Yurluey Mendoza, 33, a guerrilla fighter in the Revolutionary Armed Forces of Colombia (FARC) who joined the rebels at age 14, was photographed during the 10th Conference of the FARC in the Sabanas del Yari, Colombia. (Joao Pina for The Washington Post)

On Oct. 2, Colombians will vote in a referendum to decide the fate of the peace agreement with the Revolutionary Armed Forces of Colombia (FARC). This peace accord would conclude 52 years of brutal civil war and four years of determined negotiations.

If the “yes” campaign wins, the FARC will have 180 days to disarm and demobilize. Many observers worry about what will then happen to the ex-combatants. Will the former FARC soldiers become law-abiding civilians — or will large numbers turn to crime?

Crime is the choice for many former rebels around the world. Why?

These fighters have spent their lives specializing in violence and illegal activities. After civil conflicts, many turn to that line of work — and become a serious threat to peace and political order.

Why do so many ex-combatants turn to crime? Theories abound. Some posit that it is because they can’t find jobs in a legal economy plagued by high rates of unemployment. Others suggest it results from psychological factors such as post-traumatic stress syndrome.

Our new research strongly supports an alternative theory: Former fighters who stay close to one another — and, especially, to their former commanders — are at a significantly higher risk of turning to crime.

Here’s how we investigated this question

In 2013-2014, we measured the percentage of ex-combatants that committed crimes in Colombia using two instruments: first, confidential administrative data from the Colombian attorney general’s office, and second, a survey of 1,158 ex-paramilitaries and guerrillas in 47 municipalities of Colombia. For the survey, we randomly sampled these ex-combatants both in and out of prison with support from the Colombian government and the Organization of American States.

The survey is an important complement to the administrative data because many who resort to crime are not detected by the criminal justice system. We used sensitive question techniques and a self-administered survey to elicit honest responses about engagement in crime, generating a wealth of information on criminal activity.

For details on the methodology, see our research paper.

This is what we found.

We found a gap between the two sources of data. The attorney general’s data suggest that 11 percent of ex-combatants get involved in crime. But when we put the attorney general’s data together with our survey data, we found that 24 percent of ex-combatants had committed crimes.

We didn’t find any evidence that ex-combatants turned to crime because they couldn’t find work. Nor did depression or trauma appear to tip someone into behaving criminally. What we did find, quite strongly, is that when former fighters stayed closely tied with their former comrades and commanders, they were more likely to commit any type of crime, and violent crime in particular. It didn’t matter whether we were studying ex-paramilitary fighters or ex-guerrillas: What made the difference between law-abiding civilians and criminals was whether they stayed closely tied to their brigade.

Here’s why. Wartime networks work like gang networks: They exert social pressure to go along with what everyone else is doing; they make criminal behavior seem normal and acceptable; and they bring status, esteem and a sense of belonging.

And this is key: The top and mid-ranking commanders are the ones who lead the rest to crime.

What do our findings mean for demobilizing FARC rebels?

If the referendum passes, the FARC rebels will hand over their weapons and attempt to re-enter society. Yes, they need jobs, income and psychological counseling, all of which are cornerstones of disarmament, demobilization and reintegration (DDR) programs. But while that may be necessary, it’s not going to be sufficient. Those working to reintegrate the FARC will also want to consider the ex-rebels’ wartime networks of companions, friends and superiors.

These ties can’t just be dismantled, of course. The FARC leadership wouldn’t accept that. Nor would ex-fighters so easily relinquish their loyalties and communities. Trying to dissolve these connections would risk remilitarization.

But the networks could be redirected by persuading the commanders to have confidence in the government. That might include re-socializing commanders into legal norms of society and bringing commanders into legitimate political and social leadership spots. They would then pull their followers along with them into licit civilian lives.

Of course, that’s an enormous project. It involves changing the mind-set of rebels who have spent decades in the jungle and who profoundly mistrust the government. But the alternative is to leave a practiced military force at large to become a criminal gang.

Sarah Zukerman Daly is assistant professor of political science at University of Notre Dame.

Laura Paler is assistant professor of political science at University of Pittsburgh.

Cyrus Samii is assistant professor of politics at New York University.