Are Politically Diverse Teams More Effective?

C. J. Burton/Getty Images

Political polarization in the United States and around the world is on the rise. Voting patterns, surveys, and discussions on Twitter all indicate that we now agree on fewer and fewer issues and like each other less and less. A 2015 study revealed that Americans were more comfortable having children marry into a family of another race than another political party. Moreover, U.S. politicians from the political right and left have become more polarized in how they vote and how they talk on the floor of Congress, resulting in fewer compromises and more insurmountable conflicts. Political ideology has thus become a growing faultline across teams and organizations with which today’s managers must contend and yet, research findings suggests that we tend to not handle these divisions well.

Nevertheless, decades of social science research has found that the presence of diverse perspectives is beneficial for creative companies and teams. Diverse perspectives enable groups to search a wider space of solutions to their problems, even if that creative process can be uncomfortable and taxing. In our own prior research, we examined millions of book co-purchases online and found that those on the political right and left — those who bought conservative versus liberal books — purchased very distinct books of science and literature. This may make conversation across political divides more difficult as interlocutors cannot even appeal to the same “facts,” but it also increases the possibility that political conversations can unlock diverse perspectives on more and more issues.

In our research, recently published in Nature Human Behaviour, we sought to discover whether these potential benefits of diversity could be realized in practice. To do so, we turned to a particularly important setting, one in which teams collaborate to create collective knowledge products consumed by billions around the world: Wikipedia, the online, crowd-constructed encyclopedia. Behind each Wikipedia article — each information product — is a team of volunteer editors who created it, often with tremendous individual effort and collective coordination. When editors produce articles on particularly divisive topics, such as abortion, minimum wage laws, or President Donald Trump, does political diversity increase or decrease their effectiveness in creating high quality, balanced treatments? Moreover, how does polarization influence how they collectively agree on what content should be included, excluded, and how it should be framed?

First, we estimated the political affiliation of more than 600,000 Wikipedia contributors through their intensity of contribution to liberal versus conservative articles. We assumed that editors who make frequent contributions to pages associated with American liberalism would hold left-leaning opinions, and vice versa. This was a controversial hypothesis, and many in the Wikipedia community felt that perhaps the opposite would be true, with liberals correcting conservative pages and conservatives returning the favor. Nevertheless, a survey we conducted of active Wikipedia editors found that millions of editor contributions to liberal versus conservative pages was a significant predictor of whether they identified as liberal or conservative and how they voted.

Following this validation, we then measured the political diversity or polarization of each team of editors behind 232,000 different Wikipedia pages, considering editor groups with a broader range of ideological alignments to be more “polarized.” When this measure of polarization was used to predict Wikipedia’s internal ratings of article quality (ranging from “stub” up to “featured article”), we found that higher team political polarization was strongly associated with higher page quality, far exceeding the quality of similarly sized biased, neutral, or moderate editor teams. This was especially true for political articles, but also those on social issues and science.

In order to understand the drivers behind this polarization premium, we analyzed team discussion on article “talk pages” where editors debate over what should be included and how the article should be written. For popular topics, these debates read like the transcripts from a high-stakes academic conference or product development meeting. Analyzing the content of these discussions, we found that polarized teams engage in more debates but with less toxic conflict than ideologically unbalanced teams, where the efforts of lone, contrarian editors to “de-bias” articles sometimes provoked charged disputes. Moreover, polarized teams focused on similar content as moderates, but discussed it with many more words in many more ways. Finally, polarized teams more frequently appealed to Wikipedia’s norms encoded in consistent policies and guidelines, such as that articles should manifest a neutral point of view (NPOV), cite respected sources for support (CITE), and avoid or disclose conflicts of interest (COI). Adherence to these norms buffered against the raw emotions and abuse found in many less-regulated online communities. When surveyed, editors engaged on polarized teams described grueling discursive work that led to superior products.

The findings are surprising. Political polarization is typically regarded as negative, but we reveal that if the power of diverse, polarized perspectives can be unleashed, it can positively influence quality productivity. Beyond documenting a “proof of concept,” it is even more important to understand why Wikipedia succeeds where so many other platforms fail. This is partly because Wikipedia has an encyclopedic monopoly and if people want a voice on the web, they need to play ball. But other patterns are clear as well.

Frequent appeal to Wikipedia guidelines and policies in polarized teams, the violation of which could lead to an editorial crack-down on participation, suggest that increased oversight and bureaucracy might be beneficial. In a community or media environment without laws, or with weak or attenuated norms, the potential increases for that environment to turn toxic, with shorter conversations, less collaboration, leading to lower quality. Another way in which Wikipedia is different is its well-known and publicized commitment to discourse and consensus. Strongly signaling such a mission upfront may induce self-selection from those individuals willing to cooperate for the common good. Wikipedia’s guidelines and policies generate not only high-quality articles, but a sustained collaborative culture. Guidelines and policies are not simply rules concerning what may be written; they constitute a culture by stipulating appropriate social conduct, such as how editors treat one another in talk-page debates, and even how we, as researchers, needed to engage with the community. It required enormous effort to earn the right to conduct this study. We had to become members of the community in order to understand the community. But these protections preserved a space that unleashed the power of diversity.

Perhaps the biggest takeaway from our work is a new perspective on bias. At the individual level, bias is generally undesirable. It leads to bad investments and wrong conclusions. But bias is sometimes driven by passion, which leads those we deem biased to work longer and harder for something they believe in. Bias also enables access to useful information and perspectives forged deep within ideological echo chambers. Collectively, teams with mixtures of bias that are willing to engage and collaborate can yield superior performance. This reveals a “silver lining” of political diversity and disagreement in these polarized times. Even as political polarization rises in the U.S. and around the world, if we realize the other side has something worthwhile to say, the arguments become smarter than the participants.

Read More

Leave a Comment