Kenya: “Slate Says IRI Is Sitting on Exit Poll”

USAID’s gift from the American people to the Elections Commission of Kenya included a modern communications network designed to make the vote-count faster and more efficient in 2007. It apparently didn’t. One more letdown from the rhetoricians of the technological sublime?

At the present time there has been a proliferation in the abuse of such public opinion polls whose results are not based on the use of serious methodologies. As a result, surveys have been publicized that were undertaken without the use of representative samples or properly formulated questionnaires or that were conducted by bogus companies that impeccably report on the methodology employed, but whose results clearly reflect political bias. It is common that some companies send results to the communications media, which, in turn, release the results without verifying their authenticity or quality. It is the ease at [sic] which the results of the surveys can be manipulated that has led campaign strategists to circulate apocryphal results, simulations that present their candidates in the electoral races as the winners, for the purpose of generating doubts or confusion among the voters. –Carolina Pacheco Luna & Mauricio Velásquez Posada, “Electoral Surveys and Their Regulation,” Comparative Media Law Journal (UNAM, Mexico), No. 8 (December 2006)

Ignoratio Elenchi (Ignoring the Counterevidence): A one-sided case presents only evidence favoring its conclusion, and ignores or downplays the evidence against it. In inductive reasoning, it is important to consider all of the available evidence before coming to a conclusion.

Once upon a time, we knew how to do clandestine regime change. We need to reacquire that capability. –John Bolton, United States Ambassador to the United Nations (Foreign Policy, article undated on Web site)

Once upon a time, the term diplomatic implied a certain quality of discretion — such as not announcing to the world that you planned to start engineering other people’s political orders in a clandestine manner.

Does John “I am the Walrus” Bolton not get that acquiring a clandestine capability might imply that the first rule of Fight Club is, you don’t talk about Fight Club?

And speaking of the International Republican Institute in Kenya, about which I have argued

the role of the International Republican Institute in this election … needs to be probed harder than the prostate of a 70-year-old life-long chainsmoker …

… Robert Baird notes:

Today at Slate, Alex Halperin wonders why a Kenyan exit poll sponsored by the International Republican Institute hasn’t been released to the public:

Halperin, wondering:

The International Republican Institute, a democracy-fostering nonprofit funded by the U.S. government—and despite the name, officially nonpartisan* — commissioned an Election Day exit poll but has declined to release the results. Two people familiar with the results told me that they showed [Raila] Odinga with a substantial lead over President Kibaki—one reported eight points, the other nine points.

Major world news wires created even more dramatic expectations than that:

• Grey Lady-Globo: “Kenyan Opposition Wins With 57% of the Vote!”

Speculating on why that might be:

Why would the IRI withhold a poll that showed Odinga in the lead? … The answer to Halperin’s question might therefore be depressingly simple: the IRI won’t release their polling data because they don’t want the wrong guy to win.

Mr. Baird’s footnote as to “officially nonpartisan”:

*Note: It may be true, as Halperin argues, that the IRI “missed an opportunity to advance its mission of promoting democracy and fair elections,” but if so, it wouldn’t be the first time. In fact, a person familiar with the organization’s activities in Haiti a decade ago might be excused for doubting the sincerity of that mission in the first place. Saying that the International Republican Institute is “officially nonpartisan” is a little like saying that Iran is “officially democratic”: it’s true but pointless.

IRI’s March 2007 opinion polling showed high favorables for the incumbent government, it seems to me, looking it over.

I am in the process of making PNGs of it for posting so we can consider its methodology carefully. It is one of those surveys where “the methodology is available upon request.”

I would guess that deciding to conduct it in English probably created quite a demographic skew to begin with.

On which more later. Let me actually read the Slate article before jumping, inductively, to a conclusion here …

The paragraphs in question:

Unfortunately, one bit of data has not surfaced. The International Republican Institute, a democracy-fostering nonprofit funded by the U.S. government—and despite the name, officially nonpartisan—commissioned an Election Day exit poll but has declined to release the results. Two people familiar with the results told me that they showed Odinga with a substantial lead over President Kibaki—one reported eight points, the other nine points. One has only to remember the United States’ 2004 elections to know how fallible exit polls are, but a U.S.-sponsored survey would have weight here and could have given the ECK pause before it called the election so disastrously.

Ken Flottman, an official in the IRI’s Nairobi office, said the data would serve additional purposes, such as studying voter demographics. The organization issued a statement criticizing the vote counting but does not mention its data. It missed an opportunity to advance its mission of promoting democracy and fair elections.

The Kenyan statistics body should subpoena those results or throw somebody in a Kenyan jail trying.

As I was just reading, that is precisely what the Statistics Act was designed to require:

• Kenya: “Of Perception Surveys and the Statistics Act”

On the dueling exit poll wars worldwide, see also

• O Globo: “Chávez Won the Referendum Because He Manipulated the System!”
• “Judy Millerist Globovision Was a Victim of Infinite Guile!”: Diário de América
• Venezuela: “Exit Polls Favoring YES Were Nonexistent”
• Dueling Polls in Costa Rican Referendum
• Mexican Elections Defended by Idiots: Lies, Damned Lies, and the Critique of Pure Polling
• Good Enough for Maracatú: The Polling Margin of Error in the Primeiro Turno

I am not quite sure about this statement by the man from Slate:

Divisive, culture-war-type issues played hardly any role in the campaign.

The Kenyan National Civil Rights Commission reported otherwise: