Full Length Article
The long reach of education: Early retirement

https://doi.org/10.1016/j.jeoa.2015.08.001Get rights and content

Abstract

The goal of this paper is to draw attention to the long lasting effect of education on economic outcomes. We use the relationship between education and two routes to early retirement – the receipt of Social Security Disability Insurance (DI) and the early claiming of Social Security retirement benefits – to illustrate the long-lasting influence of education. We find that for both men and women with less than a high school degree the median DI participation rate is 6.6 times the participation rate for those with a college degree or more. Similarly, men and women with less than a high school education are over 25 percentage points more likely to claim Social Security benefits early than those with a college degree or more. We focus on four critical “pathways” through which education may indirectly influence early retirement – health, employment, earnings, and the accumulation of assets. We find that for women health is the dominant pathway through which education influences DI participation. For men, the health, earnings, and wealth pathways are of roughly equal magnitude. For both men and women the principal channel through which education influences early Social Security claiming decisions is the earnings pathway. We also consider the direct effect of education that does not operate through these pathways. The direct effect of education is much greater for early claiming of Social Security benefits than for DI participation, accounting for 72% of the effect of education for men and 67% for women. For women the direct effect of education on DI participation is not statistically significant, suggesting that the total effect may be through the four pathways.

Introduction

The central goal of this paper is to draw attention to the long lasting influence of education. It is of course not news that education is an important determinant of a person’s life course. The focus in this paper is the relationship between the level of education and two routes to early retirement. One is through the Social Security Disability Insurance program (DI), with very few people leaving DI once accepted. The second is through the early claiming of Social Security retirement benefits by those who have not already retired through the DI program. These routes are used disproportionately by those who are ill-prepared to work longer because of health or other reasons. The analysis brings to the fore just how important and long-lasting the influence of education can be. The magnitude of the “education effect” on these retirement outcomes is likely to be surprising to many readers. The results illustrate not only the enormous influence of education but also that change in the breadth and depth of education may play an important role in improving preparation for retirement in the future. To confront a wide range of problems that we face it will likely be necessary to address the critical role played by education. Early retirement is simply an example to bring attention to the far-reaching influence of a key foundation for well-being throughout the life course.

Although the focus of this paper is on the far reaching effect of education, the outcome we consider – early retirement – is itself an important policy issue for at least two reasons. First, delaying retirement can have a positive effect on the financial sustainability of the Social Security system. In addition, the nearly 8% increase in benefits for each year that claiming is delayed provides a substantial benefit to those for whom delaying claiming is feasible.

We begin by considering the relationship between education and the receipt of DI benefits for persons between the ages of 50 and 62. Then we consider the early claiming of Social Security benefits by persons between the ages of 62 and 65 who are not receiving DI benefits at 62. Education may affect DI participation and early claiming of Social Security benefits in many ways. For both routes to retirement we emphasize four critical pathways – health, employment, earnings, and the accumulation of assets – through which education may indirectly influence early retirement decisions. Education may affect DI decisions or the early claiming of Social Security benefits indirectly through each of these pathways. But education may also have an additional direct effect on both routes to retirement that does not operate through the designated pathways. We estimate both the direct and indirect influence of education on these routes to retirement. In doing so we use two different estimation methods to provide estimates of the upper and lower bounds for the direct and indirect effects of education.

Several recent papers – Autor et al., 2008, Goldin and Katz, 2008, Acemoglu and Autor, 2012 for example – emphasize the changing education composition of the workforce and its lasting effects in the labor market. They consider the relationship between educational trends and the restructuring of the U.S. labor market in recent decades. In particular, they highlight the concern that the growth in the education of the workforce has failed to keep pace with the growth of high-skill jobs. One widely studied consequence has been growing earnings inequality or “job polarization.” Here, we emphasize another critical aspect of the effect of education on labor market experience: the relationship between education and routes to retirement.

We recognize that the pathway approach that we present is only one possible way of exploring the relationship between education and DI participation and between education and the early claiming of Social Security benefits. There are at least two issues that arise in this regard. One is that we focus attention on four pathways, but there may be others. For example one of the pathway variables to DI participation (and perhaps more so to early claiming of Social Security benefits) might be life expectancy. That is, education affects life expectancy which in turn affects the decision to delay receipt of Social Security benefits. We do not include life expectancy but we do include health which is strongly related to life expectancy

A second, and related issue, is the extent to which the relationship between education and each of the pathways is causal. For example, education and earnings are strongly related, but the extent to which this relationship is causal has been a long-standing issue in economics. Card (1999), in his survey of the literature on the effect of education on earnings, puts it this way: “it is very difficult to know whether the higher earnings observed for better-educated workers are caused by their higher education, or whether individuals with greater earning capacity have chosen to acquire more schooling.” Another example is that childhood health may affect both educational attainment and late-life health. Thus part of what may appear to be an effect of education may actually be the effect of health. In the analysis that follows we measure the association between education and each pathway (and the association between each pathway and early retirement), but we make no attempt to determine the proportion of the association that might be considered causal. In this paper “education” is taken to be a marker for all that accompanies education without attempting to explore the mechanisms underlying the strong positive association between education and pathways to retirement. Rather, the goal is to highlight the magnitude of the relationship between education and an important life event – early retirement.

For ease of exposition, however, we often use the term “effect” to describe the relationship (either indirectly through the pathways or directly) between education and DI or early claiming of Social Security benefits.

The remainder of the paper is in four sections. Section 1 presents descriptive data that help to motivate and support the more formal analysis that follows. Section 2 presents the analysis of DI participation. Section 3 presents the analysis of the early claiming of Social Security benefits. Section 4 is a summary and discussion.

Section snippets

Descriptive data

The descriptive data emphasize the substantial relationship between education and the pathway variables – health, employment, earnings, and assets – through which education is assumed to influence DI participation and the early claiming of Social Security benefits. We begin by describing the striking relationship between education and DI participation and early claiming of Social Security benefits and then turn to the relationship between education and the pathway variables.

Disability Insurance participation

The analysis pertains to persons between the ages of 50 and 62. We exclude persons over the age of 62 because of eligibility for early Social Security benefits at that age. The analysis is based on the 1996–2010 waves of the Health and Retirement Study (HRS). There are approximately two years between each wave of the HRS. In each wave we include only those persons who have not previously received DI. We determine (by using the date benefits were first received) whether a person is a first-time

Estimated marginal effects of three specifications

The analytic approach we follow to understand the relationship between education and early claiming of Social Security benefits is the same as the approach followed in the analysis of DI participation. We begin by estimating three specifications analogous to the specifications estimated for DI participation. The dependent variable is whether a person begins receiving early Social Security benefits (at ages of 62, 63, or 64). The pathway variables are measured in the most recent HRS wave prior

Summary and discussion

The goal of this paper is to draw attention to the long lasting influence of education. To illustrate this influence we focus on the relationship between the level of education and two routes to early retirement – the Social Security Disability Insurance program (DI) and the early claiming of Social Security retirement benefits. These routes are followed disproportionately by those who are typically ill-prepared to work longer because of health or other reasons. Of men with less than a high

Acknowledgements

Support for this research was provided by the Met Life Foundation through the MetLife Foundation Silver Scholar Award, administered by the Alliance for Aging Research, to David Wise. The research was also supported by the U.S. Social Security Administration through grant # RRC08098400-06-00 to the National Bureau of Economic Research as part of the SSA Retirement Research Consortium, and by the National Institute on Aging, through grants #P01 AG005842 and #P30 AG012810. We have benefited from

References (17)

  • David Card

    “The causal effect of education on earnings”

  • Daron. Acemoglu et al.

    What does human capital do? a review of Goldin and Katz’s ‘the race between education and technology’

    J. Econ. Lit.

    (2012)
  • Autor et al.

    The polarization of the U.S. labor market

    Am. Econ. Rev. Papers Proc.

    (2008)
  • Banks, James, Carl Emmerson and Gemma Tetlow. 2014. “Effect of Pensions and Disability Benefits on Retirement in the...
  • David. Blau

    Retirement and consumption in a life cycle model

    J. Labor Econ.

    (2008)
  • Jeffrey Brown

    Differential mortality and the value of individual account retirement annuities

  • Jeffrey Brown et al.

    Estimating life tables that reflect socioeconomic differences in mortality

  • John Y. Campbell

    Household finance

    J. Finance

    (2006)
There are more references available in the full text version of this article.
View full text