observations (Third+ –generation Asian Americans). Sample sizes were not sufficient to investigate other groups shown in Table 2.4. Figure 2.1—Percentage Distribution of the Educational Attainment of Young California Adults by Race, Ethnicity, and Immigrant Generation, Ages 25 to 34 _____________ 12Educational attainment by immigrant generation is measured using the 1996–2004 Current Population Survey. The CPS tends to measure higher high school completion than does the Census, in part because the Current Population Survey does not include institutionalized populations. 13Several studies at the national level have found higher educational achievements among the second generation relative to the first or third generations (Smith, 2003; Farley and Alba, 2002; Chiswick and DebBurman, 2004; Kao, 2004; Kao and Tienda, 1995). 12 Table 2.4 and Figure 2.1 describe educational attainment for young adults in California by immigrant generation, but these statistics should not be used to assess intergenerational progress. The intergenerational progress of these young adults should be measured against the achievement of their own parents or against people of their parents’ generation (as opposed to other young adults). We turn to the analysis of intergenerational progress in the next chapter. 13 3. Intergenerational Progress and Educational Attainment The previous chapter described the low educational attainment of immigrants from Mexico, Central America, and Southeast Asia. Because immigrants from these regions and their children constitute about onethird of all California youth, it is important for all Californians that these children acquire more education than their parents to gain the skills required by the state’s economy and the income necessary to provide for their families. In this chapter, we examine the recent history of intergenerational progress in the United States. We demonstrate substantial educational progress between second-generation immigrants and their foreign-born parents from all major immigrant-sending regions—progress that continues between the second and third generations for Mexicans, the state’s largest immigrant group. Because of the recent record, we project a high degree of educational progress for the third generation (grandchildren) of today’s Mexican immigrants. Despite this progress, Latino generations in years to come will still fall far short of the educational levels currently attained by U.S.-born whites. The analysis of intergenerational progress in this chapter is based on immigrants and their descendents nationwide. Use of a national sample is appropriate and avoids potentially biasing measures of intergenerational progress. For example, if immigrant families who have been successful have been more likely to remain in California, then limiting the analysis to those in California would overestimate the intergenerational progress of immigrants. Measuring Intergenerational Progress Improvements in educational attainment from the first generation to the second generation are well documented (Chavez, 1991; Card, 15 DiNardo, and Estes, 1990; Perlmann and Waldinger, 1997; and Farley and Alba, 2002), but some researchers have found little if any progress from the second to the third generations, particularly for Latinos or Mexicans (Borjas, 1994; Grogger and Trejo, 2002; Trejo, 1997, 2003; Chapa, 1990; Camarota, 2001; Kao and Tienda, 1995; and Kao, 2004). Measuring progress from the second to the third generation is difficult because of a lack of data on grandparents’ place of birth, and the research evidence that calls into question the intergenerational progress of recent immigrants is not definitive.1 Findings that point to weak progress between second and third generations often fail to measure the appropriate type of progress—progress in relation to one’s own parents (or progress from one generation of adults to an older previous generation)—or fail to consider nationally representative data. One study that compares current young adults with those of their parents’ generation finds strong progress from the first to the second generation and continuing progress to the third generation (Smith, 2003).2 In this chapter, we measure intergenerational progress in education using two methods.3 In both methods, we use nationally representative data from the Current Population Survey to estimate changes in educational attainment from one generation to the next, comparing attainment for adults with that of their parents. The CPS includes questions on the place of birth of all household members as well as that of the parents of those household members. We measure intergenerational change first by comparing the educational attainment of one entire generation to that of a previous generation and, second, by directly comparing an adult’s educational attainment with that of his/her parents. The first approach, a comparison of specific age groups from entire generations, uses a large sample but suffers from not directly linking parents with their descendants. In this sample, we cannot identify the third generation apart from fourth and later generations. We are careful to identify the age groups and immigrant generations that allow the most appropriate _____________ 1See Johnson (2005) for a more complete discussion. 2Smith and Edmonston (1997) also find progress across the first three generations. 3See Appendix A for details of our methodology. 16 comparison of one generation with a previous one. In particular, we compare educational attainment for young adults (ages 30 to 39) from one generation with the attainment of older adults (ages 57 to 66) from a previous generation. Young adults in their thirties have largely completed their formal education. On average, their parents are 27 years older. Thus, older adults ages 57 to 66 from a previous generation will include many of the parents of these young adults, and this group therefore serves as a benchmark for measuring intergenerational progress. The second approach has the advantage of explicitly linking parents with their adult descendants by examining households in which adult children and their parents live together. This allows us to precisely measure true (parent-to-child) generational change. Moreover, unlike the first approach, we are able to specifically identify the third generation. The disadvantage of this approach is that the sample is not large and is subject to selection bias because it requires that adult children be living with their parents.4 Because of sample size limitations, we use this multiple-generation household approach only for the state’s largest immigrant groups. Intergenerational Progress Has Been Impressive We find strong intergenerational progress in education. Across all generations for all groups, we consistently find that children acquire more education than their parents. This is true for U.S.-born children of immigrants (the second generation compared to the first generation), as well as for third-generation descendants of immigrants. The results are consistent for both approaches. That is, we find similar levels of progress in both the smaller sample that explicitly links parents with their adult children and the larger sample that compares one entire generation to another entire generation. _____________ 4See Appendix A for a more complete discussion of the sample. NELS includes information on parental education and young adult educational attainment (to age 26) regardless of living arrangements. Findings on intergenerational educational progress based on the NELS are consistent with the results shown in this chapter (see Appendix Tables A.6 and A.7). 17 First Generation to Second Generation There are dramatic increases in educational attainment from the first generation to the second generation for all ethnic groups. These increases are evident in comparisons of the entire second generation (ages 30 to 39) with the entire first generation (ages 57 to 66) (Table 3.1) as well as in comparisons of individuals specifically with their own parents (Table 3.2).5 The increases are evident for highly educated immigrant groups as well as for less educated immigrant groups. Overall, for immigrants from all regions, second-generation descendants born in the United States are much better educated than their parents’ generation. Whereas over one-third of the first generation have not graduated from high school, more than nine of 10 secondgeneration descendants have graduated from high school (Table 3.1). College graduation rates are much greater for young second-generation adults than for their parents’ generation, with over one-third of the second generation graduating from college compared to fewer than one in four in the first generation. These improvements from the first generation to the second generation are much larger than for other generations of U.S. natives. Among U.S.-born individuals with both parents born in the United States, 22 percent of the older generation has completed college compared to 28 percent of the younger generation (Table 3.1 bottom panel, third+ generation). Moreover, the young second generation in the United States is more likely than other U.S. natives to have graduated from college. Results for the intergenerational household sample of parents and adult children are similar (Table 3.2). In this sample, second-generation adult children are twice as likely as their first-generation parents to complete college. Although patterns of progression vary with ethnicity and region of origin, intergenerational progress is observed for all groups. The largest _____________ 5We focus on changes from the first generation to the second generation, but results are similar for changes from the first generation to the 1.5 generation (individuals born abroad but who came to the United States at age 10 or younger and therefore primarily if not wholly educated in the United States). Results for the 1.5 generation are available from the authors. 18 Table 3.1 Percentage Distribution of the Educational Attainment of Groups in the First and Second Generations Less Than High School Ethnic Group Diploma All ethnic groups First generation, ages 57 to 66 37 Second generation, ages 30 to 39 9 Mexican First generation, ages 57 to 66 78 Second generation, ages 30 to 39 18 Central Americana First generation, ages 57 to 66 54 Second generation, ages 30 to 39 16 Southeast Asianb First generation, ages 57 to 66 38 Second generation, ages 30 to 39 13 Other Asian First generation, ages 57 to 66 21 Second generation, ages 30 to 39 3 White (non-Hispanic) First generation, ages 57 to 66 20 Second generation, ages 30 to 39 4 Black or African American First generation, ages 57 to 66 35 Second generation, ages 30 to 39 11 All ethnic groups Third+ generation, ages 57 to 66 18 Third+ generation, ages 30 to 39 8 High School Diploma 26 26 12 33 26 26 31 22 22 14 32 24 34 23 38 35 Some Bachelor’s College Degree 13 23 30 35 64 34 15 9 11 32 26 12 19 24 41 12 44 25 58 18 29 29 43 14 17 33 33 22 22 29 28 SOURCES: Authors’ analysis of the Current Population Survey (1996–2004). NOTE: See Appendix Table A.1 for sample sizes. aCentral Americans are identified as Latinos born in Central America (Belize, Guatemala, Honduras, Costa Rica, El Salvador, or Panama) or with at least one parent born in Central America. The second generation includes the 1.5 generation. Our sample of Central Americans for the 1.5 and second generation ages 30 to 39 is 310. bSoutheast Asians are identified as Asians born in Southeast Asia (Vietnam, Cambodia, or Laos) or with at least one parent born in Southeast Asia. The second generation includes the 1.5 generation. Our sample of Southeast Asians for the 1.5 and second generation ages 30 to 39 is 169. 19 Table 3.2 Percentage Distribution of the Educational Attainment of First-Generation Parents and Their Second-Generation Adult Children Ethnic Group All ethnic groups First-generation parents Second-generation descendants Mexican origin First-generation parents Second-generation descendants Central Americana First-generation parents Second-generation descendants Southeast Asiana First-generation parents Second-generation descendants Other Asian First-generation parents Second-generation descendants White First-generation parents Second-generation descendants Less Than High School Diploma 38 10 75 14 45 10 29 2 16 1 25 8 High School Diploma 30 27 15 36 33 32 22 25 26 10 38 26 Some College 16 33 7 38 11 43 11 21 18 23 20 33 College Graduate 15 30 3 12 11 15 37 52 41 66 17 34 SOURCES: Authors’ analysis of the Current Population Survey (1996–2004). NOTES: Data in the table include only households with parents and adult children in which the adult children are ages 25 to 49. Results are the average of mother-to-child and father-to-child comparisons. See Appendix Table A.2 for sample sizes. aFor Central Americans and Southeast Asians, we combine the 1.5 and second generations to increase sample sizes. group of immigrants in California is from Mexico (see Chapter 2).6 Mexican immigrants are among the least educated populations in the United States. In the sample of older first-generation Mexican immigrants, almost four in five have not graduated from high school. Although first-generation immigrants from Mexico tend to have very _____________ 6Mexicans are also the largest group of immigrants to the United States (U.S. Census Bureau, 2002). 20 low levels of education, their children show remarkable gains. Among young adults in the second generation, over four in five have graduated from high school (Table 3.1, Mexican panel). Progress is also evident at the other end of the educational spectrum. The share of college graduates is about four times greater among the second generation than the first. Results for the multiple-generation household sample are nearly identical, also showing strong gains from the first generation of Mexican parents to their adult children (Table 3.2). Even with this progress, however, educational attainment levels for second-generation Mexican Americans are lower than those of other U.S. natives. Although Central American immigrants show higher levels of educational attainment than Mexican immigrants, a majority of older Central American immigrants have not graduated from high school and only one in nine is a college graduate (Table 3.1, Central American panel). By the second generation, the proportion of Central Americans who are college graduates is near the average for all U.S.-born, and the share who do not have a high school diploma drops tremendously (from 54% for the first generation to 16% for the second generation). The sample of multiple-generation households for Central Americans also shows that adult second- and 1.5-generation children are much better educated than their parents, but progress does not appear to be as dramatic for college graduation.7 In contrast to immigrants from Mexico and Central America, Asian immigrants (not including those from Southeast Asia) tend to be better educated than other immigrants; indeed, they are more likely than U.S. natives to graduate from college. This is particularly the case for immigrants from East Asia, the Philippines, and India. It is not surprising then that their second-generation descendants have especially high levels of education, although the increase across generations is remarkable (Table 3.1, Other Asian panel). Over half of secondgeneration descendants of Asian immigrants have graduated from college and not completing high school is extremely uncommon. _____________ 7The 1.5 generation is made up of foreign-born individuals who came to the United States by age 10. The 1.5 generation in this sample is less likely than the second generation to graduate from college. We combine these groups for Central Americans to increase our sample sizes. 21 Even adult children of Southeast Asian immigrants, the least educated Asian immigrant group, show tremendous increases in educational attainment compared to their parents. Still, their high school dropout rates are higher than those of whites and other Asians. College graduation rates are remarkably high for second-generation Southeast Asians.8 The results for Southeast Asian immigrants and their descendants outlined here are dominated by Vietnamese, the largest and best educated of the Southeast Asian groups. Results for Hmongs, Laotians, and Cambodians could be very different. Results from the multigenerational household sample confirm these general patterns but should be interpreted with caution because of the small sample size. Non-Hispanic white immigrants in California and the United States come from many different countries and regions, including Canada, Europe, and the Middle East. The older first generation has an educational profile that is similar to that of third-and-later-generation older adults in the United States, with the share not completing high school almost as large as the share completing college. The second generation that includes their children is far better educated, with 43 percent graduating from college and only 4 percent not completing high school. Finally, blacks also show strong intergenerational progress from the first generation to the second, with the proportion graduating from college almost twice as high in the second generation as in the first. However, California is home to relatively few black immigrants. Second Generation to Third Generation Measuring educational attainment is difficult for the third generation. Because the CPS does not include information on grandparents’ place of birth, we are able to identify only third-and-later generations in our large sample. In the smaller multigenerational households sample, we are able to identify those of the third generation precisely, because they live with their second-generation parents. Having _____________ 8Note that the strong educational progress of Southeast Asians is consistent with their relatively high level of school attendance shown in Chapter 4. Because of the small sample size, we combined the 1.5 and second generations for Southeast Asians. 22 arrived primarily in the late 1970s and the 1980s, the immigrant population from Southeast Asian refugee-sending countries has not been in the United States long enough to develop a sizable third generation.9 Similarly, third-generation Central American immigrants are too few to report in our multigenerational household sample and cannot be identified in other households of the CPS.10 Black immigrants are an even more recently arrived group, on average, and have not yet developed a sizable second generation of older adults. Intergenerational progress continues from the second generation to the third generation, although progress is not as dramatic. For every group, third-generation adults are better educated than their secondgeneration parents (Tables 3.3 and 3.4). For less educated groups, these improvements are similar in scale, although generally not quite as large, as between first-generation immigrants and their second-generation children. Among all third-and-later-generation young adults, only 8 percent were high school dropouts, compared to 14 percent for the second generation of older adults (Table 3.3). Among adult children living with their second-generation parents, nine of 10 had graduated from high school and more than one in four had graduated from college, compared to only 75 percent and 16 percent (respectively) of their parents (Table 3.4). Third-generation Asians show strong intergenerational progress in the multigenerational household sample that allows comparisons of adult children with their parents (Table 3.4) but only modest progress in the larger sample that compares the entire third-and-later generation with the older second generation (Table 3.3). However, both samples show that third- or third-and-later-generation Asians are much better educated than their peers of other ethnicities, with high college graduation rates and low levels of high school dropouts relative to other third-generation groups. _____________ 9There are too few third-generation Southeast Asians to measure progress from the second to the third generations. In our multigenerational household sample, we observed no third-generation Southeast Asians. 10The CPS combines Central Americans with South Americans in its ethnic categories. 23 Table 3.3 Percentage Distribution of the Educational Attainment of Groups in the Second and Third-and-Later Generations Ethnic Group Mexican Second generation, ages 57–66 Third+ generation, ages 30–39 Asian Second generation, ages 57–66 Third+ generation, ages 30–39 Black or African American Second generation, ages 57–66 Third+ generation, ages 30–39 White (non-Hispanic) Second generation, ages 57–66 Third+ generation, ages 30–39 All ethnic groups Second generation, ages 57–66 Third+ generation, ages 30–39 Third+ generation, ages 57–66 Less Than High School High School Diploma Diploma 47 28 20 38 7 27 4 26 n/a 12 40 10 36 7 33 14 35 8 34 18 38 Some College 18 29 29 26 32 23 29 23 29 22 Bachelor’s Degree 7 12 37 44 16 30 31 28 28 22 SOURCES: Authors’ analysis of the Current Population Survey (1996–2004). NOTES: Asian includes Pacific Islanders. See Appendix Table A.3 for sample sizes. Similarly for whites, the comparison of third-and-later-generation young adults with the entire second-generation of older adults shows only slight improvements in educational attainment (Table 3.3), whereas the comparison of third-generation adults with their second-generation parents who live in the same household shows much more substantial improvements (Table 3.4). In either case, the third (or third-and-later) generation is well educated, with high school dropout rates quite low and college completion quite high. Third-generation Mexicans also show substantial improvements in education over the second generation. Among third-generation Mexicans who live with their parents, college graduation rates are twice those of their second-generation parents (although still relatively low at only 11 percent); and although over half of their U.S.-born parents did 24 Table 3.4 Percentage Distribution of the Educational Attainment of Second-Generation Parents and Their Third-Generation Adult Children Ethnic Group All ethnic groups Second-generation parents Third-generation descendants Mexican origin Second-generation parents Third-generation descendants Asian Second-generation parents Third-generation descendants White (not Hispanic) Second-generation parents Third-generation descendants Less Than High School Diploma 25 9 54 18 13 2 21 8 High School Diploma 39 36 28 39 42 21 42 36 Some College 19 29 13 33 30 41 20 27 Bachelor’s Degree 16 26 5 11 15 36 18 29 SOURCES: Authors’ analysis of the Current Population Survey (1996–2004). NOTES: Data in the table include only households with parents and adult children in which the adult children are ages 25 to 49. Results are the average of mother-to-child and father-to-child comparisons. See Appendix Table A.2 for sample sizes. not graduate from high school, more than eight in 10 third-generation descendants of Mexicans did graduate from high school (Table 3.4). Results from the larger sample (which compares entire generations but does not compare parents directly with their children) are consistent with the smaller parent-to-child sample findings presented above, showing strong although not quite as dramatic intergenerational progress from the second to the third generations (Table 3.3). This is an important finding: Increases in educational attainment do not cease between the second and third generations among Mexican Americans. Nonetheless, sizable proportions of the third generation have not completed high school and the proportion of college graduates is quite low. Although third-generation Mexicans are better educated than their parents, they are not better educated than second-generation Mexicans of similar age. This conundrum is important to understand and decipher, as it has led to confusion even among scholars who study 25 intergenerational change. Second-generation parents (Table 3.4) are much less educated than second-generation adult children (Table 3.2). Indeed, second-generation parents are only slightly better educated than first-generation parents, with very few college graduates among either group. This is because today’s second-generation parents were educated (or not) at a very different time in the United States. Not only were educational attainment levels much lower overall many decades ago for the entire U.S. population, certain groups—including Mexican Americans—had less access to public education. For example, in the 1930s the California legislature withheld state funds to local school districts that did not segregate Mexican children (Surace, 1982, citing Heizer and Almquist, 1977). Finally, the parents of today’s secondgeneration parents were less educated than today’s first-generation immigrants. In many of the primary sending countries of immigrants to the United States, including Mexico, educational attainment levels are on the rise. For example, recent immigrants from Mexico are much better educated than Mexican immigrants who came to the United States 30 or 40 years ago, a reflection of rising education levels in Mexico. Because parental educational attainment is such an important determinant of educational attainment for their children, these low levels of educational attainment of second-generation parents have important implications for third-generation educational attainment. In sum, cross-sectional comparisons of all third-generation adults with the entire second generation show only marginal differences for those of Mexican descent. However, when we compare the third generation with their parents (who form the older, least-educated part of the second generation), we see dramatic gains. Thus, intergenerational progress for Mexican Americans from the second to the third generations is strong when it is properly measured by comparing attainment for individuals with attainment of their own parents. Improvement Is Still Needed for Some Groups Increases in educational attainment from one generation to the next, although impressive, have not been sufficient for all groups to reach parity with other U.S. natives. This disparity can be partly explained by the increasing educational attainment among U.S. natives, as well. For 26 example, younger U.S.-born adults are better educated than the older cohort that includes their parents: Among third-and-later-generation young adults, the proportion who did not complete high school is less than half the rate of third-and-later-generation older adults (8% for those ages 30 to 39 versus 18% for those ages 57 to 66). Of particular concern is the state’s largest immigrant group, Mexicans. Second- and thirdgeneration Mexicans have not reached parity with other U.S.-born groups. Rates of college graduation are low, and rates of high school dropout are higher for Mexican-origin descendants than for other U.S. natives. The pattern for Asians is very different, with remarkably high levels of educational attainment—much higher than for U.S.-born whites. Even many first-generation Asian immigrants are well educated: 50 percent of the youngest adult Asian immigrants are college graduates. Second-generation Asians have even higher levels of education. Thirdand-later generations still exceed the educational levels of their parents and other U.S.-born whites. Levels of college education for U.S.-born Southeast Asians are also higher than for U.S.-born whites, even though their parents are the least educated Asian group. One cautionary note is evident, however, in the relatively high rate of high school dropouts among 1.5-generation Southeast Asians—a rate that is twice that of U.S.born whites. It is possible that Vietnamese immigrants and their descendants have met with much success educationally, whereas Laotians and Cambodians—among the least educated of any immigrant groups in the United States—have not.11 In other words, the diversity of educational attainment might be a reflection of the diversity of the immigrant groups included in the category, with the relatively numerous Vietnamese composing the best educated Southeast Asians and the less numerous Hmong, Laotians, and Cambodians composing the least educated Southeast Asians. _____________ 11Vietnamese immigrants tend to have higher levels of education than other Southeast Asians. Chapter 4 shows the higher family economic resources of Vietnamese compared to other Southeast Asians. 27 Implications for the Future Using historical educational progress across generations as a basis, we develop projections of educational attainment for the descendants of the most recent immigrants to the United States. These projections allow us to consider the future path of intergenerational progress. To develop the projections, we used transition probabilities based on an individual’s generation, maternal education, paternal education, and region of origin. Details of the method are included in Appendix A. The premise of the projections is that within an immigrant group, the historical relationship between an individual’s educational attainment and that of his/her parents will remain static over time. This relationship is, of course, a strong one—for example, the children of parents who are college graduates are much more likely than children of parents who did not complete high school to graduate from college. Nonetheless, other factors also play a role in the determination of educational attainment, and this set of projections implicitly assumes that those other factors are both embodied in the historical patterns and will remain unchanged in the future. The purpose of these projections is not to forecast what will happen, rather it is to suggest what would happen if historical patterns persist. Because Mexicans are the single largest group of immigrants and have the lowest levels of education, we focus on that group exclusively in this discussion. The large majority of Mexican immigrants have not completed high school. Indeed, well over half have not graduated from high school. Nonetheless, the vast majority of their children and grandchildren will graduate from high school. We project that only 13 percent of the children of today’s immigrants will not graduate from high school (Table 3.5), and only 11 percent of their grandchildren (the third generation) will not complete high school. A small but not insignificant proportion of the second generation will even graduate from college. That proportion is projected to increase from the second generation to the third generation but, at 17 percent, will remain far lower than that of other U.S. natives. 28 Table 3.5 Percentage Distribution of the Projected Educational Attainment of Second- and Third-Generation Descendants of Mexican Immigrants Less than high school diploma High school diploma Some college Bachelor’s degree FirstGeneration Mothers 66 21 8 5 SecondGeneration Children 13 34 39 13 ThirdGeneration Grandchildren 11 30 42 17 SOURCE: Authors’ projections. NOTES: Data in the table are based on the parent-to-child education transition matrix (see Appendix A). First-generation numbers are based on the Current Population Survey (2004) for Mexican immigrants ages 30 to 39. These projections suggest strong intergenerational gains in educational attainment from the first generation to the second generation and continuing if less dramatic gains from the second generation to the third generation. However, they also suggest that the gap between U.S.-born whites and the third-generation descendants of Mexican immigrants will continue into the future. Of particular concern are the relatively low rates of completing college. These patterns persist into the third generation because the tremendous gains into “some college” by the second generation do not translate into college graduation for the third generation. Of course, these projections are based on historical patterns and future patterns could be different. Indeed, it is the goal of public policy in general and education policy in particular to improve these outcomes. 29 4. Determinants of the Educational Attainment of Foreign-Born Youth The education and socioeconomic conditions of the first generation set the stage for the subsequent progress of the second generation. In this chapter, we explore the determinants of differences in educational attainment for those of the first generation, beginning with the recent historical context of their migration. We then turn to today’s foreignborn youth, suggesting that the low educational attainment of firstgeneration Mexican youth can be understood as a continuation of the historical pattern whereby youth from families with low socioeconomic resources come to the United States as young adults primarily for work opportunities. This analysis provides a background for discussing policy challenges and possible directions for improving educational attainment among first-generation Mexican youth—topics we address in the concluding chapter. Recent Historical Context During the 20th century, Mexican immigration to the United States was primarily made up of low-skilled workers seeking positions in agriculture, production, and service jobs.1 Even during the 1940s and 1950s, when foreign immigration to the United States was severely restricted, many Mexicans came to California as agricultural workers _____________ 1See the introduction to Waldinger and Bozorgmehr (1996) for a brief history of migration to California. The volume also provides an account of the ethnic transition of Los Angeles since the 1970s, including socioeconomic conditions of specific ethnic groups. See López and Stanton-Salazar (2001) for a brief discussion of the history of Mexican immigration to California and an analysis of second-generation Mexican Americans in San Diego. See Massey, Durand, and Malone (2002) for a history of Mexican immigration and related U.S. policies. 31 under the Bracero program (1942–1964). Still others came as unauthorized workers in production and service sectors. The easing of immigration restrictions by the Hart-Cellar Act (1965) brought additional flows of Mexican workers and their family members. Flows increased after the passage of the Immigration Reform and Control Act of 1986 (IRCA), which legalized the status of those who had been U.S. residents since 1982 and “special agricultural workers.” These residents and workers were then joined by family members under family reunification provisions, which were further expanded in 1990. Immigration from Central America has primarily been more recent, beginning in earnest during the late 1970s with Salvadorans and Guatemalans fleeing political unrest at home. Thus the historical context for immigration from Central America is not as closely tied to work opportunities in low-wage sectors as is Mexican immigration. The recent history of Asian migration has been profoundly different from that of Mexican migration. The Immigration Restriction Acts (1921, 1924) severely limited Asian immigration through very low country-of-origin quotas. The Hart-Cellar Act lifted these quotas and allowed immigration to reunite families and for those with scarce and needed skills. In the years that followed, many new Asian immigrants, notably Koreans and Asian Indians, came as skilled workers and paved the way for later immigrants to come as kin. The relatively high educational attainment of recent Asian immigrants can be understood partially as a consequence of this “skill-based” reason for entry into the United States. During the late 1970s and the 1980s, a different wave of Asian immigrants arrived: refugees fleeing communist regimes in Vietnam, Cambodia, and Laos. The first Southeast Asian immigrants were typically elites but were followed by refugees with relatively low educational attainment. Family Characteristics Among today’s immigrant youth, family characteristics follow a pattern that might be expected given the recent context of immigration: Immigrants from Mexico, Central America, and Southeast Asia tend to 32 have low resources, whereas those from other parts of Asia tend to have higher economic resources. These family resources are important because they help determine the educational attainment of firstgeneration youth. Parental education, particularly maternal education, is strongly linked to a child’s cognitive development and school success (Haveman and Wolfe, 1995). Research has shown that higher educational attainment is also associated with having a mother who speaks English, living with both parents, and having family income above the poverty level.2 To examine family resources among immigrant youth, we focus on those ages 13 to 17 who live with their families (Table 4.1).3 Immigrant youth from Mexico have particularly low family resources. Among those who live with their mother, only 17 percent have a mother with a high school diploma. Among Central American immigrant youth, family resources are also low, but not as low as for Mexican youth. For example, 31 percent have a mother who has completed high school. Among Southeast Asian immigrants, especially those from Cambodia and Laos, maternal education is low and the share living above poverty is also particularly low (only 46%). In contrast, Filipino youth are in families with high resources: 91 percent have a mother who completed high school, and 92 percent live above the poverty line. Among Asians from other backgrounds, there is a fair amount of diversity, but family resources tend to be relatively high. _____________ 2See Haveman and Wolfe (1995) for a review of the literature on the determinants of youth outcomes. See also Pirog and Magee (1997), Denton, West, and Walston (2003), Mare and Winship (1988), Grogger and Trejo (2002), and Chuang (1997). Mayer (1997) finds that family income has only a small effect on educational attainment. She calculates that if family income were doubled for poor families, the overall high school dropout rate would fall from 17.3 to 16.1 percent. She argues that the effect is small because many basic necessities are met through government programs, including Food Stamps, Medicaid, and housing subsidies. 3See Beavers and D’Amico (2005) for a study of the conditions of children in immigrant families nationally with state comparisons. 33 Table 4.1 Percentage Distribution of the Family Resources of Foreign-Born California Youth by Race and Ethnicity, Ages 13 to 17, 2000 Mother Mother Youth Youth Family Family Has High Has Mother Lives Lives Income Receives Share School Bachelor’s Speaks with Both with Any Above Public of Total Diploma Degree English Parents Relative Poverty Assistance White Mexican Central American Vietnamese Other Southeast Asian Filipino Other Asian 9 56 7 4 4 6 11 83 17 31 38 19 91 79 35 72 2 27 5 42 2 25 2 24 55 92 34 50 81 64 59 79 77 77 81 97 76 93 64 93 69 98 65 97 46 97 92 96 81 21 9 7 47 66 5 5 SOURCE: Authors’ calculations from the 2000 Census. NOTES: Maternal education and English ability include only youth who are co-resident with their biological, adoptive, or stepmother. “Mother speaks English” is the percentage who speak only English at home plus those whose mother speaks English “well” or “very well.” The share living with a relative includes only those who live in a household headed by a relative. The 2000 Census does not describe family relationships for people who are unrelated to the household head. See Appendix A for further details. Age at Arrival A growing body of research has noted that age at arrival is an important determinant of socioeconomic outcomes for immigrants. Hill (2004) finds that among immigrants in California, those who arrived as young children have similar educational outcomes to native-born youth of the same race and ethnicity.4 Among young adults from Mexico, a large share arrive at ages 15 and older. Of those who arrived by age 18, about half arrived after age 14 (Figure 4.1).5 Of those who arrived by age 24, 72 percent arrived after _____________ 4See Chiswick and DebBurman (2004) for a review of studies of age at arrival and post-immigration schooling. See also Rumbaut (2004) and Gonzalez (2003). 5Figure 4.1 reports the age of migration for two age cohorts to show that the major patterns are consistent. For a single age cohort, the age pattern could be dominated by an event such as IRCA. This act allowed legalization of immigrants who were working in the United States in 1982 and earlier. Following IRCA, there was a spike in immigration from Mexico in 1988–1990 as people joined newly legalized family and friends in the United States. For the older cohort, this spike begins about age 14. For the younger cohort, the spike begins at about age 8. However, for both cohorts, the highest levels of immigration are after age 14. 34 20 Mexicans 15 Ages 25 to 30 in 2000 Ages 19 to 24 in 2000 Percentage 10 5 0 20 Southeast Asians 15 Ages 25 to 30 in 2000 Ages 19 to 24 in 2000 Percentage 10 5 0 20 Other Asians 15 Ages 25 to 30 in 2000 Ages 19 to 24 in 2000 Percentage 10 5 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Age at arrival SOURCE: Authors’ calculations from the 2000 Census. NOTES: Data in the figure include foreign-born youth who arrived in the United States before age 19. Age at arrival does not differ substantially between men and women. Figure 4.1—Age at Arrival of Foreign-Born Mexicans and Asians 35 age 14 (not shown). In contrast, Southeast Asian immigrants were more likely than Mexican youth to arrive at early ages. Among Southeast Asians who arrived in the United States by age 24, about half arrived after age 14.6 Among other Asians, immigration is fairly evenly spread across all ages but tends to be slightly higher at ages 17 and 18. Among other Asian immigrants who arrived by age 24, just over half arrived after age 14. The high share of Mexican immigrant youth who arrive in their late teens is particularly important because these youth are much less likely than other youth to enroll in school. Among Mexican immigrant men, almost 90 percent of those who came to the United States as young children were enrolled in school at ages 16 to 18, compared to only about a third of those who arrived at age 17 (Figure 4.2, first panel). Enrolled in school 100 80 60 Working 100 Mexican 80 Southeast Asian Other Asian 60 Percentage 40 Other Asian 40 Southeast Asian 20 Mexican 20 00 2 5 8 11 14 17 2 5 8 11 14 17 Age at arrival Age at arrival SOURCE: Authors’ calculations from the 2000 Census. NOTES: Data in the table include foreign-born youth who arrived in the United States before age 19 and who have not completed high school. Reported statistics are three-year moving averages. Figure 4.2—Activities of Foreign-Born Mexican and Asian Men by Age at Arrival, Ages 16 to 18 _____________ 6Arrival by age 24 is measured for immigrants ages 25 to 30. Immigration from the Southeast Asian refugee-sending nations peaked around 1980. For the older cohort, this peak occurred when they were ages 6 to 11. For the younger cohort, this peak occurred when they were age 5 and under. 36 Among Asian and Southeast Asian immigrant men ages 16 to 18, school enrollment was over 90 percent even for those arriving at age 17. Mexican men arriving in their late teens are much more likely than other youth to be working.7 About half of those who arrived after age 14 were working at ages 16 to 18 (Figure 4.2, second panel). Among Mexican men who arrived at younger ages and among Asians, work participation at ages 16 to 18 was only around 20 percent (with the exception of Southeast Asian men who arrived in their late teens, for whom work participation was about 33 percent). College-age youth, those ages 19 to 21, had a similar pattern (not shown in the figure). Among Mexican men who arrived at age 19, only 11 percent were enrolled in school and 70 percent were working. Among Southeast Asian men who arrived in their late teens, 64 percent were enrolled in school and 44 percent were working.8 Among other Asian men who arrived at age 19, 74 percent were enrolled in school and 39 percent were working. Mexican women who arrived in their late teens were also less likely than other youth to be enrolled in school at ages 16 to 18. Among those who arrived at age 17, only 43 percent of Mexicans were enrolled in school compared to 87 percent of Southeast Asians and 94 percent of other Asians (Figure 4.3). Work participation was fairly similar between the groups at about 20 percent, but Mexican women were much more likely to be married (21% were married compared to less than 2% of Southeast Asian and other Asian women) and much more likely to be raising young children (7% compared to less than 2% of Southeast Asian and other Asian women). Among immigrant women of college age, ages 19 to 21, Mexicans were also far less likely to be in school (not shown in the figure). Only 14 percent of Mexican women who arrived at age 19 were enrolled in school, 36 percent were working, 41 percent were married, and 22 percent were raising a young child. Southeast Asian women who arrived _____________ 7In Figures 4.2 and 4.3, the share working includes those who are working while in school. 8Because of small sample sizes, Southeast Asian statistics are averaged over those who arrived between ages 16 and 21 (but, similar to other groups, are limited to men ages 19 to 21 at the time of the 2000 Census). 37 Enrolled in school 100 80 60 Working 100 Mexican 80 Other Asian Southeast Asian 60 Percentage 40 Other Asian 40 Southeast Asian 20 Mexican 20 00 2 5 8 11 14 17 2 5 8 11 14 17 Age at arrival Age at arrival SOURCE: Authors’ calculations from the 2000 Census. NOTES: Data in the table include foreign-born youth who arrived in the United States before age 19 and who have not completed high school. Reported statistics are three-year moving averages. Figure 4.3—Activities of Foreign-Born Mexican and Asian Women by Age at Arrival, Ages 16 to 18 in their late teens were much more likely to be in school (70%) but were also fairly likely to be working (40%), married (29%), and raising young children (17%).9 In contrast, other Asian women who arrived at age 19 were very likely to be in school (81%), fairly likely to be working (30%), and very unlikely to be married (10%) or raising a child (3%). The analysis of age at arrival shows that Mexican immigrant youth are more likely than Southeast Asian youth or other Asian youth to have come to the United States in their late teens. Mexican youth who arrive in their late teens have low levels of school enrollment, high levels of work participation among men, and high levels of marriage and childrearing among women. These findings suggest that many Mexican immigrant men come to California for work and many women come to work and to join their spouses or start a family. _____________ 9Because of small sample sizes, Southeast Asian statistics are averaged over those who arrived between ages 16 and 21 (but, similar to other groups, are limited to women ages 19 to 21 at the time of the 2000 Census). 38 In contrast, Asian immigrants are not particularly likely to arrive in their late teens and even among those who do, school enrollment is high, work participation among men is substantially lower than that of Mexican men, and family formation among women is lower than among Mexican women. These findings suggest that many Asian youth come to California to attend school. Many Southeast Asians came to California as refugees to escape oppressive conditions in their native countries. School enrollment is fairly high among Southeast Asians, even among those who arrived in their late teens. This chapter has identified two factors that contribute to the lower educational attainment of Mexican immigrant youth relative to Asian immigrant youth. First, family characteristics associated with youth success, such as parental education, are much lower among Mexican immigrant youth than among Asian youth (other than those from Southeast Asia). Second, for immigrant youth from Mexico, many arrive after age 14 and a large share do not appear to enroll in school in the United States. Their educational attainment is therefore quite low. Among Southeast Asian youth, family resources are also very low. Southeast Asian immigrant youth are more likely than Mexican immigrant youth to be living in poverty. However, Southeast Asian immigrants are much more likely than Mexican immigrants to enroll in school. One partial explanation for this difference may be the refugee status of many Southeast Asian immigrants. As refugees, Southeast Asian youth tend to come to the United States with their families and their families are eligible for public assistance. More than half of Southeast Asian youth receive some form of public assistance income, whereas among Mexican youth the share is less than 10 percent (Table 4.1). Public assistance income likely makes it possible for Southeast Asian youth to choose school because their family income is less dependent on their work earnings. 39 5. Determinants of the Educational Attainment of U.S.-Born Youth Low educational attainment among Mexican immigrants helps explain the low educational attainment of second-and-later-generation Mexican Americans because parental education is a key determinant of attainment (see Chapter 4). In this chapter, we explore the relationships between parental education and other family characteristics and educational attainment among U.S.-born youth. In particular, we examine the extent to which differences in family characteristics can explain racial and ethnic differences in educational attainment. In addition to providing a better understanding of the factors that explain racial and ethnic differences, the analysis is used to estimate the share of California children at risk of not finishing high school or college and to suggest mechanisms for targeting educational support programs. Family Characteristics Perhaps not surprisingly, the family characteristics of U.S.-born (second-and-later-generation) youth follow similar patterns to those shown for immigrant youth in the previous chapter. Mexican American youth tend to have low maternal education: Fewer than half have a mother with a high school diploma and only 5 percent have a mother with a bachelor’s degree (Table 5.1). Maternal education is also low among Southeast Asians. Southeast Asian and Latino youth are less likely than Asian and white youth to be growing up in a family with income above the poverty level.1 Family income is particularly low among Southeast Asians with ancestry other than Vietnamese. _____________ 1The 2000 Census measures poverty in 1999. Since 1999, a recession and unstable economic conditions have led to mild growth in poverty (Reed, 2004). 41 Table 5.1 Percentage Distribution of the Family Characteristics of U.S.-Born California Youth by Race and Ethnicity, Ages 13 to 17, 2000 Mother Mother Youth Family Has High Has Mother Lives with Income School Bachelor’s Speaks Both Above Share Diploma Degree English Parents Poverty White 44 93 29 99 71 92 Mexican American 29 46 5 70 66 78 Central American 2 44 7 66 63 79 Vietnamese 1 55 10 55 78 76 Other Southeast Asian 1 21 2 35 77 52 Filipino 2 95 53 99 81 95 Other Asian 4 84 44 77 87 92 African American 8 84 14 99 37 73 Family Income Above Low Income 81 49 47 54 22 86 80 51 SOURCE: Authors’ calculations from the 2000 Census. NOTES: See the notes to Table 4.1. The low-income threshold is twice the federal poverty threshold for each family. Factors Contributing to Racial and Ethnic Differences in Educational Attainment To explore the extent to which these differences in family characteristics explain racial and ethnic differences in educational attainment, we use longitudinal data that measure family resources in the 8th grade and follow youth until roughly age 26. We use a national sample because such longitudinal data are not available for California. However, because of similarities in family resources and youth outcomes by race and ethnicity, the national results appear to be applicable to California (see Appendix A). We focus our discussion on findings for Mexican American youth because of their relatively low educational attainment. Findings for African Americans and Asian Americans are 42 reported in the figures. The survey has too few Southeast Asians for analysis.2 Compared to third-and-later-generation white men (hereafter referred to as “white men” for ease of exposition), third-and-latergeneration Mexican American men were 7 percentage points less likely to complete high school (Figure 5.1).3 Using a statistical model, we adjust for parental education to compare Mexican Americans with whites who have similar parental education backgrounds. After this adjustment, third-and-later-generation Mexican American men have high school completion rates that are not statistically distinguishable from those of white men.4 In other words, parental education and any factors associated with parental education (such as family income, family size, or even neighborhood characteristics) explain the lower high school completion rates of third-and-later-generation Mexican American men. For third-and-later-generation Mexican American women, the gap with white women is substantially reduced when we adjust for parental education and the gap is eliminated when we further adjust for family income, single-parent families, speaking English in the home, family size, and region of residence. Among second-generation Mexican American men, their lower high school completion rates are within the sampling error of the survey. For second-generation Mexican American women, low parental education explains their gap with white women. Overall, then, the lower high school completion rates of Mexican Americans relative to whites is explained by their lower family resources, particularly parental education and related factors. _____________ 2We do not report results for second-generation whites because the NELS sample appears to be not representative. For third-and-later-generation Asians, the NELS sample is too small and also appears to be not representative. See Appendix Table A.4. 3Figure 5.1 and related figures show marginal effects from logistic regressions (measured at the means of other variables). Statistical significance is based on the coefficient estimate. See Appendix C for model results. 4We adjust for parental education by including dummy variables for mother having a high school diploma, attending college, and having a bachelor’s degree plus a dummy for both parents not completing high school and a dummy for both parents having bachelor’s degrees. 43 Percentage point difference with whites 30 Baseline Adjust for parental education Add adjustments for additional family factors 20 Add adjustments for school and neighborhood poverty 10 0 –10 –20 Men –30 Women ssttttseeshhhiihcceieirrMMMMrrcoocAddAeeeedfdfonnor++ArxxxxA+i+iiiiinddnssciccccciggddggggaaaaaaaaeeeegeegnnnnnnnnnnnnenneeeeeAAAAAAAAnrreenrrrraaeeammammmmmttmraartiitttiieeeeeeeeiiaaoorrrrrrtrrtooiiiiiiiiooiinnnccncccccconnoaaaaaaaannnnnnnnnn,,,,,,,, SOURCE: Authors’ calculations from NELS. NOTES: High school completion defined as diploma by about age 20 (six years after 8th grade). Figure shows percentage point differential with third-and-higher-generation whites based on a logistic regression. White-filled bars signify difference with whites is not statistically significant at the 10 percent level. Figure 5.1—High School Completion Rates Relative to Third-and-LaterGeneration Whites In contrast, second-generation Asian men are more likely than white men to complete high school and less than half of the difference can be explained by parental education and other family factors.5 When we add adjustments for whether they attend a high-poverty school and whether they live in a low-income neighborhood, about half of the gap can be explained.6 The high school completion rate among second-generation _____________ 5Several studies have shown the importance of “attitudes” and parenting styles for educational outcomes, especially among Asians (Fuligni, 1997, 1998; Kao and Tienda, 1995; and Kao, 2004). 6We use the share of students who are eligible for a free or reduced price meal as a measure of school poverty. We use poverty rate in the neighborhood of the school. For 44 Asian American women is not statistically different from that of white women. Among young Californians who leave school before finishing high school, the age at leaving school is not substantially different across racial and ethnic groups. For most groups, the share of youth who are not enrolled in school at ages 13, 14, 15, 16, and 17 is 3 percent or less.7 For Mexican Americans, the share out of school is 3 percent or less through age 15, 4 percent for age 16, and 5 percent for age 17. Among those who do not complete high school, the majority have completed 11th or 12th grade (Figure 5.2).8 The share is slightly lower for Mexican Americans (62%) than for whites (70%). Similar to high school completion, enrollment in postsecondary education tends to be lower for Mexican Americans than for whites but the gaps disappear when we adjust for parental education (Figure 5.3).9 Indeed, second-generation Mexican Americans have higher postsecondary school attendance than whites with similar parental education backgrounds. Mexican Americans are substantially less likely than whites to complete a bachelor’s degree (Figure 5.4). The college completion gap disappears after adjusting for family characteristics, except in the case of third-and-later-generation Mexican American women for whom adjustments cut the gap by only about half. For Mexican Americans, gaps with whites in bachelor’s degree completion are much larger than gaps in postsecondary school enrollment (comparing Figures 5.3 and 5.4). Findings from the 2000 ______________________________________________________________ research on neighborhood and school effects, see Duncan and Hoffman (1990), Brewster (1994), Patterson (2000), Wolfe, Wilson, and Haveman (2001), and Light and Strayer (2000, 2002). 7We rely on the 2000 Census for this analysis because the NELS sample for youth not completing high school is too small. 8The 2000 Census allowed respondents to indicate that they completed the 12th grade but did not achieve a high school diploma. For example, a person may have attended an entire year of schooling after the 11th grade but not completed all requirements for a high school diploma. 9Postsecondary school attendance is measured as enrollment in any postsecondary institution. Postsecondary school attendance and achieving a bachelor’s degree are measured in 2000, when most NELS respondents were age 26. 45 50 White 40 Mexican American 30 African American Percentage 20 10 0 8th grade or less 9th grade 10th grade 11th grade 12th grade SOURCE: Authors’ calculations from the 2000 Census. NOTE: The sample is too small to calculate the grade level distribution for Asians and Central Americans. Figure 5.2—Grade Level Completion Rates of U.S.-Born Young California Adults with No High School Diploma, Ages 19 to 24, 2000 Census confirm that not only do Mexican Americans have relatively low college enrollment, but of those who have attended college, the share that graduated with a bachelor’s degree is particularly low for these groups in California and the rest of the nation (Table 5.2).10 For most major racial and ethnic groups, young adults from California are slightly more likely than those from other states to have attended college, although they are not more likely to have completed a bachelor’s degree (Table 5.2).11 Perhaps the large size of the community college system helps explain the higher college enrollment of Californians. In 2002, the California Community College (CCC) system enrolled over 146,000 first-time freshmen compared to less than _____________ 10Measures of bachelor’s degree attainment in the Census (Table 5.2) exceed those of the CPS (Figure 2.1) for each major racial and ethnic group in California and nationally. 11Migration to and from other states to attend college or shortly after finishing college tends to increase college attendance and completion rates for California. To estimate college rates for those who were educated in California, Table 5.2 reports statistics by state of birth regardless of state of current residence (see Reed, 2005). 46 ssttttseeshhhiihcceieirrMMMMrrcoocAddAeeeedfdfonnor++ArxxxxA+i+iiiiinddnssciccccciggddggggaaaaaaaaeeeegeegnnnnnnnnnnnnenneeeeeAAAAAAAAnrreenrrrraaeeammammmmmttmraartiitttiieeeeeeeeiiaaoorrrrrrrtrtooiiiiiiiiooiinnnccncccccconnoaaaaaaaannnnnnnnnn,,,,,,,, Percentage point difference with whites Baseline 40 Adjust for parental education Add adjustments for additional family factors 30 Add adjustments for school and neighborhood poverty 20 10 0 –10 –20 Men Women –30 SOURCE: Authors’ calculations from NELS. NOTES: Postsecondary school attendance is defined as enrollment in any postsecondary educational institution as of about age 26. See the notes to Figure 5.1. Figure 5.3—Postsecondary School Attendance Relative to Whites 30,000 in the University of California (UC) system and less than 40,000 in the California State University (CSU) system (Table 5.3).12 The representation of Latino students in the CCC system matches that of public high school graduates in California (Table 5.3), whereas Latinos are underrepresented in the CSU and UC systems. Owing to the large size of the CCC system and the relatively large share of Latinos entering CCCs, almost 80 percent of all Latino freshman in public _____________ 12Enrollment data include first-time freshman enrollees from private high schools and from out of state. If limited to graduates from California public high schools, the share of whites in UC would be slightly lower (34%) and the share would be higher for Latinos (15%) and Asians (35%). 47 Percentage point difference with whites Baseline Adjust for parental education Add adjustments for additional family factors 40 Add adjustments for school and neighborhood poverty 30 20 10 0 –10 –20 Men –30 Women ssttttseeshhhiihcceieirrMMMMrrcoocAddAeeeedfdfonnor++ArxxxxA+i+iiiiinddnssciccccciggddggggaaaaaaaaeeeegeegnnnnnnnnnnnnenneeeeeAAAAAAAAnrreenrrrraaeeammammmmmttmraartiitttiieeeeeeeeiiaaoorrrrrrtrrtooiiiiiiiiooiinnnccncccccconnoaaaaaaaannnnnnnnnn,,,,,,,, SOURCE: Authors’ calculations from NELS. NOTES: College completion is defined as attaining a bachelor’s degree as of about age 26. See the notes to Figure 5.1. Figure 5.4—College Completion Ratio Relative to Whites higher education in California are in the CCC system.13 Thus, from a policy perspective, the success of students in the CCC system is an important determinant of the success of Latino youth, an issue we return to in the concluding chapter. Transfer rates from CCC are fairly low, with about 48,000 students per year transferring to CSU, about 11,000 students per year transferring to UC, and about 8,000 students per year transferring to private institutions (California Postsecondary Education Commission, 2002). _____________ 13See also Lopez, Puddefoot, and Gándara (2000). Fry (2002) documents the disproportional enrollment of Hispanics in community colleges at the national level. 48 Table 5.2 Percentage Distribution of College Attendance and Completion Rates of U.S.-Born Youth by Race and Ethnicity, Ages 25 to 29, 2000 Born in California Born Elsewhere in the U.S. Some Bachelor’s Graduation Some Bachelor’s Graduation College Degree Rate College Degree Rate White 69 31 44 65 32 49 Mexican American 48 12 26 45 13 28 Central American 65 23 35 65 27 41 Vietnamese 63 31 50 Other Southeast Asian Filipino 82 39 48 84 48 57 Other Asian 91 66 72 88 70 79 African American 54 15 28 48 15 31 SOURCE: Authors’ calculations from the 2000 Census. NOTES: The column labeled “Graduation Rate” shows the percentage with a bachelor’s degree of those who attended college. The sample size is too small to calculate statistics for Vietnamese in California and for other Southeast Asians nationally. Table 5.3 Percentage Distribution of Students in Higher Education Institutions in California by Race and Ethnicity, 2002 Public High School CCC CCC CSU CSU UC UC Graduates Freshmen Transfers Freshmen Graduates Freshmen Graduates White 43 39 42 39 43 37 40 Latino 33 33 20 23 20 14 12 Filipino 3 4 36 4 5 4 Other Asian 12 10 14 14 14 33 30 African American 7 7 57 5 3 3 Number 323,573 146,240 49,029 38,710 59,025 29,664 34,716 SOURCE: Reed (2005), based on California Postsecondary Education Commission Data. NOTES: CCC transfers include transfers to private institutions. Columns do not sum to 100 percent because some students do not identify their race or ethnicity. 49 The share transferring to four-year institutions is particularly low for Latinos (Table 5.3). One explanation for low transfer rates is that students attend CCCs for other programs such as workforce training, vocational or occupational education, and remedial education. Thus, the CCC system provides important opportunities for educational attainment beyond simply the transfer function. Research on the efficacy of community college provides mixed results. In an analysis of research based on the NELS, Fry (2004) finds that Latino students are more likely than similarly prepared white students to choose less-selective and “open” admissions colleges such as community colleges. Fry argues that because selective colleges tend to have higher graduation rates, more Latino students would earn bachelor’s degrees if they were to choose the same colleges.14 In contrast, Sandy, Gonzalez, and Hilmer (forthcoming) find that differences in student quality (and not institutional quality) explain most of the lower bachelor’s degree completion rates for students who attend community college. Furthermore, Gonzalez and Hilmer (forthcoming) find that access to a community college increases the chances that Latino students finish a bachelor’s degree. Apparently, although some Latino students and others may have been hindered in their progress at community colleges (e.g., by lack of counseling, inability to transfer units, or limited availability of courses), many others have found that community colleges promote their progress by providing greater access through low fees, flexible schedules, and local availability. California Youth At Risk of Low Educational Attainment Estimates of the relationship between family resources and young adult educational attainment can be used to describe the size and nature of the California population at risk for low attainment in the future. If the relationships measured in the NELS were to continue to hold, we would _____________ 14Fry (2004) also finds that Latinos have lower graduation rates than whites when comparing those in the same colleges. He argues that lower completion rates may be related to Latinos’ delayed enrollment in college, greater financial responsibility for family members, and living at home rather than in campus housing. 50 expect about 17 percent of California’s children ages 0 to 9 to become high school dropouts (Table 5.4). Of course, California has had many educational reforms over the last decade and further reforms are likely.15 However, conditions for higher education may erode, particularly in light of current fiscal constraints and the growing number of college-age youth (Hanak and Barbour, 2005). Other changes, such as income growth or large waves of new immigrants, may also affect educational and other outcomes. Therefore, the statistics in Table 5.4 do not tell us what will happen but do provide a sense of the size and nature of the child population most at risk for low attainment. The analysis suggests that the children most at risk of low attainment are those whose parents have not completed high school. Roughly 30 percent of California children have parents with no high school diploma. Table 5.4 Percentage Distribution of Educational Attainment by Family Characteristics, Ages 0 to 9, 2000 Family Characteristic All Neither parent has a high school diploma Mother has a high school diploma Mother attended college Mother has a bachelor’s degree Both parents have a bachelor’s degree Parents do not speak English Family income < $25,000 Raised by single parent Share of Children 30 17 26 7 11 12 29 23 No High School Diploma 17 36 15 7 2 <1 29 31 29 No Bachelor’s Degree 68 95 82 61 15 4 91 88 83 SOURCE: Authors’ calculations from the NELS (1994, 2000) and the 2000 Census. NOTES: Children living with a single parent who has not completed high school are included in “Neither parent has a high school diploma.” “No high school diploma” is estimated at age 20 and “No bachelor’s degree” is estimated at age 26. See Appendix A for a discussion of measurement issues. _____________ 15See Reed (2005) for a review of recent educational reforms in California. 51 If the relationships measured during the 1990s were to hold, 36 percent of these children would drop out of high school and 95 percent would not receive a bachelor’s degree (by age 26). The risk of low attainment is also very high among children whose parents do not speak English, families with low income, and children being raised by a single parent. Among California children at risk of low attainment, Mexican Americans make up a disproportionate share. For example, among children with parents who do not have a high school diploma, Mexican Americans make up 68 percent, whereas among all children that number is 40 percent. Thus, this analysis provides further evidence that we should expect the relatively low educational attainment of Mexican Americans to continue in the next generation (see also Chapter 3). 52 6. Conclusion Our results have much to say about what we can expect for the future of California. We find evidence of substantial educational progress between second-generation immigrants and their foreign-born parents—progress that continues between the second and third generations. These findings are particularly important because prior research comparing same-age cohorts of second- and third-generation immigrants has found that second-generation immigrants tend to have higher educational attainment, suggesting the possibility that educational progress stalls between the second and third generations. Using the recent record as a basis, we project a high degree of educational progress for the third-generation grandchildren of today’s Mexican immigrants. Despite continued intergenerational progress, educational attainment for third-and-later-generation Mexican Americans still lags that of U.S.-born whites. Without future improvements, Mexican American generations in years to come will fall far short of the educational levels currently attained by U.S.-raised whites.1 The stakes are high—13 percent of California’s youth are immigrants from Mexico and another 21 percent are U.S.-born Mexican Americans. In addition, Central American immigrants, whose educational attainment is somewhat similar to that of Mexican immigrants, are growing in number (currently representing 2% of the youth population). Low educational attainment is of particular concern in light of the growing importance of education in the California labor market. Today, more so than in past decades, economic success is strongly related to educational attainment. Education can also be linked to health outcomes and civic participation. Thus, education is arguably one of the most important factors for the integration of immigrants into California. _____________ 1See López and Stanton-Salazar (2001) for a stark perspective on the future for Mexican Americans. 53 Policy Directions The findings of this study have important implications for education-related policy in California. Other areas of public policy, such as federal immigration policy and state health care policy, are relevant for immigrant integration, but these broader policy issues were not the focus of this study and we do not discuss them here.2 Improving educational attainment and other conditions among Mexican immigrants who arrive as youth is particularly challenging. Among young adult Mexican immigrants in California, almost three of every four arrived after age 14. Many of these youth are not enrolled in California schools. Thus, youth programs that operate in and through schools will not reach these Mexican youth, who have some of the lowest educational attainment in California. One approach is to encourage them to enroll in high schools. For those who do, English language training and educational counseling may be particularly important. However, a substantial share of those who are not in school are working and raising families. For these youth, adult education programs in school districts and community colleges may provide better schedules for part-time, evening, and weekend coursework in vocational or academic subjects.3 Another approach would be to target the workplace, potentially working with employers who may see advantages to workers developing English language or literacy skills. In addition, as these youth become parents, policies could be devised to engage them through their young children, by offering parental support and literacy courses.4 For second and third generations, and for immigrants who do enter California schools, the quality of the K–12 public education system is clearly a key factor in success. Recent research suggests that the quality of public schools in California is relatively low (Carroll et al., 2005; The Civil Rights Project, 2005) and that Latino and African American students are more likely than other groups to be in low-performing _____________ 2The Little Hoover Commission (2002) describes a broad range of state and local policy approaches for integrating immigrants in California. 3See de Cos (2004) for a study of adult education in California. 4For example, the First Five commissions in California support a variety of parenting-support services, as do the federal Head Start programs. 54 schools (Reed, 2005). Several recent reforms have sought to improve California schools, particularly in the areas of student achievement, teacher quality, and adequate facilities (Reed, 2005). In addition to these basic elements of school quality, English language learning is of particular concern for the children of immigrants.5 One additional approach would be to target students from poor families for educational counseling, tutoring, and other mentoring.6 For example, such programs could target children whose parents have not completed high school. We find that about 30 percent of California’s children are growing up in families where neither parent has completed a high school diploma and that as many as 95 percent of these children might not achieve a bachelor’s degree.7 Mexican American children make up 68 percent of these at-risk children but they constitute only 40 percent of all children. Almost 80 percent of Latinos who enroll in public higher education enter through the CCC system. Through open admissions policies, low fees, flexible schedules, and local availability, community colleges provide greater access to a college education. Of great concern, however, is the low transfer rate to four-year institutions; transfers are especially low among Latino students.8 In addition to preparing students for transfers, community colleges provide English language, remedial, and vocational courses. As the value of education and skills in the California economy continues to grow, these courses will become increasingly important to _____________ 5See Jepsen and de Alth (2005) for analysis of English language learners in California schools. See also Rumbaut and Cornelius (1995). 6The Advancement Via Individual Determination (AVID) program is an example of a program that targets middle and high school students for additional mentoring and support. See Torres and Marquez (2005) for a discussion of parental outreach programs for postsecondary education. 7Head Start is an example of a program targeting poor children early in their educational careers. Wald and Martinez (2003) suggest an alternative targeting approach focused on youth who have poor outcomes during their teens, including those not completing high school, those deeply involved in the criminal justice system, young unmarried mothers, and adolescents in the foster care system. 8See California Postsecondary Education Commission (2002) for a summary of recent policies to address student transfer from CCC. 55 workforce training, especially for those who do not go on to complete a bachelor’s degree. Directions for Future Research The analysis of this report suggests important areas for future research. Although education policy is clearly a target area for promoting the integration of immigrants in California, research has not provided clear direction on specific policies that will improve school quality and further educational attainment, particularly among disadvantaged youth. Ongoing and future research at PPIC is working to address specific policy options in the areas of K–12 financing and reform, community college education, and adult English language training. Policy research on out-of-school immigrant youth who arrive in their teenage years is important as they represent a substantial share of California’s youth and their children will be a substantial share of the next generation of Californians. Their activities, attainment, and resources are different from those of youth who were born in the United States. Future PPIC research will examine resources, needs, and policies for these immigrant youth. Finally, California is undergoing major demographic transitions as the second-generation population continues to grow. PPIC will be developing projections of the demographic and socioeconomic conditions of California’s future population. These projections have important implications for California policy in the areas of education and workforce training, health and human services, and civic participation. 56 Appendix A Notes on Data and Methods Throughout this report we have relied on several sources of data as well as a variety of methods. In this appendix, we provide further details on the data and methods. The appendix is arranged by chapter and, within chapters, generally follows the order that data and methods are discussed in the main text. Chapter 2 Information on the number of youth by country of birth, ancestry, race, and ethnicity comes from the 2000 Census (5-percent Public Use Microdata Sample).1 We include as “U.S.-born” those who were born in one of the 50 United States and those who were born abroad of American parents. Youth born in Puerto Rico or other outlying areas of the United States are included as foreign-born. For detailed ancestral backgrounds (Table 2.1 and Appendix Table B.2), we used the first of two ancestries recorded. In some cases, we aggregated similar responses (for example, combining “Mexican” with “Mexican American”). In cases where youth identified a single Hispanic or Asian identity in response to “Hispanic origin” and “Race” questions, we used these to determine detailed ancestry. Table 2.2, which shows aggregated race and ethnicity, is based mainly on responses to “Hispanic origin” and “Race” questions with two exceptions. First, because a substantial number of Hispanics did not identify a specific Hispanic background, we used detailed ancestry to allocate these youth first to Mexican American, then to Central American, and finally to “all other Latinos.” Second, for Southeast Asians, we used ancestry to include _____________ 1Although the Census Bureau attempts to include undocumented immigrants in the Census and Current Population Survey, the number of immigrant youth is likely to be underreported. 57 Cambodian, Khmer, Hmong, Laotian, Meo, Western Lao, Vietnamese, Katu, Ma, and Mnong. The 2000 Census does not have information that allows us to separately identify second-generation U.S.-born youth from later generations. We use the CPS (March Annual Social and Economic Supplement, 1996–2004 combined)2 to estimate these shares among U.S.-born youth, separately for each row of Table 2.2.3 We identify the second generation as those with either parent born outside the 50 United States. We identify “third-and-later generations” as those with both parents born in one of the 50 United States. We then applied these shares to the count of U.S.-born youth in the Census by race and ethnicity. All analysis with the Census and CPS uses the sample weights (individual-level weights). Rumbaut (2004) provides a detailed discussion of the complexity of using the CPS to identify and describe the first and second generations. A serious limitation of the CPS is that it includes only the noninstitutionalized, civilian adult population. As noted by Rumbaut (2004) and Blumstein (2001), experiences with the criminal justice system differ substantially across racial, ethnic, and immigrant groups. Another issue that affects our analysis is that later generations and youth of mixed parentage are less likely to identify a specific ancestry. Chapter 3 Developing estimates of immigrant intergenerational progress is made difficult by the lack of large datasets that record an individual’s nativity, his/her parents’ nativity, and his/her grandparents’ nativity. Immigration researchers have been frustrated by this lack of data on the _____________ 2We do not use the CPS for 1994 and 1995 because of concerns raised by Jeffrey Passel at the Urban Institute regarding the representativeness of race, ethnicity, and immigrant generation in these years. 3The CPS does not identify detailed ancestry. We use “Hispanic origin” to identify subgroups of U.S.-born Hispanics. We separate Central American immigrants from South American immigrants using country of birth. For Southeast Asian immigrants, we use country of birth of youth and parents (Vietnam, Laos, and Cambodia). Many Southeast Asian youth whose parents were refugees from these countries were themselves born in Thailand and Malaysia. We identify these youth from the birthplace of their parents. 58 place of birth of a U.S. native’s parents and grandparents (Hirschman, 1994). The primary sources of information on the population of the United States, the Decennial Censuses and the CPS (March Annual Social and Economic Supplement), do not include information on place of birth of an individual’s grandparents. The last three decennial Censuses have not even collected information on the place of birth of an individual’s parents. Beginning in 1994, the CPS has included information on the place of birth of an individual’s parents but still does not collect information on grandparents’ nativity. Such information is necessary to identify the third-generation descendants of immigrants. The General Social Survey (GSS), a much smaller survey than the CPS and conducted only in English, does collect information on foreign birth of an individual’s parents and grandparents and also collects information on ancestry but not specific country of birth. We used the GSS to estimate intergenerational change for Mexican Americans and found that the results are consistent with those presented in this report (see Johnson, 2005). In this chapter, we use recent CPS data to estimate changes in educational attainment across three generations of immigrants and their descendants. We use the following definitions of generations: The first generation consists of anyone born outside the United States (not of American parents) and who came to the United States at age 11 or older; the 1.5 generation consists of those foreign-born individuals who came to the United States as young children; the second generation consists of U.S.-born individuals with at least one foreign-born parent; and the third generation consists of individuals with at least one foreign-born grandparent. We develop two samples from the CPS data to estimate intergenerational changes. In one sample, we consider only those adults who live with their parents. This multigenerational household sample links parents with their children and provides for direct identification of first, second, and third generations. Specific countries of origin can be identified in the public use data for an individual and his/her parents, allowing for the identification of the first- and second-generation descendants of immigrants by country of origin. Third-generation descendants of immigrants can be directly identified by examining the place of birth of the parents of the adult child’s parents, something that 59 can be done only if the adult child lives with his/her parents. In addition, because the CPS includes data on the educational attainment of all household members, information on the educational attainment of the adult child and his/her parents is available in the multigenerational samples of the CPS. This permits an analysis of generational changes that compares an individual with his/her parents. Moreover, aggregation bias in ethnic categories is not a problem with this sample: Because we compare adult children with their parents, ethnic groups are consistently represented in the different generations. The tables show the average of educational attainment for mothers and fathers of a previous generation in comparison to that of their adult children. The primary concern with the multigenerational sample is whether residents of such households are representative of all immigrants and their descendants. On the one hand, we might think that living in an extended family is an indication that an individual has not adopted U.S. norms, as such households are uncommon among U.S. natives. In addition, adult children might live with their parents because they lack the resources to establish their own households. Both of these factors suggest that such individuals would have lower educational attainment. On the other hand, living in an extended family might be a marker of a strong and supportive family environment or an indication that an adult child has the resources to care for his/her parents. If this is the case, then we might expect such individuals to have higher levels of education. Results from the data suggest that adult children who live with their parents have similar educational attainment profiles as adults who do not live with their parents. In the other sample, we compare all individuals of one generation with all individuals from another generation. In this much larger sample, parents and children are not linked, and the third generation must be identified indirectly. Indeed, third-and-later-generation descendants of immigrants must be grouped together. Any adult born in the United States to parents born in the United States is assigned to the third-and-later generation. Many, if not most, of the third-and-later generation are fourth-and-later generations. Third-and-later-generation descendants of Mexican and Central/South American immigrants are identified on the basis of ancestry; specifically, those U.S. natives who 60 report one of five Mexican or Central/South American ethnicities and whose parents were born in the United States constitute the third-andlater generation.4 Third-and-later-generation descendants of Asian immigrants are identified on the basis of race; specifically, those U.S. natives who report their race as Asian or Pacific Islander and whose parents were born in the United States constitute the third-and-later generation. Aggregation bias could be a problem with the Asian category, with Asian subgroups possibly disproportionately represented in the different generations. This approach produces a much larger sample than the multigenerational sample but suffers from not directly linking parents with their descendants. In addition, third-and-latergeneration assignment is dependent on responses to race or ancestry questions, and some descendants of first-generation immigrants might not identify with that immigrant’s race or ancestry. In other work on descendants of Mexicans, we show that this most likely causes a downward bias of third-generation educational attainment. We are careful to identify the age groups in combination with the immigrant generation that provide for the most appropriate comparison of one generation with a previous generation. We do so by comparing older members of earlier generations with younger members of subsequent generations—that is, by comparing a generation that includes parents with a subsequent generation that includes their children. In both of our samples, we consider the largest immigrant groups to California and the United States. Data are from the 1996–2004 CPS (combined surveys). Sample sizes are shown below (Tables A.1 to A.3). Samples for some groups are so small (less than 100) that we did not publish results. For Central Americans and Southeast Asians, we combine the 1.5 generation with the second generation to evaluate changes that occur between immigrants and their children. This is necessary because of the small samples for these groups. _____________ 4The CPS groups Central Americans with South Americans in its reporting of ancestry. 61 Table A.1 Sample Sizes for Data Used in Table 3.1 All ethnic groups First generation, ages 57–66 Second generation, ages 30–39 Mexican First generation, ages 57–66 Second generation, ages 30–39 Central American First generation, ages 57–66 Second generation, ages 30–39 Southeast Asian First generation, ages 57–66 Second generation, ages 30–39 Other Asian First generation, ages 57–66 Second generation, ages 30–39 White (non-Hispanic) First generation, ages 57–66 Second generation, ages 30–39 Black or African American First generation, ages 57–66 Second generation, ages 30–39 All ethnic groups Third+ generation, ages 57–66 Third+ generation, ages 30–39 15,004 14,191 2,708 2,825 503 310 455 169 2,635 809 4,035 7,076 875 384 86,973 176,757 NOTE: Central American and Southeast Asian second-generation samples include the 1.5 generation. Table A.2 Sample Sizes for Data Used in Tables 3.2 and 3.4 First:second generation Second:third generation Total 3,577 8,645 Mexican 815 890 Central American 176 Insuf. Southeast Asian 113 Insuf. Other Asian 352 430 White 1,303 4,984 NOTES: The Central American and Southeast Asian second-generation samples include the 1.5 generation. “Insuf.” indicates that the sample size was less than 100, and no results are provided. 62 Table A.3 Sample Sizes for Data Used in Table 3.3 Mexican Second generation, ages 57–66 Third+ generation, ages 30–39 Asian Second generation, ages 57–66 Third+ generation, ages 30–39 Black or African American Second generation, ages 57–66 Third+ generation, ages 30–39 White (non-Hispanic) Second generation, ages 57–66 Third+ generation, ages 30–39 All ethnic groups Second generation, ages 57–66 Third+ generation, ages 30–39 Third+ generation, ages 57–66 1,228 6,030 384 1,287 Insuf. 20,230 7,723 143,591 9,795 176,757 86,973 NOTE: “Insuf.” indicates that the sample size was less than 100, and no results are provided. Finally, using the historical progress across generations with respect to educational attainment as a basis, we develop projections of educational attainment for the descendants of the most recent immigrants to the United States. To develop the projections, we use transition probabilities based on an individual’s generation, maternal education, and region of origin. These transition probabilities are derived from the CPS multigenerational household sample that links parents with their children. Thus, we have education levels for both the adult child and his/her parents. The transition probabilities are developed separately for first- to second-generation changes in educational attainment, and second- to third-generation changes. We report the results only for descendants of Mexican immigrants. Five educational attainments are measured—less than 9th grade, 9th to 12th grade but no high school diploma, high school diploma, some college, and bachelor’s degree—but only four are reported. The accuracy of the projections depends on the stability of the transition probabilities over 63 time. In addition, the method assumes that the transition probabilities we observe in multigenerational households are an accurate representation for all second- and third-generation individuals, regardless of whether they live in a multigenerational household. Chapter 4 Analysis in this chapter, focused on foreign-born youth, is based on the 2000 Census. We chose the 2000 Census over the CPS because of the larger sample size. Identification of race and ethnicity follows the same strategy as in Chapter 2. However, among these immigrant youth, Hispanic subgroups, Vietnamese, and Filipinos are identified by country of birth. Southeast Asians are identified by ancestry and country of birth (including, for example, immigrants of Vietnamese descent who were born in Thailand). Information on maternal education and English language ability is available only for those youth living with their mother. For this reason, we report family characteristics for youth ages 13 to 17—ages when most immigrant youth (79%) are living with their mother. The share living with their mother is lower among Mexican and Central American immigrants (about 75%) and higher among white immigrants (90%). Chapter 5 Analysis of family characteristics and other topics, focused on U.S.born youth but not by immigrant generation, is based on the 2000 Census. We chose the 2000 Census over the CPS because of the larger sample size. Identification of race and ethnicity follows the strategy described for Chapter 2. Vietnamese and Filipinos are separately identified based on “Race” (using any combination that includes Vietnamese or Filipino, respectively) and ancestry. Information on maternal education and English language ability, available only for youth living with their mother, is available for 86 percent of U.S.-born youth ages 13 to 17. The share living with their mother is lower among African Americans (77%). 64 To examine the role of family characteristics in explaining differences in youth attainment, we use the 1988 NELS.5 The NELS is a nationally representative sample of 8th-graders who were first surveyed in the spring of 1988. Follow-up surveys were conducted in 1990, 1992, 1994, and 2000. The NELS is a valuable data source for this study because it is longitudinal—we observe characteristics (e.g., parental education) at early ages that can then be related to attainment of the same youth at later ages. In addition, the NELS covers enough years to allow observation of young adult attainment, such as completion of a bachelor’s degree by age 26. For this reason, we chose not to use more recent longitudinal surveys such as the National Longitudinal Survey of Youth from 1997 and the Adolescent Health Survey. However, in follow-up work on this issue, we do expect to use these newer surveys. Measures from the NELS are not defined exactly the same as those from the Census or CPS. In general, the NELS has more detailed information for the specific cohort. For example, to measure high school completion in NELS, we use high school diploma received by 1994 when most respondents are approximately age 20.6 In comparison, using the Census we measured high school diploma or GED received for those ages 25 to 29. Relative to the 2000 Census and the CPS, the NELS measures unusually high numbers of students who report having enrolled in postsecondary education. The difference appears to be due to a difference in the surveys. In the Census and CPS, the question on educational attainment refers specifically to college and to college degrees. In the NELS, the question refers to “college, university, or vocational, technical, or trade school for academic credit.” The NELS _____________ 5For other studies of determinants of youth outcomes, see Haveman and Wolfe (1995) and Pirog and Magee (1997). For studies of differences between the foreign-born and the native-born, see Mare and Winship (1988), White and Kaufman (1997), and Fligstein and Fernandez (1985). For studies of the relationship between outcomes and behaviors such as drug and alcohol use, crime, gang activity, and youth employment, see Chuang (1997), Mensch and Kandel (1988), Eckstein and Wolpin (1999), Schoenhals, Tienda, and Schneider (1998), and Moore and Zaff (2002). 6This is the same definition of high school graduation used in Grogger and Trejo (2002). Graduation is defined as of 1994 because most respondents were still in high school when the previous survey was administered in the spring of 1992. Graduation does not include those with a General Equivalency Diploma. 65 also has several follow-up questions and thus is more likely to capture enrollments that respondents may fail to mention (e.g., because they were of short durations) in the Census and CPS. Therefore, the magnitudes of group differences as measured in NELS are not directly comparable to those shown in Chapters 2 and 5. Following Grogger and Neal (2000), we use NELS sample weights in our estimates (sample weights are specific to each year of the survey). Grogger and Trejo (2002) compare the NELS with the CPS and conclude that the NELS is representative of the U.S.-born population and immigrants arriving by age 5 for whites, African Americans, and Mexican Americans. They do not consider Asian Americans or separately consider second versus third-and-later generations. We compared the 2000 NELS and the CPS and found the NELS to be fairly representative of third-and-later-generation whites, second- and thirdand-later-generation Mexican Americans, second-generation Asian Americans, and African Americans (Table A.4).7 Among Asians, the NELS does not have adequate samples of any subgroup. We therefore model an aggregated second-generation Asian category, including Southeast Asians.8 For third-and-later-generation Asian Americans, the sample size was quite small and appeared to be not representative. For second-generation whites, the sample was also apparently not representative. We did not analyze “Other Hispanics” because the sample size was relatively small and the group of the greatest relevance for California, Central Americans, was not separately identified. Kao (2004) and Kao and Tienda (1995) use the NELS to study the academic achievement of high school students by immigrant generation, including first-generation immigrants. However, the NELS is not representative of most first-generation immigrant youth (who never attended the 8th grade in the United States, with many arriving after age 14). We include immigrants arriving before age 5 with the second generation based on their similar characteristics (Hill, 2004) and to _____________ 7We do not include second-generation blacks or Native Americans because of their small sample sizes in the NELS. 8The sample of Southeast Asians is small enough that dropping them from the analysis does not change the results. 66 Table A.4 Sample Sizes and Percentage Distribution of College Completion Rates in the NELS (2000) and CPS (1996–2000) Men White, second generation White, third+ generation Mexican American, second generation Mexican American, third+ generation Other Hispanic, second generation Other Hispanic, third+ generation Asian American, second generation Asian American, third+ generation African American, third+ generation Women White, second generation White, third+ generation Mexican American, second generation Mexican American, third+ generation Other Hispanic, second generation Other Hispanic, third+ generation Asian American, second generation Asian American, third+ generation African American, third+ generation NELS Sample Bachelor’s Size Degree 203 5,368 211 235 128 91 240 69 855 215 4,710 261 253 132 96 274 69 731 35 28 7 13 37 22 62 18 11 61 36 14 14 27 13 59 28 23 NOTE: CPS results are for those ages 25 to 29. Difference CPS in Sample Bachelor’s Bachelor’s Size Degree Degree 1,498 38 3 25,836 31 3 1,355 12 5 1,488 12 –1 966 20 –16 506 18 –5 614 55 –7 259 44 26 3,585 14 2 1,615 44 –17 28,335 35 –1 1,530 12 –2 1,653 12 –2 1,139 21 –6 603 18 4 570 57 –2 274 46 18 5,243 16 –7 increase the sample sizes for the second generation. We remove immigrants arriving at age 5 or older. To examine high school completion, we use the 1994 survey to measure “high school diploma” at a time when most respondents were about age 20. For postsecondary schooling and bachelor’s degree completion, we use the 2000 survey, when most respondents were about age 26. For detailed model results from the NELS logistic regression analysis, see Appendix C. We use the NELS at the national level because the California subsample is small and not necessarily representative. We include a California indicator (dummy variable) in the models along with other 67 regional indicators. We argue that the model results are likely to be relevant for California because racial and ethnic differences in education are similar for California youth and youth from the rest of the nation (see Table 5.2 and Reed, 2005).9 To verify the applicability of NELS model estimates to California youth, we use model coefficients from the NELS to project the bachelor’s degree completion rates of Californiaborn children ages 6 to 9 in the 1980 Census.10 We then compare these projections to bachelor’s degree attainment for California-born young adults ages 26 to 29 in the 2000 Census. There are many reasons why the projections would not match the 2000 Census, including that measures of education and income differ in the two surveys. However, the results confirm that the NELS models are applicable to California (Table A.5). For the estimations in Table 5.4, we followed the same approach, using California children under age 10 in the 2000 Census. The NELS allows us to assess parent-to-child educational progress without limiting the analysis to young adults who live with their parents (as was necessary for the analysis in Chapter 3). The analysis in Chapter Table A.5 Percentage Distribution of NELS Model Projections Versus Actual Bachelor’s Degree Attainment for Californians by Race and Ethnicity White Mexican American Asian African American NELS Projection 35 11 53 19 2000 Census 31 13 58 16 SOURCES: Authors’ calculations from NELS (2000) and the 1980 and 2000 Censuses. _____________ 9Asians, especially Filipinos, from California are somewhat less likely to achieve a bachelor’s degree than are their counterparts nationally. 10For these NELS models, we combined immigrant generations because the Census does not distinguish between second and third-and-later generations. Instead, we add English language proficiency of the parents to the NELS models. 68 3 relies on the CPS to study more racial and ethnic groups than are represented in the NELS. In particular, as shown in Table A.4, the NELS is not representative of third-and-later-generation Asian Americans. However, as a check that our findings on intergenerational progress are robust, we show here that all of the groups represented in the NELS demonstrate substantial educational progress from parent to child (Tables A.6 and A.7). Table A.6 Percentage Distribution of Educational Attainment by Generation: Mother-to-Child Ethnic Group All ethnic groups Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Mexican Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Asian Foreign-born parents Second-generation youth White (non-Hispanic) Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Black or African American Second+-generation parents Third+-generation youth Less Than High School Diploma High School Diploma Some College Bachelor’s Degree 31 19 31 19 6 8 48 39 11 29 42 18 5 16 46 33 65 15 18 2 14 14 61 11 25 31 40 4 10 15 59 16 16 15 28 41 1 2 33 64 11 21 44 24 2 6 41 51 9 30 41 20 5 15 43 36 15 24 52 10 6 18 57 20 SOURCES: Authors’ calculations from the NELS (2000). NOTE: In our analysis of the NELS, second-generation youth also include foreign-born youth who arrived in the United States by age 5. 69 Table A.7 Percentage Distribution of Educational Attainment by Generation: Father-to-Child Ethnic Group All ethnic groups Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Mexican Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Asian Foreign-born parents Second-generation youth White (non-Hispanic) Foreign-born parents Second-generation youth Second+-generation parents Third+-generation youth Black or African American Second+-generation parents Third+-generation youth Less Than High School Diploma High School Diploma Some College Bachelor’s Degree 30 15 27 28 6 8 46 40 12 22 37 30 4 15 44 38 72 6 20 2 13 19 57 12 28 22 39 12 8 11 63 18 4 17 29 50 5 2 31 63 10 20 32 39 1 6 39 53 10 22 36 32 4 15 41 40 25 21 41 14 4 16 56 25 SOURCE: Authors’ calculations from the NELS (2000). NOTE: In our analysis of the NELS, second-generation youth also include foreign-born youth who arrived in the United States by age 5. 70 Appendix B Detailed Nativity and Ancestry Table B.1 Number of Youth and Educational Attainment for the Top 26 Countries of Birth Reported by Foreign-Born Californians, Ages 13 to 24, 2000 Mexico Philippines El Salvador Vietnam Guatemala Korea Taiwan India Thailand China Iran Hong Kong Armenia Japan Nicaragua Honduras Laos Canada Russia Peru Indonesia Cambodia Ukraine South Korea England Pakistan Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Population 783,124 76,753 59,612 58,701 42,795 28,228 25,859 23,576 22,822 22,337 16,536 14,819 13,588 12,731 11,479 11,370 11,283 9,250 7,733 7,614 7,114 7,096 6,769 5,982 5,941 5,011 High School Diploma (%) 36 94 46 83 38 97 99 96 90 90 95 98 86 96 70 40 71 97 95 87 94 81 91 97 97 85 Bachelor’s Degree (%) 4 34 5 33 5 56 73 81 52 59 48 70 24 45 12 4 13 62 59 20 60 23 37 59 55 44 SOURCE: Authors’ calculation from the 2000 Census. NOTES: Includes all countries of birth with California populations over 5,000. Education is reported for adults ages 25 to 29. 71 Table B.2 Number of Youth and Educational Attainment for the Top 25 Ancestries Reported by U.S.-Born Californians, Ages 13 to 24, 2000 Mexican African American German Irish English Italian American Filipino White Chinese Hispanic Scotch Irish French European American Indian Norwegian Portuguese Japanese Vietnamese Polish Salvadoran Swedish Korean Dutch Puerto Rican Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Population 1,228,338 310,810 279,195 210,186 178,050 161,383 158,956 107,742 94,380 82,943 72,515 66,807 46,820 46,422 38,530 37,770 37,754 34,913 34,878 34,659 32,360 31,934 31,781 27,848 26,601 High School Diploma (%) 76 86 95 94 96 94 88 95 85 97 83 95 93 97 84 98 94 97 78 96 81 98 96 94 79 Bachelor’s Degree (%) 12 17 40 39 45 41 25 39 20 70 19 43 30 54 13 43 26 55 34 54 15 48 62 42 18 SOURCE: Authors’ calculation from the 2000 Census. NOTES: Includes all ancestries with California populations over 25,000. Specific race or Hispanic identifiers were used in place of “ancestry” when available. For example, a U.S.-born Californian identifying Hispanic background as “Mexican” is included as Mexican regardless of ancestry identification. Ancestry was used when the race or Hispanic identifier was not specific (e.g., “Other Hispanic”). Education is reported for adults ages 25 to 29. 72 Appendix C Detailed Model Results The tables in this appendix provide coefficient estimates from logistic regressions for high school completion by age 20, postsecondary school attendance by age 26, and bachelor’s degree completion by age 26. For a description of the methods, see Appendix A. 73 Table C.1 Logistic Probability Model for High School Completion Men Women Mexican American, second generation Mexican American, third+ generation Black, third+ generation Asian American, second generation Mother has a high school diploma Mother has some college Mother has a bachelor’s degree Neither parent has a high school diploma Both parents have a bachelor’s degree Family income Family income squared Raised in single– parent home Number of siblings Family size School offers vocational education % receiving free lunch at school Neighborhood poverty Constant Observations Pseudo R-squared (1) (2) (3) (1) (2) 0.358+ 0.663** 1.358** 0.129 0.389+ (1.65) (3.00) (5.42) (–0.68) (1.90) –0.205 0.041 0.294 –0.513** –0.141 (–0.97) (–0.19) (–1.16) (–2.68) (–0.71) –0.509** –0.382** –0.144 –0.069 0.17 (6.44) (4.59) (–1.48) (–0.71) (–1.63) 0.87+ 0.924+ 0.935+ 0.065 0.104 (1.85) (1.93) (1.81) (–0.2) (–0.31) 0.452** 0.264* 0.171 0.794** 0.622** (3.55) (2.02) (–1.13) (6.31) (4.83) 1.039** 0.834** 0.83** 0.693** 0.48** (8.14) (6.34) (5.46) (5.96) (3.96) 1.94** 1.342** 1.442** 2.033** 1.471** (10.69) (6.94) (6.75) (9.77) (6.76) –0.945** –0.747** –1.046** –0.741** –0.651** (5.72) (4.42) (5.32) (4.77) (4.11) 0.644** 0.405* 0.442* 1.729** 1.509** (3.78) (2.29) (2.29) (6.02) (5.18) 0.023** 0.016** 0.028** (6.78) (4.20) (8.42) –0.005** –0.002 –0.008** (2.62) (–0.89) (5.15) –0.241* –0.359** 0.026 (2.30) (3.04) (–0.26) –0.079** –0.079** –0.05* (2.99) (2.67) (2.03) –0.026 0.006 –0.015 (–0.89) (–0.19) (–0.52) 0.107 (–0.49) –0.716** (3.36) –0.766+ (1.88) 1.123** 0.495* 2.134** 1.419** 0.331 (9.29) (2.30) (6.56) (12.23) (–1.58) 7,331 7,331 7,331 6,671 6,671 0.0886 0.1145 0.2963 0.1067 0.1423 (3) 0.608* (2.51) –0.113 (–0.48) 0.373** (3.01) 0.123 (–0.33) 0.659** (4.38) 0.559** (3.92) 1.495** (6.27) –0.527** (2.84) 1.594** (5.26) 0.025** (6.46) –0.006** (3.78) 0.066 (–0.59) –0.041 (–1.46) –0.002 (–0.07) –0.319 (–0.78) –0.461+ (1.92) –0.478 (–1.07) 2.526** (5.37) 6,671 0.3415 SOURCE: Authors’ calculations from the NELS (1994). NOTES: T-ratio is given in parentheses. Models 2 and 3 also include regional and California indicators. +Significant at 10 percent. *Significant at 5 percent. **Significant at 1 percent. 74 Table C.2 Logistic Probability Model for Postsecondary School Enrollment Men Women (1) (2) (3) (1) (2) (3) Mexican American, second generation Mexican American, third+ generation Black, third+ generation Asian American, second generation Mother has a high school diploma Mother has some college Mother has a bachelor’s degree Neither parent has a high school diploma Both parents have a bachelor’s degree Family income Family income squared Raised in singleparent home Number of siblings Family size School offers vocational education % receiving free lunch at school Neighborhood poverty Constant Observations Pseudo R-squared 0.531* 0.697** 1.058** 0.411* 0.821** (2.24) (2.89) (4.20) (2.23) (4.00) 0.404+ 0.443+ 0.664* –0.13 0.314 (1.65) (1.76) (2.53) (–0.65) (–1.48) –0.161+ –0.05 0.101 0.388** 0.77** (1.78) (–0.52) (–1) (3.28) (6.03) 2.709** 2.582** 2.73** 0.218 0.342 (3.02) (2.86) (3.00) (–0.66) (–0.99) 0.45** 0.279+ 0.258+ 0.756** 0.53** (3.09) (1.85) (1.67) (6.28) (4.18) 1.263** 0.996** 0.964** 1.432** 1.12** (8.63) (6.57) (6.21) (11.86) (8.73) 3.503** 2.726** 2.703** 3.334** 2.566** (13.08) (9.89) (9.71) (12.63) (9.41) –0.652** –0.444* –0.433* –0.476** –0.323* (3.39) (2.22) (2.12) (3.06) (1.98) 1.964** 1.631** 1.701** 1.607** 1.324** (8.59) (6.98) (7.19) (6.70) (5.36) 0.035** 0.03** 0.041** (9.53) (7.74) (11.63) –0.009** –0.007** –0.012** (4.87) (3.61) (7.94) 0.199 0.224+ –0.055 (–1.57) (1.74) (–0.48) –0.111** –0.107** –0.082** (3.60) (3.45) (3.15) 0.019 0.04 –0.112** (–0.56) (–1.18) (3.74) –0.594** (3.45) –0.953** (4.63) –0.383 (–0.93) 0.108 –0.594* 0.491 0.442** –0.127 (–0.78) (2.43) (–1.6) (3.88) (–0.56) 5,626 5,626 5,626 5,613 5,613 0.1254 0.1528 0.1790 0.1286 0.1834 0.7** (3.29) 0.298 (–1.37) 0.743** (5.54) 0.348 (–1) 0.509** (3.91) 1.118** (8.50) 2.493** (9.06) –0.226 (–1.36) 1.34** (5.34) 0.04** (10.95) –0.011** (7.54) 0.009 (–0.08) –0.084** (3.16) –0.104** (3.39) –0.386+ (1.68) –0.513* (2.23) 1.327** (2.79) 0.572+ (1.73) 5,613 0.2100 SOURCE: Authors’ calculations from the NELS (2000). NOTES: T-ratio is given in parentheses. Models 2 and 3 also include regional and California indicators. +Significant at 10 percent. *Significant at 5 percent. **Significant at 1 percent. 75 Table C.3 Logistic Probability Model for a Bachelor’s Degree Men Women (1) (2) (3) (1) (2) (3) Mexican American, second generation Mexican American, third+ generation Black, third+ generation Asian American, second generation Mother has a high school diploma Mother has some college Mother has a bachelor’s degree Neither parent has a high school diploma Both parents have a bachelor’s degree Family income Family income squared Raised in singleparent home Number of siblings Family size School offers vocational education % receiving free Lunch at school Neighborhood poverty Constant Observations Pseudo R-squared –0.845* –0.303 0.097 –0.325 0.105 (2.08) (–0.73) (–0.23) (–1.38) (–0.42) –0.643* –0.247 –0.093 –0.899** –0.605* (2.02) (–0.75) (–0.27) (3.60) (2.34) –0.809** –0.666** –0.444** –0.209+ –0.156 (6.18) (4.96) (3.20) (1.87) (–1.31) 0.883** 0.872** 0.888** 0.608** 0.677** (3.07) (2.92) (2.87) (2.63) (2.73) 0.57* 0.341 0.283 0.723** 0.67** (2.50) (–1.44) (–1.17) (4.48) (4.01) 1.055** 0.798** 0.706** 1.22** 1.161** (4.77) (3.44) (3.00) (7.83) (7.11) 2.85** 2.267** 2.224** 2.727** 2.353** (12.52) (9.33) (9.01) (16.10) (13.05) –1.11** –0.866* –0.871* –0.954** –0.734** (2.85) (2.14) (2.13) (3.78) (2.82) 0.965** 0.836** 0.838** 1.083** 0.865** (8.68) (6.88) (6.77) (10.62) (7.99) 0.021** 0.014** 0.017** (7.12) (4.69) (6.87) –0.005** –0.003* –0.003** (4.20) (2.30) (3.15) –0.362** –0.357* 0.302** (2.38) (2.30) (2.59) –0.131** –0.112** –0.131** (3.51) (2.94) (4.72) –0.067+ –0.057 –0.02 (1.66) (–1.38) (–0.64) –0.544** (4.40) –1.303** (4.82) –1.33* (2.52) –1.951** –2.128** –0.829* –1.717** –2.326** (8.98) (6.86) (2.38) (11.17) (9.61) 5,626 5,626 5,626 5,613 5,613 0.1748 0.2070 0.2291 0.1614 0.1949 0.18 (–0.7) –0.541* (2.04) –0.014 (–0.11) 0.66** (2.64) 0.635** (3.74) 1.126** (6.78) 2.267** (12.39) –0.677* (2.57) 0.794** (7.21) 0.015** (5.88) –0.002* (2.43) 0.338** (2.87) –0.128** (4.51) –0.014 (–0.45) –0.228+ (1.80) –1.051** (4.75) 0.454 (–1.05) –1.687** (5.90) 5,613 0.2082 SOURCE: Authors’ calculations from the NELS (2000). NOTES: T-ratio is given in parentheses. Models 2 and 3 also include regional and California indicators. +Significant at 10 percent. *Significant at 5 percent. **Significant at 1 percent. 76 References Arias, Omar, and Walter W. McMahon, “Dynamic Rates of Return to Education in the U.S.,” Economics of Education Review, Vol. 20, No. 2, 2001, pp. 121–138. Beavers, Laura, and Jean D’Amico, Children in Immigrant Families: U.S. and State-Level Findings from the 2000 Census, The Annie E. Casey Foundation and the Population Reference Bureau, Washington, D.C., January 2005. Betts, Julian R., The Changing Role of Education in the California Labor Market, Public Policy Institute of California, San Francisco, California, 2000. Blumstein, Alfred, “Race and Criminal Justice,” America Becoming: Racial Trends and Their Consequences, Vol. 2, 2001, pp. 21–31. Borjas, George J., “Long-Run Convergence of Ethnic Skill Differentials: The Children and Grandchildren of the Great Migration,” Industrial and Labor Relations Review, Vol. 47, No. 4, 1994, pp. 553–573. Brewster, K. L., “Race Differences in Sexual Activity among Adolescent Women: The Role of Neighborhood Characteristics,” American Sociological Review, Vol. 59, 1994, pp. 408–424. California Postsecondary Education Commission, Student Transfer in California Postsecondary Education, Sacramento, California, February 2002. Camarota, Steven A., “Immigration from Mexico: Assessing the Impact on the United States,” Center for Immigration Studies, Washington, D.C., 2001. Card, David, John DiNardo, and Eugena Estes, “The More Things Change: Immigrants and the Children of Immigrants in the 1940s, 77 the 1970s, and the 1990s,” National Bureau of Economic Research Working Paper No. 6519, Cambridge, Massachusetts, 1990. Carroll, Stephen J., Cathy Krop, Jeremy Arkes, Peter A. Morrison, and Ann Flanagan, California’s K–12 Public Schools: How Are They Doing? The RAND Corporation, Santa Monica, California, 2005. Chapa, Jorge, “The Myth of Hispanic Progress: Trends in the Educational and Economic Attainment of Mexican Americans,” Journal of Hispanic Policy, Vol. 4, 1990, pp. 3–18. Chavez, Linda, Out of the Barrio: Toward a New Politics of Hispanic Assimilation, Basic Books, New York, New York, 1991. Chiswick, Barry R., and Noyna DebBurman, “Educational Attainment: Analysis by Immigrant Generation,” Economics of Education Review, Vol. 23, No. 4, 2004, pp. 361–379. Chuang, Hwei-Lin, “High School Youths’ Dropout and Re-Enrollment Behavior,” Economics of Education Review, Vol. 16, No. 2, 1997, pp. 171–186. The Civil Rights Project, Harvard University, “Confronting the Graduation Rate Crisis in California,” 2005, available at www. civilrightsproject.harvard.edu/research/dropouts/dropouts05.pdf. de Cos, Partricia, Educational Opportunities for Adults in California, California Research Bureau, Sacramento, California, February 2004. Denton, Kristin, Jerry West, and Jill Walston, Reading—Young Children’s Achievement and Classroom Experiences, U.S. Department of Education, National Center for Education Statistics, NCES 2003070, Washington, D.C., 2003. Duncan, Greg J., and Saul Hoffman, “Welfare Benefits, Economic Opportunities, and Out-of-Wedlock Births among Black Teenage Girls,” Demography, Vol. 27, No. 4, 1990, pp. 519–535. Eckstein, Zvi, and Kenneth I. Wolpin, “Why Youths Drop Out of High School: The Impact of Preferences, Opportunities, and Abilities,” Econometrica, Vol. 67, No. 6, 1999, pp. 1295–1339. 78 Farley, Reynold, and Richard Alba, “The New Second Generation in the United States,” International Migration Review, Vol. 36, No. 2, 2002, pp. 669–701. Fligstein, Neil, and Roberto M. Fernandez, “Educational Transitions of Whites and Mexican-Americans,” in George J. Borjas and Marta Tienda, eds., Hispanics in the U.S. Economy, Academic Press, Orlando, Florida, 1985. Fry, Richard, “Latinos in Higher Education: Many Enroll, Too Few Graduate,” Pew Hispanic Center, Washington, D.C., 2002. Fry, Richard, “Hispanic Youth Dropping Out of U.S. Schools: Measuring the Challenge,” Pew Hispanic Center, Washington, D.C., June 2003. Fry, Richard, “Latino Youth Finishing College: The Role of Selective Pathways,” Pew Hispanic Center, Washington, D.C., June 2004. Fuligni, Andrew J., “The Academic Achievement of Adolescents from Immigrant Families: The Roles of Family Background, Attitudes, and Behavior,” Child Development, Vol. 68, No. 2, April 1997, pp. 261–273. Fuligni, Andrew J., “The Adjustment of Children from Immigrant Families,” Current Directions in Psychological Science, Vol. 7, No. 4, August 1998, pp. 159–183. Gonzalez, Arturo, “The Education and Wages of Immigrant Children: The Impact of Age at Arrival,” Economics of Education Review, Vol. 22, No. 2, 2003. Gonzalez, Arturo, and Michael J. Hilmer, “The Role of Two-Year Colleges in the Improving Situation of Hispanic Post-Secondary Education,” Economics of Education Review (forthcoming). Griliches, Zvi, “Estimating the Returns to Schooling—Some Econometric Problems,” Econometrica, Vol. 45, No. 1, 1977, pp. 1–22. 79 Grogger, Jeffrey, and Derek Neal, “Further Evidence on the Effects of Catholic Secondary Schooling,” Brookings-Wharton Papers on Urban Affairs, Washington, D.C., 2000. Grogger, Jeffrey, and Stephen J. Trejo, Falling Behind or Moving Up? The Intergenerational Progress of Mexican Americans, Public Policy Institute of California, San Francisco, California, 2002. Hanak, Ellen, and Elisa Barbour, Sizing Up the Challenge: California’s Infrastructure Needs and Trade-offs, Occasional Paper, Public Policy Institute of California, San Francisco, California, 2005. Haveman, Robert, and Barbara Wolfe, “The Determinants of Children’s Attainment: A Review of Methods and Findings,” Journal of Economic Literature, Vol. 33, No. 4, 1995, pp. 1829–1878. Heizer, Robert F., and Alan J. Almquist, The Other Californians: Prejudice and Discrimination under Spain, Mexico, and the United States to 1920, University of California Press, Berkeley, California, 1977. Hernandez, Donald, ed., From Generation to Generation: The Health and Well-Being of Children in Immigrant Families, National Research Council and Institute of Medicine, National Academy Press, Washington, D.C., 1998. Hernandez, Donald, ed., Children of Immigrants: Health, Adjustment, and Public Assistance, National Research Council and Institute of Medicine, National Academy Press, Washington, D.C., 1999. Hill, Laura E., The Socioeconomic Well-Being of California’s Immigrant Youth, Public Policy Institute of California, San Francisco, California, 2004. Hill, Laura E., and Christopher Jepsen, “Positive Outcomes from Poor Starts: Predictors of Dropping Back In,” Working Paper, Public Policy Institute of California, San Francisco, California, 2005. Hirschman, Charles, “Problems and Prospects of Studying Immigrant Adaptation from the 1990 Population Census: From Generational Comparisons to the Process of ‘Becoming American,’” International Migration Review, Vol. 28, No. 4, 1994, pp. 690–713. 80 Jaeger, David A., and Marianne E. Page, “Degrees Matter: New Evidence on the Sheepskin Effects in the Returns to Education,” The Review of Economics and Statistics, Vol. 78, No. 4, 1996, pp. 63–84. Jepsen, Christopher, and Shelley de Alth, English Learners in California Schools, Public Policy Institute of California, San Francisco, California, 2005. Johnson, Hans P., “Maternity Before Maturity: Teen Birth Rates in California,” California Counts, Vol. 4, No. 3, Public Policy Institute of California, San Francisco, California, February 2003. Johnson, Hans P., “Increases in Educational Attainment: Measuring the Progress of the Third Generation,” Working Paper, Public Policy Institute of California, San Francisco, California, 2005. Johnson, Hans P., and Joseph Hayes, The Central Valley at a Crossroads: Migration and Its Implications, Public Policy Institute of California, San Francisco, California, 2004. Kane, Thomas, and Cecilia Rouse, “The Community College: Educating Students at the Margin Between College and Work,” Journal of Economic Perspectives, Vol. 13, No. 1, 1999, pp. 63–84. Kao, Grace, “Psychological Well-Being and Educational Attainment Among Immigrant Youth,” in Donald Hernandez, ed., Children of Immigrants: Health, Adjustment, and Public Assistance, National Research Council and Institute of Medicine, National Academy Press, Washington, D.C., 1999. Kao, Grace, “Parental Influences on the Educational Outcomes of Immigrant Youth,” International Migration Review, Vol. 38, No. 2, Summer 2004, pp. 427–449. Kao, Grace, and Marta Tienda, “Optimism and Achievement: The Educational Performance of Immigrant Youth,” Social Science Quarterly, Vol. 76, No. 1, March 1995, pp. 1–19. Light, Audrey, and Wayne Strayer, “The Determinants of College Completion: School Quality or Student Ability?” Journal of Human Resources, Vol. 35, No. 2, 2000, pp. 299–332. 81 Light, Audrey, and Wayne Strayer, “From Bakke to Hopwood: Does Race Affect College Attendance and Completion?” The Review of Economics and Statistics, Vol. 84, No. 1, 2002, pp. 34–44. Little Hoover Commission, We the People: Helping Newcomers Become Californians, Sacramento, California, 2002. López, David E., and Ricardo D. Stanton-Salazar, “Mexican Americans: A Second Generation at Risk,” in Rubén G. Rumbaut and Alejandro Portes, eds., Ethnicities, University of California Press, Berkeley and Los Angeles, California, 2001. Lopez, Elias, Ginny Puddefoot, and Patricia Gándara, eds., A Coordinated Approach to Raising the Socio-Economic Status of Latinos in California, California Research Bureau, Sacramento, California, 2000. Mare, Robert D., and Christopher Winship, “Ethnic and Racial Patterns of Educational Attainment and School Enrollment,” in Gary D. Sandefur and Marta Tienda, eds., Divided Opportunities: Minorities, Poverty, and Social Policy, Plenum Press, New York, New York, 1988. Massey, Douglas S., Jorge Durand, and Nolan J. Malone, Beyond Smoke and Mirrors: Mexican Immigration in an Era of Economic Integration, Russell Sage Foundation, New York, New York, 2002. Mayer, Susan E., What Money Can’t Buy: Family Income and Children’s Life Chances, Harvard University Press, Cambridge, Massachusetts, 1997. Mensch, Barbara S., and Denise B. Kandel, “Dropping Out of High School and Drug Involvement,” Sociology of Education, Vol. 61, No. 2, 1988, pp. 95–113. Moore, Kristin Anderson, and Jonathan F. Zaff, “Building a Better Teenager: A Summary of ‘What Works’ in Adolescent Development,” Child Trends Research Brief, Publication #2002-57, Washington, D.C., 2002. Myers, Dowell, John Pitkin, and Julie Park, California Demographic Futures: Projections to 2030, by Immigrant Generations, Nativity, and Time of Arrival in U.S., School of Policy, Planning, and 82 Development, University of Southern California, Los Angeles, California, 2005. Neumark, David, The Effects of School-to-Career Programs on Postsecondary Enrollment and Employment, Public Policy Institute of California, San Francisco, California, 2004. Neumark, David, California’s Economic Future and Infrastructure Challenges, Occasional Paper, Public Policy Institute of California, San Francisco, California, 2005. Patterson, Rhiannon, “Neighborhood Effects on High-School Drop-Out Rates and Teenage Childbearing: Tests for Non-Linearities, RaceSpecific Effects, Interactions with Family Characteristics, and Endogenous Causation Using Geocoded California Census Microdata,” unpublished paper, January 2000. Perlmann, Joel, and Roger Waldinger, “Second Generation Decline? Children of Immigrants, Past and Present—A Reconsideration,” International Migration Review, Vol. 31, No. 4, 1997, pp. 893–922. Pirog, Maureen, and Chris Magee, “High School Completion: The Influence of Schools, Families, and Adolescent Parenting,” Social Sciences Quarterly, Vol. 78, No. 1, 1997, pp. 710–724. Portes, Alejandro, and Rubén G. Rumbaut, Legacies: The Story of the Immigrant Second Generation, University of California Press, Berkeley and Los Angeles, California, 2001. Ramakrishnan, Karthick, “Children of Mixed Nativity: Accounting for the 2.5 Generation,” Working Paper, Public Policy Institute of California, San Francisco, California, 2003. Ramakrishnan, S. Karthick, and Hans P. Johnson, “Second Generation Immigrants in California,” California Counts, Vol. 6, No. 4, Public Policy Institute of California, San Francisco, California, May 2005. Reed, Deborah, “The Growing Importance of Education in California,” Occasional Paper, Public Policy Institute of California, San Francisco, California, 2003a. 83 Reed, Deborah, Racial and Ethnic Wage Gaps in the California Labor Market, Public Policy Institute of California, San Francisco, California, 2003b. Reed, Deborah, “Recent Trends in Income and Poverty,” California Counts, Vol. 5, No. 3, Public Policy Institute of California, San Francisco, California, February 2004. Reed, Deborah, “Educational Resources and Outcomes in California, by Race and Ethnicity,” California Counts, Vol. 6, No. 3, Public Policy Institute of California, San Francisco, California, February 2005. Reed, Deborah, Christopher Jepsen, and Laura E. Hill, “Transitions to Work for Racial, Ethnic, and Immigrant Groups,” Working Paper, Public Policy Institute of California, San Francisco, California, 2005. Reyes, Belinda I., ed., A Portrait of Race and Ethnicity in California: An Assessment of Social and Economic Well-Being, Public Policy Institute of California, San Francisco, California, 2001. Rumbaut, Rubén G., “The New Californians: Comparative Research Findings on the Educational Progress of Immigrant Children,” in Rubén G. Rumbaut and Wayne A. Cornelius, eds., California’s Immigrant Children: Theory, Research, and Implications for Educational Policy, Center for U.S.-Mexican Studies, University of California, San Diego, La Jolla, California, 1995. Rumbaut, Rubén G., “Passages to Adulthood: The Adaptation of Children of Immigrants in Southern California,” in Donald Hernandez, ed., Children of Immigrants: Health, Adjustment, and Public Assistance, National Research Council and Institute of Medicine, National Academy Press, Washington, D.C., 1999. Rumbaut, Rubén G., “Ages, Life Stages, and Generational Cohorts: Decomposing the Immigrant First and Second Generations in the United States,” International Migration Review, Vol. 38, No. 3, Fall 2004, pp. 1160–1205. Rumbaut, Rubén G., and Wayne A. Cornelius, eds., California’s Immigrant Children: Theory, Research, and Implications for 84 Educational Policy, Center for U.S.-Mexican Studies, University of California, San Diego, La Jolla, California, 1995. Rumbaut, Rubén G., and Alejandro Portes, eds., Ethnicities: Children of Immigrants in America, University of California Press, Berkeley and Los Angeles, California, 2001. Sandy, Jonathan, Arturo Gonzalez, and Michael J. Hilmer, “Alternative Paths to College Completion: The Effect of Attending a Two Year School on the Probability of Completing a Four Year Degree,” Economics of Education Review (forthcoming). Schoenhals, Mark, Marta Tienda, and Barbara Schneider, “The Educational and Personal Consequences of Adolescent Employment,” Social Forces, Vol. 77, No. 2, 1998, pp. 723–761. Setterson, Richard A., Jr., Frank Furstenberg, Jr., and Rubén G. Rumbaut, eds., On the Frontier of Adulthood: Theory, Research, and Public Policy, University of Chicago Press, Chicago, Illinois, 2005. Smith, James P., “Assimilation across the Latino Generations,” American Economic Review, May 2003, pp. 315–319. Smith, James P., and Barry Edmonston, eds., The New Americans, National Research Council, National Academy Press, Washington, D.C., 1997. Surace, Samuel J., “Achievement, Discrimination, and Mexican Americans,” Comparative Studies in Society and History, Vol. 24, No. 2, 1982, pp. 315–339. Torres, Celina, and Amalia Marquez, Reaching Higher Ground: Parental Outreach Programs at the Post-Secondary Level, Tomás Rivera Policy Institute, University of Southern California, Los Angeles, California, 2005. Trejo, Stephen J., “Why Do Mexican Americans Earn Low Wages?” Journal of Political Economy, Vol. 105, No. 6, December 1997, pp. 1235–1268. 85 Trejo, Stephen J., “Intergenerational Progress of Mexican-Origin Workers in the U.S. Labor Market,” Journal of Human Resources, Vol. 38, No. 3, Summer 2003, pp. 467–489. U.S. Census Bureau, Foreign-Born Population of the United States, Current Population Survey—March 2002, Detailed Tables (PPL-162), 2002, available at www.census.gov/population/www/socdemo/ foreign/cps2002.html. Wald, Michael, and Tia Martinez, “Connected by 25: Improving the Life Chances of the Country’s Most Vulnerable 14–24 Year Olds,” Working Paper, William and Flora Hewlett Foundation, Menlo Park, California, November 2003. Waldinger, Roger, and Mehdi Bozorgmehr, eds., Ethnic Los Angeles, Russell Sage Foundation, New York, New York, 1996. White, Michael J., and Gayle Kaufman, “Language Usage, Social Capital, and School Completion among Immigrants and Native-Born Ethnic Groups,” Social Science Quarterly, Vol. 78, No. 2, 1997, pp. 385–398. Wolfe, Barbara, Kathryn Wilson, and Robert Haveman, “The Role of Economic Incentives in Teenage Nonmarital Childbearing Choices,” Journal of Public Economics, Vol. 81, 2001, pp. 473–511. 86 About the Authors DEBORAH REED Deborah Reed is a research fellow and director of the Population Program at the Public Policy Institute of California. She is a specialist in labor economics with research interests in labor markets, income distribution, public policy, and poverty. A recipient of fellowships from the Mellon Foundation and Yale University, she has served as a consultant to the World Bank in addition to her teaching and research activities. She holds a Ph.D. in economics from Yale. LAURA E. HILL Laura E. Hill is a research fellow at the Public Policy Institute of California, where she researches immigrants, international migration, race and ethnicity, and youth. Before joining PPIC as a research fellow, she was a research associate at The SPHERE Institute and a National Institute of Aging postdoctoral fellow. She has a Ph.D. in demography from the University of California, Berkeley. CHRISTOPHER JEPSEN Christopher Jepsen is a research fellow at the Public Policy Institute of California. His primary research area is the economics of education, with a particular focus on student achievement. Specific research topics include English learners, class size reduction, immigrant intergenerational progress in education, and student demographics in the Central Valley. He holds a Ph.D. in economics from Northwestern University. HANS P. JOHNSON Hans P. Johnson is a demographer whose research interests include international and domestic migration, population estimates and projections, and state and local demography. Before joining PPIC, he was senior demographer at the California Research Bureau, where he conducted research for the state legislature and the governor’s office on population issues. He has also worked as a demographer at the California Department of Finance, specializing in population projections. He holds a Ph.D. in demography from the University of California, Berkeley. 87 Related PPIC Publications Falling Behind or Moving Up? The Intergenerational Progress of Mexican Americans Jeffrey Grogger and Stephen J. Trejo The Socioeconomic Well-Being of California’s Immigrant Youth Laura E. Hill English Learners in California Schools Christopher Jepsen and Shelley de Alth “Second-Generation Immigrants in California” California Counts: Population Trends and Profiles Volume 6, Number 4, May 2005 S. Karthick Ramakrishnan and Hans P. Johnson Racial and Ethnic Wage Gaps in the California Labor Market Deborah Reed and Jennifer Cheng “Educational Resources and Outcomes in California, by Race and Ethnicity” California Counts: Population Trends and Profiles Volume 6, Number 3, February 2005 Deborah Reed PPIC publications may be ordered by phone or from our website (800) 232-5343 [mainland U.S.] (415) 291-4400 [outside mainland U.S.] www.ppic.org 89" ["post_date_gmt"]=> string(19) "2017-05-20 09:36:34" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(8) "r_905drr" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2017-05-20 02:36:34" ["post_modified_gmt"]=> string(19) "2017-05-20 09:36:34" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(50) "http://148.62.4.17/wp-content/uploads/R_905DRR.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) }