Sunday, July 23, 2006
The went fishing and all they caught was minnows
As most probably know the opposition has been trying to find flaws with the Venezuelan electoral system which it can then use as an excuse to pull out of the elections and save themselves the embarrassment of what looks sure to be a huge defeat. One of the early excuses has been – “we can’t have fair elections because the electoral registry is all messed up and packed with pro-Chavez ghost voters”. Never mind that other outside agencies such as CAPEL performed an audit of the registry and found it acceptable.
Last month three major universities proposed an audit, the methodology of which had some rather significant problems of its own. So much so that the electoral authorities refused to sanction that proposed audit.
The universities were not deterred and went ahead with the audit on their own using public copies of the registry and census data from the National Statistics Institute. Just a couple of days ago they made public their findings which were that while there were a great many errors in the registry, which CAPEL also found, most of these errors were long standing (ie, they were there before Chavez even came to power) and they found they were randomly distributed, that is they really were just errors which would not favor any particular side in an election. In other words, there was no manipulation of the registry to gain electoral advantage and no reason why this registry couldn’t be used for national elections.
Here is a chart that was published that showed the distribution of errors in 1998 before Chavez came to power, where the errors are now (essentially in the same place) and where the errors have increased most:
Now, you might ask, how could they determine that these errors were random in nature and didn’t give advantage to anyone in particular. Simple, they correlated the errors with voting patterns in previous elections and found that the errors were pretty evenly distributed between regions where Chavez did well and where the opposition did well. In fact, one telling statistic given in the El Universal article is that in locations where there were significant errors in the registry the NO (the Chavez side) won with 55% of the vote whereas in locations where there were few errors the NO won with 62% of the vote. That shows that voting tendancies were similar between the areas with many errors and those with few errors but Chavez actually did better where there were fewer errors. This led the mathematicians involved in the study to state on Globovision:
Clearly, the registry has problems but it is not a reason to consider the December elections unfair. So there is yet another excuse out the window for the opposition.
A final note on this. From what I can see a big part of the problem with the errors in the registry is that once people are registered they are not ever properly de-registered when they should be – i.e. when they die, or move to another location. This is why a large number of people of over 100 years of age are on the registry and a number of municipalities have more registered voters than people over age 18. It would seem one possible solution to this would be that if a person goes more than a certain period, say 10 years, without voting they should be automatically purged from the voter rolls. If they really are still around and want to vote they would have to re-register. That would be a slight inconvenience for some but it would help solve some of these problems.
|
Last month three major universities proposed an audit, the methodology of which had some rather significant problems of its own. So much so that the electoral authorities refused to sanction that proposed audit.
The universities were not deterred and went ahead with the audit on their own using public copies of the registry and census data from the National Statistics Institute. Just a couple of days ago they made public their findings which were that while there were a great many errors in the registry, which CAPEL also found, most of these errors were long standing (ie, they were there before Chavez even came to power) and they found they were randomly distributed, that is they really were just errors which would not favor any particular side in an election. In other words, there was no manipulation of the registry to gain electoral advantage and no reason why this registry couldn’t be used for national elections.
Here is a chart that was published that showed the distribution of errors in 1998 before Chavez came to power, where the errors are now (essentially in the same place) and where the errors have increased most:
Now, you might ask, how could they determine that these errors were random in nature and didn’t give advantage to anyone in particular. Simple, they correlated the errors with voting patterns in previous elections and found that the errors were pretty evenly distributed between regions where Chavez did well and where the opposition did well. In fact, one telling statistic given in the El Universal article is that in locations where there were significant errors in the registry the NO (the Chavez side) won with 55% of the vote whereas in locations where there were few errors the NO won with 62% of the vote. That shows that voting tendancies were similar between the areas with many errors and those with few errors but Chavez actually did better where there were fewer errors. This led the mathematicians involved in the study to state on Globovision:
Bernardo, who is a statistics expert, told the TV station Globovision that his analysis of the voter registry has shown that there are many errors, but these are administrative and lack political intentionality that could alter the results of an election. Errors that the opposition has found recently, such as the large number of over 100 year olds are true, he said, but, “In practice [these] do not favor the government and their distribution is uniform.”
Similarly, USB mathematician Raúl Jimenez, who is also involved in the audit, said, “One must be responsible. The electoral registry is a disaster and the CNE has done nothing to improve it, but there is nothing to indicate a political intention in the anomalies.”
Jimenez explained that his audit estimated the electoral results from the recall referendum without the erroneously registered voters and found that the President would have won the vote in any case. “The newly inscribed are not being placed so that they vote for the President. We have an aspect that is conclusive: there are no [political] tendencies [to the data],” said Jimenez.
Clearly, the registry has problems but it is not a reason to consider the December elections unfair. So there is yet another excuse out the window for the opposition.
A final note on this. From what I can see a big part of the problem with the errors in the registry is that once people are registered they are not ever properly de-registered when they should be – i.e. when they die, or move to another location. This is why a large number of people of over 100 years of age are on the registry and a number of municipalities have more registered voters than people over age 18. It would seem one possible solution to this would be that if a person goes more than a certain period, say 10 years, without voting they should be automatically purged from the voter rolls. If they really are still around and want to vote they would have to re-register. That would be a slight inconvenience for some but it would help solve some of these problems.
|
