Subscribe to E-Bulletin Donate to CEGA

Using Smartphones to Reduce Electoral Fraud in Afghanistan and Uganda

Development Challenge

Free and fair elections are central to democracy and provide a vital means of empowering citizens to hold politicians accountable.  Election fraud commonly undermines this critical function in many young democracies largely due to weak electoral institutions, often characterized by limited oversight of election officials.  The opportunity to gain power and money incentivizes politicians to bribe election officials to illegally alter vote totals.  This research project evaluates the use of smartphones as a novel, low-cost electoral monitoring technology as a means to improve the fairness and legitimacy of elections in developing countries. 

Context

The first iteration of this project took place during the September 2010 parliamentary elections in Afghanistan.  The electoral environment was characterized by a large number of candidates competing with close margins of victory, which created real opportunities to shift electoral outcomes by tampering with vote counts.  Electoral institutions were weak and characterized by limited accountability of election officials who may choose to assist corrupt candidates.  Finally, there was a history of candidates activating patronage networks to manipulate vote counts.  

The researchers replicated the study on a broader scale during the February 2011 Ugandan presidential elections.   

Evaluation Strategy

The experimental sample for the Afghanistan study included 471 polling centers (7.8 percent of all polling centers operating on Election Day) in 19 of the country’s 34 provincial capitals.  238 of these polling stations were assigned to receive the intervention.  The treatment constituted a letter sent by hand to the 238 Polling Center Managers during polling hours on Election Day, which announced that researchers would photograph election returns forms the following day.  The researchers refer to this technology as “photo quick count”.  The letter also explained that the purpose of photo quick count is to document any discrepancies between photographed returns forms at the polling center and results certified by Afghanistan’s Independent Election Commission.  Polling Center Managers were asked to sign the letter as acknowledgement of receipt. 

Researchers then visited all treatment polling stations on the morning after the election, using smartphones to take picture of the polling center election return forms.  Taking independent photographic records of election returns forms and separating them from the electoral chain of custody provides a means of detecting returns form manipulation.

In the Ugandan replication, researchers expanded the sample size to 1,000 polling stations across the country.  The treatment again comprised letters sent to half the sample, reminding polling center managers about the requirement to post election return forms and that researchers would photograph these forms.  This time, however, the smartphones were equipped with an application that allowed field researchers to immediately submit the data for rapid analysis of whether number-patterns revealed electoral fraud. 

Results and Policy Implications

In Afghanistan, the photo quick count monitoring technology reduced the incidence of theft or damaging of election materials at polling centers from 18.9 to 7.9 percent (a reduction of about 60 percent) and reduced the number of votes cast for powerful candidates by 25 percent.  In Uganda, the researchers also found a significant decrease in vote rigging among treatment polling centers compared with those that did not receive the letter or have photographs taken of their results forms. 

The researchers argue that ICT-based corruption monitoring technologies such as photo quick counts represent a potentially highly cost-effective method to reduce electoral fraud.  They suggest future research collect data on the changes in bribe prices to election officials as a test of the core mechanism behind the decrease in electoral fraud.  In the medium- and long-term, research should investigate the effects of reduced electoral fraud of governance, social welfare, and citizen attitudes toward government.  Finally, the researchers state that innovative uses of ICT provide a new direction for political economy and development research. 

Timeline

2010 - 2011