Work-Study: Time Use Tradeoffs, Student Work Hours and Implications for Youth Employment Policy

Abstract

Does paid employment during high school and college displace the time students spend in educational activities? Most enrolled college students in the US now work in paid jobs, almost half of whom work 25 or more hours per week. An economic approach suggests that students consider the tradeoffs involved with work versus study time allocation in terms of both current income and future earnings capacity and well being. There may be some complementarity, not just substitutability, between work and education time, regarding educational outcomes. Previous research tends to find that when paid hours exceed some threshold level, typically somewhere between 15 and 25 hours per week, various indicators of students‘ academic performance are lower. Longer work hours also undermine certain aspects of mental health. This research applies the pooled 2003-2005 American Time Use Survey (ATUS) data (n=47k#) to empirically investigate four main questions: (1) Are paid work hours of students associated with time spent doing homework or research and/or attending class? (2) If so, at what threshold point of paid work hours are hours of student work displaced? (3) Are there differences between college and high school students in the above relationships? They are addressed with econometric analyses of the ATUS sample of college (n = 1,314, with 1,121 full-time) students and high school students. (4) Is student employment in certain industries or occupations associated with more time spent studying? Work hours are found to be inversely related to hours in educational activities among those aged 16-24. Moreover, there are nonlinearities by the number of actual hours. In contrast to previous studies and samples, students who work as little as 5 or more hours spend a statistically significant lesser amount of time studying than their cohorts who are not employed. The extent to which work displaces time spent studying is consistent across levels of weekly work hours, but becomes largest when hours are 40 or more, even when controlling for various demographic and occupational characteristics, but not time spent in class. The conclusion explores how to investigate whether students who work during the school year have a relatively lower well being, as indicated by available estimates of net affect associated with particular uses of time? It also explores implications for policies, such as extending youth employment regulatory protections to students if it is warranted by clear threats to their mental or physical well being.