Can Single-Sex Education Make Women Less Risk-Averse?

(Photo: Eric James Sarmiento)

Research indicates that women are generally more risk-averse than men, and this risk-aversion is often cited as a partial explanation for the shortage of women in high-level corporate positions. A new essay by Alison Booth, Lina Cardona Sosa, and Patrick Nolen suggests that single-sex education may change women’s risk preferences.  In a recent paper, the researchers conducted a controlled experiment:

[W]e designed a controlled experiment using all incoming first year economics and business students at a British university. The subjects were asked to make choices over real-stakes lotteries at two distinct dates – the first week of term and the eighth week of term…

Prior to the start of the academic year, students were randomly assigned to classes. Our ‘nurturing’ environment is the experimental peer-group or class to which students were randomly assigned by the timetabling office. The class groups were of three different types – all female, all male, or mixed gender.

The results were striking:

[W]omen are significantly less likely to make risky choices than men at both dates. However, after eight weeks in the single-sex class environment – within the larger coeducational milieu – women were significantly more likely to choose the lottery than their counterparts in coeducational groups. No such result was found for men in the single-sex groups. In other words, after eight weeks, the women in the single-sex classes were no more risk averse than men.

The authors say their research indicates that “gender difference in behaviour under uncertainty found in previous studies might actually reflect social learning rather than inherent gender traits.”


I'm inclined to trust this experiment and its results. One question though: how were the instructors assigned to the different class groups? It's conceivable that an instructor who was inclined to choose an all-female class might also be better at making students less risk-averse.


Greg, the point you are bringing up is very valid, and I'm curious to hear the answer.
Was the same person running the three classes or not? Was it a male or female instructor class? Nowhere in the article this is explained and it could definitely have a great impact on the results, of course always assuming that the topics discuss and all the basics where the same in the 3 environments.


I wish the article had mentioned the nature of this "lottery." For some reason, I formed the impression - not supported by anything I could find upon re-reading - that the lottery involved assignment to a new class. If that were the case, the experiment could simply show that the women were more likely to want to escape a single-sex environment.


The lottery was all just for money. You have a sheet with 20 scenarios. Each scenario is a choice between getting $1-20 no strings attached, or having a 50-50 chance of winning $30.

They didn't blind the instructors giving the test, so maybe they were coaching the all woman class to be more aggressive but how likely is that?


Thank you for the description -- I even looked around some to see if I could find the description of the lottery procedure.

It doesn't seem to me to be a very robust measure of "risk." The only actual risk seems to be coming out exactly where you already are -- right? Off the top of my head I can't think of a quick and easy measure like this is to administer, but "risk" to me implies some sort of noticeable downside risk -- putting up my own money, say, not just the experimenter's! I get that the differences between the group are still there regardless of what's measured, so that's something, but still...I don't see this as any real plus for single sex ed!

Eric M. Jones.

I have always thought that segregating sexes was an easy way to improve education--especially when the kids are suffering from sex-hormone poisoning--like in their teens and twenties.

So maybe "risk-averse" is just one difference of many.

Next question: Uniforms....


These two comments seem contradictory:

"[W]omen are significantly less likely to make risky choices than men at both dates. " and " In other words, after eight weeks, the women in the single-sex classes were no more risk averse than men."

If the women always were significantly less likely to make risky choices, how can they say they were no more risk-adverse than the men?


Good find. The part about women closing the risk gap after eight weeks of single sex ed isn't in the paper. I think freakonomics put it in there on accident. It could be true, I couldn't tell you though since all the data in the paper is completely beyond me.