Parental job loss and the education enrollment of youth☆
Introduction
There is considerable evidence that job loss is followed by large and persistent falls in the earnings of individuals (Jacobson et al., 1993, Stevens, 1997). Oreopoulos et al. (2008) also found that parental job loss can affect the earnings and welfare receipt of their children when those children are in the early stages of adulthood.1 These effects were interpreted by the authors as potentially causal intergenerational income effects. This research was unable, however, to uncover the pathways by which parental job loss and subsequent income reductions may affect the subsequent earnings of children.
This paper investigates one particular pathway by which parental job loss may affect the subsequent earnings of children: via lower levels of education attainment. Specifically, the paper investigates the effect of parental job loss when youth are 16 to 18 years old on enrollment in university and community college in the 2 years after high school. Job loss is identified as job ends due to permanent layoff (redundancy) and employer business failure. The effects of job loss are estimated while controlling for a number of important parental background characteristics, such as initial parental income and parental education levels.
One main avenue by which parental job loss may affect the education outcomes of youth is via an income effect. The large and persistent reduction in parental income following job loss may lower the ability of parents to support their children in their education pursuits. Attending university is costly, via both direct costs (tuition fees and books) and income foregone. Parental transfers may assist youth to pay these costs. Such transfers are important if student loans are unavailable, if loan limits are not large enough to cover costs, or if students are unwilling to borrow, given risk preferences or debt aversion. If parental job loss is affecting youth enrollment via a parental income effect, it implies that youth face financial constraints.
Given that the focus in this analysis is on parental job losses occurring when the youth is at the end of high school, early learning outcomes of youth should not be affected. Only potential financial constraints on post-secondary education attendance should be identified here. Recent research on credit constraints in the US have concluded that credit constraints are quite small: Carneiro and Heckman, 2002, Keane and Wolpin, 2001, Cameron and Taber, 2004. Ellwood and Kane (2000), however, suggest that credit constraints may be sizable. Belley and Lochner (2007) also find that credit constraints may have become more binding on US youth over time, being more important in 2000 than in the early 1980s. In addition, Belley et al. (2009) estimate less binding financial constraints in Canada than in the US in 2000.
Job loss may have effects on families in addition to income loss, including loss of self esteem, lower mental health, changes in attitudes toward risk, even marriage breakdown and substance abuse. These additional effects may directly influence the education outcomes of youth. It is thus difficult to interpret the effects of parental job loss on youth education outcomes as due to income effects alone. Even so, if parental job loss does causally affect youth education outcomes, it highlights significant negative effects of job dislocation that may not be fully understood.
The Canadian Survey of Labour and Income Dynamics (SLID) is employed in this analysis. The advantage of this data set is the ability to identify both parental job losses and the subsequent annual education enrollment of youth. Education enrollment of youth is analyzed from age 16 until age 19 or 20.2 The analysis begins with estimation of standard discrete choice models of post-secondary education enrollment. I find that parental job losses correspond with significant reductions in the probability of university enrollment (10 percentage points), after controlling for initial parental income, parental education and other observable demographic characteristics.
Some readers may be concerned that job losses may be correlated with certain unobserved characteristics of parents that also affect the education enrollment of their children. If so, the estimates I construct in the analysis may not reflect causal effects. In order to allay such fears, I also construct estimates for the set of annual education attendance outcomes for youth from age 17 to age 19 separately. This estimation allows me to construct one particular test of the exogeneity of parental job loss. To implement the exogeneity test, the effects of job losses that occur after an education enrollment outcome has already occurred are estimated. Such future losses should have no actual effect, but may have non-zero estimated effects if job losses reflect unobserved heterogeneity. These future job loss indicators were found not to be statistically significant, providing comfort that causal effects are being estimated.
Although studies of the relationship between family background and education outcomes are numerous, only a few have attempted to identify the causal effect of parental income on the level of schooling completed.3 Acemoglu and Pischke (2001) estimate a significant causal parental income effect, while several other studies have concluded that the causal effect is small or even zero (Mayer, 1997, Shea, 2000, Levy and Duncan, 2000, Blanden and Gregg, 2004). A variety of techniques were employed in these latter studies to identify causal effects, with many relying on changes in income over time within families for identification.4 The majority of such income changes may, however, only be temporary or chosen by parents, hampering the ability of such techniques to identify significant causal effects. In this paper, I identify parental shocks (job losses) that are most likely to be exogenous, and to have persistent effects on income, and examine whether these particular shocks have significant effects on youth education outcomes.
This study is closely related to recent studies on the effect of parental job loss on their children's subsequent education attainment in the US. Shea (2000) attempted to estimate the causal effect of parental income on years of schooling completed by age 25 using information from the Panel Study of Income Dynamics (PSID). Job loss of fathers, along with occupation and union status, was employed as instruments for long-run income levels. Page et al. (2009) use the PSID to analyze the effect of parental job loss before a child turns 16 on subsequent outcomes for their children, including income, earnings, and education attainment when the children are in their mid 20s. The authors only find negative effects of job loss on children's subsequent education attainment among children of lower income and black parents. Kalil and Wightman (2009) also use PSID data and estimate a significant negative effect of parental job loss on the probability of youth from middle income families obtaining any post-secondary education by age 21, with much larger effects for black youth than white youth.
This analysis is also related to a number of recent studies of the effect of parental job loss on children's earlier education outcomes. Rege et al. (2007a) find a negative effect from father's job loss on the Grade Point Average of Norwegian children at the end of high school, if the family is located in a poor job market. Kalil and Ziol-Guest (2008) find a negative effect of father's involuntary job separations (from layoff, plant closure, firing, health problems, etcetera) on grade repetition and expulsion of children from school in the US Survey of Income and Program Participation (SIPP) data. Stevens and Schaller (2009) also find that parental job loss increases the probability of their children repeating a grade at school using SIPP data. They find the effects are concentrated among children whose parents have a high school education or less.
The outline of the paper is as follows. A brief description of the Canadian post-secondary education system is provided in Section 2. The SLID data is described in Section 3. Estimates of the effect of parental job loss on post-secondary education enrollment outcomes are presented in Section 4. Section 5 concludes.
Section snippets
Canadian post-secondary education system
The majority of universities and community colleges in Canada are financed and regulated by provincial governments. University is the Canadian equivalent of four year college in the US. In 2006/07, university tuition fees for undergraduate arts programs ranged from $1668 to $6200 per year across Canadian provinces, with fees increasing considerably in many provinces over the preceding 15 years. Fees are thus a little below levels charged by US public four year colleges. Entrance to university in
Survey of Labour and Income Dynamics (SLID)
The Survey of Labour and Income Dynamics (SLID) is a longitudinal survey of Canadian households. Approximately 15,000 households are sampled for inclusion every 3 years. Once a household is chosen, all its original members are interviewed annually for 6 years, even if some members leave the household. This survey feature is especially important here as youth completing high school may leave the parental home to either work or study. Youth who leave the household are tracked to their new residence
Estimated models
The linear probability model is employed to investigate two binary outcomes for youth: (1) attended any post-secondary institution (university, community college, trade school and business school), and (2) attended university.11
Conclusions
Oreopoulos et al. (2008) found that job losses by fathers can have important intergenerational effects, lowering the subsequent earnings levels and raising the welfare receipt rates of their male children when they reached adulthood. This analysis identifies a very important potential pathway by which parental job loss can have such effects on the future labour market success of children: via educational attainment. Higher levels of educational attainment are strongly related to higher earnings
References (42)
- et al.
Changes in the wage structure, family income, and children's education
European Economic Review
(2001) - et al.
Job losses and child outcomes
Labour Economics
(2008) Linear regression after selection
Journal of Econometrics
(1981)- et al.
Parental employment circumstances and children's academic progress
Social Science Research
(2008) Does parents' money matter?
Journal of Public Economics
(2000)- et al.
What is happening in the youth labour market in Canada?
Canadian Public Policy
(2000) - et al.
The changing role of family income and ability in determining educational achievement
Journal of Human Capital
(2007) - et al.
Post-Secondary Attendance by Parental Income: A Canada—U.S. Comparison. manuscript, University of Western Ontario
(2009) - et al.
Family income and educational attainment: a review of approaches and evidence for Britain
Oxford Review of Economic Policy
(2004) The effect of income on child development
The Review of Economics and Statistics
(1999)
The dynamics of educational attainment for Black, Hispanic and White males
Journal of Political Economy
Estimation of educational borrowing constraints using returns to schooling
Journal of Political Economy
The evidence on credit constraints in post-secondary schooling
The Economic Journal
Job displacement, disability, and divorce
Journal of Labor Economics
Tuition fees and equality of university enrolment
Canadian Journal of Economics
Parental job loss, income shocks and the education enrolment of youth
The impact of family income on child achievement: evidence from the earned income tax credit
Organisational downsizing and work stress: testing synergistic health effects in employed men and women
Journal of Epidemiology and Community Health
Who is getting a college education? Family background and the growing gaps in enrollment
Measuring the load, easing the burden: Canada's student loan programs and the revitalization of Canadian postsecondary education
Access to college and university: does distance matter?
Canadian Public Policy
Cited by (121)
Gender differences in intergenerational effects of laid-off parents
2023, Economic SystemsStudent loans: Credit constraints and higher education in South Africa
2023, Journal of Development EconomicsShort-run shock, long-run consequences? The impact of grandparental death on educational outcomes
2022, Economics of Education ReviewFinancial relaxation of Islamic rural banks and private schools resilience during COVID-19
2024, International Journal of Evaluation and Research in Education
- ☆
Note: The statistical analysis presented in this document was produced from Statistics Canada micro-data. The interpretation and opinions expressed are my own and do not represent those of Statistics Canada. I am indebted to David Green, Nicole Fortin, Thomas Lemieux, Craig Riddell, and Lee Grenon. Many thanks also to participants at numerous university workshops and seminars. All errors are my own.