on Facebook
on Twitter
on YouTube
Research on Labor Markets and Human Capital Investments
Labor markets and their outcomes have been a long-standing focus of demographic research, both in their own right as a source of social and economic stratification and for the ways in which decision-making about labor market outcomes influences and is influenced by demographic behaviors such as marriage, fertility, and migration. Moreover, since domestic labor markets are influenced by an array of government policies and regulations, it is important to also understand how those policies affect decisions about whether to work and how much to work, and how those decisions, in turn, affect demographic outcomes.
These decisions are relevant not only at a point in time, but also as they change throughout the life course or change over time in response to external shocks, say to the economy, or to policy shifts. Labor market outcomes such as earnings are determined in part by investments in human capital, including education, training, health, and migration. While such investments take place throughout the life course, they are concentrated in early and middle childhood, adolescence, and the transition to young adulthood, as families and individuals make choices about schooling, labor supply, fertility and child care, health inputs, and residential location. The timing and quality of investments are likely to affect both well-being during childhood and success in adulthood.
Across and within different countries and time periods, there is wide variation in the nature of the human capital investments made in children and adults. Investment decisions and the associated outcomes also vary widely for different demographic groups defined by age, sex, health status, immigration status, and so on. Investment decisions are shaped by resource constraints of the household or family, the opportunities for investment offered by the local community and economic environment, and the incentives and disincentives embodied in government policies and private institutions. Investments in child schooling, for example, are associated with family income, the education of the child’s parents, the structure of the family, and the characteristics of the local community. Understanding these varied influences is vital for implementing future strategies to raise the level of human capital acquisition and improve its distribution. As with research on children and families, analyses of labor market behavior and human capital investment must also confront the dynamic nature of these outcomes and their joint determination with other demographic behaviors.
Contributions by PRC Staff
Recent research of PRC staff has used a variety of data sources to explore a number of issues, including:
Labor markets
Van Soest has contributed methodological and substantive advances to the study of labor supply of individuals and married couples, accounting for important policy dimensions such as tax policy and benefit rules (Van Soest and Das, 2001; Van Soest, Das and Gong, 2002). He focuses in particular on the effects of tax reforms in the Netherlands on married women’s labor supply. Gong and Van Soest (2002a) analyze the labor supply of married women.
In addition to Smith’s work on immigrant economic outcomes, Van Soest has contributed to our knowledge of the labor market performance of immigrant workers in the U.K. and Germany (Dustmann and van Soest, 2001, 2002a, 2002b). Using state-of-the-art econometric methods, he demonstrates that the estimated relationship between fluency and labor market outcomes can be biased downward if the endogeneity of language fluency is not accounted for. Dustmann and Van Soest, forthcoming) analyze models of self-reported speaking fluency, with an emphasis on the relation between ethnic concentration and years since migration.
Several PRC staff have examined labor force issues for older workers. Loughran considers the incentives for older workers, those past normal retirement age at 65, to remain in the labor market (Haider and Loughran, 2001). His results indicate that labor supply is concentrated among the most educated, wealthiest, and healthiest elderly and that non-pecuniary considerations play an important role in determining elderly labor supply decisions. In related work, Maestas examines re-entry into the labor market after retirement. She finds that 24 percent of retirees eventually return to work, usually within two years of their initial retirement. She examines whether “unretirement” is a response to financial shocks or part of a planned multi-stage retirement process, finding overwhelming evidence in support of planned unretirement. She also notes that only half of workers appear to follow the traditional retirement path of permanent exit (Maestas, 2004).
Smith and his collaborators have considered the impact of Indonesia’s economic crisis on labor market outcomes for men and women in that setting (Smith et al., 2002). Kapteyn and collaborators have studied the effect of worksharing policies, which are particularly popular in Europe, on national employment rates in OECD countries and found that such policies have no significant effect on the number of employed (Kapteyn, Kalwij, Zaidi, 2004).
Karoly and Zissimopoulos (2004) have studied older self-employed workers. Although the 1990’s witnessed a downward trend in self-employment rates, because self-employment rates rise at older ages, the approaching retirement of the baby boom cohort may halt or reverse that trend.
Klerman and his colleagues have studied the timing of return to work for new mothers (Klerman and Leibowitz, 1999). Longitudinal data from the NLSY have been used to document the timing and prevalence of return to work after childbirth, with differences explored based on pre-pregnancy employment intensity and for teenage mothers. Klerman is now updating that work, tracking the recent reversals of trends toward increased maternal work and the emergence of paternity leave in the United States (Klerman, 2003). Loughran and Zissimopoulos study the impact of age at first marriage on the wage growth of women using longitudinal data from the NLSY (Loughran and Zissimopoulos 2004).
Back to Top
Human Capital Investments
Van Soest has used data from the U.K. National Child Development Survey to examine the effect of school quality on the decision to stay in school at age 16 and subsequent wages in adulthood (Dustmann, Rajah and van Soest, 2003). He and his collaborators find that a measure of school quality, the child-teacher ratio, has a positive effect on staying in school, which, in turn, has a positive effects on subsequent wages. Datar has studied how age at entry into kindergarten affects children’s academic and other school outcomes (Datar, 2004). She finds that delaying kindergarten entry raises achievement significantly among all children. However, delaying entrance increases the value-added from schooling only among at-risk children.
Back to Top
Welfare Reform
Broadly speaking, in the past few years, the primary interest among PRC staff has been the role of social welfare policy in affecting the full range of demographic and economic behavior in both developed and developing countries. In the United States, the 1996 Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) represents the most significant departure in the structure of the U.S. social safety net since the War on Poverty began. This national policy shift and related changes at state and local levels has been the subject of a considerable body of research by PRC staff, research that will stay on the agenda for the foreseeable future.
In addition, the Medicare program in the U.S. is the focus of much current policy interest as policymakers look for ways to expand coverage while containing costs. Equally significant policy initiatives are implemented in settings outside the United States in the developed and developing world in such areas as family planning, education, health care, social insurance, and pensions. At the broadest level, PRWORA transformed the Aid to Families with Dependent Children (AFDC) entitlement program into a state-administered block grant program, now called Temporary Assistance to Needy Families (TANF). PRC staff have contributed to our understanding of how PRWORA and the earlier changes in state welfare policies have affected key demographic and economic behaviors and outcomes for families and children. These contributions derive from research that spans the national, state, and local levels.
At the national level, Karoly, Klerman and Rogowski, have undertaken one of the only evaluations of the impact of the changes in the Supplemental Security Income (SSI) program embodied in the 1996 welfare reform law (Karoly, Klerman, and Rogowski, 2001, 2002). This work identifies impacts on program caseloads and costs, as well as on affected families and children. In other work, Klerman has provided important insights into the dynamics of welfare caseloads and the impact of policy reforms (Haider and Klerman, 2001; Klerman and Haider, 2001). Their work points to a number of biases in most prior studies of welfare caseloads that overlooked the importance of dynamics. Karoly and Klerman, together with Jeff Grogger at UCLA, have undertaken the most comprehensive review to date of the impact of PRWORA on a wide range of outcomes, including the caseload, employment and earnings, income and poverty, fertility and marriage, and child well-being (Grogger, Karoly, and Klerman, 2002a, 2002b). Their work demonstrates the inherent trade-offs in designing a welfare system that seeks to both promote self-sufficiency while also improving the standard of living for the poor.
More recently, Klerman and colleagues have used more recent national administrative data to reexamine the effect of special welfare policies on the welfare caseload. With more recent data and more post-reform data, they show strong and separate effects of the independent components of reform, in the directions predicted by economic theory and with lag patterns consistent with the nature of the policies. However, consistent with Klerman’s previous work on California data (Haider and Klerman, 2001; Klerman and Haider, 2001), the effects of the economy are also very important, so that the expected effect of time limits pushing down the welfare caseload in 2002 (five years after federal welfare reform) is offset by the simultaneous recession. Klerman and colleagues are now extending this line of analysis to other outcomes (employment, earnings, income, and poverty) using the CPS.
At the state and local level, Klerman has led a team of investigators that is evaluating the impact of California’s welfare reform program (known as CalWORKs) by combining administrative and survey data (collected as part of the evaluation) to measure the impact of CalWORKs on welfare participation, employment and earnings, income, other government support, child well-being, household composition, marriage, and fertility. In work currently under way that exploits the variation in welfare policy across California’s 58 counties, the team is addressing how specific components of reform—such as time limits, sanctions, diversion, job search, training, child care, and other support services—affect outcomes. Additional work has focused on outcomes among refugees and on health insurance among current and former recipients (Gresenz and Klerman, 2002; Cox and Klerman, 2002a; Cox, Klerman and Aguire-Happoldt, 2003).
Back to Top
Future Directions and Scientific Objectives
Our research in this area will be guided by four key scientific objectives:
- To increase our knowledge of the factors that encourage or constrain human capital investments by individuals and families in the areas of health, education, labor market skills, and residential choice
- To continue to extend static and dynamic models of labor market outcomes to include the role of health status and disability
- To deepen our understanding of the impact of the 1996 welfare reform law on families and children, with a particular focus on vulnerable populations
- To use state-of-the-art methods to identify the causal impact of other policy reforms in the United States and overseas
Future work will examine labor markets and human capital investments in both low- and high-income settings. Smith, using the NIS, will investigate static and dynamic perspectives of labor market outcomes for immigrants in aggregate and for immigrant subgroups, while van Soest will continue his analysis of European immigrant groups. Loughran will continue to explore the determinants of labor force behavior for older workers, considering the role of workplace accommodations and health, while Rohwedder will focus on the role of pensions and retirement, including the role of information and uncertainties in the process of retirement planning. In related work, Karoly, Zissimpoulos, and Maestas will study self-employment among older workers in the U.S. and the U.K., particularly as a bridge to retirement, while Maestas will examine whether women who retire jointly with their husbands forego valuable earnings growth and accruals to their private pension and Social Security benefits.
These and other analyses are made possible by the rich information included in the RAND-collected surveys (e.g., IFLS, L.A.FANS, NIS), as well as analyses that draw on other survey data (e.g., HRS/AHEAD data prepared by RAND’s Center for the Study of Aging, ELSA in the U.K,. and SHARE in ten continental European countries) and administrative data. Studies in this area will be guided by an interest in outcomes for diverse segments of the population defined by demographic group, family structure, and socio-economic status. Projects will consider labor market choices and investment decisions in communities with vastly different opportunities and constraints that may promote or hinder optimal investment strategies. Considerable attention will be paid to sorting out the contribution of both observable and unobservable determinants at multiple levels—individual, family, and community. The impact of policy changes and economic shocks on short-run and longer-run investment and labor market decisions will also be of interest.
Much of the future work on welfare work will continue to revolve around issues of welfare reform as natural extensions of work already under way. Grogger, Karoly, Klerman (2002a) identify a number of holes in the existing research literature, a number of which will be the focus of investigation by PRC staff. These include funded analyses of the health insurance coverage, Food Stamp use, and use of the Earned Income Tax Credit by former welfare recipients (Klerman), and the impact of organizational type on the structure and effectiveness of welfare-to-work services (Klerman). Planned future analyses include the impact of adolescent decision-making (Karoly, Klerman), fertility (Klerman) and child well-being (Klerman, Pebley, Sastry). The PRC’s focus on the policy implications of demographic phenomena and the impact of policy on demographic decision-making will exploit RAND’s broader capabilities in policy analysis.
Together, the projects on welfare reform will inform future decisions about the direction of social welfare policy in the United States and other social welfare policy reforms in both high- and low-income countries. The PRC’s future work will use several strategies to overcome some of the challenges in assessing policy impact identified at the outset of this discussion. Several studies will use a common methodological approach that compares outcomes before and after the policy change for multiple jurisdictions with different policies at a point in time and over time, accounting for both observed and unobserved factors that may also influence the outcomes of interest. This will increase the confidence that measured effects of the economic and policy changes are causal rather than spurious. A number of these projects will continue to make methodological advancements in the use of administrative data alone and in combination with survey data. Projects will also be able to draw on the L.A.FANS, which was designed to capture the effects of welfare reform at the neighborhood level.
Top