Academic journal article Research in Applied Economics

The Screening Hypothesis and the Returns to Schooling in Argentina

Academic journal article Research in Applied Economics

The Screening Hypothesis and the Returns to Schooling in Argentina

Article excerpt


We conduct four tests of screening in education. We apply our methodology to labor market data from Argentina, using two points in time, a period during which the returns to schooling in Argentina increased. There do not appear to be significant increases associated with years of schooling that would represent the attainment of a primary or secondary certificate. The only signal that there might be screening occurs at 17 years of schooling which could be argued represents the attainment of a tertiary education degree. However, 15 years of schooling also represents a significant threshold in 2002. The returns to schooling are higher in the private sector. Much of the increase in the returns to schooling overall is due to the increase in the returns to tertiary education. The returns to complete university are higher in the private sector. This provides no evidence of screening since the private sector seeks to maximize profits and recognizes the higher productivity of the more educated. Overall, there is little evidence of screening driving the returns to schooling.

Keywords: Returns to schooling, Argentina, screening

1. Introduction

According to screening theories, those with more schooling tend to earn more not solely because schooling makes them more productive, but because schooling acts as a credential. This view was early on dismissed due to a lack of evidence (Layard and Psacharopoulos 1974). More recent analyses, exploiting data that allow researchers to disaggregate earnings by years of completed schooling, has questioned the linearity of the earnings function approach, suggesting that there are significant discontinuities associated with diploma years, thus suggesting evidence of sheepskin effects in the returns to schooling and hence screening (Hungerford and Solon 1987; Jaeger and Page 1996; Gibson 2000; Bedard 2001; Bauer, Dross and Haisken-DeNew 2005; Brown and Sessions 2006; Pons 2006; Skalli 2007; Bitzan 2009). This returning theme in the literature, that is, the extent to which estimated returns to schooling reflects productivity differences, is something we turn to empirically using recent data from a country where it is documented that there is a strong relationship between demand for education and earnings over the last decade: Argentina. Central to human capital theory is that education directly augments individual productivity and, therefore, earnings (Becker 1964; Schultz 1961). The alternative to this explanation - that education serves to "sort" individual into higher productivity jobs, by "signaling" or "screening" more productive individuals (Spence 1973; Arrow 1973; Stiglitz 1975) - is what we test.

In this paper four tests are applied to ascertain the extent to which screening drives the observed returns to schooling. One test for screening looks at the returns for people with a complete education versus those who dropped out (Layard and Psacharopoulos 1974); hereafter referred to as the LP-test. If education is acting as a signal, then certification from a course should convey more information to prospective employees about an applicant's ability. Returns for completers, therefore would be higher than returns for non-completers.

We also estimate a variant of the LP-test in order to control for the fact that those with incomplete levels of schooling have attained various years of schooling. Therefore, we also control for both levels and years. That is, we estimate a spline function where we control for both years of schooling within each level of schooling (that is, primary, secondary and tertiary), and a dummy variable indicating completion of each level of schooling.

A third test, and variant of the above, accounts for the timing of the dropout decision by examining year-to-year returns and looking for discontinuities in the estimates at years corresponding to the termination of levels. The so-called "sheepskin effects" of education ask whether it is years of schooling or highest qualifications that are more important (Jaeger and Page 1996; Arabsheibani and Manfor 2001). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.