The Returns to Academic and Vocational Qualifications in Britain.Pdf

(2002) - Lorraine Dearden, Steven McIntosh, Michal Myck, Anna Vignoles

Journal: Bulletin of Economic Research
Link::
DOI::
Links::
Tags:: #paper #NCDS #Attainment #Transition #school-to-work
Cite Key:: [@Dearden2002]

Abstract

This paper uses data from the 1991 sweep of the National Child Development Study (NCDS) and the 1998 Labour Force Survey (LFS) to provide a comprehensive analysis of the labour market returns to academic and vocational qualifications. The results show that the wage premia from academic qualifications are typically higher than from vocational qualifications. However, this gap is reduced somewhat, when we control for the amount of time taken to acquire different qualifications. This is particularly important for vocational courses, which generally take shorter time periods to complete. In the paper we also investigate how returns vary by gender, subsequent qualifications, and the natural ability of individuals. Finally, by comparing the NCDS results with those from the LFS, we estimate the bias that can result from not controlling for factors such as ability, family background and measurement error. The results reveal that the estimated returns in the NCDS equations controlling for ability, family background and measurement error are similar to the simple OLS estimates obtained with the LFS, which do not control for these factors. This suggests that the biases generally offset one another.

Notes

“The results show that the wage premia from academic qualifications are typically higher than from vocational qualifications. However, this gap is reduced somewhat, when we control for the amount of time taken to acquire different qualifications.” (Dearden et al., 2002, p. 249)

“Finally, by comparing the NCDS results with those from the LFS, we estimate the bias that can result from not controlling for factors such as ability, family background and measurement error. The results reveal that the estimated returns in the NCDS equations controlling for ability, family background and measurement error are similar to the simple OLS estimates obtained with the LFS, which do not control for these factors. This suggests that the biases generally offset one another.” (Dearden et al., 2002, p. 249)

“A final advantage of our analysis is that we can compare the results obtained using different data sets, with different strengths and weaknesses. For example, we use the NCDS cohort study, which contains a huge amount of information about the respondents, covering their entire lives. This means that we are able to control for many characteristics that are usually excluded from studies of the returns to education, including individuals’ early ability.” (Dearden et al., 2002, p. 252)

“The consensus from the work using standard OLS models is that the return to an additional year of schooling in the UK is about 6 –10 percent (Blundell et al., 1999), although estimates that allow for the endogenous nature of schooling using an instrumental variable estimator,2 such as Harmon and Walker (1995) can be twice as high” (Dearden et al., 2002, p. 252)

“The wage data (real gross hourly wages (1995 = 100)) are taken from the 1991 survey” (Dearden et al., 2002, p. 254)

“drop from our sample individuals who are employed but who have missing observations on wages, those who did not sit ability tests at the age of seven and those for whom we do not have information on qualifications. We also drop individuals who are self-employed or in fulltime education.” (Dearden et al., 2002, p. 255)

“A benefit of using the two data sets is that we can compare the results estimated with the NCDS, which control for ability and family background, to those estimated with the LFS, which do not.” (Dearden et al., 2002, p. 261)

“To the extent that high ability and good family background are likely to be positively correlated with both education level and earnings, their omission will bias the estimates upwards.” (Dearden et al., 2002, p. 261)

“The upward bias on the estimates caused by the omitted ability and family background variables seems to be particularly large in the case of vocational qualifications.” (Dearden et al., 2002, p. 261)

“Another potential source of bias to the estimates is measurement error bias, caused by the misreporting of qualifications. Misreported data are most frequently encountered when respondents have to recall events that have occurred in the past” (Dearden et al., 2002, p. 261)

“The extensive information available in the NCDS allows us to obtain instruments for the qualifications variables, and so investigate this issue.” (Dearden et al., 2002, p. 262)

“The results reveal that measurement error does appear to be a problem in the qualifications data in the NCDS” (Dearden et al., 2002, p. 262)

“Another question we examined is whether the returns to qualifications varied according to individuals’ ability.32 Such an analysis was only possible with the NCDS. We split our sample into two ability groups; a high ability group, consisting of all individuals in the top two quintiles of either the maths or reading ability distribution at age seven, and a lower ability group, containing all other individuals. We then interacted the ability measures with the qualification variables to see if there was evidence of heterogeneity by ability. The results suggest that the wage premia to academic qualifications do not differ significantly between low and high ability individuals. However, the premia from vocational qualifications are significantly higher for low ability individuals. This suggests that vocational qualification paths may have much larger returns for those in the bottom end of the ability distribution” (Dearden et al., 2002, p. 268)