Professor Yan Chen studies behavioral market design at the University of Michigan. In conversation with BiGS, Professor Chen explores the impacts of social media on women in the job market, how powerful actors can address hiring biases by advocating for women candidates on digital platforms, and how to scale such interventions for widespread progress.

This interview has been edited for length and clarity:

Tell us a little bit about your research

So, we are trying to address the fact that women are disadvantaged at various stages of their academic careers, but especially in economics. We thought that, you know, maybe social media promotion could be one way to improve female candidates’ job market outcomes. And in the 2022 to 2023 job market season, we conducted a field experiment on Twitter (now called X). We sent job market candidates a pre-survey. We also ask them to send us a tweet about their job market paper. So, we posted every tweet on our research account on Twitter to promote them, and also to get a baseline of how much attention their job market paper gets for a random subset of the papers.

We then asked influencers, economist influencers, to re-tweet (now known as a repost on X) the job market papers in their subfield. And...we sort of disproportionately advantaged the underrepresented groups, you know, women, underrepresented minorities, LGBTQ+ job candidates. And then we look at how that extra exposure changes the perception, the attention received on social media, and the downstream job market outcomes.

We found that having an influencer re-tweet their job-market paper has 300% to 400% increase in likes and views. So, their job market paper received higher views and more likes on Twitter. The effect was very large and robust. Later, in the summer of 2023, we'd send a post-job-market interview to learn whether they received interviews, fly-outs, or offers. And we found that there was a larger effect for women. Women whose job market papers were re-tweeted by influencers had one more job offer compared to the control group. In the control group, women, on average, got three job offers. And [in] the treatment group, women got, on average, four job offers. So, it does seem like this is a scalable intervention. And it did seem to benefit women in the end.

So, your research is showing the positive side of social media.

Yeah. That [is] one thing that we wanted to look at. So, first of all, the economist influencers on Twitter (X) are predominantly (~80%) from the top 30 institutions. And they tend to tweet their own students. So, what we try to do is to ask them to tweet students from lower-ranked institutions in their field, disproportionately tweeting women and other underrepresented groups. And we show that it does have a causal effect on the number of job offers that women get. This method [can] be applied to other areas where inequality exists.

We can, for instance, identify the experts, the influencers in other fields, let's say physics or chemistry, where women are also underrepresented. And, we can ask these influencers, expert influencers, to publicize research or job market papers from other underrepresented or lower-ranked institutions. And we anticipate that this would be helpful.

At this point, we are only looking at short-term outcomes. I think there's one recent research paper in AI and machine learning where they look at five-year outcomes of social media promotion and they find that it increases citation. There's also another study that looks at economics, which has a longer publication cycle. So that's nine years out and they also find that it increases citation. So, I think if we, you know, go back to these candidate job market candidates ten years out and track where they are in their career ladder, we might find longer-term effects. At this point we're only one year out, so we find the short-term effects.

How we can apply some of these learnings to business leaders. Can you offer two or three takeaways for business leaders?

So, for business leaders, you know, one takeaway is mentoring. You can think of this as a very light touch, low-cost mentoring method, which is ‘I'm just going to, you know, try to promote these new people in my field by paying attention,’ either by publicizing their output, their research, or just...promoting those young people in my field who are underrepresented, let's say, women in technology.

Did you encounter any resistance from the economic influencers when you asked them to do this?

Not 100% of the influencers agreed to participate. So, we're hoping that when the result is published, [we] will have wider acceptance of our intervention.

If you have any ideas or guidance or tips from leaders on how they might implement something like this, we'd love to hear it.

Well, light touch, meaning, you know, if a junior colleague just published a paper or, you know, came up with a new product, you want to publicize them, especially when they're underrepresented. You want to send out that email to your entire group and say, ‘Look, so-and-so just published a new paper, Kudos’ or ‘so-and-so just finished this wonderful new product for our organization.’

So, you want to send your praise early on and in a very positive way. Yeah. And publicize them. Promote them. Publicize them. Yeah.

What do we still need to learn or understand about inequality in the digital age? What do you think might be the next research frontier?

The next research frontier would be looking at where inequalities exist and how [they] affect our understanding of science and our practice. And, you know, evaluate how effective various types of interventions are in promoting equity and publicize them and make them scalable. One example is the American Economic Association's mentoring program for female assistant professors.

I was among the first cohort. And once they figured out, through a randomized controlled trial, that [it] is highly effective, they stopped the randomized controlled trial. So, the young new assistant professors can, you know, apply for the program and be mentored.