Colombians in Bogota react Sunday after learning about the rejection of a peace deal with the FARC following a plebiscite. (Nicolo Filippo Rosso/Bloomberg News)

On Sunday, Colombians narrowly defeated a peace agreement that would have ended more than 50 years of war between the government and the Revolutionary Armed Forces of Colombia (FARC). The outcome was largely unexpected, both at home and abroad. Pollsters predicted that supporters of the deal would win. Many public figures supported peace.

But one prominent voice opposed the agreement: Colombia’s former two-term president Álvaro Uribe. A populist, far-right leader who supported a strong-arm approach to ending the conflict, Uribe argued that the deal was too lenient on the rebels and insisted that a tougher agreement was possible.

Uribe’s active campaigning against the peace deal and his success in helping to defeat it has helped secure his position as one of Colombia’s most important political figures.

One source of Uribe’s importance is his impassioned group of followers. These supporters, called uribistas, constitute a growing political movement with serious political and social implications for Colombia, as we can see in the defeat of the peace agreement.

Here’s how we know that Uribe is leading a movement

We took a nationally representative survey in May 2015 with 1,000 randomly selected participants. Measuring only in the conventional way, we found that a minority of Colombians openly affiliate with any party. The “uribista” party, Democratic Center (CD in Spanish), has the largest proportion of those who said they sympathized with a political party. But this represented only 18.5 percent of the electorate. (Consider that in the United States, 32 percent of citizens are currently affiliated with with the Democratic Party.) The Liberal Party followed, with 13.3 percent of the electorate. The incumbent party, called the U Party, claimed 7.2 percent. But a large plurality of voters (45.3 percent) did not identify with any party.

In other words, Colombians today are largely apartisan. But we suspected that many uribismo supporters do not openly affiliate with CD. CD was founded in 2013 and has participated in only one general election and one subnational election. What’s more, in general, Colombians reject and mistrust political parties. This is driven by a disconnect between citizen demands and partisan elites’ self-interests. That mistrust may make Colombians reluctant to openly endorse or participate in any party, even if they identify with what uribismo stands for — which includes a military defeat of the FARC and imprisonment of its leaders, as well as a conservative view of family values.

To test our theory, we asked respondents if they would vote for an uribista candidate for National Congress; for regional government; and for local government. This way we could measure support for (and opposition toward) the movement (uribismo) rather than that of the party created to sustain that movement (CD).

Survey respondents could answer definitely yes, probably yes, probably no, definitely not or not sure.

Those who said they would “definitely” vote for an uribista candidate for all three public positions were categorized as “hard-core uribistas,” or what Colombians have come to call pure-blooded uribistas. These made up 23.4 percent of the sample, or almost five points more than the CD sympathizers (18.5 percent).

Those who responded that they would “probably vote” for an uribista candidate for all three public positions (or any combination of “definitely yes” and “probably yes”) were cataloged as “leaning uribistas.” These made up 17.1 percent of the sample.

Finally, those who answered they would “definitely not” vote for an uribista candidate for all three public positions were defined as “anti-uribistas.” These made up 31.7 percent of the sample.

Firm opponents are rarely considered when studying partisanship. But our data suggest that, in less institutionalized and/or polarized situations, negative partisans can affect their fellow voters and therefore who wins elections. By our measure, hard-core and leaning supporters outnumbered those who opposed uribismo.

We also asked how individuals ranked themselves ideologically, on a scale where 1 is extreme left and 7 extreme right. That’s how we learned that hard-core uribistas perceived themselves as further to the right (5.27) than CD sympathizers (5.17) did. Anti-uribistas were further to the left (4.13), although they nearly aligned with the median voter (4.37). Leaning uribistas fell between hard-core followers and the median voter, at 4.86. These results were statistically significant.

Here’s why we conclude that ‘uribistas’ defeated the peace agreement

What do these identities mean at the ballot box? Are these points of view why the peace deal was voted down? Only 37 percent of the electorate actually voted on Oct. 2. If hard-core uribistas made up only 23.4 percent of the electorate, to defeat the deal, they’d have had to vote at a higher rate than other Colombians. Did they?

We believe that is quite plausible. In our 2015 survey, we asked:

Some people are willing to participate in a diverse set of activities to promote their political beliefs. I am going to list a series of activities. Tell me how acceptable or not each of these activities is to you in promoting your own political beliefs. Please use a scale of 1 to 7, where 1 is not at all acceptable and 7 is very acceptable. [The activities included are: Participate in a political party meeting; Convince others to vote like you; Participate in a legal protest; Invade or damage private property; Take over a building; Block streets or highways].

Hard-core uribistas thought that participating in party meetings and persuading others to vote were more acceptable than did other Colombians. On average, core uribistas ranked participating in party meetings at 4.53 and persuading others to vote at 4.34. Anti-uribistas, by contrast, ranked those at 2.78 and 2.42, respectively.

From this, we conclude that hard-core uribistas are more politically active than their fellow citizens. And in a short campaign like that of the peace deal plebiscite – which lasted only five weeks – that matters. Those who campaigned for a “yes” vote came from a group that that’s less likely to be actively involved in parties and political persuasion. Uribistas were ready to organize, argue, and act for their beliefs. That may have pushed the vote over to “No.”

Jennifer Cyr is assistant professor of political science and Latin American studies at the University of Arizona. Carlos Meléndez is a post-doctoral researcher at Universidad Diego Portales in Chile.