Voting behavior and electoral irregularities in Kenya's 2013 Election

Abstract

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.

Keywords:

• voting behavior
• ethnicity
• performance
• campaigns
• electoral irregularity
• exit poll
• Kenya

Acknowledgements

We dedicate this manuscript to Joel Barkan, whose untimely death occurred during its production. Joel was fiercely committed to the study of Kenya and African politics; he was also a loyal friend and mentor to us and his many colleagues and students around the world. We will miss him. The authors recognize funding from the University of California and UC-San Diego's Yankelovich Center for Social Science Research, and Long recognizes funding from the Harvard Academy for International and Area Studies. We thank Joel Barkan, Eric Bjornlund, Nicolas Cheeseman, Glenn Cowan, Danielle Jung, Karuti Kanyinga, Peter Lewis, Gabrielle Lynch, Justin Willis, and seminar participants at the Johns Hopkins School of Advanced International Studies for comments. Nathan Combes, Sara Kerosky, Lauren Pearson, and Alex Verink provided excellent research assistance. All mistakes remain with the authors.

Notes

1. Bates, “Modernization, Ethnic Competition”; Ferree, “Explaining South Africa's Racial Census”; Ferree, Framing the Race; Horowitz, Ethnic Groups in Conflict; Posner, Institutions and Ethnic Politics.

2. Gibson and Long, “Presidential and Parliamentary Elections.”

3. Long et al., “Kenya's 2013 Election.”

4. Barkan, “Comment: Further Reassessment of ‘Conventional Wisdom’”; Bratton and Kimenyi, “Voting in Kenya”; Gibson and Long, “Ethnicity, Performance, and Elections.”

5. Long, “Voting, Fraud, and Violence.”

6. For example, Agence France-Presse, “Kenyatta Turned ICC Indictment.”

7. Long, “Voting, Fraud, and Violence”; Gibson and Long, “Ethnicity, Performance, and Elections.”

8. Gibson and Long, “Presidential and Parliamentary Elections.”

9. Long, “Voting, Fraud, and Violence”; Gibson and Long, “Ethnicity, Performance, and Elections.”

10. Gibson and Long, “Presidential and Parliamentary Elections.”

11. Chandra, Why Ethnic Parties Succeed; Ferree, “Explaining South Africa's Racial Census”; Ferree, Framing the Race; Horowitz, Ethnic Groups in Conflict; Posner, Institutions and Ethnic Politics; Gibson and Long, “Ethnicity, Performance, and Elections.”

12. Chandra, Why Ethnic Parties Succeed; Posner, Institutions and Ethnic Politics.

13. Mattes, Election Book; Ferree, “Explaining South Africa's Racial Census.”

14. Long, “Voting, Fraud, and Violence.”

15. Horowitz, Ethnic Groups in Conflict.

16. Hechter, “Group Formation.”

17. Mattes, Election Book; Posner and Simon, “Economic Conditions and Incumbent Support”; Bratton et al., Public Opinion, Democracy, and Market Reform; Lindberg and Morrison, “Are African Voters Really Ethnic”; Ferree, “Explaining South Africa's Racial Census”; Bratton et al., “Voting Intentions in Africa.”

18. Barkan, “Comment: Further Reassessment of ‘Conventional Wisdom’”; Bratton and Kimenyi, “Voting in Kenya”; Kimenyi and Gutierrez Romero, “Identity, Grievances, and Economic Determinants”; Long, “Voting, Fraud, and Violence”; Gibson and Long, “Ethnicity, Performance, and Elections.”

19. Van de Walle, “Presidentialism and Clientelism”; Ferree, Framing the Race.

20. Horowitz, “Campaigns, Ethnic Polarization, and Violence”; Long, “Voting, Fraud, and Violence.”

21. Long, “Voting, Fraud, and Violence.”

22. Bratton and Kimenyi, “Voting in Kenya”; Kimenyi and Gutierrez Romero, “Identity, Grievances, and Economic Determinants”; Ferree and Horowitz, “Ties that Bind?”; Bratton et al., “Voting Intentions in Africa”; Gibson and Long, “Ethnicity, Performance, and Elections.”

23. Ferree, Framing the Race; Long “Voting, Fraud, and Violence.”

24. Mattes and Piombo, “Opposition Parties and the Voters”; Ferree, “Explaining South Africa's Racial Census”; Ferree, Framing the Race; Gibson and Long, “Ethnicity, Performance, and Elections.”

25. Anderson and Lochery, “Violence and Exodus.”

26. Barkan, “Kenya: Lessons from a Flawed Election”; Barkan and Ng'ethe, “Kenya Tries Again.”

27. Kenya's average annual growth rate was 5.4% from 2003 to 2007 and 3.75% from 2008 to 2013, showing a large drop-off for the period after 2008's post-election violence. For Kenya, see http://databank.worldbank.org/data/views/reports/tableview.aspx. Accessed June 20, 2013.

28. See http://www.worldbank.org/en/country/kenya/overview. Accessed June 20, 2013.

29. Ferree and Long, “Violating the Secret Ballot”; Gibson and Long, “Presidential and Parliamentary Elections”; Hoffman and Long, “Party Attributes, Performance, and Voting”; Long et al., “Kenya's 2013 Election.”

30. Gibson and Long, “Presidential and Parliamentary Elections.”

31. The results are robust to the use of linear probability and probit models.

32. The three are included in the same model for Model 4 since they do not highly correlate with each other.

33. Rerunning these models with the inclusion of full ethnic controls (results not shown) generates significant results for Kalenjin and Kamba.

34. All Online Supplemental Tables 1–4 are available on the journal website.

35. Ideally, we would also like to weight the survey by true turnout to account for any discrepancies between the voter registry (which we used for sampling) and who actually voted. However, given claims of irregularities including double-voting, ballot-stuffing, and tally fraud, we are skeptical of official turnout statistics. Nonetheless, weighting the survey by official turnout figures still produces statistically tied candidates (Kenyatta with 41.6% and Odinga with 38.5%) with neither garnering enough of the vote to pass the 50% + 1 threshold to avoid a run-off.

36. See note 34.

37. A total of 145 respondents RTA the question on the trustworthiness of both TNA and ODM. All respondents told us if they were close to a party, and all but 87 were willing to tell us which party it was.

38. King et al. “Analyzing Incomplete Political Science Data.”

39. The standard errors for the Amelia estimates are as follows: for the RTA split, Odinga 0.402% with a standard error (SE) of 0.025; Kenyatta 0.405% with SE = 0.023; other candidates took 0.19% with SE = 0.019. For the overall results, Odinga took 0.453% with SE = 0.007; Kenyatta took 0.456 with SE = 0.006. These uncertainty estimates are for the imputations only and do not reflect uncertainty due to sampling. We ran ten imputations of the model and otherwise used the default Amelia settings. In addition to provincial dummy variables, we included survey variables measuring: overall government performance evaluations, their opinions regarding why the government was underperforming, economic evaluations, security evaluations, the most important problem facing the country, partisanship, the strength of the partisanship, a view on the trustworthiness of CORD, a view on the trustworthiness of Jubilee, whether they were offered money for their vote, whether they were intimidated, whether the election was rigged in their area, views on the IEBC, age, gender, education, employment, religion, and ethnicity.

40. Elections Observation Group in Kenya ‘The Historic Vote.’