Professional networks and discrimination on Twitter – 04/21/2023 – Policies and Justice

To stand out professionally, talent, good ideas or hard work are not always enough. In many cases, having access to good contact networks is crucial for professional success.

However, people have access to different networks of contacts. This can happen for a variety of reasons, but one plausible cause is discrimination: do people belonging to certain groups—blacks, women, graduates from less prestigious colleges—have less access to networks because they suffer discrimination based on these characteristics?

To test this hypothesis, researchers Nicolás Ajzenman (McGill University and FGV-EESP) and Bruno Ferman (FGV-EESP) and I conducted a randomized experiment on Twitter, a social network widely used to establish professional contacts in academia. We created fictitious accounts that looked like real people and claimed to be PhD students at American colleges. The accounts were identical to each other except for three characteristics: gender (male or female), racial identity (white or black), and the college they claimed to attend (one of the top 10 most prestigious colleges in the US or relatively less prestigious colleges). . The following figure helps to illustrate the experiment.

The experimental accounts randomly followed profiles of several economists from around the world on Twitter. After a few days, we measured the proportion of economists who reciprocally “followed them back.” By comparing the proportion of “follows back” for each type of account, we can identify the existence (or not) of discrimination in this behavior, since, with the exception of the characteristics we manipulated, our accounts were identical.

The number of followers is essential to benefit from a presence on Twitter, as they are the ones who will see what we share. The more followers, the greater the reach and potential impact of an account on this social network. For this reason, we chose this metric to study the formation of professional networks.

Our choice to focus on the community of academic economists on Twitter was no accident either. First, networks are particularly critical in academia. Second, the economics profession, both internationally and in Brazil, has been widening the debate about its lack of diversity (as discussed, for example, in the testimonies of these inspiring black Brazilian economists).

The lack of diversity has negative effects not only for marginalized groups, but also for the profession as a whole: when most researchers have homogeneous backgrounds, there is potentially less innovation and important topics are not studied. Third, the economist community is quite active on Twitter, which is ideal for our experiment.

In the experiment, we created 80 dummy accounts and followed around 8,000 economists on this social network. What we found is that, in fact, the number of incoming connections varies significantly depending on the student’s characteristics. Comparing the most extreme cases, we saw that white women studying at highly prestigious universities were “followed back” in 23.9% of cases, while black men from relatively less prestigious colleges, in only 14.4%; a difference of more than 65%.

The main results are represented in the graph below. Black students were 10% less likely to receive “follow back” than white students, while students who said they studied at colleges in the “top 10” American received almost 22% more “follows”. Also, women had 25% more “follows” than men.

The differences found are substantial and demonstrate quantitatively that there are disparities in access to academic networks between members of these different groups, which may contribute to the lack of diversity in the profession. The results regarding race and college affiliation are consistent with economists’ perceptions and with evidence of disparities in these dimensions.

However, the result for gender is surprising because it goes against the numerous evidences of discrimination against women in academia. Different mechanisms can explain this result, with potentially different consequences. For example, some Twitter users may be actively seeking to engage with women more to try to reduce gender barriers in the profession. However, it is also possible that some are using the platform with the aim of establishing relationships of a social (and not professional) nature. The fact that we found suggestive evidence that male users prioritize following female experimental accounts at a higher rate than female users may indicate which of the mechanism types is dominant, although the experiment is not able to elicit the motivation behind each” followed”.

In our work, we discuss other results and implementation details. In general, the experiment we carried out is relevant as it is the first to quantify the existence of discrimination in the formation of professional networks. This evidence can contribute to subsidize policies and programs that seek to reduce disparities and barriers against marginalized groups. However, much research on the subject is still needed to understand, for example, the motivation behind each “followed” and to investigate how discrimination suffered at different career stages add up and interact.

Even so, an experiment like ours is a first step in showing that, even in an online environment — potentially more democratic — different groups face different barriers to being heard and developing professional contacts.

The editor, Michael França, asks each participant of the space “Politics and Justice” of the Sheet suggest a song to the readers. In this text, the one chosen by Pedro C. Sant’Anna was “Pela Internet 2”, by Gilberto Gil.

PRESENT LINK: Did you like this text? Subscriber can release five free hits of any link per day. Just click the blue F below.