With Donald Trump playing the old hits to an enthusiastic Conservative Political Action Conference (CPAC) crowd and congressional Republicans making post-election pilgrimages to Mar-a-Lago to pay respect, many observers wonder when the former president’s star will fade within the party. While political celebrity typically disappears quickly, Trump’s support from his most fervent disciples doesn’t appear to be vanishing any time soon. In fact, supporting Trump has become a substantial part of how they publicly define themselves.
This phenomenon — incorporating political allegiance into individual people’s personal identity — has been measurably increasing in recent years. Since 2015, we have been tracking the Twitter bios of millions of its users. We see when someone adds words, and when someone takes them away. That way we can witness millions of identities shift almost in real time.
Our findings, published this month in the Journal of Social Computing, suggest that at least on Twitter, Americans are treating partisan ideology as more central to their personal identities. Perhaps most striking for a historically religious country: Starting in 2018, Americans’ Twitter bios are more likely to include their politics than their faith.
How we used social media bios to measure personal identity
Social scientists are very interested in personal identity — how a person fundamentally defines themselves and wishes to be seen. But is notoriously difficult to measure. We think that social media gives us a handy tool for doing so: the online “bio,” wherein a person writes a brief self-description for others to read. These bios are statements of identity, in which someone summarizes themselves as they wish others to see them.
We used an automated process that randomly takes data “snapshots” of publicly visible Twitter bios at certain intervals and then catalogues the words that appear in them. This way we can analyze trends in how people define themselves and how those definitions change.
We randomly selected a large number of public tweets at certain intervals and added those users to our database. Our automated program captures many of the same users more than once. For those users, we can see how their bios change over the years. In all, we’ve collected bio snapshots on more than 20 million unique Twitter users. For about 1.5 million, we have a bio snapshot of each year between 2015 and 2020.
Next, we can analyze this trove of data by searching bios for certain categories of words. For example, we look for terms that indicate sports affinity, such as football, athlete or MMA; words that suggest an identity as an artist, like dancer, musician, performer; and words of popular religious affiliation, such as atheist, Christian, Catholic or Muslim. And we count common political identifiers, like Republican, liberal or progressive.
Biden may be getting rid of the Authorizations for Use of Military Force. Whoa!
On Twitter, politics is increasing as a notable point in Americans’ identities
As you can see in the figure below, in 2020, Twitter users were about 15 percent more likely to refer to Trump or “MAGA” in their bio than to mention Jesus or Christianity. This may not surprise commentators who have compared the former president’s following to that of a religious movement.
But that’s just one subset in the overall increase in identifying one’s Twitter avatar with a political position. All the political words we thought to measure, ranging from “anarchist” to “woke,” have been mentioned more often, inching up in usage rate every year. As you can see in the figure below, in 2015, the bio of a random tweet would be about 50 percent more likely to convey a religious affiliation than a political affiliation. Five years later, that was has reversed. Increasingly, Twitter users don’t just discuss politics, they are politics.
It’s important to note that relatively few Twitter users convey affiliation with politics or religion in their bios. In 2020, such bios constituted just 2 percent and 1 percent of the year’s sample, respectively. Certain other categories are still more likely to appear in a bio, like sports fandom or arts appreciation (each appearing in about 3.5 percent of 2020 bios), but the gap is narrowing — those bios are steadily decreasing in rate, while the political bios become more common every year.
And, of course, the typical Twitter user is not a perfect stand-in for the typical American. Users of the site are somewhat younger, more educated, higher-earning and more politically engaged than the average citizen as a whole, according to the Pew Research Center. Twitter users may be more likely than non-tweeters to adopt political identities. But 20 million people is a fairly large sample size, and there is ample evidence that political activity and animus are rising among the general public more broadly.
Are these new identities steps on the way to a more tribal United States?
Twitter users’ identity shift toward politics is notable, because adult identities are very resistant to change. To abandon an aspect of one’s identity is to discard a fundamental part of oneself — a painful process, sometimes called an “identity crisis,” which most people go to great lengths to avoid. Once we begin to define ourselves by something, we organize our lives around it and defend it fiercely against threats. If your position as “pro-life” or “pro-choice” becomes an identity rather than a mere policy preference, an attack on that position can be felt as an assault on you as a person.
Americans might find it shocking, but in Europe, former presidents do get sentenced to prison.
Putting politics as a central aspect of identity represents a rapid shift in how Americans think of themselves. Despite the benefits of a politically engaged populace, there is reason for concern. The health of a pluralistic society depends on the ability to communicate and work together across differences.
When those differences become written into a person’s very identity, however, conditions are ripe for the dysfunctions of tribalism, where citizens see themselves as belonging to a distinct and rigid cultural group in a zero-sum competition with rival groups for power and resources. While the United States is not there yet, Americans may wish to consider how to remain a cohesive society across differences.
Don’t miss any of TMC’s smart analysis! Sign up for our newsletter.
Nick Rogers is a PhD candidate in sociology at Stony Brook University, and teaches courses in statistics and research methodology at New York University’s Center for Global Affairs.
Jason J. Jones is an associate professor in the department of sociology and the Institute for Advanced Computational Science at Stony Brook University.