The Wrong Snowball: Reintegration Signals and Demobilization in Colombia

Question and contribution.

Do reintegration programs persuade fighters beyond those enrolled? In Colombia, the most visible reintegration milestone is an ex-combatant opening a state-funded business. Using municipality-quarter administrative data and a stacked difference-in-differences design over 2005-2016, we find these openings were followed by fewer subsequent demobilizations nearby — roughly 734 fewer over three years, or 5.1 percent of observed FARC demobilizations. Two explanations fit the evidence: the businesses often struggled in plain view, and former fighters remained targets for intimidation. The visibility that demobilization policy counts on to attract the next defector can just as easily warn them off.

Abstract.

Reintegration programs can influence demobilization behavior beyond enrolled participants by transmitting public signals about post-exit life. In Colombia, one of the most visible milestones in reintegration is the state-funded opening of a business: after laying down arms, ex-combatants who complete a transition that often lasts years become eligible for seed capital, so each opening signals whether the state is delivering on a central promise. Whether such signals encourage or discourage future defections is theoretically ambiguous: successful openings may generate a positive demonstration effect, but fragile post-opening trajectories and intimidation risk may reverse it. Using municipality-quarter administrative data and a stacked difference-in-differences design, we find that business openings by ex-combatants are followed by lower demobilization counts in the same area, implying about 734 fewer demobilizations over the next three years, or 5.1 percent of observed FARC demobilizations in the 2005-2016 study period. Quantitative heterogeneity and coded focus-group evidence point to two complementary mechanisms: an expectation gap between the program's formal promise and observed post-opening fragility, and persistent intimidation risk that raises the perceived security cost of exit. Effects are largest for commerce and service businesses — the most publicly visible sector types — and in municipalities with higher ex-combatant death rates, consistent with a mechanism that operates through the local observability of post-exit trajectories and perceived security risk. The results point to a broader lesson for ongoing and future DDR processes: reintegration generates a positive snowball only when implementation quality and post-exit security are sufficient to shift beliefs favorably; otherwise, the same visibility on which DDR policy relies can backfire.

How the project is framed.

BibTeX.

@unpublished{aparicio2025wrong,
  title = {The Wrong Snowball: Reintegration Signals and Demobilization in Colombia},
  author = {Verónica Abril and Juan P. Aparicio},
  note = {Working paper},
  year = {2025},
  url = {/assets/docs/Aparicio_Abril_Wrong_Snowball.pdf}
}