This article summarizes the Babcock et al. (2017) 1, recently published in the American Economic Review.
Among the different reasons to explain the gender gap in the labor market one is the process by which men and women advance in the workplace, and among the variables that can affect this advance is the time women, relative to man, dedicate to tasks that benefit the organization, but are less likely to affect their evaluation and career advancement.
Field studies have systematically shown that women indeed spend more of their time on these tasks. Not only that, the studies show that women volunteer more to perform these tasks and also that they are asked to volunteer more often than their male coworkers. For instance, Porter (2007) 2 finds in the National Survey of Postsecondary Faculty (NSOPF) that female faculty spend 15 percent more hours on committee work than do men and Polachek (1981) 3, Goldin and Rouse (2000) 4, and Black and Strahan (2001) 5 find that women hold a different portfolio of tasks in the workplace.
In order to understand the reasons for working on low-promotability tasks, Babcock et al. (2017) study the differences in the response to requests and in the offers to perform those tasks, and the possible causes for them. There is already some field evidence about this hypothesis. Using the data in Tannenbaum et al. (2013) 6 Babcodk et al. (2017) find that 7% of female faculty volunteer to be on a committee after receiving an email request, compared to 2.6% of male faculty. The novelty of the study in Babcock et al. (2017) is the experimental approach to isolate the variables to get a more precise perspective of the causality. For that purpose they conduct a series of five experiments that are variations of the following baseline.
Experimental subjects are randomly assigned to groups of three. Members of the group are then given two minutes to make an investment (volunteering) decision. Individual earnings are 1 USD in the event that no one invests before the end of the two minutes. If one group member makes the investment, the round ends, and the individual making the investment secures a payment of 1.25 USD, while the other two group members each receive 2 USD. The experimental subjects participate in ten rounds, and they know that the group composition will change in each round.
The first experiment is the baseline setting. The results show that groups succeed in investing 84.2% of the time, although this rate decreases with time from 88.4% in the first five rounds to 80% in the last five. The most important finding for the hypothesis is that the average investment rate for men is 23 percent and that for women is 11.2 percentage points higher.
The second experiment conducts sessions where only men or only women participated. If gender differences in investing are caused by women being more conforming, more altruistic or more risk averse, then one will expect higher investment rates in the all-women sessions than in the all-male sessions. The results show a lower investment rate compared to the first experiment (80.8% vs. 84.2%). As before, the rate decreased, this time from 90.3% in the first five rounds to 71.3 in the last five. The most remarkable difference, however, is that women are not more likely than men to invest. The success rate was 81% in the all-female groups and 80% in the all-male groups, with no statistically significant differences. This suggests that beliefs rather than preferences drive the gender difference documented in the first experiment.
In the third experiment the authors extend their design to include an outside requestor who, after seeing pictures of the three group members, must ask one of them to invest, and receives 1 USD if no one invests and receives 2 USD if any member of the investment group invests. For that purpose, subjects were assigned to groups of four, and were told that one of them would be designated the requestor, and the other three will be playing as before, after reading the request. To gather more data, subjects were asked to write their requests for the case they were selected to be the requestors. The group members may or may not follow the suggestion the requestor, and their payoffs are the same as in the baseline experiment. Now groups succeeded in investing 93.3% of the times, substantially more than in the absence of the requestor. Further, the investments were made earlier in the session. Recall that subjects play in 10 rounds. The average number of requests a subject receives is thus 10. However, the mean and median number of requests for men are 8.7 and 9, respectively, while for women they are 11.1 and 12, respectively. A statistically significant difference consistent with participants holding the belief that women are more likely to invest than men. In terms of probabilities of being asked, these numbers translate in a 39% chance that a woman is asked and a 27% chance that a man is asked. The differences are confirmed when the authors control for other observable characteristics. Further, the numbers do not change much when the requestor is male or female. Finally, overall participants who are asked to invest have an investment rate of 65.5%, a percentage that is only 14% for those who are not asked. The response to a request, however, does differ by gender, as the investment rate becomes 51% for men and 76% for women.
Experiment 4 explores the role of beliefs. Experimental subjects participated in a five round version of Experiment 1 and then where asked to predict the outcomes of the original Experiment 1. Results in the shorter version of Experiment 1 were similar to the results of the first five rounds of the original Experiment 1. When asked to predict the behavior of the other group, they were informed of individual characteristics of the members, including age, sex, years of schooling, whether they were US born, and choice of major. The answer had to be given in probabilistic terms. Depending on the accuracy of their guesses participants received either 1 USD or 2 USD per round. The result is that while third parties anticipate the direction of the gender gap in investment rates, they underestimate the magnitude of the difference, estimating a gap of 2.3 percentage points instead of the real 11.2 percentage points.
The belief that women invest more than men is consistent with Experiment 1, but not with Experiment 2. A fifth experiment is performed to show whether individuals are more altruistic towards one of the genders. The specifics of this last experiment are a little more complicated, but the results are simple: there are no differences.
The main conclusion of the experiments is that the gender gap is not due to preferences, but to the belief that women more than men will volunteer. In particular, these beliefs make women more likely to invest, men less likely to do so, and both men and women more likely to ask women to volunteer.
- Babcock, L.; Recalde, M.P.; Vesterlund, L. and Weingart, L. 2017. “Gender Differences in Accepting and Receiving Requests for Tasks with Low Promotability.” American Economic Review 107:3, 714–747. ↩
- Porter, S.R. 2007. “A Closer Look at Faculty Service: What Affects Participation on Committees?” Journal of Higher Education 78:5, 523–41. ↩
- Polachek, S.W. 1981. “Occupational Self-Selection: A Human Capital Approach to Gender Differences in Occupational Structure.” Review of Economics and Statistics 63:1, 60–69. ↩
- Goldin, C., and Rouse C. 2000. “Orchestrating Impartiality: The Impact of ‘Blind’ Auditions on Female Musicians.” American Economic Review 90:4, 715–41. ↩
- Black, S.E., and Strahan P.E. 2001. “The Division of Spoils: Rent-Sharing and Discrimination in a Regulated Industry.” American Economic Review 91:4, 814–31. ↩
- Tannenbaum, D.; Fox, C.; Goldstein N., and Doctor, J. 2013. “Partitioning Option Menus to Nudge Single-Item Choice.” Paper presented at the Society for Judgement and Decision, Annual Conference, November. ↩