How do international, armed, humanitarian interventions shape gang violence, and by extension, influence civilian mobility on the ground? Focusing on the U.N. Multinational Security Support (MSS) Mission in Haiti (2023–present), we exploit a daily panel of commune-level data on gang violence, using data from ACLED, Haitian press reports, and mobile-phone mobility records, using estimates derived by Flowminder from Digicel call detail records, to trace the impact of five exogenous information shocks on the mission's likelihood and timing. We find that as the intervention appeared more imminent, gangs reduced inter-gang clashes and forged tactical alliances, redirected violence toward the state, and escalated remote attacks on infrastructure. These anticipatory shifts had immediate civilian consequences, as heightening violence reduced both intra- and inter-communal movement, revealing how violence shapes everyday mobility. By integrating high-frequency measures of both violence and movement, this study contributes to research on the consequences of external security governance, showing that the pre-deployment phase of intervention is a politically charged period in which armed actors and civilians adapt in ways that can entrench insecurity before the first foreign boots touch the ground. 

A large body of literature documents a positive relationship between legislative representation and redistribution, particularly in the United States. Standard legislative bargaining models further reinforce this association. Yet both theory and empirical findings rely on two critical assumptions: that legislatures possess sufficient budgetary authority and that legislators face strong electoral incentives. This paper shows that, in a context where those conditions do not hold, the link between representation and redistribution disappears. We study the effect of representation on intergovernmental transfers in Chile, a unitary country with a Congress with limited budgetary power, elected in multi-member districts. We exploit two quasi-natural experiments - the authoritarian origins of the electoral system and its recent reform - to provide causal identification. Across both sources of exogenous variation in representation, we find no significant effects on redistribution. Our findings highlight the mediating role of institutions in shaping the consequences of legislative malapportionment.