In trying to understand what drives the Occupy Wall Street protesters, certain commentators have resorted to stereotyping: casting them as unemployed youth with liberal arts degrees, disappointed that their degree in puppetry or medieval French didn’t pay off. According to this rhetoric, there’s plenty of demand for engineers and scientists, students are just too stupid or lazy to specialize in these fields. Setting aside the condescending tone, how true is this claim?
Analyzing data on unemployment by major, as featured in the Wall Street Journal, we find that it’s a half-truth, up for interpretation. In terms of unemployment rates alone, technical fields seem to have no advantage over more generalist education. Rather, while some science or engineering majors are in high demand and others face up to 15-20% unemployment, the liberal arts majors seem to occupy a happy medium around 6-7% unemployment. But if we dig a little deeper and look at expected returns in terms of wage, then the advantages of a mathematically oriented major become evident. Readers of this blog might appreciate that majoring in Military Technologies provides one of the highest returns despite a 10.9% unemployment rate. Sometimes it makes sense to take a risk and hold out for your dream job, especially when you’re specialized.
For those of you curious to see the numbers, I included the full table of data here, highlighting specific majors I refer to in the text.
Based on data from the Georgetown University (Hoya Saxa!) Center on Education and the Workforce the employment rate for college grads varies from 0.0% for School Student Counseling to 19.5% for Clinical Psychology. This should already ring some alarm bells, as within the same field the least technically trained are far more likely to find a job than the most technically trained. And the list goes on. The much maligned English majors have an unemployment rate of about 6.7%, which puts them above environmental engineering (2.2%) but below industrial engineering (9.25%). Engineers clearly face a far greater variance in their chances of employment than do those majoring in the humanities.
Why? Because fields in science, math and engineering require so much more specialization. This data set has 26 categories for engineering and 33 for science. Hence students will emerge with a deeper but much narrower skill-set. The demand for their skills will depend on the vagaries of the economy, with some fields in much higher demand relative to others. For example, the discrepancy between environmental and industrial engineering makes sense when we consider that the US is downsizing its manufacturing sector while promoting environmental sustainability. For an English major on the other hand, the ability to analyze text and write well applies to many white-collar jobs, providing a far greater pool of potential employers.
At the same time, this simple statistic can mislead. It doesn’t differentiate between job types, so counts the professor and burger flipper as equally employed. Perhaps graduates with a more specific skill-set will hold out longer for their dream job, while generalists take what they can get. Though they face more uncertainty, the returns to specialization in terms of wages might make it worthwhile.
Expected Returns to Education
To address this concern, I took the existing data and computed an expected return to education by major. It multiplies the likelihood of employment by the mean wage if employed, assuming that those unemployed receive zero wage. When ranked according to those expected returns, it clearly pays to specialize in the hard sciences, engineering and mathematics. Often the high pay more than compensates for the high unemployment rate, as with Military Technologies, mentioned earlier, with $84,422.25 in expected annual returns. In such fields, it clearly pays to not get the job at the local McDonalds and hold out for the right opportunity.
Does this make sense? According to economic theory, unemployment rates and wages are negatively correlated along a ‘wage curve’. A large supply of available workers bids down the price of labor. From the chart, we see that though this holds true in general, there are many notable exceptions. These outliers might suggest that certain sectors face a high level of frictional rather than cyclical unemployment; jobs are out there, but it’s difficult to match candidates to positions because of location and lack of information.
For policy makers, this might suggest that too many students do liberal arts at the expense of math and science. This makes perfect sense if risk-averse students prefer the relative certainty of getting a decently paid job to the risk of getting either a high paid job or nothing. More science grants and scholarships would certainly help students who hesitate to choose the riskier, more rewarding option.
Students choosing a major face a choice between what Nassim Taleb termed ‘mediocristan’, low risk & low returns, and ‘extremistan’, high risk & high returns on average. The decision of course depends on personal preference, as life is about more than having a job and a lot of money. But as economic foundations tremble and students who once though they had a guaranteed job find themselves in the street, perhaps it’s worth taking the risk to do something quantitative.
–  (1-unemp%)(.25* 25th%earnings + .5 * Median earnings + .25 * 75th%earnings) + unemp% * 0