Archive for August 13th, 2012

Protecting The Votes in Venezuela Part I: The Evolution Of The Electoral Registry

August 13, 2012

Last week I was invited to the Comando Venezuela to see what they have been doing about insuring the votes for the opposition are counted in the Oct.7th. Presidential election and that only votes that should be counted are part of the results. The person I talked to has been involved with many of the statistical studies done up to now but you could say that his role today is that of auditing all processes going forward, insuring that things are done, that the quality of the work done is a good one and coming up with tests for all these processes.

As you walk through the halls of the comando, the walls are covered with maps of all states, showing critical voting centers, ranked according to their importance. These maps are more than decoration, while I waited to be met, twice people went up to them to talk or discuss some point about them.

The Comando Venezuela has been working on essentially three fronts to guarantee a clean election on October 7th., these three main areas have been the subject of a lot of work (there is some overlap):

1) The cleaningness of the Electoral Registry

2) The voting system and the new fingerprint system

3) Protecting the vote on October 7th.

In this first part, I will discuss their study of the evolution of the Electoral Registry since the 2010 National Assembly election and what this shows. You can find the presentation for this part here.

The first concern is how the electoral registry has evolved, whether it makes sense or not. Below is the table for the evolution since the last election (which the opposition won numerically).

Clearly, it makes sense that 86.6% of new voters registered are under 25. There has been a request from the CNE for more information about the newly registered people in the over 40 group, some 40 thousand voters, which has yet to be satisfied. However, for those that believe that the Electoral registry is a mystery of sorts, political parries receive monthly data about the newly registered voters and migrations.

The data of newly registered voters has been examined to see its consistency. For example, below you can see the distribution of new voters and changes in voting center distributed according to socio-economic level, showing that there are more new voters in the poorer strata of the population, which is consistent with birth rates 20 years ago. :

In terms of migrations, close to 1.7 million people requested a change in his or her voting center. A study was made of the patterns of these migrations, the summary is shown in the table below:

The data was found to be consistent in terms of migrations, most of them occurring between centers which are similar in terms of socio-economic characteristics. But more importantly, a random telephone poll found that 97% of those that had migrated, had requested the migration and those errors in the migrations (or those whose center had been closed) did not appear to have any political bias in terms of who they affected, that is, both pro-Government and position voters were affected by it. Finally, during the opposition primaries in February, the Electoral Registry was used and there was no evidence of involuntary migration affecting voters.

There are two problems that have yet to be cleared up in the Electoral Registry. First, there continues to be roughly 18 thousand id’s associated to 9 thousand people and dead people have not been removed sufficiently fast from it. Both these issues are being discussed with the CNE and particularly the first on is expected to be solved before October 7th.

The Catholic University has made a study of the registry and has found that it is consistent with the population. The only inconsistency is that of an over representation of the group above 70 years of age, which I already noted.

All in all, the Comando Venezuela believes that with adequate, well prepared witnesses at the tables, the areas of concern with the Electoral Registry can be minimized and the voteĀ  affected by it in only minimal fashion. This is the main thrusts of point 3) in this discussion which I will try to present before the end of this week.