Exploring the Scope and Limits of Higher Education In Reducing Social Disparities In Vulnerability to Precarious Work: Evidence from the U.S.

Background and purpose: Precarious work (PW) refers to uncertain and insecure employment arrangements, including the absence of minimum-wage coverage, workplace health insurance, standard work schedules (35-48 hours), and collective worker representation. PW has received scholarly attention due to its adverse social effects and its growing prevalence worldwide.

Previous research identifies social factors shaping vulnerability to PW: gender, race, age, disability, and education. Educational attainment is highlighted for its relevance to reducing such vulnerability. Nevertheless, few studies have analyzed whether higher education reduces social disparities in vulnerability to PW. More specifically, few studies examine whether higher education statistically moderates the effects of other socio-demographic variables on PW vulnerability. Existing work has examined the moderating role of higher education in gender gaps in vulnerability to part-time employment, but not across other key inequality gaps or PW dimensions. Therefore, a knowledge gap remains regarding the moderating effect of higher education on inequality gaps in vulnerability to multiple aspects of PW. Addressing this gap is critical to evaluate the potential of education to reduce labor market inequalities.

This study, therefore, examines the moderating effect of higher education on the effect of other social determinants of PW in the United States. The research question of this study is:

What is the moderating effect of higher education on the relationship between other socio-demographic factors and PW in the U.S.?

Methods: Using data from the 2023 Current Population Survey (n = 55,529 paid workers), I estimate binary logistic regression models to predict three PW features: lack of minimum-wage coverage, lack of workplace health insurance, and lack of union affiliation (all coded 1 = yes, 0 = no). Key predictors include being a woman, being a racial minority, experiencing disability, having higher education (all the above coded 1 = yes, 0 = no), and age groups (23-30, 31-64, 65+). I included interaction terms involving higher education and key predictors. Moderation is assessed using predictive margins for experiencing PW features across groups defined by the interaction terms. Statistical significance of the moderation was evaluated through pairwise comparisons of average marginal effects.

Results: Higher education substantially reduces the greater vulnerability young workers have relative to adults in lacking workplace health insurance. Among lower-educated workers, young workers (aged 23-30) are more likely than adult workers (aged 31-64) to lack coverage (52% vs. 43%, gap = 9 percentage points). Among higher-educated workers, this gap is nearly eliminated (39% vs. 38%, gap = 1 percentage point). The reduction (8 percentage points) is statistically significant and amounts to about 88% of the original gap. However, higher education does not substantially alter disparities in minimum-wage coverage or union affiliation.

Conclusions and implications: These findings contribute to debates on the role of education in promoting equitable access to decent work. By identifying which disparities are most responsive to higher education, the study offers a nuanced assessment of its potential and limits in reducing exposure to PW, and provides a basis for targeting interventions toward the most education-sensitive PW dimensions.