Breaking the cycle? The effect of education on welfare receipt among children of welfare recipients

https://doi.org/10.1016/j.jpubeco.2007.02.001Get rights and content

Abstract

We examine the impact of high school graduation on the probability individuals from welfare backgrounds use welfare themselves. Our data consist of administrative educational records for grade 12 students in a Canadian province linked with their own and their parents' welfare records. We address potential endogeneity problems by: 1) controlling for ability using past test scores; 2) using an instrument for graduation based on school principal fixed effects; and 3) using a Heckman–Singer type unobserved heterogeneity estimator. Graduation would reduce the probability of welfare receipt of drop-outs by 1/2 to 3/4. Effects are larger for individuals from troubled family backgrounds and low income neighbourhoods.

Section snippets

Income Assistance and education in British Columbia

Income Assistance (IA), BC's form of welfare payments, is one part of a larger system of social security programmes.2 Unemployed individuals who meet employment related eligibility requirements can collect benefits from a federally administered unemployment insurance system for defined durations. In contrast, IA is

Previous literature

Our study is at least indirectly related to a number of literatures studying welfare and educational effects. Our motivation for studying this topic stems from the literature on the inter-generational transmission of welfare use. Several studies have established a strong positive correlation in welfare use across generations within a family for the US (Antel, 1992, An et al., 1993, Gottschalk, 1990, Gottschalk, 1996, Levine and Zimmerman, 1996). The issue of whether that correlation reflects a

Data

The data we use in our investigation is compiled from a linkage of three data sources. The first source is BC Ministry of Education records on grade 12 students for the years 1991 through 1996. These records contain information on all individuals enrolled in grade 12 in a BC high school at the start of November in each year. For each student we know whether they failed a previous year of school and whether they graduate from high school. We also observe the high school the individual attended.

The identification problem and solutions

The identification issues we seek to address can best be explained in the context of a model of the joint determination of high school graduation and future IA recipiency status. The approach we take is similar to that specified in Aakvik, Heckman and Vytlacil (2000) and builds on Heckman (1981) and, to some extent, Heckman and Singer (1984).

We begin by specifying the process determining high school graduation. We represent the choice of graduating from high school with an index function that

Ordinary least squares and two stage least squares estimates

In order to present the data patterns in the most transparent way possible, we begin with estimates from Ordinary Least Squares (OLS) estimation of the graduation process and OLS and two stage least squares (2SLS) estimates of the process for IA use at age 19.5. Given that we include the test score, past failure and family background variables, which we interpret as proxies for ωi, OLS estimation of the IA process yields consistent estimates of the graduation effect under the assumption that

Defining treatment effects

Our ultimate goal in estimating the model just described is to obtain estimates of the impact of graduating from high school on IA use for various groups. In the terminology of the training impact literature, we are interested in estimating treatment effects, where the treatment in this case is graduation and the outcome we focus on is IA receipt. As Aakvik et al. (2000) point out in deriving what is essentially this model in a different context, if we have consistent estimates of all of the

Conclusions

In this paper, we study the impact of high school graduation on the probability that grade 12 students from welfare backgrounds use welfare themselves at each age between 19 and 24. We make use of a unique dataset linking high school and welfare receipt records for a group of individuals entering grade 12 in the province of British Columbia in the period 1991 to 1996. We address potential endogeneity issues using a combination of proxy variables, instruments and an econometric framework

Acknowledgements

We are grateful for comments from participants at seminars at the 2003 Canadian Economics Association meetings, Western Research Network on Education and Training, the University of British Columbia, the University of Melbourne, and the Institute for Fiscal Studies, UCL.

This research paper is not an official document of the BC Government. The opinions expressed herein are those of the authors, and do not necessarily represent the official policies or positions of the BC Ministry of Social

References (22)

  • G. Barrett

    The effect of educational attainment on welfare dependence: evidence from Canada

    Journal of Public Economics

    (2000)
  • R.M. Blank

    Analyzing the length of welfare spells

    Journal of Public Economics

    (1989)
  • P. Gottschalk

    Is the correlation in welfare participation across generations spurious?

    Journal of Public Economics

    (1996)
  • A. Aakvik et al.

    Treatment effects for discrete outcomes when responses to treatment vary: an application to Norwegian vocational rehabilitation programs

    Journal of Econometrics

    (2000)
  • C. An et al.

    Teen out-of-wedlock births and welfare receipt: the role of childhood events and economic circumstances

    Review of Economics and Statistics

    (1993)
  • J. Antel

    The inter-generational transfer of welfare dependency: some statistical evidence

    Review of Economics and Statistics

    (1992)
  • G. Barrett et al.

    An untold story: the characteristics of welfare use in British Columbia

    Canadian Journal of Economics

    (1998)
  • N. Beaulieu et al.

    Intergenerational reliance on social assistance: evidence from Canada

    Journal of Population Economics

    (2005)
  • S. Boggess

    Family structure, economic status, and educational attainment

    Journal of Population Economics

    (1998)
  • D. Card

    The causal effect of education on earnings

  • P. Carneiro et al.

    2001 Lawrence R. Klein Lecture estimating distributions of treatment effects with an application to the returns to schooling and measurement of the effects of uncertainty on college choice

    International Economic Review

    (2003)

    • Cited by (16)

    • Pathways of Disadvantage: Unpacking the Intergenerational Correlation in Welfare

      2021, Economics of Education Review
      Citation Excerpt :

      less likely to complete high school.22 Given the established link between failing to complete high school and future welfare reliance (e.g., Coelli et al. 2007), it is not surprising that they have higher rates of welfare receipt as a result. Parental welfare receipt is also transmitted to young people through the curriculum choices and educational achievements of those who do complete high school.

    • Leadership effects: School principals and student outcomes

      2012, Economics of Education Review
      Citation Excerpt :

      The data we employ is BC Ministry of Education administrative data on all youth enrolled in grade 12 in public schools at the start of November in the period 1995–2004. Private schools are not a substantial factor in BC, with only 7.5% of grade 12 students attending private high schools in 2002 (Coelli, Green, & Warburton, 2007). Students in BC obtain a high school diploma if they complete 80 credits at the grade 10 or higher level, including a certain number of required credits, and with 16 credits being at the grade 12 level.

    • Understanding adolescent responses to differently worded suicide attempt questions: Results from a large US pediatric sample

      2022, Psychological Medicine

    • Antecedents of Islamic welfare: productivity, education, and the financial aspect

      2024, International Journal of Islamic and Middle Eastern Finance and Management

    • Employment success of social assistance recipients: A provincial analysis by industry

      2023, Canadian Review of Sociology

    • Predictors of becoming not in education, employment or training: A dynamic comparison of the direct and indirect determinants

      2022, Journal of the Royal Statistical Society. Series A: Statistics in Society
    View all citing articles on Scopus
    View full text
    Copyright © 2007 Elsevier B.V. All rights reserved.