Quality of Election Administration, Part 7

Left Bank has published an analysis of first round results, pointing out some anomalies in the distribution of data. The authors use Benford's Law to assess returns, much like I did in Elections and Democracy after Communism. They point to unusual results associated with both Yanukovych and Tymoshenko. The article implies that falsification took place and the tests identify fraud. But, results from the analysis of data using Benford's principles must be assessed with care. First, they point only to anomalous results and do not demonstrate causation. Second, the distribution of election data differs from the kind of data that Benford assessed (especially because polling station sizes are constrained and vary from one to another). Third, forensic accountants who use Benford's Law to assess financial data often use a battery of "data interrogation tests" to assess quality; one test is not enough. Other issues associated with the analysis of election data underscore the need for a circumspect interpretation of results. However, in past elections, analysis using Benford's Law has identified anomalies in areas where other evidence of falsification emerged.

On a related note, controversy has erupted over administrative practices for the second round. The Verkhovna Rada was convened for its 6th session this week, and amended aspects of the presidential election law which the president quickly signed. Changing the law days ahead of an election, and between rounds, is an unusual practice. But, Ukrainian politicians have amended election rules regularly ahead of the vote. The changes focused on minor administrative procedures, but may have serious implications for the vote count. The quorum rule for commissions was altered, allowing for decisions to be made without the previously mandated 2/3 attendance (instead, quorum is now a majority). Commission members may be replaced, and the rules do not require that they represent the same presidential candidate (only that they have approval of local government and the higher-level electoral commission).

Tymoshenko issued a statement about the rule changes, asserting that they were improper and undermine the electoral process. The Party of Regions countered that Tymoshenko's supporters were planning to undermine vote counting in areas where Yanukovych is strong by not showing up and preventing results from being counted and certified. Tymoshenko was lying about the effect of the new rules, according to the Party of Regions. Both parties have been willing to disrupt government activities in the Verkhovna Rada by blocking the Speaker's access to the dais, and creating noise that prevents deliberation and decision-making. In that context, the Party of Regions' allegations that BYuT might disrupt electoral practices are not out of the question. Also, a majority quorum rule is standard for bodies to convene (see Robert's Rules of Order). Nevertheless, last minute rule changes are generally inadvisable as they may lead to a perception of impropriety and may be difficult to implement.

[Hat tip to Maksym Palamarenko for the Left Bank citation and Mark Nigrini for the "data interrogation tests" phrase.]

Popular Posts