Do students choosing liberal arts degrees regret it?

One way to measure students' satisfaction with their educational experience is to ask graduates, if they could choose again, would they select the same program.

Canada's National Graduates Survey (NGS) has been asking some variation on that question since 1982. The data has many limitations, as I explain in this post. But the little we can find out from readily-available data tells us this: most graduates, when asked, would select the same field of study again again, but graduates of some programs are more likely reaffirm their choices than others.

The first snapshot comes from 1982, when students were asked, two years post-graduation, "Given your experience since completing the requirements for the ____, would you have selected the same educational program, a different program, or not taken any post-secondary program?"  My calculations using the NGS public use microfile suggest that the students most likely to say that they would choose the same program again were those in business, engineering, and the health professions, including medical school. Graduates of "other social sciences" and "biological sciences" were particularly unlikely to select the same program again. Humanities grads had some regrets, but not as many as social science grads [Oops - vertical axis mislabeled.]

The class of 1982 graduated into one of the worst labour markets since the Great Depression, so their experience might not be typical. Unfortunately the field of study information has not been included in the NGS public use microfiles since 1982. Fortunately, Employment and Social Development Canada has published selected tabulations from the 2013 NGS on Canada's open data portal here. 

In 2013, graduates were asked, three years after graduation, "If you could choose again, would you select the same field of study or specialization that you completed?" Here are the results:

Thirty years on, the satisfaction patterns are remarkably consistent - graduates of fields such as social science, psychology, and biology are the least likely to make the same educational choices again. Among humanities graduates, 71.3% would chose the same field of study again, which is somewhat below the average for all bachelor's and first professional degree holders (73.9%). Business and management graduates have a fairly typical rate of satisfaction with their education (74.4% would choose the same again), but the people who are really pleased with their educational choices are graduates of health, education, engineering, and computer science programs. For example, 85.0% of engineering graduates would choose the same field of study again.

One possible reason why graduates of business schools, professional health programs, and engineering might be happy with their choice of program is that graduates from these programs typically have relatively high earnings (see, for example, Ross Finnie here, Marc Frenette and Kristyn Frank here, and Council of Ontario Universities here). Data from the 2005 NGS finds that there is a strong correlation between a graduate's earnings and the likelihood that they respond that they would select the same program again.

The graph above includes no controls. Further analysis with the 2005 data found that, after controlling for income, variables such as gender, language spoken at home, or parental education had no statistically significant effect on a graduate's likelihood of selecting the same program again (though as I will discuss in a later post, gender might have mattered more if I had used a different year's NGS data).

Earnings and employment prospects can explain why engineers and computer scientists would pick the same field of study again. But why might humanities graduates appear be more likely to say that they would select the same field of study again than graduates of, say, social and behavioural sciences, or even physical or life sciences? [caveat: I say "appear to be" because I only have tabulations for the 2013 data, and have not tried to manually crank through various tests to see the differences here are statistically significant.]

It could be that studying philosophy or history is a privilege that is associated with other forms of privilege - and thus subsequent success in the job market. For example, it is less risky to invest in a humanities degree if mom or dad can help with the job search after graduation. A very crude way of examining this hypothesis is with the 2011 National Household Survey, which provides information on immigrant status, visible minority status, home language, and a person's field of study in their highest degree. Looking only at bachelor's degree holders between the ages of 25 and 34, there is a strong relationship between field of study and family background:

(Note that this table excludes graduates who have gone on to, say, law school). Humanities graduates might be somewhat protected in the labour market relative to, say, physical and life science graduates, by their ethnicity or language skills. There are, however, only small and generally insignificant differences between humanities and social science graduates with respect to characteristics that can be observed in the National Household Survey.

Alternatively, it could be that humanities programs such as history or philosophy do a better job of teaching writing and critical thinking skills than do social science programs, and that explains any satisfaction differences between the two types of programs.

Another question raised by this analysis is who is to  blame: whose fault is it that over a third of social science graduates are so dissatisfied with their field of study that, if doing their degree again, they would select a different field of study? That one third of physical and life science graduates are? Perhaps professors are to blame: we could teach different material, or teach the same material differently. Perhaps universities are to blame for selling innocent 17 and 18 year-olds on low-value credentials, or offering a hundred places in pre-med-type programs for every one spot in medical school. Or perhaps the blame lies with the restrictive admission practices of, for example, business schools. If my university's business school had a more open admissions policy, there would be fewer - and perhaps fewer unhappy - economics majors. Or perhaps this is all an artifact of the National Graduate Survey, which had a response rate below 50 percent. It might have over-sampled malcontents living in parental basements.

I don't know. What do you think?

Comments

[Linda Welling]

I think I like this campaign for a "more open admissions policy" in business schools, but I am still thinking about it.

[Frances Woolley]

Linda - here's a hypothetical to think about. Imagine a student with a 79.5% high school average who would like to study business but falls just below the 80% business school admission average at your university or mine.

What are their options - community college? Or a social science degree, e.g. economics, where they will learn theory and metrics and some econ appications.

If that person's goal in life is a mid-level administrative job that involves, say, spreadsheets or purchasing or sales or conference organization - wouldn't business be a better preparation for that than my economics of the public sector course? Why divert that person away from what he or she wants to do?

Unless you're a business school hoping to build a reputation for quality graduates by restricting who you will admit...

[Livio Di Matteo]

Well Frances, what can I say? In hindsight, if I had to do it over again, I think I would become an ophthalmologist and set up a practice in Arizona.

[Frances Woolley]

Livio - Still lots of snow on the ground in Thunder Bay? Still a pile in my front yard, but it's melted enough for a snowdrop to emerge yesterday...

[T. Chen]

Have you considered that career satisfaction is due to personality types? STEM/Med people tend to skew as objectivists. They will favor merit-based professions that have clearly defined measuring sticks. Conversely, other personality types will find these fields dry and boring and self-select into the other careers. By nature of established power hierarchies, existing members will not favor 'clearly defined measuring sticks'. These are fields that social skills and soft people skills play a stronger part. These are fields where the distribution of winners and losers is far more harsh. Subjective measuring sticks are often boolean. Is it surprising that new entrants are thus dissatisfied?

This dovetails into your surmise that existing privileges enhances the comparative success rate of one entering these fields.

And in case someone gets the wrong idea, these are just generalities. Successful STEM/Med people need soft-skills to and vice versa. We're talking population samples and there's some inkling of correlation between Holland's Big Six domains of vocational interest and the standard big five personality.

[Frances Woolley]

T. Chen "Have you considered that career satisfaction is due to personality types?"

This is an interesting theory, and I strongly suspect that you are right, personality types matter. The people who find themselves drawn to the critical theory paradigms common in some social sciences/humanities may have a tendency to apply that critical lens to their university education too. A couple of things that the theory needs to explain, however:

- what explains the physical and life sciences results? Wouldn't they tend to be objectivist, like STEM/Med types, too?
- I'm not sure which fields you have in mind when you think about ones where the distribution of winners and losers is far more harsh. I think Ross Finnie's work finds that the distribution of earnings for business grads is pretty unequal - some business grads end up managing Starbucks, others have mega-star careers. Any winners/losers/personality story would need to fit the income distribution facts.

[T. Chen]

1. With life(and physical) sciences at the BS level there's one thing that comes to mind. I find that the type of entry-level jobs offered are very rote. That requires a low openness to experience [OtE] (i.e. perseverance). University marketing however, targets life sciences as an 'emerging field' which ironically draws in more people oriented towards a higher OtE and Investigative (on the Holland's six). Modern day guidance counselling based on Holland's vocational types pushes Investigative students into life/physical science. In the real world, actual investigative jobs are reserved for PhDs.

Agriculture/resource fields do not have this marketing mismatch. Entry level jobs there are also low OtE ones. Holland's vocational matches these fields with "Realistic" which won't be surprising for any kid who likes hands-on jobs.

An interesting thing to note is that the self selection isn't symmetric. Kids with low OtE will filter out fields that portray themselves as dynamic. Kids with high OtE have preference, but are less likely filter and will end up in ill-fitting 'dead-end' (from their perspective) jobs.

2. For my comment on the distribution of winners and losers, I was looking at fine arts, humanities, education, and ‘other’ social sciences in particular. Graduates here move on to many various occupations – most of which are not related to their field of study. Of those who stay in-field, competition for the few vocation related spots tends to be fierce. In resource constrained environments, human nature is to hoard rewards to be doled out as fief to consolidate security. The phrase ‘who you know’ comes to mind as do glass ceilings.

One additional thing to watch out for on the satisfaction/earnings graph is that the field averages don’t necessarily map to it. Earnings spread within a field confounds it. Entry level social workers for instance have a median of approx. $47k The fact that most graduates from behavior social sciences don’t become social workers (most other related jobs require a masters) and end up on the < $40k in degree unrelated work.

Without this knowledge, one might suggest that we need to pay behavior social science jobs more. That may move the mean for that cohort, but does very little for the majority of graduates who did not have the means to compete with the winners. In fact, such a response would merely entrench existing members.

Contrast to the high satisfaction for Business Management despite their distribution of earnings. The difference might be that they’re still applying their degree, even if it may be in a smaller pond. Holland’s vocational matches for them even at entry-level jobs so that’s a plus. And compared to the arts and social sciences, I would say a smaller pond is still better than being culled out altogether.