Skip to main content.

Sunday, November 09, 2008

Votes continue to be counted and thus states have not released figures for the turnout for this election. Yet we see analyses like this one at Politico asserting confidently that turnout was no higher than in 2004. The point is to call into question the success of the Obama campaign's strategy to increase voter participation.

The Democratic increase struck some analysts as modest, considering the party’s immense get-out-the-vote operation, strong anti-Bush sentiment and Obama's popularity.

The argument is based on a painfully obvious fallacy, however.

It happens that I've been working for several days on research for a post about an election irregularity I uncovered in Pennsylvania. So I was able to spot the fallacy immediately in the Politico article, when the author David Paul Kuhn tried to produce some evidence to support his argument that turnout failed to increase this year:

In Pennsylvania, 5,851,730 voters cast ballots with 99.8 percent of votes counted — a rise of nearly 690,000 voters over 2004, according its secretary of state. But due to higher registration, the percent of eligible voters who cast ballots dropped from 68.96 in 2004 to 66.8 this year.

The first statement is false. The figure 5,851,730 is the number of votes that have been tallied for president in PA. Even without looking at PA results precinct by precinct, as I've done, Kuhn ought to have known that a significant number of voters did not cast a vote for president. That's especially going to be true in a year when so many Republicans were dissatisfied with their party's nominee.

We'll just have to wait for the official numbers for turnout. In the heavily Republican rural towns in my region of PA, the presidential vote generally was as high or slightly higher than in 2004. These are areas where, I believe, few new voters were added to the rolls. Thus the turnout increased or at least held steady in this demoralized Republican area.

In any case, we also need to consider the impact of lost votes – especially those not tallied correctly by electronic voting systems (I'll have more to say on this later), as well as the voters who were repulsed by long lines on Election Day.

At the polling place where I was working, the line became ridiculously backed up for much of the day. A Republican party "striker" showed up and, whether deliberately or through some combination of stupidity and incompetence, created a bottleneck that slowed the voting nearly to a halt. I recognized the problem immediately and called in for help. Obama campaign workers and lawyers came out repeatedly to help put a stop to this snarl-up, but even so we could barely contain the damage that this one Republican operative was doing. Thus as the day dragged on, scores of potential voters just walked away without casting their vote. In the space of just 20 minutes in late afternoon, I counted a dozen voters who left the line or turned away when they realized they'd have to wait for hours.

Poor planning, unnecessary delays, and lousy technology – which we continue to suffer from - all contribute in some degree to holding down the turnout. It's not reasonable to assume, on the basis of limited evidence, that the Obama campaign failed to get its voters to turn out on Election Day.

Comments

9 comments

[1]
To get the most accurate figure for turnout, you need to have the number of ballots counted, not just the number of votes tallied for the highest office. That's doubly important in years when voters are unenthusiastic about one of the candidates - as they were this year.

Also worth noting that the turnout rate should be calculated using the number of eligible voters (rather than the number of residents who are older than 17, since that figure also includes non-citizens and felons who've lost the right to vote). By that measure, turnout in PA in 2004 was somewhat above 62.6% (PA didn't give a total for ballots counted, thus the uncertainty).

http://elections.gmu.edu/Tu...

Posted by smintheus at Sunday, November 09, 2008 14:33:07

[2]
I really appreciate your tenacity in pointing out this glaring fallacy that enables the Villagers to diminish Obama's victory.

Another issue worth noting that appears to have escaped Kuhn and others — the significant number of voters purged from the rolls prior to the election. In some cases duplicate records, deaths and felony convictions of voters were correctly removed as eligible voters.

In other cases, the Katherine Harris' run amok attempted to disenfranchise voters with easily curable registration errors.

The UK Independent reports:

"Around 13 million US voters have been purged from the electoral rolls since 2004. That's 10 per cent of the 120 million votes cast in 2004 and twice as many voters as have been added through recent massive voter registration drives.

The proportion of electors dropped from the voters' lists is staggering: 17 per cent in Colorado, 15 per cent in Washington State, 14 per cent in New York, 13 per cent in Nevada and 10 per cent in Missouri."

More here: http://blogs.independent.co...

So in essence, any talk of year-to-year turnout comparisons must take into account those voters removed from the rolls.

I suspect that Kuhn's figures do not and therefore the impact of active voters plus newly registered voters would actually be much higher than those who turned out in 2004.

Posted by em dash at Sunday, November 09, 2008 15:04:26

[3]
All of that's true. I probably should have mentioned the large purges of voter rolls that took away some voters even while others were being added by registration drives.

The George Mason elections website also points out that the widespread belief that turnout has become appallingly bad during the last generation is based on miscalculations of turnout rather than reality. That's because the widely reported turnout stats have been based on the entire population - which includes an increasing number of people who are NOT eligible to vote. If you calculate turnout based on who's eligible, you see that the rate dipped during the 1970s but has bounced back since 1988. In 2004, the turnout rate was close to the high levels we saw during the '50s and '60s.

http://elections.gmu.edu/vo...

Kuhn is a reliable "concern troll" with regard to things Democratic. Over at The Impolitic, Libby maintains that this Politico report downplaying the size of the turnout should be read in light of Kuhn's book, The Neglected Voter. It argues that the Democratic Party has lost the support of men, esp. white men, and thus the GOP has become the predominant party. That would be hard to square with voter enthusiasm this year for the Democratic ticket.

http://theimpolitic.blogspo...

Posted by smintheus at Sunday, November 09, 2008 17:09:07

[4]
The word is that there were serious problems all over in terms of the process of voting.

I know the Obama folks are very busy now getting ready to clean out the Augean stables that are the Bush Administration, but someone needs to focus on what happened and clean this mess up.

I strongly advocate a push to replace those incompetent county clerks across the country with people who can do the job. This position is an important squeeze point for preserving democracy. We need to take it seriously and put very good people in these positions

Posted by shirah at Sunday, November 09, 2008 19:10:41

[5]
Good point. I read some line from Captain Ed claiming a ridiculously low turnout nationally - which Howard Kurtz quoted without fact-checking, of course.

Posted by Batocchio at Monday, November 10, 2008 11:55:44

[6]
Yes, I agree with comment #4. I can speak a bit for Illinios, or Cook County anyway, where Chicago is.

Myself and friends who voted at several different polling places were all asked for id. I considered myself a very informed voter, having read about what can go wrong with the machines, to call Election Protection or other hotlines if I'm offered a provisional ballot as it probably wouldn't be counted... So when I was asked for id, I figured that that was unfortunately the law in Chicago or Illinois. Later I found out from an attorney friend working for Election Protection that it is not the law in IL!!! My friends who voted in other polling places were asked as well.

I wonder how many people without ids were turned away or given provisional ballots? My attorney friend said when he told the poll workers that they had to let people without ids vote, they argued with him. They were totally misinformed.

I also have one friend whose name was purged because she changed her address recently, and another who experienced the Sequoia machine break down while she was voting.

I don't know why any precinct would even purchase Sequoia products. Well, I guess there's something wrong with all of them, but here's what I read about Sequoias,

"To comply with a federal court order [HAVA], Nassau County purchases ballot-marking devices from Sequoia Voting Systems. Of the first 240 devices delivered, 85% are too defective to be usable, placing the county in jeopardy of violating the court order."--
http://www.issuelab.org/res...

Posted by Vaness at Monday, November 10, 2008 12:22:15

[7]
The typical measure of turnout when comparability is needed, is either percentage of the voting aged population, or the percentage of eligible to register voters. Either way registration is viewed as simply a barrier to voter turnout (one of the most significant barriers, although hardly all of the problem as states with election day voter registration illustrate).

The percentage of people who vote, but not for President is a theoretically limiting factor, but the percentage of voters who do not vote for President is far lower than for any other office, so it is a reasonable number to use for a good estimate. In Colorado, the number of votes cast for U.S. Senate was about 3% less than the number cast for President, and the number of votes cast for the most voted upon ballot issue was about 4% less than the number of voters cast for President. Also, do we care that much about voters who show up but don't even vote for President?

Also, there is no good reason to suspect that the number of people who don't vote for President varies much from election to election. So the numbers should be fairly comparable.

Posted by ohwilleke at Monday, November 10, 2008 13:34:26

[8]
Vaness, the poll workers in my precinct actually asked me to identify my party registration on Election Day. An Obama campaign worker told me that was being done elsewhere too, and that poll workers were certain (wrongly, of course) that they had to ask voters for party ID.

Ohwilleke, as I said higher up, there's a big difference in turnout figures if you calculate by eligible voters as opposed to voting age population.

I'd like to see some stats before I assume that the proportion of people who vote for president remains as high in years when one or both major candidates is not very admired as in years when voter enthusiasm is high for both. The number of voters (mostly Republicans) who've told me that they'd refuse to cast a vote for president this year was much higher than in years past.

Posted by smintheus at Monday, November 10, 2008 23:59:13

[9]
Also, upon checking Kuhn's figures further, I discovered that he made an error about the 2004 "turnout" in PA. There were 5,765,764 votes cast for the five presidential candidates who appeared on the ballot in 2004. So the increase from 2004 to 2008 was not "nearly 690,000 voters" as Kuhn claims.

Posted by smintheus at Tuesday, November 11, 2008 00:08:54

Add Comment

This item is closed, it's not possible to add new comments to it or to vote on it

buy viagra viagra alternative order viagra online viagra sale buy viagra online generic viagra online cheap viagra