Do you have any questions about our projects or our site in general? Do you have any comments or ideas you would like to share? Please feel free to call us, email us or simply send us a message from here.
Copyright © 2024 The HEAD Foundations. All Rights Reserved.
Leaders and changemakers of today face unique and complex challenges. The HEAD Foundation Digest features insights and opinions from those in the know addressing a wide range of pertinent issues that factor in a society’s development.
Informed opinions can inspire healthy discussions and open up our imagination to new possibilities. Interested in contributing? Write to us at info@headfoundation
Leaders and changemakers of today face unique and complex challenges. The HEAD Foundation Digest features insights and opinions from those in the know addressing a wide range of pertinent issues that factor in a society’s development.
Informed opinions can inspire healthy discussions and open up our imagination to new possibilities. Interested in contributing? Write to us at info@headfoundation
Science, Technology
& Innovation
Arts, Culture & Society
Teaching & Learning
Inclusivity & Diversity
Crisis Management
Leadership
& Management
Entrepreneurship
& Employability
Globalisation
& Regionalisation
Policy, Governance
& Reform
Future & Sustainability
Returns to schooling: Driven by family background or education-related policies?
‘Does schooling pay off?’ – a seemingly straightforward question, but it is in fact a puzzle among economists.1 Answers would differ based on how the returns to schooling are estimated. Among the top concerns is whether such estimations have any causal connotation between the amount of schooling and its returns (i.e. earnings). The positive relationship between schooling and earnings has been established in 1962 by economics Nobel Laureate, Gary Becker, in his human capital theory. Its empirical representation comes later, in the form of Mincerian wage equations, with years of schooling as the key explanatory variable of interest in determining earnings.2 The concern is the endogeneity issue of schooling, i.e. how would one know if schooling does indeed have any causal impact on earnings, and how much would that returns to schooling be. The endogeneity issue arises due to ability bias, where ability is typically related with years of schooling. The impact of schooling would be confounded by ability, hence the difficulty in isolating schooling’s causal impact on earnings.
To navigate the empirical tight spot, economists have resorted to what is known as the instrumental variable (IV) estimation. Simply put, it extracts the causal parts of schooling by first estimating how much schooling depends on family background (parents’ education is an overwhelmingly popular choice as the instrumental variable), and then insert only the causal parts (or ‘exogenous’ in economics nomenclature) of schooling into the wage equation to estimate the returns to schooling. Then, in the early 1990s, along came a groundbreaking work which has since revolutionized how returns to schooling are causally estimated. Angrist and Krueger’s 1991 seminal paper is now regarded as the catalyst work that has initiated a whole new strand of literature on using education-related policies such as compulsory schooling laws or education access reforms (instead of family background), in estimating causal returns to schooling.3
Following this development, we can now ask: which one actually drives the causal returns to schooling – family background, or education-related policies? And, how much does schooling pay off? To address these questions, we conduct a meta-analysis of 74 empirical studies from which we retrieve returns-to-schooling (RtS) coefficients estimated using both the causal instrumental variable and non-causal naïve estimation approaches. All the empirical studies included in our meta-analysis employed Mincerian wage equations with earnings as the dependent variable, years of schooling as the key explanatory variable, and either a family background or an education-related policy IV – these are the main criteria for empirical studies to be included in our meta-analysis. The geographical focus of those empirical studies ranges from the US, UK, Asia (mainly China), and several European and America Latin countries. We use Angrist and Krueger’s 1991 seminal work as the threshold publication year, i.e. empirical studies published from that year onwards (until 2022) and that meet the main inclusion criteria are used in our meta-analysis.
Key findings from our meta-analysis suggest an overall impact of 0.108, meaning an additional year of schooling is associated with a 10.8% increase in earnings, on average. We also find that over the years, returns to schooling exhibit an upward trend in general. Probing deeper, our analyses provide statistical evidence that education-related policy factors are driving the results more than family background factors. That is, education-related policies such as compulsory schooling laws, have larger impact on earnings – for instance, individuals who are compelled by schooling laws to complete certain years of schooling would see a larger increase in earnings, as compared to say, individuals with highly educated parents (who would be likelier to encourage their children to acquire more years of schooling).
So, what do our key findings imply? Since there is evidence that education-related policy factors are driving earnings or the returns to schooling, governments and authorities in education matters need to be even more attentive and responsible when designing policies pertaining to compulsory schooling laws, education access, equality, and equity reforms, number and quality of schools, for instance. Education-related policies should not be made in a knee-jerk or ad-hoc manner, because years of schooling are dependent on such policies. Years of schooling in turn, would affect earnings. As fiscal positions of governments all over the world are becoming ever tighter and with national coffers threatened by global economic uncertainties, education policy designs should shift from the notion of universality to those targeted at specific groups such as the marginalised and underprivileged.
Not only that, the impact of education-related policies could also be dissected by specific types (e.g. science versus non-science) and levels of education (e.g. primary, secondary, tertiary), not just by years of schooling. Another example of targeted policies is to build more schools in rural areas to improve education access for the typically marginalised rural communities; this way, it tackles both the resource allocation issue and education equity issue simultaneously. The advocacy of nuanced and targeted education-related policies would pave way for not only better deployment of national resources, but also better realisations of those policies into earnings eventually. In essence, policy advice and recommendations stemming from research based on causal wage-schooling relationships would be more credible and useful than those based on mere associations.
Assoc Prof Dr Soon Jan Jan is an associate professor from the School of Economics, Finance and Banking (SEFB), Universiti Utara Malaysia (UUM). She earned her PhD in economics from the University of Otago, New Zealand. Her areas of expertise include applied microeconometrics, education economics, and labour economics. She publishes consistently in SSCI/Scopus-indexed journals.
Acknowledgements: This work was supported by the Malaysian Ministry of Higher Education Fundamental Research Grant Scheme [FRGS/1/2016/SS08/UUM/02/9], and UUM University Grant [SO Code: 21425].
The HEAD Foundation Commentary is a platform to provide timely and, where appropriate, policy-relevant commentary of topical issues and contemporary developments. The views expressed by the authors are solely their own and do not reflect opinions of The HEAD Foundation.
GENESIS KAI: Charting the Artificial Horizon in Art
Artificial Creativity: The Rise of Generative AI and Its Toll on Artistic Integrity
What’s to Gain from AI?