College Premium Research Articles

Is College Worth It? Class and the Myth of the College Premium

Book review of Is College Worth It? Class and the Myth of the College Premium, by Richard Ohmann and Ira Shor (2024).

Why Didn’t the College Premium Rise Everywhere? Employment Protection and On-the-Job Investment in Skills

Why Didn’t the College Premium Rise Everywhere? Employment Protection and On-the-Job Investment in Skills

Heterogeneity, selection, and the policy effect of educational expansion on college graduate earnings in China, 1981–2015

This article examines the impact of social origin on the returns to college education in the context of China’s higher education expansion since 1999. Utilizing a double-treatment setting, a marginal treatment effect framework is adopted to estimate the causal effect of college education on earnings, while a difference-in-differences methodology identifies the policy effect of educational expansion on the college premium. Analysis of data from a series of nationally representative Chinese surveys reveals that the “true” college earnings premium is rather small for the post-expansion cohort, and that much of the observed earnings gap between college and high school graduates after the expansion is due to returns to unobserved abilities. Further analysis shows that the college earnings premium after the expansion declines more for rural-origin children with schooling probabilities in the top percentiles than for their urban counterparts.

Comment on “What Can Students Gain from China's Higher Education?”

Over the past four decades, China has built up one of the largest higher education systems in the world which is no small achievement. The Chinese higher education system is very diverse and complex, and varies a great deal in terms of ease of access for students, the quality of teaching and research, and the returns to education. Therefore, summarizing its progress or the lack of it is a daunting task. Li et al. (2023) offer fresh findings on the returns to higher education in China and their determinants. Their main conclusions are that the returns to college education in China have stabilized after a significant rise and the relatively high returns are largely explained away by signaling and social networks channels as opposed to the human capital mechanism. However, Li et al. (2023) can be improved by talking to the broader Chinese literature and incorporating the institutional details of the higher education sector in China into its analysis. First of all, as Li et al. (2023) rightly point out in the beginning of their paper that the enrollment in Chinese higher education has experienced a 10-fold expansion since 1999 and the resultant system is very diverse and highly stratified. The elite colleges, 211 project colleges, regular 4-year teaching colleges and 3-year vocational colleges operate in utterly different orbits with the elite colleges catching up quickly with the leading universities in the world on many fronts. The expansion concentrated disproportionally on the lower-tier universities, and therefore drove down the returns to an education in these universities as well as the overall returns to higher education to some extent. Another paper coauthored by the lead author of Li et al. (2023), Prof. Hongbin Li, clearly shows that the college premium for young workers declined while the college premium for senior workers increased over the period 1990–2019 (Li et al., 2022). This trend is also evident in work of Chinese scholars, for example, Ding et al. (2012, 2013). Although I understand the sampling is not necessarily representative at the different tiers and further analysis of the four groups of universities may not be possible, the overall conclusion of Li et al. (2023) needs to be qualified. Second, the Chinese labor market is also quite diverse. Dual labor market theory would not adequately capture its complexity. Public sector jobs couple strong stability and high benefits with relatively low nominal salaries. The private sector mostly relies on salaries but leading firms can attract even graduates of elite colleges. Therefore equation (1) in Li et al. (2023) at least needs to control for the sector the graduates landed in. The same goes for the city/region/industries of choice as these greatly shape the kind of work–life balance college graduates will enjoy. Third, while I appreciate the Li et al.'s (2023) efforts to categorize the determinants into the three mechanisms of human capital, social network, and signaling, it might be difficult to pinpoint their independent effects using existing data. For example, it might be premature to draw the conclusion that human capital accumulation in college does not generate the proper returns for students because human capital proxied by the grade point average (GPA) may affect students' long-term socio-economic status through the possibility of securing a spot for graduate school abroad or in China. Li et al.'s (2023) table 3 shows exactly that students who intend to go to graduate school have much higher GPAs. Even among students who enter the labor market immediately upon graduation, their GPAs are somewhat correlated with other variables in the other two dimensions. In particular, when local party organizations consider who to recruit as their member, GPA is definitely one of the criteria. On the social network front, is parental income social capital or economic capital? Other researchers have looked at this question using whether or not parents work in the public sector, whether or not they are party members or whether or not they hold public offices. Parental education is usually used as a measure of cultural capital. Moreover, is a student's score on a college entrance exam a signal or a measure of the student's cumulative human capital prior to college?

Non-College Occupations, Workplace Routinization, and the Gender Gap in College Enrollment

Women used to lag behind men in college enrollment but now exceed them. We argue that changes in non-college job prospects contributed to these trends. We first document that routine-biased technical change disproportionately displaced non-college occupations held by women. We next employ a shift-share instrument for the impact of routinization to show that declining non-college job prospects for women increased female enrollment. Results show that a one percentage point decline in the share of routine task intensive jobs leads to a 0.6 percentage point rise in female college enrollment, while the effect for male enrollment is directionally smaller and insignificant. We next embed this instrumental variation into a dynamic model that links education and occupation choices. The model finds that routinization decreased returns to non-college occupations for women, leading them to shift to cognitive work and increasing their college premium. In contrast, non-college occupations for men were less susceptible to routinization. Altogether, our model estimates that workplace routinization accounted for 63% of the growth in female enrollment and 23% of the change in male enrollment between 1980 to 2000.

A special issue on housing affordability: An introduction

A casual search of “housing affordability” (HA) on Google delivers more than 190,000,000 results. While it is an imprecize measure, it confirms that HA is a global concern. Recent studies on the topic have been discussed by Ben-Shahar et al. (2020), Deng et al. (2019), Gabriel and Painter (2020), Leung (2022), among others. Motivated by the vast demand for affordability research, we organized a virtual workshop titled “Post-Crisis Housing Market and Financial Stability: Recovery without Affordability?” in July 2020. We thank Yin-Wong Cheung, who co-organized the workshop, and the City University of Hong Kong for the support. When planning the workshop in 2019, we considered the world after the Global Financial Crisis in 2008. Little did we know that the COVID-19 pandemic would impact the world on such, and probably larger, scale. Therefore, we appreciate the effort by Malpezzi (2022), who surveys the housing affordability issue post-COVID-crisis. While the impact of COVID may still be evolving (Wang et al., 2022) and the global economic situation could deteriorate with the Russia-Ukraine War (United Nations, 2022), we believe that Malpezzi (2022) will serve as a useful reference for future research. Previous research on housing affordability focuses on the demand side. Recent efforts pay attention to the supply side, including the real estate developer behaviors (e.g., Fan et al., 2022; Leung et al., 2020). In this special issue, Sun and Yiu (2022) examine the default probabilities of real estate developers in five ASEAN countries (Indonesia, Malaysia, Philippines, Thailand, and Vietnam) with firm-level data. If many real estate developers default at the same time, the new supply of housing will be affected, and house prices will also be affected. The firm-level approach is also in line with the current research trend. A distinct feature of this special issue is our balance of theoretical analysis, empirical evidence, and policy implications. The three other papers make use of structural models to produce sharp interpretations. For instance, Leung and Tang (2022) construct a dynamic equilibrium model where the house price-to-income ratio (PIR) evolves endogenously. PIR is a commonly used metric, but its equilibrium dynamics were not explored. Leung and Tang (2022) fill this gap. Based on their model, they derive a statistical test for detecting persistent deviations of PIR from the steady-state values. They also apply their model to OECD data. Yao (2022) documents that the drop in American youth homeownership is mainly driven by non-college-educated workers. She then develops an overlapping generations model to rationalize such a phenomenon. She finds that the root reasons are the rising college share and the widening of college premium. Based on her model, she recommends that eliminating mortgage interest tax deduction can increase the homeownership rate of non-college-educated workers. Last but not least, Yılmaz and Yeşilırmak (2022) build a spatial equilibrium model and compare the housing vouchers with the transportation vouchers. Un-like the traditional model with a unique employment center, the central business district (CBD), their model features multiple employment centers. They find that the housing vouchers affect the spatial ordering of households through the income effect, while transportation vouchers affect the spatial ordering through the marginal cost of commuting. The “losers” lose less in terms of welfare under transportation vouchers. All these analyses and results presented in this special issue are far from the “final words” on the housing affordability issue. Instead, we hope they will stimulate further investigations on the topic, and we look forward to learning more from future research.

Higher education expansion and the changing college wage premium in Hong Kong, 1976–2016

This article examines the trends and patterns of returns to college education in Hong Kong in the context of educational expansion. Using the data from nine waves of population censuses/by-censuses from 1976 to 2016, we employ age–period–cohort models to investigate the trends in college wage premium in Hong Kong. The descriptive statistics show that, for those aged 22–26, college wage premium declined over time. In addition, there are substantial variations in college wage premiums among different birth cohorts, which cannot be explained by age and period main effects. More specifically, college wage premiums declined almost monotonically for those born after 1950, and two younger birth cohorts (1975–1979 and 1980–1984) experienced cumulative disadvantage in terms of the college premium over the life course. We discuss social and political ramifications of these findings.

Higher education expansion and supply of teachers in China

We examine the teacher labour market in China using the 2005 mini-Census, in the context of the transformation of the world's largest education system. We first document a significant increase not only in quantity, but also in quality of teachers during 1990–2005. Instrumental Variables results based on the natural experiment of a substantial expansion of higher education in 1992/93 indicate a large positive causal effect of the expansion on supply of teachers. Consistent with differential opportunity costs across graduate occupations, the supply effect is more pronounced for women and those living in less developed regions. Further analyses of differential college premiums in earnings and non-pecuniary benefits between teaching and non-teaching occupations suggest that teacher recruitment has become more market-oriented and flexible, in attracting low to lower-middle ability college graduates into teaching in an increasingly decentralized and competitive graduate labour market.

College and Lifecycle Earnings Inequality

Has a rising college earnings premium contributed to increased earnings inequality over the past 50 years? I answer this question with a model of college linked to a Ben-Porath human capital problem. The model is estimated from data on college attainment rates and lifecycle income profiles, which I document by education level. I find that lifetime earnings inequality within cohorts has risen significantly since the 1970s but little of this is directly attributable to the college premium. Higher tuition is also not particularly important. Increasing earnings risk seems to be the main driver, with higher college return dispersion a contributing factor.

Effects of cohort size on college premium: Evidence from China's higher education expansion

In this paper, we document the lesser-known heterogeneous trends of college/non-college earnings premium across age groups from 1995 to 2013 in China. Specifically, the college premium in 2013 for the younger group (age 25–34) was about 30 percentage points, similar to the level in 1995, while the college premium in 2013 for the older group (age 45–54) increased to 50 percentage points, nearly double that of 1995. To attribute these divergent trends of the college premium to the changes in the relative size of college workers, we use the model by Card and Lemieux (2001), which incorporates imperfect substitution between similarly educated workers in different age cohorts. Due to the distinctions of these trends in China, our identification is free of the overestimation issue that the existing studies suffer. Our results are similar to those in the U.S., U.K., Canada, and Japan. Holding the age cohort and survey year constant, a one unit increase in log relative size of college workers is associated with about 10 percentage points decrease in college/non-college premium and about 18 percentage points decrease in college/high school premium. We further find that the negative effect is much more substantial among the new entrants (age 25–29) than experienced workers (age 30–54). By this pattern, we demonstrate that the new labor market entrants are more sensitive to their own cohort size and argue that the confounding ability composition effect should not be a serious issue.

Wages, Skills, and Skill-Biased Technical Change

The canonical supply-demand model of the wage returns to skill has been extremely influential; however, it has faced several important challenges. Several studies show that the standard approach sometimes produces theoretically wrong-signed elasticities of substitution, yields counterintuitive paths for skill-biased technical change (SBTC), and does not account for the observed deviations in college premia for younger vs. older workers. This paper shows that these failings can be explained by mis-measurement of relative skill prices and supplies (based on standard demographic composition-adjustments) and by inadequate ad hoc functional form assumptions about the path for SBTC. Improved estimates of skill prices and supplies that account for variation in skill across cohorts within narrowly defined groups help explain the observed deviation in the college premium for younger vs. older workers, even with perfect substitutability across age. Re-estimating the model with these prices and supplies produces a good fit with better out-of-sample prediction and robustly yields positive elasticities of substitution between high and low skill workers. The estimates suggest greater substitutability across skill and a more modest role for SBTC. We implement two new approaches to modelling SBTC. First, we study the extent to which recessions induce jumps or trend-adjustments in skill bias and find evidence that both features are important (but differ across recessions). Second, we link SBTC to direct measures of information technology investment expenditures and show that these measures explain the evolution of skill bias quite well. Together, these approaches suggest that the ad hoc assumptions for SBTC previously employed in the literature are too crude to fit the data well, leading to the incorrect conclusion that SBTC slowed during the early-1990s and under-estimates of the elasticity of substitution between high and low skill workers.

Risk, the College Premium, and Aggregate Human Capital Investment

Despite increases in the college earnings premium to persistently high levels, investment in college education remains low. We can understand this apparent puzzle by considering the risk of attending college and, in particular, the possibility of failing to graduate. Students with a reasonable probability of completing college already enroll, and for those who do not enroll, the low chance of completion blunts the impact of the rising college premium. In the absence of improved college readiness, our quantitative results suggest that continuing long-standing trends in skill-biased technological change can be expected primarily to increase earnings inequality rather than college attainment. (JEL E24, I22, I23, J24, J31, O33)

Earnings returns to tertiary education in urban China, 1988–2008*

Employing eight data sets from national surveys conducted between 1988 and 2008, I examine earnings returns over time to tertiary education relative to upper secondary education in urban China. A trend of an increasing college premium is found. However, this trend reversed after the mid-2000s, when the cohort of college graduates who were affected by the dramatic expansion of higher education after 1999 entered the labor market. The increasing share of employment in the market sector contributes to the increasing college premium, while the relative oversupply of graduates from tertiary education lower earnings returns to university or above education.

Wages, Skills, and Skill-Biased Technical Change: The Canonical Model Revisited

The canonical supply-demand model of the wage returns to skill has been extremely influential; however, it has faced several important challenges. Several studies show that the standard approach sometimes produces theoretically wrong-signed elasticities of substitution, yields counterintuitive paths for skill-biased technical change (SBTC), and does not account for the observed deviations in college premia for younger vs. older workers. This paper shows that these failings can be explained by mis-measurement of relative skill prices and supplies (based on standard demographic composition-adjustments) and by inadequate ad hoc functional form assumptions about the path for SBTC. Improved estimates of skill prices and supplies that account for variation in skill across cohorts within narrowly defined groups help explain the observed deviation in the college premium for younger vs. older workers, even with perfect substitutability across age. Re-estimating the model with these prices and supplies produces a good fit with better out-of-sample prediction and robustly yields positive elasticities of substitution between high and low skill workers. The estimates suggest greater substitutability across skill and a more modest role for SBTC. We implement two new approaches to modelling SBTC. First, we study the extent to which recessions induce jumps or trend-adjustments in skill bias and find evidence that both features are important (but differ across recessions). Second, we link SBTC to direct measures of information technology investment expenditures and show that these measures explain the evolution of skill bias quite well. Together, these approaches suggest that the ad hoc assumptions for SBTC previously employed in the literature are too crude to fit the data well, leading to the incorrect conclusion that SBTC slowed during the early-1990s and under-estimates of the elasticity of substitution between high and low skill workers.

The strategic overuse of student loans

This theoretical model shows that students strategically overuse student loans. In imperfect labour and credit markets, firms compensate graduates for a share of their loans via higher wages. This shifts some costs to firms and low-skilled workers who suffer from higher unemployment and lower wages. While any student loan increases unemployment, the effects on welfare and inequality depend on the purpose of the loan: A loan for tuition fees reduces human capital and credit accessibility, the college premium increases significantly, and welfare declines. A loan spent on consumption increases education and credit accessibility, while wage inequality and welfare change marginally.

Reconciling Changes in Wage Inequality with Changes in College Selectivity Using a Behavioral Model

We estimate a structural dynamic Roy model of education, labor supply and earnings on the 1979 and 1997 cohorts of males taken from the National Longitudinal Survey of Youth (NLSY) and evaluate to what extent changes in education and labor supply decisions across cohorts have been explained by changes in i) the college premium, ii) the utility of attending higher education, iii) grade progression standards, and iv) the value of non-market time. We quantify the evolution of the relative and absolute qualities of both college graduates and college attendants (associates). We find that it is impossible to rationalize changes in observed schooling decisions without appealing to a large increase in intrinsic taste for education, despite a doubling of the cost of college and its impact on debt-load. The population distribution of the college premium has shifted to the right, going from 50% to 58%, while the premium of actual college graduates has shifted to the left, going from 72% to 54%, thereby pointing toward a reduction of the relative quality of college graduates. The absolute quality (human capital) of college graduates has however remained stable. For college attendants (associates), both relative and absolute quality dropped. One implication of the relative flattening of age earnings profiles is the removal of the negative effect of late college graduation on early life-cycle wages. Our estimates indicate it moved from a 4% penalty per year of delay to an insignificant quantity by the early 2000's.

Discussion

Discussion

Why Didn&Apos;T the College Premium Rise Everywhere? Employment Protection and On-the-Job Investment in Skills

Why Didn&Apos;T the College Premium Rise Everywhere? Employment Protection and On-the-Job Investment in Skills

Why Didn't the College Premium Rise Everywhere? Employment Protection and On-the-Job Investment in Skills

Why has the college wage premium risen rapidly in the United States since the 1980s, but not in European economies such as Germany? We argue that differences in employment protection can account for much of the gap. We develop a model in which firms and workers make relationship-specific investments in skill accumulation. The incentive to invest is stronger when employment protection creates an expectation of long-lasting matches. We argue that changes in the economic environment have reduced relationship-specific investment for less-educated workers in the United States, but not for better-protected workers in Germany.

Genes, Education, and Labor Market Outcomes: Evidence from the Health and Retirement Study.

Recent advances have led to the discovery of specific genetic variants that predict educational attainment. We study how these variants, summarized as a linear index-known as a polygenic score-are associated with human capital accumulation and labor market outcomes in the Health and Retirement Study (HRS). We present two main sets of results. First, we find evidence that the genetic factors measured by this score interact strongly with childhood socioeconomic status in determining educational outcomes. In particular, although the polygenic score predicts higher rates of college graduation on average, this relationship is substantially stronger for individuals who grew up in households with higher socioeconomic status relative to those who grew up in poorer households. Second, the polygenic score predicts labor earnings even after adjusting for completed education, with larger returns in more recent decades. These patterns suggest that the genetic traits that promote education might allow workers to better accommodate ongoing skill biased technological change. Consistent with this interpretation, we find a positive association between the polygenic score and nonroutine analytic tasks that have benefited from the introduction of new technologies. Nonetheless, the college premium remains a dominant determinant of earnings differences at all levels of the polygenic score. Given the role of childhood SES in predicting college attainment, this raises concerns about wasted potential arising from limited household resources.