Earnings penalty of educational mismatch: a comparison of alternative methods of assessing over-education

Abstract

In this paper we systematically evaluate the impact of using the alternative methods conventionally used in the international literature on the measured incidence of educational mismatch and its earnings effects. We use a rich Australian longitudinal data set for a controlled group of full-time employed workers. Using panel data estimation, we address individual heterogeneity and measurement error, which are important in educational mismatch analysis. We show that alternative methods of measurement result in a range of estimates, with the Mode measure providing the most stable results across instrumental variables (IV) selections in panel fixed effects instrumental variables (FEIV) estimations. Based on the Mode measure, the incidence rate of over-education is 32.3%. The earnings penalty for each year of over-education is 2.5%, which is larger than 0.6% in fixed effect estimation and also larger than 1.9% in OLS estimations.

KEYWORDS:

• Over-education
• education mismatches
• earnings
• measurement error
• instrumental variable

JEL CLASSIFICATIONS:

• J24
• J31

Acknowledgements

This paper uses unit record data from the Household, Income and Labour Dynamics in Australia (HILDA) Survey. The HILDA Project was initiated and is funded by the Australian Government Department of Families, Housing, Community Services and Indigenous Affairs (FaHCSIA) and is managed by the Melbourne Institute of Applied Economic and Social Research (Melbourne Institute). The findings and views reported in this paper, however, are those of the authors and should not be attributed to either FaHCSIA or the Melbourne Institute.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 For example, government tertiary education expenditure in Australia was AU$33.03 billion, accounting for 2% of GDP in 2015–2016 (ABS, 2016).

2 Occupational growth indicates whether a worker has moved into a higher-ranked occupation over time.

3 The earnings penalty in this literature refers to the difference between the earnings returns to each year of over-education (and under-education), in comparison to the returns for each year of required education. Therefore, the penalty estimate also provides a measure of the difference between the earnings returns to each year of over-education (and under-education), compared to workers with similar years of education, but who are engaged in occupations that match their education level.

4 Signalling effects for selection indicate that workers with higher levels of education are likely to have higher productivity than workers with lower levels of education, based on human capital theory. Therefore, based on signalling hypotheses, workers with higher levels of education are more likely to be hired by employers.

5 We further examined the potential impact of selection into full-time employment, and also employment, using Heckman (1979) selection adjustments. We found the coefficients in the earnings models stayed unchanged across the models with and without the selection adjustments. These results alleviated concerns regarding sample selection effects in our preferred fixed effects models. These results are available on request. Since the focus of the analysis is on a comparison of measurement methods, and not policy or time-sensitive periods, we decided to keep the analysis using the 2001–2009 data waves as a sufficiently long period for the current study.

6 In this study, the HILDA Survey does not include questions such as ‘What level of education do you need to get your current job?’ or ‘What level of education is required to perform your job?’. Thus, we do not have a measurement on ‘over-education’ based on the Self-Report measure. Thus, worker’s self-report is not applicable.

7 Both the Mode and the attained years of education measures are in full years, corresponding to the official degree completion year requirements. For example, if a respondent takes longer to complete a degree, such as 3.5 years to complete a 3-year degree, the recorded completed years of education is 3 years.

8 This is one of the two prominent approaches. The other approach is the standard ORU (Over-Education, Required Education, and Under-Education) earnings model proposed by Duncan and Hoffman (1981). Rather than controlling for attained years of education, the required years of education are controlled for in the model. In this case, an over-educated worker earns more and an under-educated worker earns less than matched co-workers, holding other characteristics constant. Thus, when this alternative specification is made, the coefficient of over-education is positive, but the coefficient of under-education is negative (Duncan & Hoffman, 1981).

9 In detail, over-skilling is derived from the HILDA Survey by using the responses, scored on a seven-point scale, to the question ‘I use many of my skills and abilities in my current job’, with a response of 1 corresponding with strongly disagree and 7 with strongly agree. Individuals with responses of 1, 2, 3 or 4 on the scale were classified as over-skilled and those with responses of 5, 6 or 7 as skill-matched. Sensitivity tests confirm the cut-off points for over-skilled and skill-matched are appropriate.