Online Classes and Gig Jobs Help Balance School and Work

Featured in print Digest

This figure is a scatter plot titled, Uber Drivers’ Hours while Enrolled Online at Arizona State University. The y-axis is labeled, Difference in Uber driving hours, relative to 3 weeks before starting course. It range from negative 7.5 to positive 2.5, increasing in increments of 2.5. The x-axis is labeled, weeks from course start, ranging from negative 5 to positive 11. There is a vertical dotted line at 0 labeled, start of 8-week course. There another vertical dotted line at 8 labeled, end of 8-week course.  The graph shows data points representing the difference in Uber driving hours before, during, and after an 8-week course. Prior to the course start, the points fluctuate between 0- and 1-hour increase in driving hours. At week 0, when the course begins, the value drops close to negative 2.5 hours, indicating a slight decrease in Uber driving hours. The values generally remain around this level until the end of the 8-week course. After the course completion, the data points hover around a negative 1 hour decrease in Uber driving hours, but they are not statistically significantly different from 0. The note on the figure reads, Bars represent 95% confidence intervals. The source line reads. Source: Researchersʼ calculations using data from Arizona State University and Uber.

A growing share of US college students work part- or full-time to support themselves while studying, a trend that seems likely to continue as tuition costs rise and workers in sectors affected by technological change return to school to retrain and upskill. According to the US Department of Education, 64 percent of full-time undergraduates and 86 percent of part-time students work at least 20 hours per week.

In Assessing the Costs of Balancing College and Work Activities: The Gig Economy Meets Online Education (NBER Working Paper 32357), Esteban Aucejo, A. Spencer Perry, and Basit Zafar find that the flexibilities afforded by online study and gig work — two relatively recent developments — can mitigate the frictions students have typically faced when combining work with a college education.

They study Uber drivers enrolled in tuition-free online classes at Arizona State University (ASU) through a partnership between the university and the rideshare company, focusing on the extent to which participants shift between study and work as returns to these activities change, for example, when class workload intensifies or the wages for driving decline. Tuition-free classes are an important incentive for Uber driver-students to participate in the partnership, as these students show greater sensitivity to education costs than their peers. In a survey where student Uber drivers were asked about their enrollment choices, enrollment probabilities would decrease by about 50 percent in a scenario where costs rose from $0 to $6,000, the latter reflecting the regular yearly fees for an ASU working student.

By assembling data on participants’ online class-related activities and their trips for Uber, the researchers can examine detailed time allocation (which otherwise is almost impossible to observe). For the average student-driver, studying costs about $180 a month in foregone earnings, 8 percent of the average monthly earnings of a non-student Uber driver. This suggests that enrolling in online classes has a low opportunity cost for participants.

The researchers use temporary shocks to potential earnings from Uber, such as extended periods of surge pricing resulting from a large local event, to analyze how a 10 percent increase in Uber hourly wages affects participants’ study time. They find a decline in time spent on classroom activities of only 1.7 percent, or about one minute for every $1 increase in Uber hourly wages. They also have information on the weekly workloads in various courses. A 10 percent increase in workload in a class a student is enrolled in reduces that student’s time spent driving by about 1 percent. The findings show that drivers adjust their labor market engagement when college activities become more demanding, but by only modest amounts. This elasticity is low, in part due to drivers adjusting their typical driving behavior when taking the courses. In addition, the researchers do not find any evidence that during busier academic times, drivers who are also students receive lower driver ratings or tips. Similarly, they do not detect any negative impact of driving hours on academic performance.

Finally, the researchers survey participants to elicit their perceptions of the program. Participants expect the degree program to substantially increase their long-run earnings and report that the flexibility of online learning makes them significantly more likely to attempt to attain a bachelor’s degree. Only 37 percent of participants said they would have pursued a degree if the Uber-ASU program were not an option, indicating their valuation of the program’s flexibility.

—Shakked Noy