Votes and Violence: Evidence from a Field Experiment in Nigeria

Paul Collier, Pedro Vicente


This paper investigates the 2007 Nigerian election violence based on a nationwide field experiment based on anti-violence campaigning. The campaign was randomised across neighbourhoods and villages of six states of Nigeria. These states represent the main socio-economic regions of the country.

The campaign was conducted in half of those locations before the 2007 elections by a major international NGO, ActionAid, which specialises in community participatory development. It included town meetings, popular theatre, and the distribution of campaign material, standardised across all locations. It was aimed at empowering citizens to counteract local violence. Its activities were designed to reduce the costs of protest and collective action through electoral participation. In a complementary manner, it appealed to “voting against violent politicians.” Measurements were based on representative surveys, interviews and a compilation of violence-event diaries by independent local journalists.

The study finds that the campaign decreased violence perceptions and increased empowerment to counteract violence. It observes a rise in voter turnout, and infers that the intimidation was dissociated from incumbents. These effects are accompanied by a reduction in the intensity of actual violence, as measured by journalists.

Key Findings:

  • The anti-violence campaign decreased the intensity of real violent events, implying that the behaviour of politicians who use intimidation as an electoral strategy was influenced. The paper suggests that the campaign worked through increased perceptions of local safety and empowerment of the population. It also led to boosted voter participation and electoral penalisation of candidates perceived to use intimidation (violence was dissociated from incumbents).
  • Relatively insignificant but targeted events can indeed mobilise citizens to collective action. Specifically, more participation at the polls together with improved security and empowerment of the population may be mutually reinforcing, in a context in which violence is associated with small political groups. Anti-violence campaigns may then be an especially effective form of voter education, working mainly as a coordination mechanism, and relatively undemanding on the amount of information that is passed to voters.


  • Future empirical research should not lose sight of the likely joint determination of the different electoral strategies of politicians. These may include other types of illicit behaviour such as vote-miscounting and vote-buying.
  • An anti-violence campaign cannot be the sole remedy for problematic elections; attention should be devoted to political accountability and to all illicit strategies in an integrated manner. It is in this context that voter education, broadly construed, and electoral observation may be invaluable policy tools for the improvement of elections and democracy in the developing world.


Collier, P. & Vicente, P. (2014). Votes and Violence: Evidence from a Field Experiment in Nigeria.