Thursday, July 20, 2006

What the report from the three universities really concluded While the press has concentrated on saying that the problems with the Electoral Registry were found to have no incidence on a possible presidential election, this is not exactly what these institutions concluded. What was actually concluded had a lot of caveats to it. First of all, it was explained and clarified that the study had limited access to all of the information that would be necessary to perform a proper audit of the REP. This included the absence of addresses for those registered or the information of when the person either registered or changed the data in the registry. Within these limitations, the study concluded that the anomalies found in the REP could npot possibly have an incidence on a presidential election to teh extent of chaging its outcome. However, it could change the outcome of regional elections. Additionally the study found: --There are 107 municipalities (one third of the total) where there are more voters than inhabitants over the age of 18 as projected by the Government run National Institute for Statistics. This anomaly appears to be uniformly distributed over teh whole country. --The irregularities increased since the Government registered more people to vote in 2004 as well as giving them ID cards. --Anomalies by age group, particularly in the range 45 or older. --There are significant anomalies when one looks at the statisical distribution of bithdates across the board. --Four states show anomalies which are more significant now than they were before the three million new voters were registered: Miranda, Portuguesa, Zulia and Lara. --No correlations were detected within the limitations that implied any correlation by political preferences. --The results will only be conclusive if other tests could be performed with a complete set of data.A more exhaustive study with more data provided by the CNE could be conclusive. --Many of the errors have been around for years, are significant, but appear to have a uniform distribution and thus no material impact on a presidential election. --There is no explanation for the excess voters that is reasonable. Thus, the conclusion is that within the limitations of the study, the problems of the REP detected could not change a Presidential election. 12:29:36 AM    comment []