Mini Post #14: Miss Informed
Mini Post #14: Miss Informed
Do people listen to female economists less/more than to male economists?
Lately, I’ve been writing a shorter, narrower post focusing on one specific paper once a week. Last week, I wrote about whether people with unique fashion senses are more likely to invent new technologies. This is the tab with all the previous posts.
Long (or short) time readers might be familiar with my interest in gender and gender disparities. One such dimension could be influence: do people listen more to women than to men? Someone who is very overtly sexist would not, and someone who is “woke” would listen equally. However, if considering that sexism tends to hold women back from positions of influence, then it would be possible that women deserve to be listened to more - since a woman who reaches position X had to do more than a similarly capable man. To quote Hillary Clinton “I’m not asking people to vote for me simply because I’m a woman. I’m asking people to vote for me on the merits. I think one of the merits is I am a woman.” Regardless of your opinions about Mrs Clinton, her statement is actually illustrative of a genuinely relevant question: do people listen to women less than they listen to a similarly qualified man?
The paper I’m writing about, titled “Do Female Experts Face an Authority Gap? Evidence from Economics” (2024) by Hans Henrik Sievertsen and Sarah Smith seeks to answer this query - using, of course, economists.
The University of Chicago’s Clark Center for Markets has an Economics Experts Panel (EEP), where 43 economists get asked questions on various (somewhat) topical issues and they provide commentary. For example, recently the State of New York recently proposed, an later reversed, a measure to charge tolls into New York City at certain times of the day (known as “congestion pricing”). The EEP gave their takes - weighed by confidence, 61% of the experts agreed that congestion pricing “is likely to lead to a substantial reduction in traffic congestion in the targeted area”, whilst 7% were not sure, and the remainder (32%) disagreed. The answers each economist provides are listed like this (where the number measures “certainty” from 1 to 10):
Riveting stuff (to be fair, Austan Goolsbee’s answers to a predecessor survey, the IGM Survey, are pretty iconic1). The list of authors in question has 43 people, of which 7 are women, and all of which belong to top economics departments in the United States (Harvard, Yale, MIT, Chicago, etc.). For the 43 experts, the authors of the paper find their relative age, current institution, and gender identity. Then, they collate ten statements pertaining to various policy areas (AI, Twitter content moderation, price gouging, climate change, semiconductor policy, greedflation, financial regulation, policy credibility, windfall taxes on oil, and banning junk food).
The study first asks a sample of around 100 members of the US public to give their own opinions about a topic - for example “Financial regulators in the US and Europe lack the tools and authority to deter bank runs”. 58% of the public sample agrees, with 52% confidence, while the expert sample agrees only at 44%, with 61% confidence. On average, the difference between the experts and the public is around 30%, but without any relationship to how high (or low) agreement is among either group.
Then, for each statement they build a gender and opinion balanced sub-panel, where the opinion of each woman economist is matched with one other male economist who shares the same opinions - for instance, Marianne Bertrand agrees that banning junk food advertisement would reduce childhood obesity, as do Eric Maskin and 15 other men. Choosing one man out of the matched set at random, the authors of the study produced 72 gender-balanced opinion pairs for the ten issues.
Then, the authors sent a survey to 3,027 people, who received blurbs on what the various economists believed and were later asked for their own opinions. So, person A got “Marianne Bertrand agrees that banning junk food ads reduces childhood obesity”, whereas person B got “Eric Maskin agrees that banning junk food ads reduces childhood obesity”. The answers, alongside the confidence expressed, was measured in the same manner as both the EEP and the public opinion sample.
Let’s start with the results: on average, the difference in opinions between the sub-panel of economists and public opinion shrinks from 39% (in the baseline sample) to 33% (in the experimental sample). This isn’t a lot, but controlling for the strenght of the opinion expressed by the economists, we find a larger effect, but heterogenous across statements: economists make basically no difference in greedflation, but a lot of difference in, say, price gouging or financial regulation.
Now, let’s look at who is most influenced by economists. Since the survey respondents gave their own demographic information (gender, age, ethnicity, level of education, economics knowledge, and political affiliation), we can test whether different groups of people are more influenced by economists as a whole. Men and non-White individuals see much larger effects from seeing expert opinions, as do educated individuals, and people knowledgeable of economics. Surprisingly, Republicans (the right wing party) are influenced by the opinions of economists more than Democrats (the left wing party) or nonpartisan individuals, which contradicts the stereotype that conservatives are more skeptical of experts, but may be explained by the stereotype that economists are right wing. There are no significant age effects.
Now, let’s look at female versus male experts: are female economists likely to share the opinions of the public? Looking at the public that agrees or disagrees with an economist, the authors find that the public is 1.1% more likely to agree with a female economist than with a man, but this effect is not statisitically significant.
Looking at whether female economists are more persuasive than male economists (that is, whether the public agrees with the expert opinion more often if it is presented by Amy Finkelstein than Larry Samuelson), the authors find that female economists are around 20% more persuasive than men - but these results aren’t very significant, hovering between 0.05% and 0.1% significance. Controlling for the relative prestige of each institution listed, as well as for aspects in the photo of the economist deduced by ChatGPT (smiling, business attire, background, etc.), the authors find that the results are still positive, and show a similar magnitude and significance. The size of the effect is bigger after controlling for the age of the economist and for the number of media appearances they had done in a series of US newspapers.
Looking at the likelihood that each demographic will be more persuaded by a woman than a man, the authors find stronger effects among women, minorities, educated individuals, non-conservatives, and younger respondents, as well as people more knowledgeable about economics. The effects are each statistically significant, but the differences between demographics are not - except for older individuals and with Republicans, who disagreed with the women at higher rates and were around as likely to be persuaded by them as by a man.
So, to sum up, there is some evidence that expert opinions are persuasive on the public, and there is very little statistically significant evidence of gender splits, educational splits, or economics-level splits (although the coefficient for women is larger, it is also not significantly larger). However, women economists are seen as somewhat more persuasive than men. Why is this? The authors propose two reasons: first, that strong credentials may benefit women more than men if women in general are seen as less credible; second, that the stereotypes pertaining to the relative reluctance of women to speak about topics outside their areas of expertise may benefit the credibility of women when they do talk about their own subject matter. While the latter phenomenon is true, women are still more likely to be persuasive when they express an uncertain opinion, meaning that the first explanation might be dominant: people consider that women may have had to work harder to achieve prestigious positions, which means that their capability is overestimated. However, if women are equally persuasive and equally qualified, why are they less likely to give opinions? That is an open question, per the authors.
To finish the post, some links
The paper in question (and a link to the tweet where I found it)
A paper by William Nordhaus and Douglas Rivers about opinion mismatches between expert economists and the public at large
A New York Times article by Justin Wolfers (from 2015) about the relative (lack) of recognition for female economists in the media
A paper by Christine Exley and Kirby Nielsen about the confidence gap between men and women and the explanations for it
A paper by Heather Sarsons and Guo Xu finding that men and women with similar educational attainment have different levels of confidence and certainty in policy-relevant questions. Plus an LSU writeup of the study by the authors
A paper by Billur Aksoy, Christine Exley and Kessler finding that trans, nonbinary, and gendernonconforming individuals have even lower self-confidence than men or women with similar ability
A blog post by Alice Evans regarding the influence of women in office settings
A paper by Stefano Gagliarducci and Daniele Paserman about whether women participate more in bipartisan negotiations than men
A paper by Friederike Mengel, Jan Sauermann, and Ulf Zölitz finding that women are ranked significantly lower than men in websites for ranking professors - plus a Twitter thread about the study, and a website where you can find gender disparities in the use of specific words for professors by field of study.
Who can forget the time when 17% of economists agreed with the statement “giving cash as a Christmas gift is better than giving a present”
It would be interesting to see if this holds across other disciplines because my leading theory is that the stereotypical economist is an obnoxious white man, and non-white and women economists break that stereotype and therefore aren't assumed to be obnoxious.