Friday, June 15, 2012

New Data on The Generosity of Philosophy Students

In 2007, I analyzed data on student charitable giving at University of Zurich, broken down by major.  When paying registration fees, students at Zurich could choose to give to one or both of two charities (one for foreign students, one for needy students). Among large majors, philosophy students proved the most charitable of all majors. However, philosophy majors' rates of charitable giving didn't rise over the course of their education, suggesting that their giving rates were not influenced by their undergraduate training.

I now have some similar data from University of Washington, thanks to Yoram Bauman and Elaina Rose.  At UW from 1999-2002, when students registered for courses, they had the opportunity to donate to two charities: WashPIRG (a broadly left-leaning activist group) and (starting in fall 2000) "Affordable Tuition Now" (an advocacy group for reducing the costs of higher education in Washington). Bauman and Rose published an analysis of economics students' charity and they kindly shared their raw data with me for reanalysis. My analysis focuses on undergraduates under age 30.

First, looking major-by-major (based on final declared primary major at the end of the study period), we see that philosophy students are near the top. The main dependent measure is, in any given term, what percentage of students in the major gave to at least one of the two charities. Among majors with at least 1000 student enrollment terms, the five most charitable majors were:

Major: percent giving
Comparative History of Ideas: 31%
International Studies: 24%
Philosophy: 24%
Physics: 22%
Anthropology: 20%
The five least charitable were:
Construction Management: 7%
Business Administration: 7%
Sociology: 7%
Economics: 8%
Accounting: 9%
These numbers compare with a 14% donation rate overall.

As reported by Bauman and Rose and also in Frey and Meier 2003 (the original source of my Zurich data), business and economics majors are among the least charitable. The surprise here (to me) is sociology. In the Zurich data, sociology majors are among the most charitable. (Hypotheses welcome.)

But what is the time course of donation? Bauman and Rose and Frey and Meier found that business and economics students were among the least charitable from the very beginning of their education and their charity rates did not decline further. Thus, they suggest, their low rates of charity are a selection effect -- an effect of who tends to choose such majors -- rather than a result of college-level economics instruction. My analysis of the Zurich data suggests the converse for philosophers: Their high rates of charity reflect a fact about who chooses philosophy rather than the power of philosophical instruction.

So here are the charity rates for non-philosophers, by year of schooling:
First year: 15%
Second year: 15%
Third year: 14%
Fourth year: 13%
And for philosophers (looking at 1172 student semesters total):
First year: 26%
Second year: 27%
Third year: 21%
Fourth year: 24%
So it looks like philosophers' donation rates are flat to declining, not increasing. Given the moderate-sized data set, the slight decline from 1st and 2nd to 3rd and 4th years is not statistically significant (though given the almost 70,000 data points the smaller decline among non-philosophers is statistically significant).

One reaction some people have had to my work with Josh Rust on the moral behavior of ethics professors (e.g., here and here) is this: Although some professional training in the humanities is morally improving, one reaches a ceiling in one's undergraduate education after which further training isn't morally improving -- and philosophers, or humanities professors, or professors in general, have pretty much all hit that ceiling. That ceiling effect, the objection goes, rather than the failure of ethics courses to alter real-world behavior, explains Josh's and my finding that ethicists behave on average no better than do other types of professors. (Eddy Nahmias might suggest something like this in his critical commentary on one of Josh's and my papers next week at the Society for Philosophy and Psychology.)

I don't pretend that this is compelling evidence against that position. But it does seem to be a wee bit of evidence against that position.


Richard Y Chappell said...

Do you know if there's any comparative data on donation rates to global poverty charities? If people are influenced by the work of Singer, Unger, etc., then that would presumably be where it shows up.

Eric Schwitzgebel said...

I don't know of any such data, Richard. It would be terrific to get such data -- but I'd be very nervous about trying to do so through self-report!

Anonymous said...

I'm wondering if this is due to the incredibly large amount of wealthy or upper middle class persons who decide to major in philosophy or other humanities versus business and accounting. Is there a reason to believe that this has anything to do with the majors listed here rather than the pocketbooks of those majoring in them?

Eric Schwitzgebel said...

Anon, concerns like yours are exactly what make me more interested in the time course data than in the absolute differences by major. In the Zurich data the teacher training students donate at low rates, yes that is typically viewed as an unselfish career choice.

Anonymous said...

Can you clarify the data? It seems likely to be based on major at the time of donation - yes? If so, this would seem to be very shaky grounds for any conclusion, would it not? Because you aren't tracking the donations of the same people over 4 years.

Cincinnatus C. said...

I wonder if the Washington results might be a bit different if the charity were, say, the American Enterprise Institute instead of a PIRG.

MaHernandez said...

Could the reason for ethicists or students/professors who hit the "ceiling" be because of an overly rational or overly academic view of ethics? If I may mention Hume, you must hit someone emotionally to change their behavior, and philosophy is much more reason based. So, the types of students who join philosophy recognize that giving to charity is good, and that they should do it, but don't have an emotional pull to do more and more.

Eric Schwitzgebel said...

Thanks for the comments folks!

@ Anon Wed: My analysis of the UW data is based on declared major at the end of the study period. But the data can also be analyzed student-by-student and the conclusions are basically the same. In the Zurich data, I've also analyzed both by student and by student-semester, with no material difference in results. The analyses are conceptually easier and more statistically powerful if I treat student-semesters rather than students as the units of analysis.

@ Cinn C: I agree. PIRG is less than ideal, given the political slant. The Zurich charities are probably less politically loaded.

@ MaH: That might be right. Then the question is: When does the effect of reason upon morality kick in? In high school? Kindergarten? Never?

Anonymous said...

To expand on Richard Chappell's thought, it may be that students who are more reflective--or are more strongly influenced by philosophical arguments such as Singer's--already give to charities that they believe "do the most good" or otherwise favor for moral reasons. If these students already give to such groups, they may be less likely to respond to new solicitations. That is my own experience. (I am a recent PhD in philosophy.) Since I have already settled upon certain charities that I prefer, I rarely respond to new solicitations by mail or (for example) at retail stores. Those who give very little, or have not thought much about giving, may actually be more likely to give to charities they have not researched in these circumstances.

Eric Schwitzgebel said...

@ Anon Thurs: I share your worry. This makes the whole business difficult to study. I don't really trust self-report and I don't really trust any narrow behavioral measure! But I'd also rather not give up, and I hope that a convergence of imperfect measures might be illuminating. Further suggestions welcomed!

MaHernandez said...

The only other thing I'd like to bring up is a connection to the decrease in giving with the average student, as well as the flat rate/decrease in philosophers giving, and the financial stability of the student. Could students just feel more financially able when they start school, but as they realize they're going further into debt, or having to save money for going out into the world/grad school they decrease the amount they're giving, if they give at all?

Anonymous said...

It may be relevant that Philosophy majors learn about the difficult job prospects they face during the course of their undergraduate degree and may find no new reasons to give. Personally, I would love to be more giving, but my position in the job track at the moment is making that difficult to muster.

Anonymous said...

To give a little bit of empirical support to the thoughts of Richard Chappell and the anonymous philosophy PhD above, take a quick look down the list of members at 'Giving What We Can,' a group of people who have pledged to give a minimum of 10% lifetime income to the most cost effective charities in international development work ( Not all of the members listed as students list their focus of study, but philosophy, at least from a brief glance, is significantly overrepresented.

The study Dr. Schwitzgebel discusses doesn't seem to address 'generousity' simpliciter so much as rates of giving in response to a solicitation. What would be the problem with just doing a mass survey of students and asking how much they give, who they give it to etc., and monitoring which students were most generous in terms of charitable giving that way? The philosophy, and other, students who chose not to donate in the study may in fact be donating in high numbers or significant sums to other charities, but this study wouldn't capture that.

Chris Smith (like Richard, a GWWC member)

Steve G. said...

In agreement with Anonymous #1, above, first, the data needs to be analyzed to rule out the effect of socioeconomic status.
Other factors of interest are also present, but likely, to a considerably lesser extent. The meaningfulness of any possible inference might be limited by the actual distribution and number of cases in each grouping.

Eric Schwitzgebel said...

I agree if the target is absolute differences. Time courses are, I think, more valuable data given my own interests (though they could also be compromised by an interaction effect with SES or similar).

CLains said...

That drop from the second to third year is the sound of philosophers realizing they are screwed financially. x)