1. “Ethnicity, Information, and Strategic Voting in Multi-Ethnic Democracies: Evidence from Kenya,” Electoral Studies, 2016, 44: 351-361 (with Jeremy Horowitz)

This paper explores the conditions under which voters in emerging democracies support non-viable candidates. We argue that cognitive biases and the geographic clustering of minor-party supporters in ethno-political enclaves lead to misperceptions about the electoral prospects of minor-party candidates, weakening strategic defections both among co-ethnic and non-co-ethnic supporters. We explore these arguments using original survey data from Kenya’s 2007 presidential election, a contest that featured a minor-party candidate, Kalonzo Musyoka, who stood little chance of electoral victory. Despite this, results show that most of his supporters chose to vote for the candidate, failing to perceive that he was not a viable contender. The findings suggest that theories of political behavior in multi-ethnic settings can be enriched by drawing upon insights from the political psychology literature on belief formation.

2. “Gifts, Threats, and Perceptions of Ballot Secrecy in African Elections”  African Affairs, 2016, 115, (461): 621-645 (with Karen Ferree)

Are contingent electoral strategies, like vote buying and intimidation, viable in Africa?  No, according to recent scholarship: unlike parties in other developing regions like Latin America, African parties lack the capacity to violate ballot secrecy and force voters to stick to their end of the bargain.  Voters can therefore “defect” and vote their conscience.  We challenge this perspective. Recent Afrobarometer data show that nearly one in four Africans doubt ballot secrecy. We argue that these doubts are a deliberate product of party efforts, and show that ballot secrecy perceptions correlate with vote buying and intimidation efforts and measures of campaign intensity. Pervasive doubts about ballot secrecy, rooted in party campaign strategies, challenge the notion that African parties are too weak to implement contingent electoral strategies.   Parties can and do convince voters that their vote choices are known, opening up wide range of electoral strategies, particularly in urban areas where party capacity and community accessibility are highest.

3. Improving Electoral Integrity with Information and Communications Technology Journal of Experimental Political Science, 2016, 3, (1): 4-17 (with Michael Callen, Clark Gibson, and Danielle Jung)

Irregularities plague elections in developing democracies. The international community spends hundreds of millions of dollars on election observation, with little robust evidence that it consistently improves electoral integrity. We conducted a randomized control trial to measure the effect of an intervention to detect and deter electoral irregularities employing a nation-wide sample of polling stations in Uganda using scalable information and communications technology (ICT). In treatment stations, researchers delivered letters to polling officials stating that tallies would be photographed using smartphones and compared against official results. Compared to stations with no letters, the letters increased the frequency of posted tallies by polling center managers in compliance with the law; decreased the number of sequential digits found on tallies – a fraud indicator; and decreased the vote share for the incumbent president in some specifications. Our results demonstrate that a cost-effective citizen and ICT intervention can improve electoral integrity in emerging democracies.

 4. “Evaluating the Roles of Ethnicity and Performance in African Elections: Evidence from an Exit Poll in Kenya,” Political Research Quarterly, 2015, 68, (4): 830-842 (with Clark Gibson)

While many scholars argue that ethnicity drives voting behavior in Africa, recent quantitative work finds government performance also matters. But under what conditions do Africans use ethnicity or performance to inform their vote? We argue that the importance of ethnicity and performance is conditional on whether voters evaluate co-ethnics and incumbent candidates. We hypothesize co-ethnic voters will coordinate and form blocs, whereas non-co-ethnics are more likely to divide their support between candidates. We also hypothesize that while incumbent performance matters to all voters independent of ethnicity, citizens will forgive their co-ethnic incumbent’s poor performance. Tests using data from a nationwide exit poll we conducted during Kenya’s 2007 national election strongly support our hypotheses. Our results are robust to analyses concerning the potentially confounding relationship between ethnicity and performance.

5. “Electoral and Party System Development in Sub-Saharan Africa,” Oxford Bibliographies in Political Science, 2015, New York: Oxford University Press, (with David Backer)

6. “Institutional Corruption and Election Fraud: Evidence from a Field Experiment in Afghanistan” American Economic Review, 2015, 105 (1), pp. 354-381 (with Michael Callen) 

We investigate the relationship between political networks, weak institutions, and election fraud during the 2010 parliamentary election in Afghanistan combining: (i) data on political connections between candidates and election officials; (ii) a nationwide controlled evaluation of a novel monitoring technology; and (iii) direct measurements of aggregation fraud. We find considerable evidence of aggregation fraud in favor of connected candidates and that the announcement of a new monitoring technology reduced theft of election materials by about 60 percent and vote counts for connected candidates by about 25 percent. The results have implications for electoral competition and are potentially actionable for policymakers.

(For coverage of this research, see: Slate – The Dismal Science: A Picture of Democracy; The Economist blog: How to Save Votes; Monkey Cage blog: Digital Cameras Reduce Electoral Corruption; Foreign Policy – The AfPak Channel: Point and Shoot Elections; New Scientist: Smartphone Snaps Help Deter Election Fraud.)

7. Knowledge Without Power: International Relations Scholars and the US War in Iraq,” International Politics, 2015, 52 (1), pp. 20-44 (with Daniel Maliniak, Susan Peterson, and Michael Tierney)

In this article we present several important first steps toward understanding the role of academics in shaping US foreign policy – identifying their policy views on one of the most salient foreign policy issues of this generation, the US War in Iraq; exploring how those views differ from public opinion more generally; and assessing the extent to which scholarly opinion was reflected in the public debate. To determine how IR scholars’ views on the invasion of Iraq differed from those of the public, we compare the answers of IR scholars at US colleges and universities to those of the US public on similar opinion survey questions. To this end, we analyze data from a unique series of surveys of IR scholars conducted by the Teaching, Research, and International Policy project.

(For coverage of this research, see the Washington Post).

8. “Voting Behavior and Electoral Irregularities in Kenya’s 2013 Election,” Journal of Eastern African Studies 2014 (with Karen E. Ferree and Clark C. Gibson)

Data from a unique nationwide exit poll of 6258 voters are employed to explore two central themes of the 2013 Kenyan Election: (1) the correlates of individual vote choice; and (2) the credibility of the electoral process. The analysis reveals several striking relationships between an individual’s vote choice, personal attributes, and perceptions of the campaign and candidates. We find that the leading coalitions mostly kept their co-ethnics together, although ethnic alliances proved somewhat less certain than in the past. We find that, for the most part, voters treated Uhuru Kenyatta – not sitting Prime Minister Raila Odinga – as the incumbent. The data show that campaign issues also influenced the vote: Odinga garnered more support on issues related to constitutional implementation, corruption, and the International Criminal Court (ICC), while Kenyatta won on the economy, employment, and security. Exit poll data also reveal irregularities in the electoral process, including some evidence of inflated vote totals benefitting the Jubilee coalition and illegal administrative activities. The data, while not definitive, are highly suggestive of a deeply flawed electoral process and challenge claims that Kenyatta won a majority in the first round.

9. “Violence and Risk Preference: Experimental Evidence from Afghanistan,” American Economic Review, 2014, 104(1): pp. 123-148 (with Michael Callen, Mohammad Isaqzadeh, and Charles Sprenger)

We investigate the relationship between violence and economic risk preferences in Afghanistan combining: (i) a two-part experimental procedure identifying risk preferences, violations of Expected Utility, and specific preferences for certainty; (ii) controlled recollection of fear based on established methods from psychology; and (iii) administrative violence data from precisely geocoded military records. We document a specific preference for certainty in violation of Expected Utility. The preference for certainty, which we term a Certainty Premium, is exacerbated by the combination of violent exposure and controlled fearful recollections. The results have implications for risk taking and are potentially actionable for policymakers and marketers.

10. “Kenya’s 2013 Elections: Choosing Peace Over Democracy,” Journal of Democracy 24, 3, July 2014, pp. 140-155 (with Karuti Kanyinga, Clark Gibson, and Karen E. Ferree)

11. “Income, Occupation, and Preferences for Redistribution in the Developing World,” Studies in Comparative International Development 48, 2, June 2013, pp. 113-140 (with Stephan Haggard and Robert Kaufman)

Much of the theoretical work on preferences for redistribution begins with the influential Melzer–Richard model, which makes predictions derived both from position in the income distribution and the overall level of inequality. Our evidence, however, points to limitations on such models of distributive politics. Drawing on World Values Survey evidence on preferences for redistribution in 41 developing countries, we find that the preferences of low-income groups vary significantly depending on occupation and place of residence, union members do not hold progressive views, and inequality has limited effects on demands for redistribution and may even dampen them.

12. “Parties, Ethnicity, and Voting in African Elections,” Comparative Politics 45, 2, January 2013, pp. 127-146 (with Barak D. Hoffman)

Standard theories about elections in Africa suggest that they are little more than ethnic headcounts. Data from an exit poll conducted on Election Day in Ghana’s 2008 election callenges this view. The two main parties, the National Democratic Congress (NDC) and the New Patriotic Party (NPP), drew support from many ethnic groups; and there was little evidence of ethnic block voting. Rather, voters’ beliefs about the parties and incumbent performance were the main determinants of vote choice. Evaluations of the attributes of the NDC and NPP shaped the outcome of Ghana’s 2008 election far more than the ethnic identity of the candidates. These results hold important implications for understanding voting, parties, and government performance in multiethnic democracies.

13. “The Political Economy of Reforms in Kenya: The Post-2007 Election Violence and a New Constitution,” African Studies Review 55, 1, April 2012, pp.31-51. (with Karuti Kanyinga)

This article explores the package of “Agenda item 4” reforms undertaken by the Kenyan government in the mediation process following the 2007–8 postelection violence, including those relating to long-standing issues over constitutional revision. It situates the previous lack of reforms within Kenya’s political economy and demonstrates how political and economic interests thwarted progress and produced the postelection crisis. It also examines the more recent attempts to address reforms following the signing of the National Accord and the creation of a power-sharing government, and finds strong public support for constitutional revision. It concludes that these pressures from below, along with a realignment of political interests and institutional change from power-sharing, helped support reform.

14. “Was it Rigged? A Forensic Analysis of Vote Returns in Kenya’s 2007 Election,” in Democratic Gains and Gaps: A Study of the 2007 Kenyan General Elections, Nairobi, Kenya: Society for International Development, 2010 (with Karuti Kanyinga and David Ndii), Contact me for a copy.

This chapter addresses potential rigging in Kenya’s 2007 presidential election. We employ election forensics to measure the source and extent of fraud. We find that both the government strategically manipulated the Electoral Commission of Kenya (ECK) to rig nation-wide, and by enough to wrongly declare President Mwai Kibaki’s re-election. Specifically, large discrepancies exist within the ECK’s final results, and in particular when we compare turnout from presidential results to parliamentary and civic races. We also demonstrate anomalies using vote counts from the media, past voting patterns, and an exit poll. This chapter was published in an extensive edited volume that so far serves as the definitive account for academics, practitioners, and the public of Kenya’s 2007 election and post-election crisis.

15. “Addressing the Post-Election Violence: Micro-Level Perspectives on Transitional Justice in Kenya”, The Politics of Violence and Accountability in Kenya, Oxford Transitional Justice Research Centre, Oxford University, June 2010 (with David Backer and Joseph Lahouchuc) and other copy here

To help understand how victims of the 2007-08 episodes of post-election violence and other past political abuses in Kenya view transitional justice, we conducted a series of focus group discussions around the country in the summer 2009. This occurred against the backdrop of investigations by the International Criminal Court (ICC) and the establishment of a Truth, Justice, and Reconciliation Commission. Overall, the focus group participants were supportive of and otherwise open to a variety of transitional justice options, typically with a bottom-line focus on what is required to achieve meaningful and lasting results. The strongest sentiment favored compensation to victims of post-election violence as well as prosecution of those responsible for organizing the violence, a task that most preferred to assign to international authorities like the ICC. Progress in these respects was viewed as essential for stability and reconciliation. Yet there was considerable skepticism among the participants that these challenges would be taken up promptly and effectively and have a lasting, positive impact on the lives of victims in Kenya.

16. “The Presidential and Parliamentary Elections in Kenya, December 2007: Evidence from an Exit Poll,” Electoral Studies, Vol. 28, No. 3, 2009 (with Clark Gibson)

This paper publishes the results of a nation-wide exit poll for Kenya’s 2007 election. The e6it poll provides the only independent verification of the vote in the election. Comparing results from the survey with those published by the Electoral Commission of Kenya (ECK), it finds that the ECK wrongly declared incumbent President Mwai Kibaki the winner against his challenger Raila Odinga.



1. “The Logic of Insurgent Electoral Violence” Revise and Resubmit, American Economic Review (with Luke Condra, Andrew Shaver, and Austin Wright)

Competitive elections are essential to establishing the political legitimacy of democratizing regimes. Recognizing the symbolic value of voter participation, we argue that armed actors who lack institutional power try to undermine the state’s mandate through electoral violence. We study the logic of such violence, focusing on the crucial balance between disrupting elections and minimizing harm to potential supporters of the rebellion. We theorize when and where insurgents attack around elections. We test the observable implications of our theory using newly declassified microdata on the conflict in Afghanistan. This data allows us to track insurgent activity by hour, to within meters of attack locations. Our results demonstrate that insurgents carefully calibrate their production of violence in and around elections, and these tactics effectively depress voting. We find that insurgents significantly increase the intensity of violence on election days during the hours before (and as) voting centers open, but harm relatively fewer civilians than non-election periods. Insurgents also deploy improvised explosive devices along roads connecting voters and polling stations, but rarely bomb these roads multiple times. We introduce novel instruments for the timing and spatial distribution of attacks and show that insurgent electoral violence effectively undermines voter participation. Our results provide important insights for safeguarding at-risk elections in emerging democracies.

2. “Using Technology to Promote Participation in Emerging Democracies: VIP:Voice and the 2014 South African Election,” Under Review (with Clark Gibson, Karen Ferree, Craig McIntosh, and Danielle Jung)

How does technology shape political participation in emerging democracies? By lowering costs, technology draws new participants into politics. However, lower costs also shift the composition of participants in politically important ways, attracting more extrinsically motivated individuals and a “crowd” that is more responsive to incentives (malleable) and sensitive to costs (fragile). We illustrate these dynamics using VIP:Voice, a multi-channel information and communication technology (ICT) platform we built to encourage South African participation in the 2014 national elections. VIP:Voice allowed citizens to engage via low-tech mobile phones and high-tech social media, and  randomized incentives for participation. VIP:Voice generated engagement in over  250,000 South Africans, but saw large attrition as we switched from low to high-cost forms of engagement. Attrition was particularly large for extrinsically motivated participants. Crowds generated through ICT may therefore be large and easily generated, but comprised primarily of extrinsically motivated individuals whose participation decays with shifting costs.

3.  “The Double-Edged Sword of Mobilizing Citizens via Mobile Phone in Developing Countries,” Under Review (with Aaron Erlich, Danielle Jung, and Craig McIntosh)

New innovations in mobile technology provide an unparalleled opportunity for researchers and organizations to scale communications with citizens in the developing world, but bring new challenges in terms of how to generate and retain engaged users. We report on our experience building a bi-directional multi-channel mobile phone platform to engage citizens in South Africa’s 2014 presidential election. Specifically, we deployed the “VIP:Voice” platform at national scale to conduct opinion polling, to allow citizens to report on political activity, and to engage citizen monitors for polling stations on election day. Our platform operated across multiple device types, from flip-phones to Twitter, and consequently provides critical lessons on the most effective means of gathering and disseminating a rich variety of data depending on the user’s phone type. We compare different means of obtaining location in the absence of GPS, and show how different formats for soliciting and entering data generated very differential response rates. Our paper illustrates a number of concrete ways in which platform development driven by smartphone logic does not translate easily for users of more basic mobile phones, including whether questions are presented passively in a menu or pushed to a user’s phone, and the format in which user data is entered. This paper is intended to provide actionable guidance for researchers and organizations deploying ICT platforms to interact with citizen users at a national or cross-national scale in international development.

4. “Elections and Government Legitimacy in Afghanistan,” NBER Working Paper 19949, Under Review (with Eli Berman, Michael Callen, Clark Gibson, and Arman Razee)

International development agencies invest heavily in institution building in fragile states, including expensive interventions to support democratic elections. Yet little evidence exists on whether elections enhance the domestic legitimacy of governments. Using the random assignment of an innovative election fraud-reducing intervention in Afghanistan, we find that decreasing electoral misconduct improves multiple survey measures of attitudes toward government, including: (1) whether Afghanistan is a democracy; (2) whether the police should resolve disputes; (3) whether members of parliament provide services; and (4) willingness to report insurgent behavior to security forces. Moreover, these effects are strongest within the subsample of respondents who were not aware of the fairness-enhancing treatment, leading us to conclude that legitimacy was increased by perceptions of electoral fairness and efficacy.

(For coverage of this research, see Freakonomics; the Center for Effective Global Action; and the Washington Post)

5. “Damaging Democracy? Security Provision and Turnout in Afghan Elections” Under Review (with Luke Condra, Michael Callen, Radha Iyengar, and Jacob Shapiro)

In emerging democracies, elections are encouraged as a route to democratization. However, not only does violence often threaten these elections, but citizens often view as corrupt the security forces deployed to combat violence. We examine the effects of such security provision. In Afghanistan’s 2010 parliamentary election, polling centers with similar histories of pre-election violence unintentionally received different deployments of the Afghan National Police, enabling identification of police’s effects on turnout. Using data from the universe of polling sites and various household surveys, data usually unavailable in conflict settings, we estimate increases in police presence decreased voter turnout by an average of 30%. Our results adjudicate between competing theoretical mechanisms through which security forces could affect turnout, and show behavior is not driven by voter anticipation of election-day violence. This highlights a pitfall for building government legitimacy via elections in weakly institutionalized and conflict-affected states.

6. “Social Sanctioning, Trust, and Voter Turnout in Emerging Democracies” Under Review (with Danielle Jung)

If individual ballots are not pivotal and turning out is costly, why do many citizens in transitioning democracies vote? Prior approaches highlight the psychological and material incentives that increase turnout. We argue that beyond individuals’ utility maximization, voting also reflects a person’s investment in collective goods and communities must cooperate and coordinate to mobilize members. Citizens therefore turn out to avoid negative social sanctions resulting from perceptions of strong community expectations and monitoring capacity. We test predictions with original survey data from Afghanistan’s 2010 parliamentary elections. We find avoidance of negative pay-offs from social sanctioning drives mobilization, and sanctioning declines in areas of stronger social capital. Strength of ethnic attachments and vote-buying do not consistently increase turnout for marginal voters.