Tuesday, December 04, 2007

Transparency in the Venezuelan Referendum

There are many things to say about the voting process that just took place in Venezuela this past week, but this post will focus on only one important issue: the reliability and credibility of the actual voting process administered by the Consejo Nacional Electoral (CNE), which I think we can safely assume is 100% staffed with Chavistas.

It is first important to note that, from the 2004 referendum (when only 4% of the voting tables were audited, meaning the electronic tally was compared to the paper receipt) to the 2006 presidential elections (when some 30% were audited), to the December 2 referendum on the constitutional reforms (where some 54% of the voting tables were audited), the CNE has increasingly become more transparent in its operations in order to gain the confidence of a very skeptical citizenry.

Yet two days after the referendum, many people still feel that the numbers were tampered with, and that negotiations were held with Chavez so that the actual higher margin would not make Chavez look so bad. One group, for example, claimed today that the real vote was something like 58% for the NO, and 42% for the SI. Even though the opposition leader Manuel Rosales accepted the results in last year's December 2006 presidential election as valid -- did so again today -- rumors and desconfianza persist.

So what can the CNE do about it? One diplomat told me yesterday (Monday) that the CNE should publish the results of every voting table on the internet, so that people could check the results. This would go a long way to creating greater confidence in the electoral system.

Well, in fact, as of today, the CNE has done just that -- published the preliminary results of the votes that had been counted by Sunday night night (some 90% of all tables, with some voting stations that did not use voting machines coming in later, as well as votes from abroad). You can find the results here: http://www.cne.gov.ve/divulgacion_referendo_reforma/ (The voting results of the December 2006 presidential elections are equally disaggregated and available online, in case someone wants to do further analysis of political tendencies.) The most credible member of the CNE, Vicente Diaz, gave a press conference today in which he called upon everyone to access their website, and check the results for themselves.

You might ask, what does this prove? Let me run through the steps of tallying votes, at least in one voting center in Caracas which I observed, and it will become apparent that the Venezuelan electoral process allows for unprecedented transparency and for broad citizen participation in the verification of election results.
  1. At the end of the day, the members of each mesa (which in this case included one representative each of the SI and NO, along with the official supervisors) were present while electronic paper receipts were printed out. Anyone who wanted to could be present, the results are read aloud, and copies were made for the CNE as well as for others who wished to have one.
  2. Depending on the size of the voting center, 1-5 tables were audited through a random selection process carried out in each voting center. The auditoria process means that the box that contains the receipts of each vote (which is deposited in a sealed box after each electronic vote) is opened and counted manually, and a public tally is made to make sure that the paper receipts match the electronic printout.
  3. Once all of this is finished, the electronic numbers are transmitted electronically to the CNE, which does the final tally.
As you can see, this process allows for all sides (as well as extraneous individuals and observers) to verify the vote count of every single mesa. One can often walk away with an actual electronic copy of the voting results from each voting center, which is amply checked because of the auditing process. Now that the CNE has produced these initial results, which represents most all of the centers that used electronic voting machines, then anyone can go online and verify whether the numbers were tampered with.

This really is an amazing degree of transparency, if you ask me, and I'd like to know if there's any precedent for this anywhere else in the world. Given that the students and opposition political parties mounted a vigorous effort to monitor the results of the referendum, if they do not come forward now with firm proof of results tampering, then perhaps the Venezuelan people will finally be able to get past this issue, a step that will be key to further stimulate electoral turnout in the future.

That's the basic story. Of course, there were irregularities -- for example, some machines broke down and were not repaired, and paper ballots were not always immediately available as a backup -- but the respected domestic monitoring group, Ojo Electoral, which monitored a sample of 400 mesas, noted that there were only "isolated incidents" in an otherwise normal process.

I was able to note in one polling place where I witnessed the final vote count in Caracas (one that went 2-to-1 in favor of the NO) the professionalism and openness of the CNE workers who tried to resolve these problems. For example, here's a video I took of how they dealt with what happened when a machine failed to give the final tally:

And here's a young CNE official explaining (in English) what is to be done if a machine breaks down during the day, which happened alot:

Here's another video, that was taken during the reading of the electronic tally:

And another one that looks at how the paper audit is done:

In this last one, notice how I try to show that several people in the room are doing their own tallies as each individual vote is counted:

Friday, November 30, 2007

Random points on Sunday's referendum

Yesterday's march in favor of a NO vote on the constitutional reforms was reported by AP to be more than 100,000 strong, but interestingly the pro-government venezuelanalysis.com website reported "several hundred thousand" persons demonstrated. One person told me this was the biggest opposition rally since the ill-fated April 11, 2002 march during the coup period. The opposition clearly has a sense of momentum going into Sunday's vote, with a strong push against abstentionism emerging.

In the few conversations I've had so far of those who follow this closely, however, they are still cautious about how this will all turn out:
  • One cannot underestimate the capacity of the government to get out the vote for their position, as they've done in the past. Current estimates of 60% participation (those who will "definitely" vote) are still much lower than the December 2006 presidential elections, which had 74% participation with Chavez getting 7.4 million votes (about half of all eligible voters). The opposition has a lot of energy behind it, but no systematic get-out-the-vote campaign. The polls indicate that any additions of newly decided voters will tilt the current toss-up towards the NO, but just how many people will turn out remains to be seen.
  • The opposition is more prepared and organized to monitor and defend the results of the referendum than they have been in the past, but -- I'm not sure this extends to rural areas, which is where the government will be able to mobilize people overwhelmingly. One person mentioned voting tables that had gone 100% for Chavez in the past (indicating either a capacity for mobilization or an unfettered capacity for tinkering with the numbers in these areas.)
  • One person mentioned that there is often a secret vote for Chavez among middle and upper-middle class voters, who essentially vote their pocketbook (since they're doing quite well). I haven't heard of anyone who's really studied voting patterns, but this should be easily verifiable after the fact.
Stay tuned.

Sunday, November 25, 2007

Washington Post weighs in on WOLA drug claims

Several weeks ago, a story from AP found found Justice Department and GAO findings contrasted with the claims by drug czar John Walters of the Office of National Drug Control Policy (ONDCP) that the "unprecedented" recent spike in cocaine prices was evidence that US drug policy was working.

Now, Michael Dobbs, who writes "The Fact Checker" blog at the Washington Post website, has weighed in, spurred on by WOLA's recent press release challenging Walters' triumphant rhetoric. While Dobbs ended up saying the verdict on the facts is pending (and asking readers to provide other views), he finds that there is great skepticism on both sides of the aisle in Congress over ONDCP's methodology.

Dobbs also publishes a graph from a RAND report commissioned by the ONDCP, which finds a long-term decline in cocaine prices:

Dobb comments:

The most striking point in this graph is the long-term downward trend in retail cocaine prices, despite all the efforts at interdiction undertaken by successive U.S. administrations. By eyeballing the chart, you can see that there were significant price spikes in 1982, 1990, 1994, and 2000, which are
comparable to the recent increase. Each spike was followed by another sharp decline, as producers responded to the higher prices. Compared to historical levels, cocaine prices are still very low, particularly if you factor in inflation.

The RAND data is not strictly comparable to the latest DEA data as it measures the retail slice of the market, rather than average purchase prices. (RAND data for other slices of the market show similar peaks and troughs.) But it certainly suggests that policy-makers should be more cautious in using terms
like "unprecedented."

... One of the principal authors of the RAND study, Peter Reuter, a professor of public policy at the University of Maryland, said he was troubled by the way the DEA and ONDCP (the drug czar's office) kept changing its statistical methodology. "I don't understand why they don't run the series the same way (as RAND), just to remove any doubts that the data is solid. It would be much more convincing if they did that."

The ONDCP has its own blog, Pushing Back: Making the Drug Problem Smaller, which a week ago offered a rebuttal of WOLA's critique. However, ONDCP was obviously looking for a more definitive pat on the back from Dobbs than what they received, when they wrote:

On a more optimistic note, WOLA apparently sent their press release claim to the Washington Post reporter Michael Dobbs, who runs a "Fact Checker" feature for the paper. At Dobbs' request, we have provided the full data story correcting
the record about WOLA's claims, showing why their assertions are off the mark. We look forward to a fair accounting of the situation.