February 4, 2011

The Elections: Flawed Beyond Repair?

10 a.m.
Supreme Court
West Conference Room
1 First Street, NE
Washington, DC 20543

The Haiti Democracy Project sponsored a presentation on the "Haiti Democracy Project Electoral Mission." Participants included: CEPR Co-Director Mark Weisbrot; Professor Fritz Scheuren, the 100th president of the American Statistical Association and member of the OAS Verification Mission to Haiti; Roger F. Noriega, Secretary of State for Western Hemisphere Affairs under George W. Bush, and visiting fellow of the American Enterprise Institute; Rudolph H. Boulos, board member of the Haiti Democracy Project; James Morell, founder and executive director of the Haiti Democracy Project; and Ernest Preeg, former US ambassador to Haiti and chair at the International Business Center at the Center for Strategic and International Studies (CSIS).

In the segment posted below, CEPR co-director Mark Weisbrot challenges Fritz Scheuren, a member of the OAS' Expert Verification Mission and President of the American Statistical Association, to answer CEPR's two main criticisms of the OAS report, which changed the result of Haiti's first round Presidential election:

(1) That the OAS Mission Report did not do any statistical inference in order to estimate what the result might be if they had examined the other 92 percent of tally sheets that they did not examine.


(2) That the OAS Mission Report did not take into account the missing and quarantined tally sheets, which amounted to more than 150,000 votes, or about 12 percent of the total. These votes were in areas that were in aggregate much more pro-Celestin than the rest of the population, and hence would have put him in second place.

Mr. Scheuren' s answer confirms CEPR's criticisms. He acknowledges that the OAS Mission did not do any statistical inference because of "non-sampling error." This presumably refers to the missing/quarantined votes. This make sense as answer to the question of why they did not use statistical inference. However, the Mission could have at least acknowledged the problem of the missing votes, and – as CEPR did -- make some attempt to estimate how they might have affected the outcome.

Most importantly, Mr. Scheuren's answer is an admission that the OAS Mission had no statistical basis on which to recommend changing the outcome of the elections, which is what the Mission recommended, and what was eventually done.