Recent Research: College Majors and Underemployment Following the Great Recession

Throughout the Great Recession, numerous stories from sources like The Wall Street Journal pointed to the idea of the overeducated and underemployed recent college graduate as a symbol of the economic times. While some recent graduates were able to find work in their respective fields, many did indeed struggle. Several recent studies highlight the extent to which new graduates faced underemployment in the years following the Great Recession, how their college majors impacted this probability, and how information influences these college major decisions.

Jaison Abel and Richard Deitz, two researchers from the Federal Reserve Bank of New York, examine recent trends among graduates in Underemployment in the Early Careers of College Graduates Following the Great Recession, a December 2015 staff report. Overall, the authors find that from 2011 to 2014, the period following the Great Recession, the underemployment rate for recent college graduates grew to 45 percent, a level not seen since the early 1990s. While many of the underemployed actually worked in fairly well paying non-college jobs that still required elements of knowledge and skill, relatively few recent graduates were working in low-skilled service jobs, contrary to popular opinion. 

Abel and Deitz measure the underemployment rate as the share of employed college graduates working in jobs that do not require a college degree, using the Department of Labor’s O*NET database to distinguish between college and non-college jobs. From 1990-2015, the underemployment rate for recent college graduates is consistently higher than the underemployment rate for college graduates as a whole. To understand those jobs held by underemployed recent graduates in the years following the Great Recession, the researchers use American Community Survey (ACS) data to analyze a sample of roughly 346,000 observations representing about 44 million young workers, both college educated and non-college educated, and aged 22 to 27 years during the 2009 to 2013 period. Based on the hundreds of detailed occupation codes identified in the data, the researchers create 10 occupational categories across six tiers for underemployed college graduates, based on their knowledge and skill requirements, as well as the nature of the work performed.

Approximately 25 percent of recent college graduates working in jobs that did not require a college degree were in office and administrative support occupations such as secretaries, customer service representatives, or office clerks. A similar share – 24.5 percent – were working in either information processing and business support positions (e.g., human resource workers, computer support specialists, web developers, and paralegals), or as managers and supervisors. These positions represented the highest pay tier of the six underemployed occupational categories. The lowest pay tier, those for the low-skilled service workers that saturated recession rhetoric, represented 19.3 percent of underemployed recent college graduates. For young workers without a college degree, more than half (51.7 percent) were employed in occupations in the lowest two pay tiers, low-skilled service workers and physical laborers.

According to the report, there is a clear connection between college major and the probability of being underemployed. Those graduates with college majors that provide technical training and quantitative skills were less likely to be underemployed, as science, technology, engineering, and mathematics (STEM)-related majors tended to have among the lowest predicted probabilities of working in non-college jobs. While the probability of being underemployed for nursing majors was 9.5 percent, the lowest of all majors analyzed, six of the top 10 majors were an engineering discipline. The researchers also find that those college majors providing occupation-specific training, like accounting or special education, tended to be less likely to be underemployed than those providing a more generalized education.

Abel and Deitz conclude by noting that a key finding is that, to a certain extent, underemployment is a temporary phase for recent graduates as they transition from school to the labor market – a pattern that is especially evident for those beginning their careers in low-skilled service jobs. While the likelihood of being underemployed as a 22 year-old recent graduate is nearly 50 percent, this probability declines by 15 percent by age 27.

Given the apparent workforce demand for recent college graduates with majors that emphasize STEM, technical training, and quantitative analysis, more research is needed to determine why some majors fare so poorly upon graduation, and what drives students to these majors.  Determinants of College Major Choice: Identification Using an Information Experiment, and How Do College Students Respond to Public Information about Earnings?, two reports by economists Matthew Wiscall and Basit Zafar seek to understand those factors that influence major decisions.  

In that study, the authors use an experimental design to create panel data for major choices, which could otherwise be considered largely a one-time decision. Wiscall and Zafar conduct an experiment on undergraduate college students at New York University (NYU), where in successive rounds respondents are asked about their self-beliefs on their own expected future earnings and other major-specific aspects were they to major in different majors, their beliefs about the population distribution of these outcomes, and the subjective belief that they will graduate with each major. Afterwards, students are provided with accurate information on population characteristics of the major, and given the opportunity to update their self-beliefs and their subjective probabilities of graduating with each particular major.

The authors find that students in their sample have biased beliefs about population earnings, though there is considerable heterogeneity in errors, as students both over- and under-estimated average earnings. Furthermore, Wiscall and Zafar find evidence of substantial and logical updating of student beliefs about their own future earnings if given accurate information on the current population earnings.

In the context of major choice, Wiscall and Zafar find that earnings expectations and ability perceptions both play an important role in the choice of major, while marriage, spousal characteristics, and labor supply considerations play a relatively minor role. Heterogeneous tastes, however, are the dominant factor in the choice of major, leading the authors to conclude that analyses that ignore the correlation of tastes with earnings expectations inflate the role of earnings in college major choices.

In a companion article, Wiscall and Zafar seek to answer the question, How do College Students Respond to Public Information about Earnings? Using the same research methodology, this article finds evidence that college students are substantially misinformed about population earnings. Furthermore, they find that students are willing to logically revise their self-earnings beliefs when exposed to information, and have larger revisions when the information is more specific and could be considered “good” news. This disconnect prompts what the authors consider an immediate policy implication – a need for more accurate information on the returns of schooling.

Taken together, these research articles highlight the nexus of the employment outcomes for recent graduates and college decision choices.  Abel and Deitz find a clear connection between college major and the probability of being underemployed, noting that those graduates with college majors (like STEM) that provide technical training and quantitative skills were less likely to be underemployed. While Wiscall and Zafar suggest that heterogeneous tastes are the most important determinant in college major, they also note that earnings expectations play a role in determining major choices. Furthermore, in their companion research, they identify that college students are substantially misinformed about the future earnings associated with college majors. Ultimately, these findings could suggest that more needs to be done to inform college students about the longer-term impacts associated with choosing a particular major – for better, or for worse.   

Read Underemployment in the Early Careers of College Graduates Following the Great Recession: https://www.newyorkfed.org/medialibrary/media/research/staff_reports/sr749.pdf?la=en

Read Determinants of College Major Choice: Identification Using an Information Experiment: http://www.restud.com/wp-content/uploads/2014/12/MS16627manuscript.pdf

Read How Do College Students Respond to Public Information about Earnings?:  https://www.newyorkfed.org/medialibrary/media/research/staff_reports/sr516.pdf