Thursday, September 29, 2005

How to interpret the EDS / EAC Report

I've received a few emails regarding the EAC Election Day Survey report prepared by Election Data Services.

In particular, my comments regarding race and absentee ballot usage seem to be contradicted by Chapter 5 of the report.

I'm happy to be proved wrong, but it's very important to understand how EDS is reporting their results. An individual "case" in the report is not the individual, but the "jurisdiction".

This means that you cannot assume anything about individual behavior, and more importantly you cannot derive national averages from these data without weighting by jurisdiction size.

Thus, this paragraph (from Chapter 5 of the report) ...

Predominantly Hispanic jurisdictions reported the highest request rate for absentee ballots, 13.6 percent, slightly more than twice the lowest reported rate in predominantly non-Hispanic Black jurisdictions, at 5.7 percent. Predominantly non-Hispanic White jurisdictions reported a rate, 10.9 percent, slightly lower than predominantly Hispanic jurisdictions. Predominantly non-Hispanic Native American jurisdictions reported a rate, 6.1 percent, slightly higher than predominantly non-Hispanic Black jurisdictions.

... tells us very little about the relative rates of absentee ballot usage among whites, blacks, and hispanics.

It says what is says. For instance, among "predominantly Hispanic" jurisdictions, the average rate of absentee ballot request was 13.6 percent.

Those jurisdictions may vary in population size from densely urbanized counties in Arizona, New Mexico, and California to rural counties in Southwest Texas, northern New Mexico, southern Colorado, and northern Nevada.

Minimally, if we want to make a rough estimate of national averages, we need to weight the jurisdictions by the population in each county.

Upcoming Ohio vote on early voting

Ohioans are voting on November 8th to amend their Constitution and allow early voting.

According to the sponsors, Reform Ohio Now,

"We look at this as a no-brainer in terms of increasing voter participation," Jerse said. "Long lines in 2000 and 2004 made more people interested in having no-fault absentee ballots.

"It's more convenient for people if they have children in day care or job conflicts, things like that."

Jerse said he does not believe competing campaigns would manipulate early voting.

"I don't know that it will make all that much difference to the way campaigns operate, except campaigns will take advantage of the fact that they can send literature to people they know are going to vote," he said.

A political scientist, Herb Asher, is heading up Reform Ohio Now. I have the greatest respect for Herb's work, so I'm disappointed that his organization is misrepresenting all the scholarship on early voting.

As I've posted many times here, there is virtually no evidence that early voting increases the electorate. Yes, it does increase turnout but only in low-profile contests, and only by encouraging turnout among regular voters. It does not bring new citizens, disempowered citizens, young citizens, or disaffected citizens back into the process.

The claim that it won't alter campaigning is unbelievable--talk to anyone who runs campaigns in early voting states, as I have and as Bob Stein has down at Rice, or just call up the RNC or DNC. Early voting has a tremendous impact on campaigns. They have to run a two week rather than one day mobilization effort. They waste money because they don't know who has already turned out.

I applaud Reform Ohio Now's efforts to increase citizen awareness, remove the impact of big money on campaigns, and make redistricting a non-partisan process. They are attacking trust in government--THE key reason that many citizens don't vote.

But voting by mail has little to do with these goals, and by reducing the impact of election day as a community wide civic event, they may end up hurting turnout more than helping it.

If VBM is adopted, it should be adopted because a) so many citizens already vote absentee (as in Washington state), b) because citizens like it, and c) because the ballot counting process is less expensive and more accurate.

I'd prefer we try things like having election day holidays or 24 hour voting periods, however, before we move wholesale to voting by mail.

Story referenced is here

Monday, September 26, 2005

"Out of Country" Voting (OCV) in Conflict Zones

Just got the new copy of "Democracy (at sign) Large"in my mailbox.

In a special section on elections, there is a fine article by Ben Goldsmith, current the Deputy Chief Elections Officer in Afghanistan.

Ben writes about the challenges facing a nation that wants to encourage political participation by all of its citizens, but which has a large diaspora, in the case of Afghanistan and Iraq (the two cases considered in the article), disaporas caused by recent war and continuing civil unrest.

I can't do justice to the article here, and encourage everyone to look at it, but in brief, Ben considers:

  • The political question. Do domestic political actors want the diaspora to vote? How large is the diaspora relative to the domestic population? Will including the diaspora help stablize the new regime by encouraging participation of citizens presumably living in more stable settings, or will there be suspicion about those living abroad, presumably subject to "foreign" influences?
  • The registration question. Presuming that you have decided to allow citizens living abroad to vote, how do you confirm their qualifications?
  • The diplomatic question. Not every county will welcome campaign mobilization by a foreign nation within its borders. In the case of Iraq and Afghanistan, Ben has a nice discussion of how "Memoranda of Understanding" (MOU) were required with Iran, Pakistan, and 12 other nations.
  • The cost and administration questions. Simple enough: who pays for what will be more complicated and presumably more expensive election machinery? How will you actually administer the ballots (by mail? with local election machinery? at diplomatic missions?).

These are but a few of the important questions that are raised by OCV elections. In the specific cases of Iraq and Afghanistan, as an illustration, the country's handled the difficult issue of timing (elections were held so rapidly after the end of the conflict) by allowing citizens to "register" and "vote" at the same time, and then keeping the records available for challenge.

I've written about diaspora and OCV elections in this blog. Mexico has legalized voting by citizens living in the United States. This will in all likelihood result in massive GOTV and campaign rallies in LA, San Diego, and other American cities. Other countries, including Ghana, Phillipines, Israel, and Sri Lanka have recently relazed expatriate voting requirements.

It will be interesting to see how the administrative hurdles are overcome, but perhaps more interesting will be the new political dynamics that are created by these new, large, and potentially highly influential expatriate voting segments.

Sunday, September 25, 2005

For a real debate over REAL ID's

Is it possible to have a real debate about the REAL id?

An editorial in today's St. Petersburg Times typifies the short of shallow thinking that has surrounded the Carter Baker commission proposal an identification requirement at the polling place.

First, the Times claims that the identification requirement discriminates against individuals who do not have proper documentation (birth certificate, social security card).

But wait! How can such individuals register in the first place--don't they need such documents to even get on the rolls? Requiring an ID at the polls will not reduce turnout among people without proper Idocuments because they'll never have been registered in the first place.

Second, the Times makes the bad Katrina analogy. The survivors of Katrina, they argue, would have a hard time coming up with proper documentation and would be disenfranchised.

Of course, this is policy making of the lowest common denominator. We should not fashion our election laws based on the worst natural disaster to hit the US in a century or longer. Of course, we should figure out ways to ensure that hurricane evacuees--and anyone displaced by a natural or man made disaster--has their voting rights assured. But that is a special, particular case, and one that requires a special, particular solution. (In fact, Mike Alvarez, Thad Hall, and I are proposing to study just this kind of circumstance. )

REAL Id's should be debated on their merits, not on made up crises or extreme scenarios. The relevant considerations are:

  • Are Americans ready for a national ID card? American's historical distrust of central government says "no."
  • Can ID cards be made free? They must be in order to be equitable.
  • Can the other required documentation be made very low cost or free? Theoretically there is no reason to say "yes" but this requires coordinating across literally thousands of local jurisdictions.
  • Can we formulate a way to subject absentee ballots to the same level of scrutiny? The answer is "yes" -- there are technological solutions out there, if we'd only adopt them Thumbprint technology, for example, as high-tech as it sounds, is actually quite inexpensive and reliable.

Thursday, September 22, 2005

Race and Absentee Balloting

In a previous comment, Justin Levitt of the Brennan Center challenged my comments on race and absentee balloting.

This is an issue I've been interested in, but have been frustrated by the lack of good datasets. I've begun poking around in response to Justin's comments. Of course, what did I find but my own publicly available work does show continued racial disparities with respect to early voting.

See page 13 of my report, "Early Voting: Lessons for Progressives," delivered to the Progressive Targeting Conference at the Center for American Progress.

The data are still inconclusive, however. I have merged all types of "early" voting in this table, and the disagreement concerns absentee balloting only (in an in-person system, presumably ID would be checked no differently than on election day). Also, I did not control in those tables for geographical differences in the availability of early voting (and I don't place much credence in the tables for that reason--see my comments in the text).

I'm going to pursue this a bit further and will post any results here. If the Brennan Center has the resources to pursue this question, I'd start by going to a state that reports race in its registration and turnout datasets, and then compare the absentee percentages by race (for example, North Carolina).

The other thing to do is to analyze the NAES data (only 2000 is available at this time), predicting absentee ballot usage by race and a variety of other indicators. I'm sending off an RA to pursue this avenue.

More Carter-Baker commentary

Dissent from Richard Hasen in today's CSM. It captures much of what Hasen wrote on his blog, highlighting the positive features of the report but dissenting on the ID question. I don't think Hasen is opposed to an ID requirement, but thinks that the mechanics are problematic at this point. And by taking such a strong position on the ID, Hasen believes the commission will undermine its own effectiveness.

Dan Tokaji at OSU Law School has a similar take. Dan helpfully summarizes other commentary.

Finally,'s Doug Chapin gives a third summary of opinions (with a particularly nice list of newspaper editorials) in his weekly report (Sept 22 date).

Monday, September 19, 2005

Brennan Center folks wrong on race and absentee balloting

The Brennan Center has published a long response to the Carter-Baker report. I've been working my way through it tonight.

One thing jumped out at me: their claim that there is a "double standard" with respect to absentee balloting, that is "especially disturbing in light of data, examined by the Commission's predecessor, that white voters are about twice as likely as black voters to cast an absentee ballot."

This is not supported by the original report that they reference. Hansen reports data from the 1996 Current Population Survey regarding absentee balloting, showing that 9% of whites report voting absentee while only 4% of Blacks report voting via this mode.

Here's the problem--in 1996, a number of states had already begun to relax their absentee balloting requirements. States with large African American populations, especially those located in the South and NE, still had very restrictive rules. We don't know whether the differences found in the 1996 CPS report are due to regional differences in laws or something about race.

Stein's work in Texas is cited just below the table in the Hansen report (on pg. 4), and I quote: A better sense of the demographic composition of early and Election Day voters comes from a study of Texas by Robert Stein. The Texas investigators interviewed voters as they exited polling sites on Election Day and as they exited early voting stations during the three-week period of early voting in 1994. Stein found some significant differences between early and Election Day voters. The most important was age: Just as older voters tend to vote absentee so older voters tend to vote early. In fact, more than a third of the early voters in 1994 were over the age of 60 and more than half were over 50. In contrast, the youngest voters, who are overrepresented among absentee voters, were underrepresented among early voters. The circumstances that cause absentee usage among the youngest voters, absence from the jurisdiction during Election Day, are also the circumstances that preclude early voting. Stein also found some small differences in early voting by gender, with men more likely to be early voters than women, and by income, with poorer voters more likely to turn out early (the opposite of the pattern for absentee voters).

This is in line with other data that I've seen to this point: there are few if any racial differences in early or absentee voting, and what patterns exist seem to be driven more by partisan mobilization patterns than by anything fundamentally racial.

When to remove partisanship from partisan elections

One last reaction to the Carter-Baker report.

I agree with Mike Alvarez, Thad Hall, and Rick Hasen: the CB call for non-partisan election administrators is a good idea. It would be simple to implement, and in fact should save money in those many states and localities that elect such officials. It will enhance professional growth and development in this very important area. And it ought to result in more competent election administration.

What occurs to me is that this proposal is in the same spirit as calls for non-partisan redistricting commissions (for instance, Prop 77 in California), a position my totally unscientific survey and conversations with political science colleagues says is the conventional wisdom among scholars.

Many bemoan the growth of partisan polarization in the US Congress and many state legislatures, and somehow want to get parties out of politics. Here in Oregon, we're going to vote on a top-2 primary system in 2006; there is another proposal for a non-partisan legislature that may also make it to the ballot.

Non-partisan redistricting commissions, one that would draw lines without respect to incumbency, but would honor existing communities of interest, is a better solution. This way, we'd avoid politicians who draw lines only to benefit their own party or their own political position. We'd stop the practice of cutting up cities and neighborhoods just to allocate particular types of voters. And I suspect, even in the short run, elections would become more competitive and we'd see fewer partisan extremists.

Yes, some areas of the country have become reliably Republican or Democratic, and there is no way to force a competitive two party election in those areas. But in a state like Oregon, there is no reason that we should not have four or even five competitive contests each cycle.

We can't remove parties from politics, but that doesn't mean we can't help make party politics more competitive.

Carter Baker Report

Much reaction about the election reform community on the Carter Baker commission report. Fairness in advertising; I was commissioned to write a report on ballot integrity and vote by mail for the commission.

The sections of signature checking and vote by mail come from my work. I had quite a time convincing the folks at the Commission that Oregon really checked every single signature. In one interesting conversation, I just kept repeating "yes, they check every single signature."
"How many signatures do they skip?"
"EVERY Single Signature."

This point has been missed by some critics who, focusing on the REAL ID proposal, believe that absentee and vote by mail provisions treat certain voters differently. Not true--REAL IDs woud presumably be used when a voter first registers, from then on, signature checks would be used.

Some other thoughts and reactions:

  • Rick Hasen is, as usual, right on target, highlighting the positive aspects of the report but complaining about the scattershot recommendations. As he predicted, blog and press relsease reactions already focus on some of the less central aspets of the report.
  • Rob Richie ( , at, criticizes the report for not recommending IRV.
  • The directors of at Counterpunch, criticize the report for not suggesting ways to break the two party monopoly on elections.
  • John Conyers Jr. says that the recommendation for IDs is a thinly veiled Republican plot to disenfranchise citizens.
  • Mike Alvarez and Thad Hall provide their own reactions, noting that some of the recommendations repeat initiatives taking place on other fronts. Most importantly, the suggestion for interoperability, far from being "expensive and complex" (according to Tova Wang), could actually simplify voter registration and ballot administration nationwide.
  • Tova Wang is right that the Commission was particularly interested in ballot integrity issues. I suspect, however, that they agree with Wang that voter distrust is an important issue; they disagree with her whether ballot integrity matters to trust.
    Our own research, albeit limited to Miami-Dade, indicates that those voters who do report concerns that their ballot will be counted accurately are significantly less likely to take advantage of new balloting systems.
    All the focus on election administration won't solve the participation problem anyway. Fundamentally, the way to increase participation is to increase feelings of empowerment and efficacy in the population.


My pirate name is:

Captain Sam Bonney

Even though there's no legal rank on a pirate ship, everyone recognizes you're the one in charge. You can be a little bit unpredictable, but a pirate's life is far from full of certainties, so that fits in pretty well. Arr!

Get your own pirate name from

Sunday, September 18, 2005

Bill in US House attempts to guarantee absentee voting rights to displaced citizens

Proposed legilsation in the US House of Representatives attempts to assure absentee voting rights to citizens displaced by Hurricane Katrina.

HR3734, sponsored by Alabama Rep. Arthur Davis, extends overseas voting rights to citizens who certify that they have been displaced by katrina. It looks like all the bill requires is a "certification" or "affidavit" to a state official.

Interestingly, the impact of this is that it may allow Louisiana residents to continue to vote in LA elections through 2008, even though many may never return to the state.

Demopolis Times Story

Wednesday, September 14, 2005

Belated Blogging of APSA

This is a belated posting on APSA. Sorry that I didn't get the chance to write during the conference, or after. The first two weeks of the semester at Reed College can be overwhelming.

My panel on Early Voting and the 2004 Election went well. The panel started out with a little humor from the panel chair, John McNulty of University of Binghamton (newly minted PhD out of Berkeley). I wasn't sure if John knew that I had organized the panel and solicited all the papers. I didn't expect the section head to choose a chair and discussant, but was happy to get that duty out of my hands. But since I was scheduled to deliver one paper on Early Voting in Oregon, and had added my name to another paper on Early Voting in Florida, and had organized the panel, well, John decided I needed to be on all the papers. Ha ha.

But I deserve no credit for pushing this agenda forward. The other panelists are doing excellent work in the area. First off was Kate Kenski, newly minted PhD out of U Penn (Jamieson student, worked a lot with NAES). Kate has foolishly taken a job in the Communications Department at University of Arizona! Kate assures me that she'll continue to come to political science meetings. I hope so--her work in the NAES has been excellent. While I wish the NAES would release their data to the academic community a bit sooner (it's already been nearly a year, folks!), it's only through Kate's efforts that we have a battery of items on early voting, including the only survey that asks whether an early voter "regrets" casting the ballot early.

Kate found that early voters are somewhat older, somewhat better educated and informed, but otherwise differ little on most attitudinal measures. (I replicate this finding with different surveys here, here, and here.). Where Kate needs to go next is to push some of these results in a multivariate direction--her paper included a host of bivariate differences (or lack thereof), but we don't know how many of these are sustained.

Mike Traugott delivered the next paper--Mike has been working with a group of scholars on ballot technology issues, supported by the NSF. (I have to highlight the work of Mike Hamner, a third newly minted PhD on this panel, who, in collaboration with Mike, has published some excellent papers on vote by mail in Oregon. Unlike Kathleen, Traugott does the right thing and keeps his students in Poli Sci!). The paper seemed somewhat preliminary, but reported on some fascinating experimental and empirical comparisons between punchcard and optical voting machines. The data were pretty clear--the computer based technology is far superior by almost any measure (voting errors, roll off, etc). I expect more coming from this group over the next year.

My own paper, co-authored with Benjamin Bishin of the University of Miami and Daniel Stevens of Hartwick College, along with Eva Galanes-Rosenbaum, my RA at Reed College, can be found at the website. I won't go too deeply into the results here--I am most interested in our finding that people who trust government more are willing to vote early, an obvious finding perhaps upon reflection, but one that was unexpected and has been up to now unexamined in the literature. Mike Alvarez and Thad Hall blogged on the paper; go read this for an able critique.

Finally, Jan Leighley (newly moved to Arizona--Congrats Jan!), Bob Stein (newly returned to the Department from the Dean's Office--congrats Bob!), and a student whose name I forgot, continued their research on in-person early voting and party mobilization in Texas. This was the best and most complete paper on the panel. In it, Leighley and colleagues argue that party mobilization efforts surrouding early voting are only likely to help Democrats, because Republican early voters are committed early voters. The mobilization push does not increase their numbers.

Very interesting work for anyone interested in how early voting may affect mobilization strategies.

If I find URLs for these papers, I'll post them in a few days.

Tuesday, September 13, 2005

Absentee balloting and Hurrican Katrina

Today's story highlights something I had not considered yet in the aftermath of the tragedy that has beset Louisiana and Mississippi. Both states will be conducting elections in the next 14 months, a period in which hundreds of thousand of displaced citizens may not have moved back to their homes.

How will these citizens cast a ballot? Will they be allowed to cast absentee ballots? Louisiana law allows an absentee ballot for those "expected to be absent from their parish on election day." Are the parishes ready to handle nearly 1,000,000 absentee ballot requests?

And what does the future hold for Louisiana's second district, if the New Orleans population declines, as some predictions hold, by nearly 50%?

Elections are not on anyone's mind, and rightfully so. But within a few months, as candidates begin to consider runs for office, they will face a ew landscape in Louisiana politics.

Sunday, September 11, 2005

National Summit To Save Our Elections | A Call to Rescue Our Democracy

I've been invited to present a poster on Oregon's vote by mail system at the "National Summit to Save Our Elections" in Portland, OR.

The group seems partially affilated with Verified Voting (, a group that has been arguing (I think most ly unsuccessfully) that there was widespread fraud in 2004 (particularly in Ohio).

I have detailed political science's objections to this research in the newsletter of the Elections, Voting, and Political Behavior section of APSA (

I'll present the poster, however. There are some interesting speakers. And even if I disagree with some folks about Ohio in 2004, there is no reason not to give close scrutiny to our elections system.