Higher education and high-wage gender inequality.

Over the past 60 years, we have witnessed a relocation of gender wage inequality. Whereas the largest wage gaps were once concentrated among lower-paid, lower-educated workers, today these wage gaps sit among the highest-paid, highly-educated workers. Given this reordering of gender wage inequality and the centrality of college graduates to total inequality trends, in this article, we assess the contribution of higher education mechanisms to top-end gender inequality. Specifically, we use Census and ACS data along with unique decomposition models to assess the extent to which two mechanisms rooted in higher education-bachelor's-level fields of study and the attainment of advanced degrees-can account for the gender wage gap across the wage distribution. Results from these decomposition models show that while these explanatory mechanisms fare well among bottom and middle wages, their explanatory power breaks down among the highest-paid college workers. We conclude that women's attainment of "different" education (via fields of study) or "more" education (via advanced degrees) would do little to close the gender wage gaps that are contributing most to contemporary wage inequality trends. We suggest some directions for future research, and we also take seriously the role of discriminatory pay-setting at the top of the wage distribution.

Measuring the predictability of life outcomes with a scientific mass collaboration.

How predictable are life trajectories? We investigated this question with a scientific mass collaboration using the common task method; 160 teams built predictive models for six life outcomes using data from the Fragile Families and Child Wellbeing Study, a high-quality birth cohort study. Despite using a rich dataset and applying machine-learning methods optimized for prediction, the best predictions were not very accurate and were only slightly better than those from a simple benchmark model. Within each outcome, prediction error was strongly associated with the family being predicted and weakly associated with the technique used to generate the prediction. Overall, these results suggest practical limits to the predictability of life outcomes in some settings and illustrate the value of mass collaborations in the social sciences.

Predicting Layoff among Fragile Families.

The loss of a job is the loss of a major social and economic role and is associated with long-term negative economic and psychological consequences for workers and families. Modeling the causal effects of a social process like layoff with observational data depends crucially on the degree to which the model accounts for the characteristics that predict loss. We report analyses predicting layoff in the Fragile Families data as part of the Fragile Families Challenge. Our model, grounded in empirical social science research on layoff, did not perform substantially worse than the best-performing model using data science techniques. This result is not fully unforeseen, given that layoff functions as a relatively exogenous shock. Future work using the results of the Challenge should attend to whether small improvements in prediction models, like those we observe across models of layoff, nevertheless significantly increase the validity of subsequent models for causal inference.