Election Updates

As we get ready for our research workshop on election fraud at the end of the month, a friend passed along the following story from CBS News in Chicago, that the Republican Party there is alleging that at least 500 people voted more than once in 2004 in the Chicago area:

(CBS) CHICAGO The Republican Party is putting up evidence that the days of voting early and often in the Chicago area are not quite over.

Rather, local party officials declared, voter fraud is alive and in play.

Cook County Clerk David Orr confirmed on Thursday that there is some indication that at least a handful of voters voted twice in the general election in 2004.

The GOP evidence showed more than 500 cases "of double voting."

Both the Cook County Board of Elections and the clerk’s office promise thorough investigations.

A paper was recently published in Public Opinion Quartery (2006, 70(2), 224-234) by Matt A. Barreto, Matther J. Streb, Mara Marks and Fernando Guerra. If you are at a university or other institution that has online access to the journal, you can view the abstract and download the paper here. Here is the paper's abstract:

In the 2003 California gubernatorial recall election 2,775,785 absentee ballots were cast, representing about 30 percent of all voters statewide. Given the number of absentee ballots and the increasing propensity for voters in California and elsewhere to choose this voting method, we some basic questions: Who are absentee voters, and are they different from polling place voters? To answer these questions, we fielded a statewide survey of absentee voters in the days before the October 7 recall election, asking respondents why they voted absentee, their partisan and ideological preferences, demographic characteristics, and other relevant questions. We find that absentee voters do not differ significantly from the overall state electorate in terms of their vote preferences, despite being older and better educated. For example, 56 percent of absentee voters in our survey voted "yes" on the recall, compared with 55 percent for the entire state, according to official returns. Further, absentee voters favored Arnold Schwarzenegger over Cruz Bustamante by a considerable margin, similar to the overall election results. We found party registration among absentee voters to be nearly identical to statewide partisan registration.

Their study compares a sample of 300 absentee voters, with the more than 5,000 voters in the 2003 Los Angeles Times exit poll. For those who have followed the debates about absentee voting (especially in California) one of the more interesting results in this particular study is their finding that absentee voters in their sample appear to have nearly identical political preferences and behavior to the precinct voters in the exit poll sample. Whether the same result is true of elections other than the 2003 recall election is a question for future research.

The Caltech/MIT Voting Technology Project is organizing a conference on voter identification and registration, October 5-6, on the MIT campus. Here is the conference announcement, and here is a page where folks can register for the event.

The current conference agenda is also available.

While at APSA, I ran into a friend and former colleague, Scott Page, who is at the University of Michigan. Scott's written a great new book on diversity that should be out soon, and he is also an expert scholar in the area of complexity. While we were talking about his work on diversity and complexity, Scott handed me a fantastic booklet that he put together, "A Decision-Making Guide To The Michigan Civil Rights Initiative."

What Scott (and co-author Elizabeth Suhay) are trying to do in this innovative decision-making guide is to help Michigan voters work their way through the complexity of the upcoming ballot measure (like one that passed about a decade ago in California), the "Michigan Civil Rights Initiative", that would block state and local governments in Michigan from developing and implementing policies that allow attributes like race or ethnicity, gender, or national origin, to be used as criteria in government programs or policies.

The decision-making guide is innovative, as in the final set of chapters it breaks down the Michigan initiative into a variety of dimensions, and it tries to help guide voters through the development of their preferences on each dimension. It then seeks to help voters aggregate their opinions across dimensions, to help them decide whether in the final analysis they should support or oppose this initiative.

I found this to be a very interesting way to try to help voters wade through a complex and multidimensional issue, and one that might be useful to emulate in other states and for other complex ballot initiatives.

While we were at the American Political Science Association annual meetings last week in Philadephia (more on the conference soon!), we here at Election Updates passed two milestones that I wanted to note.

First, we've been writing this blog for exactly a year now; we started it late in August 2005.

Second, we just passed our 500th essay.

Thanks to Thad, Paul and Melissa for their contributions.

And thanks to folks out there who read our essays on elections, voting technology, and election reform. During the past year we have managed to get over 28,000 visits to our blog, and over 45,000 page views. We average about 100 visits per day, and 141 page views per day.

As is always the case, if you have ideas for things we should be writing about, don't hesitate to send them along!

Ben Adida's MIT Ph.D. thesis on voting technology, "Advances in Cryptographic Voting Systems", has been published and is now available for interested readers. Ben's Ph.D. supervisor was Ron Rivest, long-time member of the Caltech/MIT Voting Technology Project.

Here's the abstract from Ben's thesis:

Democracy depends on the proper administration of popular elections. Voters should receive assurance that their intent was correctly captured and that all eligible votes were correctly tallied. The election system as a whole should ensure that voter coercion is unlikely, even when voters are willing to be influenced. These conflicting requirements present a significant challenge: how can voters receive enough assurance to trust the election result, but not so much that they can prove to a potential coercer how they voted?

This dissertation explores cryptographic techniques for implementing verifiable, secret ballot elections. We present the power of cryptographic voting, in particular its ability to successfully achieve both verifiability and ballot secrecy, a combination that cannot be achieved by other means. We review a large portion of the literature on cryptographic voting.

We propose three novel technical ideas:
1. a simple and inexpensive paper-base cryptographic voting system with some interesting advantages over existing techniques,
2. a theoretical model of incoercibility for human voters with their inherent limited computational ability, and a new ballot casting system that fits the new definition, and
3. a new theoretical construct for shuffling encrypted votes in full view of public observers.

As Ben's thesis involves cryptography and lots of math, it gets pretty tough going when he gets into the heart of some of his new ideas for voting systems. But I strongly recommend his introductory chapter, where he goes through the existing literature on crypto and voting, as well as his chapter on the "scratch and vote" scheme (Chapter 4).

Ben's one of the brightest young minds working in the area of voting technology from this perspective, and it is exciting to see such great work being done by the next generation of scholars.

According to an initial study reported by the AP today (and recounted in a number of newspapers, here is a link to the Contra Costa Times), there is no compelling evidence yet that the immigration rallies earlier this year have led to substantial surges in Hispanic voter registration.

The methodology of the AP story was to examine voter registration statistics from a variety of urban areas where there are high concentrations of Hispanic citizens, and then to look for whether or not the 2006 voter registration figures (so far) are showing evidence of any surges. Here is a piece from the story talking about the study, and the specific findings:

...

For this story, the AP reviewed new registration numbers in metropolitan areas over several years. The areas included Los Angeles, San Francisco and San Jose; Chicago; Phoenix and Tucson, Ariz.; Dallas and Houston; Atlanta; Denver; and Jacksonville and St. Petersburg, Fla. The time frames included both January-through-July periods dating to 2004, as well as periods before statewide elections, when registration efforts are most intense.

The data provide a wide-angle look at new registrations, but they do have limitations. Any significant shift in registrations overall would stand out, but voters are not specifically identified by race or ethnicity. As a result, an increase in new registrations in Los Angeles County in the 100 days before this June's primary compared to the months before two prior statewide elections cannot be attributed exclusively to new Latino voters, despite extensive registration efforts.

Gains in new registrations were highest in 2004, when political parties spent lavishly to enroll new voters ahead of the presidential election.

New voter registrations increased in virtually every city between 2005 and 2006 -- but that would be expected because of congressional primaries and elections. The 2006 numbers were below the 2004 numbers in every city, often significantly.

In Harris County, Texas, home to Houston, new registrations through July tallied 55,657 -- an increase of 16 percent over 2005 but well below the 71,402 from 2004.

Dallas County showed more potential in attaining significant new voter registrations for 2006. Its new voter registrations totaled 35,590 through Aug. 15. With less than half of this year left, the figure was only 4,775 shy of the number of new voters who registered in 2004, a presidential election year. Last year, Dallas and the surrounding cities in the county had 27,321 new voters register.

...

In rare cases, registrations declined. New registrations in San Francisco were significantly lower in the 100 days before this year's June 6 primary than over the same period before a statewide special election in November 2005.

In Chicago and surrounding Cook County, registrations in the first seven months this year jumped about a third over 2005, but were far below the same period in 2004.

As the story notes repeatedly, this is not conclusive evidence that the immigration marches will not impact Hispanic voting behavior. It is still possible that as we get closer to close-of-registration periods in these urban areas that there will be surges in Hispanic registrations, that the marches might galvanize already-registered Hispanics to vote this fall, or that many of the young Hispancs involved in the marches will be more likely to register to vote when they reach 18, or that non-citizens involved in the marches may be more likely to follow the path to naturalization (and be eligible to vote in the future).

It is also possible that the AP analysis has some problems, for example, according to the reports (and the figures cited above) in some cases they were comparing changes in registration between a presidential election year (2004) and this midterm federal election year; a better comparison would be to have looked at the 2006 midterm, relative to previous midterm elections (2002 and 1998, for example). It is well known that midterm elections are not as interesting to voters as presidential elections, and that groups and parties tend not to put as much resources into get-out-the-vote efforts in midterm elections.

One thing is for sure --- this will be a research question of interest in the social science community for years to come!

From the New York Times.

From El Universal (Mexico City)

From CNN

The Palm Beach Post reports that early voting in Florida went off smoothly, even with the natural intervention of Hurricane Ernesto.

Election officials told the paper that the number of voters was up and lines were down, due to higher numbers of early voting stations.

The Miami Herald, in contrast, writes that early voting "needs more tweaks." The number of overlapping jurisdictions, as well as voters with language problems, confused some poll workers.

There are an array of stories about the problems in Mexico. As the Boston Globe reports today in a great piece that:

Mexico's electoral tribunal has rejected his call for a recount, dismissed his claims of widespread vote rigging, and is widely expected to declare his rival to be the country's president-elect. But Andrés Manuel López Obrador does not look like a man who is about to slink into the political shadows. Instead, two months after he lost the official count to Felipe Calderón by a razor-thin margin, López Obrador is stepping up his campaign of ``non violent civil resistance" and vowing to challenge the very legitimacy of Mexico's institutional order for months, if not years, to come.

Such fiery talk has some analysts talking about insurrection, and others fearing a violent right-wing backlash. Even those who predict the movement will fizzle out on its own expect a degree of political chaos at least until the handover of power in December. ``We have no respect for their institutions," López Obrador, the charismatic leftist and former Mexico City mayor, said during one of his recent addresses to supporters manning a huge sit-in protest in Mexico City. ``We are going to create our own institutions that belong to the people."

In the official count of the July 2 poll, López Obrador lost to Calderón of the governing National Action Party by just 244,000 votes -- a margin of less than 0.6 percent. The leftist and his supporters of his Democratic Revolution Party cried foul and lodged an appeal in the electoral tribunal. But he has put more effort into protesting on the streets than in the courts.

The protests have gotten to the point that on Friday, President Fox saw the podium taken over when he attempted to make his end of term address to parliment. As the BBC noted,

dozens of Obrador's deputies took over the podium and prevented President Vicente Fox from making his state-of-the-nation speech. Instead, Mr Lopez Obrador told his followers that he would go ahead and set up what he called a "national democratic convention" in two weeks' time. He has already hinted at establishing a parallel government.

What if something like this happened in the US? What if the 2006 or 2008 elections are razor sharp and one party or the other refused to end their control of one chamber of Congress or established a shadow presidency? Given the rash of examples internationally of parties doing things like this and the such intense partisanship we have today in Washington, it will be interesting to see if such problems arise here.

First, I wanted to point out that Paul's blog posting from APSA was our entry 500. I get to do blog 501, which is a sad one, the death of Warren Mitofsky, the founder of the exit poll. Below is part of his obituary from CBS News.

Warren Mitofsky, a pioneer in exit polling, a giant in the field of survey research and the former Executive Director of the CBS News Election and Survey Unit, died Friday in New York City of an aortic aneurysm. He was 71.

Mitofsky made seminal innovations in polling methods including the first exit poll used to measure the opinions of voters as they left the polling place that was initially only for CBS News, but the methodology was later adopted by all major news organizations. Also while Executive Producer of election night broadcasts at CBS News he started the CBS News polls in 1969.
One of the most active figures in exit polling up until the time of his passing, he was most recently conducting exit polls for a consortium of U.S. television networks, and internationally as well. He served in leadership positions of the nation’s major professional polling organizations, which awarded him their highest lifetime honors.

From 1967 to 1990, Mitofsky was Executive Director of the CBS News Election and Survey Unit, and was an executive producer of its election night broadcasts. He developed and conducted the first exit polls for CBS in 1967, and he also developed the projection and analysis system used successfully by CBS and Voter News Service.

Today, the methods behind the exit polls that give voice to America’s voters, and the mathematical models that help estimate election results, are in large part the result of his ingenuity and creativity. As Dan Rather once told the nation, as a heated election night’s results poured in, "I believe in God, Country, and Warren Mitofsky."