In transitioning democracies such as Kenya’s, the problem appears particularly nettlesome given weak institutions, a lack of independence and transparency on the part of electoral commissions, and the inability of international and domestic observers to monitor all aspects of the count from polling stations, constituencies, to the final official tally.
Tragically, political violence all too frequently accompanies claims of electoral malfeasance.
Adding to the malaise are incumbent political leaders who appear unwilling to give up power at any cost, and opposition parties that claim rigging only when they lose or boycott contests altogether. The credibility of arguments made by any side in a contest is obviously suspect.
Therefore, we use multiple kinds of statistical analysis building on
basic methodologies and official statistics to try to locate and
quantify ballot rigging in the Kenyan case. In Kenya, the predicted
closeness of presidential race before the election may have contributed
to an irresistible temptation on all sides – both the government or
Party of National Unity (PNU) and the main opposition or Orange Democratic Movement (ODM) to participate in fraud. As counting of votes progressed after polls closed on the
evening of December 27, 2007, members of political parties, accredited
observers, both domestic and international, as well as civil society organisations and ordinary citizens lodged allegations of rigging. Commissioners of the ECK
including its chairman also cast aspersions on the results. There were
instances during which the Commission had two different figures (high
and low) for the same constituency.
No one methodology or element of data that we employ definitively “proves” fraud. Nor are we able to assign blame on any party, candidate, or individual. However, we do use a variety of techniques and data sources to reveal startling anomalies in the vote count. Various kinds of discrepancies exist within the Electoral Commission of Kenya’s (ECK) final results, as well as when compared with other data sources. A problem in any one of the areas that we highlight below is worthy of scrutiny and ought to encourage a massive reform of the ECK.
Here, we focus on some areas that have received some attention in the press and scholarship before—but not rigorously analyzed—as well as new areas of investigation. We begin with an examination of “16A Forms,” or the forms indicating the final tally from polling stations at the constituency level that were supposed to be filled out by constituency returning officers and submitted to the ECK headquarters in Nairobi. We note a number of problems in how these forms were filled out and submitted. Next, we turn to results reported by media houses compared with official results and highlight of number of discrepancies. Last, we look at two problems of voter turnout. Suspiciously high turnout may have helped to inflate totals for candidates. Moreover, differences in turnout between the presidential and parliamentary elections result in a number of potentially problematic ballots, enough to have swung the result of the election.
It is important to stress what our analysis can and cannot say about the potential for fraud. Once again, statistical tests in this vein are helpful in highlighting a number of irregularities based on prior voting behavior in Kenya, as well as suggesting whether these irregularities tend to bias in favor of any one candidate and whether or not by enough to switch the result. While we do find a consistent bias in favor of President Kibaki, statistical analysis cannot source malfeasance in any one party, candidate, or individual. It does, however, underscore a number of problems that existed within the ECK and ought to urge policymakers and politicians to undertake serious reform of that institution. Moreover, for a complete understanding of what took place in Kenya’s 2007 elections, our statistical analysis should be bolstered by detailed and investigative information gathered from the people involved in all levels of the vote count, from
polling stations, constituencies, and headquarters of the ECK.
Electoral fraud is hard to prove. This is one of the reasons why political operatives frequently engage in it the world over, though not always successfully. In transitioning democracies such as Kenya’s, the problem appears particularly nettlesome given weak institutions, a lack of independence and transparency on the part of electoral commissions, and the inability of international and domestic observers to monitor all aspects of the count from polling stations, constituencies, to the final official tally. Tragically, political violence all too frequently accompanies claims of electoral malfeasance.