Other Authors: Andrew Beath, Georgy Egorov, Ruben Enikolopov
Geographical Region: Asia
Do electoral rules impact how voters choose between more competent candidates and candidates with policy preferences closer to their own?
Preparer: Damaris Colhoun, All pictures by: Andrew Beath
Voters across the world are often confronted with the same conundrum: vote for the most qualified candidate or vote for the candidate whose positions closely match their own. A single, hot-button issue can decide an election and the candidate who sides with the majority may not be the candidate most capable of designing effective policy. As a result, the efficacy of representative bodies often depends on how voters value competency. But can electoral rules actually mediate this trade-off and encourage voters to elect the most qualified candidates?
In order to answer this question, the authors of this study turned to Afghanistan and its National Solidarity Program (NSP). Devised in 2002 by the Afghan government and funded by bilateral and multinational donors, NSP has disbursed more than $1.1 billion to 32,00 villages across Afghanistan’s 34 provinces, making it the largest development program in the country. The goal of NSP is to deliver services and infrastructure to the country’s rural population, specifically the construction and rehabilitation of infrastructure, and the development of human capital, such as training and literacy courses. As per NSP’s mandate, the funds for these projects are allocated and monitored by gender-balanced village development councils. Therefore, NSP’s first step to achieving its development goals is development council elections.
Villages were divided into geographically contiguous districts. Development council elections were conducted by secret ballot with votes cast by writing in names. Every village resident, age eighteen or older, male or female, was both eligible to vote and to be elected to the development council. The only variable from district to district: the electoral rules. Determined primarily by district size, representatives were chosen either by “district election” or “at-large election.” District elections, the election process for the U.S. House of Representatives and the British House of Commons, are “single-winner systems” with voters casting a ballot for a single candidate to represent a single constituency. At-large elections, commonly used in European parliamentary elections, broaden the scope of the constituency, allowing voters to elect multiple candidates to oversee a broader area.
By employing a unique field experiment to compare development council election results and NSP project effectiveness, researchers saw the opportunity to measure whether different electoral rules impact how a voter chooses between a candidate with polarized policy positions and a more competent candidate.
The sample for the study was composed of 250 villages–each with an average population of roughly 1,000 people–selected from ten districts spanning northern, northeastern, eastern, central, and western Afghanistan (southern areas were excluded due to security concerns). Half of the villages was randomly assigned to hold district elections and the other half to hold at-large elections. Under district elections, the village was split into geographically-defined districts and each villager could only vote for a single candidate residing in the same district. Under at-large elections, each resident could cast three votes for any three candidates, with council members elected based on the total number of votes garnered across the whole village.
Prior to elections, researchers surveyed 2,387 randomly-selected, head-of-household males, asking them to indicate their preferences from a list of fifteen potential projects commonly funded by NSP. Development council elections were then held between October 2007 and May 2008 and projects were selected soon thereafter. As projects were implemented and completed, researchers conducted “midline” and “endline” surveys. These surveys measured the effectiveness of the NSP projects, improvements in general economic welfare, and increases in support for local leaders and the democratic process. Roughly 2,500 men and 2,500 women were asked a series of questions ranging from how often drinking water was thought to be contaminated to annual household income and whether they agreed with decisions made by local leaders in the past year.
The competency of elected development council members was measured by their educational background–whether they finished high school, middle school, or primary school. Because educational variation was so low for female council members–only 0.8 percent finished high school and less than 10 percent finished primary school–researchers focused their analysis on only male council members. Under district elections, 7% of male council members had finished high school. This compares to 11% elected under at-large elections, a statistically significant 57% increase.
But does a more educated development council lead to more effective governance? Researchers’ results suggest that the answer is yes. On average, projects started and finished more than a month earlier in villages under at-large elections. The midline and endline surveys also paint a vivid picture. Females in at-large election villages responded to the midline survey with more positive perceptions of the economic situation, attributing these changes to their local leaders. For the endline survey, respondents in at-large election villages reported lower levels of diarrhea among children and significant increases in harvest revenues. Overall, the evidence to suggest projects were more effectively implemented in at-large villages is weak, but at the very least, the faster implementation of projects did not come with sacrificed project quality.
In the end, researchers determined electoral rules do indeed condition voting behavior. Villagers under district elections voted more strategically, choosing candidates with extreme policy positions, hoping they would bargain more virulently for a voter’s specific interests. Voters in at-large elections chose candidates based upon a larger set of preferences. This led to greater overall competency amongst elected development council members and as a result, more effective governance overall.