Job Turnover, Wage Rates, and Marital Stability

The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.

Note: This report is available in its entirety in the Portable Document Format (PDF).

I. Introduction

The trends toward later and shorter marriages have contributed greatly to the changing family patterns of the last half of the 20th century (Wilson 2002). Not only are marriages lasting fewer years, the sources of marital dissolution have shifted dramatically from less widowhood to more divorce. Today, marital instability is common in most industrial countries but especially in the United States. In a recent cohort, 17 percent of men had separated or divorced by age 28;¹ by age 35, 20 percent of women had experienced a second divorce; and about 20 percent of marriages were dissolved during the first 5 years. One consequence of this instability is a high and rising rate of single parenthood. Between 1960 and 1996, the share of children not living with two parents more than tripled from 10 to over 30 percent. As of 1999, only about 60 percent of children lived with both biological and/or adoptive parents (Lerman 2001). The evidence is strong that growing up in one-parent families and unstable families is closely associated with long-term economic and social difficulties (Waite and Gallagher 2000; McLanahan and Sandefur 1994).

The high levels of marital instability in the U.S. have been taking place in the context of high levels of job instability. Every month, millions of workers leave one employer and take a job with another employer. It takes young workers a long time to enter a stable career and a long-term relationship with an employer. By the age of 30, high school graduates with no college have already worked for an average of eight employers. Nearly half of all male high school graduates experienced at least one spell of unemployment between ages 25-29 (U.S. Bureau of Labor Statistics 2000). Moreover, job instability is increasing among young men (Berhardt et al. 1998).

Researchers have long pointed to the potential connections between job market problems, marriage, and divorce. Wilson (1987) and others have tried to link the weakening of job market options for young men, especially young black men, to the decline in marriage rates. Recent increases in job instability (Neumark 2000; Berhardt et al. 1998; Light and McGarry 1998; and Farber 1995) might have led to negative consequences for workers and thus decreased their attractiveness as spouses. While the impact of unemployment on divorce has been studied for decades, important gaps remain. These gaps remain from both a substantive and methodological standpoint. Existing studies have typically focused on impacts running in a single direction—from a job market outcome to a marital status or vice versa. They often follow individuals through a specific transition, such as into marriage or out of marriage. In general, researchers have not yet extended their analyses to cover linkages between the long-term pathways in the job market and in marriage. As a result, existing approaches do not capture the time path of joint job-marriage decisions.

This study contributes to the existing literature by examining the connections between job instability, wages, and marital instability in a sequential framework. We analyze the entire sequence of annual job and marriage outcomes among young men by asking: 1) Do job stability and high wages promote marriage and marital stability? 2) How do marriage and marital stability affect job stability and wages? 3) How do shocks to education and initial labor market conditions affect the job market and marital status pathways of young men? To answer these questions, we develop a model in which young men make sequential choices about work and family and then estimate the model using a dynamic maximum likelihood (ML) approach that takes account of self-selection, simultaneity and heterogeneity into account.

The results shows that job instability contributes to reductions in wages and that high job turnover and low wages reduce the likelihood of getting and remaining married. In addition to identifying the causal link running from labor market outcomes to marital status, we find robust evidence that marriage leads to higher job stability and higher wage rates. The multivariate results reveal a pattern of dynamic feedbacks that imply cumulative impacts that exceed the contemporaneous impacts usually estimated in the literature. Simulations based on these findings illustrate how shocks that raise marriage or education or that reduce divorce set off virtuous cycles in which increased marriage interacts with increased job stability and earnings. The findings also suggest that previous studies understate the wage gains linked to marriage by ignoring the feedback effects on job stability and marriage duration.

This study improves our understanding of job and marital pathways and the linkages between them. It provides estimates of the effect of job instability on wages in the context of a model that accounts for changes in marital status. It captures the simultaneous impacts of job market factors on marriage and of marital status impacts on jobs and wages. Furthermore, it estimates these relationships while adjusting for unobserved heterogeneity that may affect both labor market and marital outcomes. Finally, it demonstrates how new econometric tools can estimate sequential job and marriage outcomes. Using these tools, one can project the long-term effects of marriage on job market outcomes and thus calculate a long-term marriage wage premium.

The next section reviews some of the relevant literature related to several lines of research relevant to this paper. Section III describes the data set, displays facts that motivate our study, and explains the rationale for the indicators we use to measure marital status and job stability. Section IV lays out and explains in detail the optimization model ascribed to individuals and how it connects to our empirical strategy. The multivariate results, which include both OLS estimates and those that control for unobserved heterogeneity, are presented in Section V. The simulation results in Section VI show examples of the implications of the statistical findings on the size of the feedback effects. Section VII concludes by discussing implications of our method and findings.

Notes from this section

1 These tabulations come from the author's of the 1979 National Longitudinal Survey of Youth. In subsequent sections, we do not distinguish between separation and divorce, and use the terms divorce or dissolution.

Note: This report is available in its entirety in the Portable Document Format (PDF).

The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.

Usage, posting and reprint of materials on the UI web site:

Most publications may be downloaded free of charge from the web site in PDF format. This information may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required.

Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact paffairs@urban.org.

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.