STEM Dropouts

Last year, I wrote a blog about research that explained why STEM (science, technology, engineering, math) majors in college become discouraged about their chances of persisting in their major. In this week’s blog, I look at the reasons why graduates of STEM majors drop out of STEM-related careers after they enter the workforce.

The analysis of undergraduate behavior that I reported on last year found that students who become disillusioned with non-STEM majors tend to do so slowly over several years and tend to experience loss of interest and diminishing expectations about the earnings they can expect in their field. On the other hand, students who become disappointed with STEM majors tend to do so precipitously in their freshman year, mostly because of diminished expectations of their ability to earn a decent GPA. In other words, it’s not boredom or lowered career expectations that discourages them, but the challenge of the coursework.

But what about after graduation? Those who graduate with a bachelor’s in a STEM field have successfully met the challenges of the curriculum. Having overcome this hurdle, do they face other barriers once they have left college and entered the workforce? I found some answers to this question by analyzing data from the 2003 National Study of College Graduates, a longitudinal study conducted by the Census Bureau on behalf of the National Science Foundation. This survey asked 170,000 college grads under age 76 about their education and their current work situation. Although the survey was updated in 2006, I used data from the 2003 sample because it polled graduates of the full range of college majors, whereas the 2006 survey looked only at science and engineering majors.

Using data from the 2003 survey, I was able to compare the work situations of people who held bachelor’s degrees from STEM and non-STEM majors. In compiling the set of STEM majors, I ruled out the social sciences and the health-care sciences that focus primarily on patient care.

One important finding was that STEM grads are more likely than non-STEM grads to be working in a job that is closely related or somewhat related to their major, as shown by the graph below. The differences are not great, but they are significant.

But when STEM grads are in an unrelated career, what is the reason? The next graph compares the responses of the two types of grads. You’ll note that for many of the reasons for not working in a related job, the STEM and non-STEM grads are quite similar. Evidently  the STEM grads have hardly any added tendency to lose interest in their field, because only one percentage point separates them from the non-STEM grads. The biggest difference is shown by the navy blue zone: STEM grads tend to be quite a bit more likely to be in an unrelated field because of their inability to find work in their field. With non-STEM grads, as indicated by the two lowest zones, disappointment with pay or promotion or dissatisfaction with working conditions was a more likely reason for working outside their field.

Keep in mind that this analysis is based on data collected in 2003. It may be that in 2012 and the years ahead, there will be lots of job opportunities for STEM grads, keeping them employed in work that is related to their major. But what I find from this survey tells me that if lots of STEM-related jobs don’t materialize soon, or if career development professionals fail to help connect STEM grads with appropriate jobs, we may lose STEM grads to unrelated careers.

More on the Outcomes of College Majors

In my previous blog, I cited a research study that found many college students expressing the view that they would be achieving greater success if they had chosen a different college major. I have found much reason to agree with them. The research that I did for the third edition of the College Majors Handbook with Real Career Paths and Payoffs: The Actual Jobs, Earnings, and Trends for Graduates of 50 College Majors showed me that graduates of certain majors achieve much higher rates of pay, employment, and satisfaction than graduates of other majors.

If you want to see the variations among individual majors, you’ll need to examine the book, because I can’t cover all 58 majors in a blog. (Yes, despite the subtitle, the book actually covers 58 majors.) However, I can summarize some of my findings here by dividing the majors into the 7 divisions that are used to organize the book: Behavioral and Medical Sciences; Business and Administration; Education; Engineering; Humanities and Social Sciences; Natural Sciences; and Technology.

Using data from the 2003 National Survey of College Graduates, I compared these 7 divisions in terms of the earnings of graduates (those whose highest degree is the bachelor’s). Specifically, the figures are the weighted means of the medians for each major within the divisions. (This approach gives extra weight to majors that have a large number of grads.) The dollar figures have been inflated from the year of the survey to 2010. As you can see, earnings vary considerably among the 7 divisions:

Here's another comparison, this time based on the job-growth projections of the Bureau of Labor Statistics for the occupations employing the graduates. For the graduates of each major, I inflated (or, in some cases, deflated) the number working in each occupation by the job-growth (or -shrinkage) figure that BLS projects for that occupation for the years 2010–2020. Then I collapsed the number of actual workers and the number of projected workers for each division and created a graph comparing the rates of projected growth:

Lastly, here’s a comparison of the level of job satisfaction expressed by grads of the 7 divisions. Again, I used weighted averages:

It’s interesting to note that the highest-earning divisions are not the fastest-growing divisions, and vice versa. Nor does job satisfaction correlate perfectly with either of these measures. However, it’s not necessary to choose between income, job opportunity, and job satisfaction when you choose a major: It is possible to find individual college majors that combine all of these payoffs. To identify them, read the book.

Hindsight on College Majors

I recently finished the third edition of the College Majors Handbook with Real Career Paths and Payoffs: The Actual Jobs, Earnings, and Trends for Graduates of 50 College Majors. The book will be coming out this summer, and I just came upon a survey research report that makes a good case for why this book is badly needed.

The College Majors Handbook compares 58 college majors based on the real experiences of graduates, including their different earnings, rate of unemployment, rate of employment in a career related to their major, job satisfaction, and types of work activities, among other topics. If there’s one lesson to be learned from this book, it’s that different college majors result in very different career outcomes. Some of these differences are economic: rate of unemployment, amount of earnings. It follows that as the cost of a college degree keeps rising, the importance of choosing the right college major also is increasing.

This lesson is driven home by a study issued this month by the John J. Heldrich Center for Workforce Development at Rutgers University’s Edward J. Bloustein School of Planning and Public Policy, “Chasing the American Dream: Recent College Graduates and the Great Recession.” The researchers—Charley Stone, M.P.P., Carl Van Horn, Ph.D., and Cliff Zukin, Ph.D.—surveyed a nationally representative sample of 444 recent college graduates from the class of 2006 through 2011 to see how they are faring in the workplace.

As I read the many interesting findings in this report, I zeroed in on what it said about the respondents’ choice of a college major. Perhaps the most interesting finding of all was how the grads responded to this question: “Thinking back to college, is there anything you would have done differently to be successful today?” The most popular response (at 37%) was “Been more careful about selecting my major or chosen a different major.”

Those who answered this way got a follow-up question: “With the benefit of hindsight, what type of major would you have chosen instead?” The most popular responses to this question were “Professional major, like communications, education, nursing, or social work” (41%) and “A major in the field of math, science, engineering, or technology” (29%). The College Majors Handbook confirms the wisdom of this hard-won insight; these majors do tend to result in higher earnings and likelihood of employment.

The report also reveals why the grads expressed this level of regret. It turns out that only 39 percent reported having thought about job opportunities in the field when deciding upon their major. No other reason seems to have played a significant factor in their choice of a major.

If you know someone who is deciding on a college major, don’t let this young person make an uninformed decision. Certainly, the career outcome is not the only factor that needs to count in the decision. But it is probably the most important factor. That’s why it’s informative to see what actually has happened to graduates of different majors.

College Loans: The Next Bubble?

Last year, for the first time, the level of student debt ran higher than the level of credit card debt. This happened partly because credit card use has been on a downward slope ever since late 2008 and partly because the upward slope of college tuition loans shows no sign of leveling off. The two trends are, of course, related. The recession has caused consumers to use credit cards less and has also caused more people to seek higher education to compete for scarce job openings (or to take a leave of absence from the discouraging job market), taking out loans because they and their families are less able to pay their own way.

This trend worries me for several reasons. One is the big-picture worry that we’re seeing the growth of a bubble. Like the housing bubble, this one involves people taking on lots of debt to acquire a commodity that they desire because it keeps increasing in value.

The present situation differs from the housing bubble, however, in some important ways. The value of houses kept increasing because of constant demand from home buyers and then crashed when the market became glutted with houses that were foreclosed on. The value of a college degree, however, keeps increasing because of the demands of employers. Even for positions where bachelor’s-level skills are not needed, employers tend to prefer job applicants with a degree, and it seems unlikely that this behavior will change. Unlike a house, a college degree is something the banks can’t repossess and throw on the market, so it’s hard to foresee a scenario in which bachelor’s or associate degree suddenly loses its value. Yes, there are a lot of people with college degrees who are unemployed right now, but the recession has hit those without degrees much harder.

The way this bubble might pop would be if massive numbers of college loan holders defaulted on their loans, like the homeowners who walked away from their houses after the value of the houses sank below the value of the outstanding mortgage. Many tuition borrowers have technically gone into default. Last year, the Chronicle of Higher Education reported that one-fifth of the loans that went into repayment in 1995 were in default. With college loans, however, precisely because there is no real estate that can be repossessed, lenders can continue to extract repayment from borrowers who have fallen behind in their scheduled payments. Bankruptcy does not generally free tuition borrowers from their repayment obligations. Lenders can garnish wagers; for federal loans, the government can also garnish tax refunds and even disability checks and Social Security payments. The danger is that in a future wave of college loan defaults, even these methods of extracting payments would not be sufficient to save lending institutions from a crash similar to what happened when the housing market collapsed.

Although there is reason to worry about what might happen at the macro level, I’m more concerned about what’s already happening at the micro level. Tuition borrowers who fall behind in their payments take a severe blow to their credit ratings and may be unable to borrow to buy a house or car. Even those who are able to keep up may find that the burden of loan repayments cuts into their disposable income for many years. Young people sometimes talk about the problem of when to tell a potential mate about your college loan. Second date? Third date? With rueful humor, they sometimes compare it to having a herpes infection or a family history of insanity.

From my perspective as a writer about career development, I’m particularly concerned about the effect of college debt on career choice. I fear that the burden of college debt sways too many students to elect majors that are potentially lucrative but ill-suited to them. An article on the Yahoo! site mentioned a student who graduated from Northeastern University with $200,000 in debt. She probably could have managed her college planning better (for example, starting at a community college and not taking a semester abroad), but it’s interesting to note how many people who commented on the story assailed her for majoring in sociology. For example, one person wrote, “I would pick a major with a higher ROI if I knew I was gonna rack up that much debt.” I would hate to see the day come when only students from well-off families can major in the liberal arts or social sciences.

How STEM Career Plans Get Derailed

Careers in STEM (science, technology, engineering, and math) are an interest of mine. I have blogged about them several times, have presented about them at conferences, and have written a book about them (Quick STEM Careers Guide). Next month, JIST will publish my STEM Careers Inventory.

The United States needs a steady supply of STEM-prepared college grads to fill the many technological jobs that our economy has created and will continue to create. In fact, if we fail to meet this need, our economy is threatened.

So I was intrigued to come upon a paper (PDF) by two economists, Todd and Ralph Stinebrickner (brothers?), “Math or Science? Using Longitudinal Expectations Data to Examine the Process of Choosing a College Major.” The researchers used data from a longitudinal study at Berea College to investigate how students go through the process of choosing a college major. They focused on large groups of majors, especially the group that they call “math/science.”

The usual assumption about how students choose a major is that students have a self-concept and a concept of the careers that the major leads to. Students are likely to change their planned major if they feel that one of these has changed so that what previously seemed like a good match now appears to be a bad fit. For example, students’ self-concept may change if they discover that they are no longer interested in the major or if they find that they lack the ability to do well in it. Their concept of the career outcomes may change if they learn that an industry associated with the major is not as promising as they previously thought it was or if an internship experience in a related career reveals work tasks or worksite conditions different from what they previously expected. The researchers sought to discover which of these changes were mainly responsible for students’ abandoning their plans to major in math/science.

The survey instrument at Berea College elicited four attitudes that students held toward their planned major: their percent chance of sticking with the major, their expected GPA, their expected income (in dollars) at age 28, and their interest in the major (on a five-point Likert scale). The research report notes that one unique advantage of this survey instrument as a window on the students’ career decision making was its frequency: “Each student was surveyed approximately 12 times each year while in school, with the first survey taking place immediately before the beginning of the student’s freshman year.”

A change in the first of the survey’s scales (chance of persistence) would indicate a change of heart toward the major. An accompanying change in the second scale (expected GPA) would indicate a change in self-concept, whereas if there were a better correlation with expected earnings or with interest in the major, this would indicate a change in students’ perception of the career.

The researchers found that expectations of persistence in the major changed differently for students planning to major in math/science: Expectations tended to decline precipitously in the freshman year (even though students did not have to formally declare their major that year), whereas students planning other majors, if they scaled back their expectations, did so more gradually over several years. And the decline in expectations for the math/science major showed a stronger correlation to anticipated GPA than to anticipated earnings or current interest. In other words, students’ experiences in their freshman year caused them to revise downward their estimates of their math/science abilities, and that’s why they expected to drift away from a math/science major.

I saw examples of this behavior in my own freshman year at a school with a large proportion of math/science majors (The Johns Hopkins University). Several friends of mine abandoned plans for math/science careers after experiencing the rigorous chemistry and calculus classes that freshman math/science majors at JHU are required to take.

The researchers conclude that we need “policies at younger ages that lead students to enter college better prepared to study math or science.” I agree. It’s not enough to get students interested in STEM careers. We need to be sure that in high school (and probably starting earlier than that) young people learn the skills they will need to succeed in a STEM college major.