With convention season now in full swing, national polling shows a close contest between Hillary Clinton and Donald Trump. The recent tightening of the presidential race highlights a recurring theme of this year’s election cycle: the surprising willingness of Republicans to come together behind Donald Trump.

This willingness presents a puzzle to a nation of pundits who have underestimated Trump’s viability for more than a year: How should we understand the coalescence of Republican support around a candidate recently viewed as unelectable by most analysts? Conversely, why has it taken so long for Clinton to unite Democrats despite significant advantages in resources, organization, and experience?

Several studies published in recent years in the field of political psychology suggest a new answer to these questions, namely that conservatives — the ideological base of the Republican Party — may simply be more inclined to form political consensus than liberals are. It’s a tendency that may help to explain our surprisingly tight 2016 election.

This research (by one of us, Chadly Stern, and collaborators) proposes that the conservative consensus advantage arises in part from a fundamental drive shared by all humans: the motivation to “share reality.” Past work finds that we all possess a motivation to build and maintain shared perceptions of reality with the people around us. You can think of the motivation to share reality as a sort of social glue, without which we would be unable to communicate, work in teams and achieve shared goals.

But although everyone possesses this motivation to some extent, conservatives tend to have more of it than liberals. For example, one study found in an online sample of Americans that conservatives were more likely to report that it is important to them to “see the world in a similar way” as people with whom they generally share the same beliefs. In a separate sample, participants completed a scale assessing the motivation to feel unique. Liberals scored higher than conservatives, being more likely to agree with statements such as “it’s hard to work under strict rules and regulations” and “in group activities, I am somewhat of a nonconformist.” You can easily see how these motivations could affect political organizing. All else being equal, if you had to win an election, would you prefer that your supporters deeply value commonality or uniqueness?

In addition, because people often view the world as they would like it to be -- the tendency toward “motivated reasoning” -- conservatives are more likely to perceive commonality in their ranks, where liberals view greater diversity in theirs. These perceptions don’t simply reflect reality, they exceed it. In a separate series of studies, Stern and his colleagues recruited a sample of liberal, moderate and conservative Americans online. Participants reported their attitudes on a variety of different political issues (e.g., attitudes toward abortion) and non-political issues (e.g., attitudes toward coffee), and then estimated how widely their attitudes were shared with politically like-minded others (for example, conservatives estimated how widely their attitudes were shared by other conservatives). The researchers compared these estimates with the actual percentage of like-minded others who shared participants’ own attitudes.

The figure below gives the results of these studies, showing how accurate liberals, moderates, and conservatives were in their perceptions of within-group similarity. For both the political and non-political issues, moderates and conservatives tended to overestimate how much they shared beliefs with other moderates and conservatives, whereas liberals underestimated how much they shared beliefs with other liberals. In other words, conservatives perceived greater commonality in their ranks than actually existed, while liberals tended to perceive greater diversity among themselves than was actually there.

But why should these misperceptions of others’ views matter? It turns out that perceiving consensus in one's political ranks can fuel organizing. In one series of studies, participants were asked to make non-political judgments (like trying to guess individuals’ birth month from photos of their faces) and then estimate the extent to which politically like-minded others would make similar judgments. After this, participants reported how confident they were that their preferred political party would be successful in the 2012 election.  Overall, the more people perceived that like-minded others had made judgments similar to their own, the more they believed that their party would succeed in the 2012 election. Feeling confident that one’s group can achieve its goals is important in part because it can propel individuals to engage in behaviors that benefit the group. Consistent with this, participants in the study who felt their party was likely to be successful in the 2012 election were also more likely to say that they intended to vote themselves. These findings highlight the important consequences of perceived consensus, showing how it can create confidence in one’s group and, in turn, a motivation to contribute to it.

Although a strong motivation to share reality might sound like a pattern of insularity to some, the ability to come together efficiently and completely behind a shared worldview is likely to be helpful for building a successful political campaign. Conversely, the liberal tendency to diverge from the views of other liberals may risk undermining solidarity and place a greater burden on politicians to find common ground among individuals who are often psychologically disinclined to do so. Indeed, if you are a liberal you may be skeptical of the very notion that falling in line with the views of others is somehow a virtue.

Of course there are additional factors that may help explain a conservative advantage in consensus-building. Although much is made of the demographically diverse Democratic base providing an advantage in raw numbers, this also means that the party’s politicians must form coalitions across significant demographic lines, uniting people with very different backgrounds and interests. Additionally, where there is just one status quo for conservatives to rally behind, there are infinite ways that liberals might change it, making the wrangling of liberals behind a single candidate a little like the proverbial herding of cats.

It probably shouldn't surprise us that there are psychological challenges to forging political coalitions on the left. After all, a drive to zig when others zag may be part of how liberals come to pursue change in the first place. The same psychological drive that leads people to favor changing the status quo also makes them reluctant to join large political coalitions. For conservatives hoping for success in November, the takeaway involves following that gut instinct to cohere with like-minded others. But for liberals it means something very different: thinking strategically about when it is time to dissent and, though it may feel wrong, when it is time to conform.

Robb Willer is a professor of sociology, psychology, and organizational behavior at Stanford University.

Chadly Stern is an assistant professor of psychology at the University of Illinois at Urbana-Champaign.