Parallel Monitoring of Elections with Texting

IFES did an amazing test recently using texting from cell phones to improve election observation.

Citizens armed only with cell phones took part in a landmark observation project organized by IFES during elections November 17. Their task was to visit polling sites throughout Kosovo and text election results to a central server, where the results were compiled and verified before being released to the public.

This parallel vote tabulation allowed citizens to receive unofficial, yet accurate, results within hours of the polls closing. In fact, the project’s vote totals were within half a percent of the unofficial results announced by the Central Election Commission some two days later.

Interesting results in Russian elections

The Los Angeles Times has an interesting story this morning, “Russian elections called a sham.” Most of the story focuses on questions of how Putin had personalized the election campaign, and whether that was undemocratic.

But buried deeper in the story were some interesting numbers:

In the restive northern Caucasus, voter turnout reached nearly 100%, and about 90% of the votes went to Putin’s party, according to official results. In Chechnya, election officials announced a 99.5% turnout and said more than 99% of those voters cast their ballots for Putin’s party.

Asked by a reporter whether he believed those numbers were correct, Peskov grasped for an answer.

“Well, it’s a very interesting result. I don’t think I have a right either to believe or not believe,” he said.

“At least, I don’t have a right to speak about that. But I know for sure these are official results and I don’t have a reason to distrust them.”

(Peskov is referenced in the story as “Kremlin spokesman”).

Those are some interesting results, similar to the data that our colleagues Misha Mygakov and Peter Ordeshook have written about, including their chapter for our forthcoming book on election fraud (to be published by Brookings Institution Press this coming year). I’ll see if Misha and Peter have looked at these election results yet …

Half of CA primary voters may be early voters

There is a new Field poll out that reports on registration status among California voters.

The report indicates that 47% of the voters in the last primary election (2006) were permanent absentee voters. How does this happen when only a quarter of the state’s citizens are registered permanent absentee?

I’ve written about this before on this blog:

  1. The same sort of people who tend to register permanent absentee are the sort of people who vote more frequently
  2. People who register as permanent absentee, and thus get their ballots in the mail, are more likely to vote in lower profile elections (like primaries)
  3. The combination of these effects results in an electorate with a far higher proportion of permanent absentee voters than their proportion of registered voters.

Ok. If you follow that logic, here is the kicker: we know from past work (and it is confirmed in the Field poll) that permanent absentee voters are older, whiter, and wealthier. What we didn’t know previously is that they are much more likely to live in the Bay Area (and conversely less likely to live in Los Angeles County).

That difference is almost completely driven by permanent absentee Democrats, who are a whopping 20% less likely to live in LA than the general population.

I’ve blogged before about how early voting will change the dynamics of the primary contest, but I’ve never given sufficient thought to how the composition of the overall electorate is affected by the composition of the early voters.

Based on pre-election polls, the compositional effects of permanent absentee balloting will help Obama on the Democratic side, and hurt Clinton and Edwards. On the Republican side, these numbers simply have to favor Guiliani.