Principal Investigators: Mark Buntaine, Sarah Bush, Ryan Jablonski, Daniel Nielson, Paula Pickering
Country: Uganda

Registration: 20151119AA

Dates of Intervention: February – March 2016 (March 2016 elections for local sub-county chairs and councillors)

Background: Uganda is a semi-democratic country prone to problems of corruption and poor governance. Uganda has regularized elections but they are typically not judged free and fair by external monitors. Ugandan citizens evince relatively high levels of political awareness, particularly at the national level. A nationally representative survey fielded in 2012 revealed that more than 83 percent of Ugandans could, in an open-ended prompt, state the name of their MP; 70 percent could name their female representative to parliament. Nearly two thirds of Ugandans listen regularly to the radio, where news programs are frequent (Milner et al. 2014). However, research on decentralization in Uganda has revealed that many citizens lack knowledge of the basic functions of local governments, which has hindered citizens’ ability to participate in local government decision-making and hold local governments accountable (Natamba et al. 2010, p. 16). This study is focused at the local level, to understand how different types of information delivered by mobile phones might impact political actions and vote choice. This research is informed by studies about the effects on voting of information regarding local politicians’ corrupt practices (Ferraz and Finan 2008) and on citizens’ expectations that local politicians provide quality public services (Habyarimana et al. 2009).

Research Design: The common treatment arm evaluates the effect of local budget management on voter attitudes and behavior. Voters at polling stations that are assigned to the common treatment arm will receive official information produced by the Auditor General about the percentage of local council budgets that are not accounted for. The alternative treatment arm will 1) present voters with information about how their public services compare to other districts, based on field audits of road quality, water quality, and solid waste services; 2) evaluate whether the effects of information differ when voters are informed that foreign donors have funded projects in their districts. All of the information treatments will be disseminated in SMS format via mobile phones.

Hypotheses:
  • Information on the comparative quality of public services will influence voting because information is directly linked to voters’ well-being.
  • The subjects will use the provided information when they participate in politics in other ways, including through discussing politicians’ performance, contacting their councillors, and signing petitions in favor of public policy programs.
  • Favorable information about budget audits and public services should increase votes for the incumbent and unfavorable information will increase opposition votes.
  • When budget and public-services audits reveal better than average performance, voters will have a more positive perception of the performance of their elected representatives than when audits reveal lower than average performance.
  • Informational treatments may have less of an effect in locations that already have high-quality incumbents, few budget irregularities, and good public services – which are also likely to be locations that are better governed, have more informed citizens, and have other distinct characteristics – than in locations that do not (Banerjee et al. 2011).