Don't Forget Not to Vote

The lack of accurate knowledge of the likely effect of Pennsylvania's voter ID law on election turnout has been frustrating. Estimates of the percentage of citizens lacking ID vary considerably because they are inferred indirectly from large public data bases. It's obvious that the best way to find the answer is to interview a random sample of Pennsylvania voters and ask them directly whether they have the identification required by the law.

Fortunately, a survey was commissioned by the plaintiffs in Applewhite, et al. v. the Commonwealth of Pennsylvania, an ongoing legal case brought in Commonwealth Court by a coalition of groups that are challenging the law. It was conducted by Dr. Matt Barreto, a survey researcher from the University of Washington. Dr. Barreto's report is available online, along with appendixes containing the wording of the questions and the tables of results.

The survey was conducted by telephone using random digit dialing, which ensures that all telephone numbers—listed or unlisted—have an equal chance of being chosen. A sample of 1285 people was obtained, 78% of them from landline exchanges and the remainer from cell phones. Since the impact of the law on Black and Hispanic voters was of special interest, they were deliberately oversampled from commercially-available lists, which should improve the accuracy of the estimates for these subgroups. The survey has a margin of error of +/-2.7%. That means that if 50% of the respondents say their favorite color is red, then 95% of the time the true percentage who prefer red will be between 47.3% and 52.7%.

34.6% of those who answered the phone agreed to participate in the survey. Since about 30% of the calls were not answered (in three tries), the response rate—the percentage of those called who completed the survey—was 24.2%. I started doing research when people seldom screened their calls and were more willing to answer research questions, so this response rate seems low to me. However, a response rate of 20-30% is considered acceptable by contemporary standards.

All of the respondents were asked as many questions as necessary to determine whether they had a valid photo ID under the law in their possession—either a current driver's license or any one of ten other acceptable forms of ID. To be valid, the ID has to have an expiration date and be issued in their current legal name. The major results were:

1. 14.4% of eligible voters (1,364,433 people) lack a valid photo ID under the law. Looking only at registered voters, it is 12.8%, and 12.6% of those who voted in 2008.
2. 37.3% of eligible voters, 34.3% of registered voters, and 34.2% of 2008 voters don't even know that the voter ID law exists.
3. 97.8% of eligible voters, 98.8% of registered voters, and 98.7% of 2008 voters believe they have a valid ID. Comparing these figures to the percentage who actually have a valid ID shows that many people are mistaken and unlikely to do anything about it before election day.
4. If you don't have a valid ID, in order to get one you need proof of citizenship, proof of address, and a social security card. 27.6% of those who do not have a valid ID will be unable to obtain one before election day because they lack one or more of those documents.

The impact of the law on the election depends on the differences among demographic subgroups in eligibility to vote. The following data are for eligible voters, although the differences are quite similar for registered voters or those who voted in 2008.

1. Gender. Women (17.2%) are more likely to lack a valid photo ID than men (11.5%).
2. Age. People over 75 (17.8%) and under 35 (17.9%) are more likely to lack ID than middle-aged people (10.3%).
3. Race. 14% of Whites, 13.2% of Blacks and 18.3% of Hispanics lack a valid ID. The nonsignificant difference between Whites and Blacks was a surprising finding.
4. Education. Voters without a high school diploma (18.5%) were most likely to lack ID. As education increases, lack of ID declines. Only 8.3% with a college degree lack ID.
5. Income. Income is the strongest predictor of lack of ID. 22% of those making less than $20,000/year lack valid ID, compared to only 8.2% of those making more than $80,000/year.
6. Region. Voters living in Allegheny—that is, Pittsburgh and vicinity—(18.7%) and Philadelphia (17.8%) Counties are less likely to have ID than those living in the rest of the state.

In 2008, Barack Obama defeated John McCain by 10.3% in Pennsylvania, which was considered a landslide. Early indications are that this year's election will be much closer. Obviously, disenfranchising 14% of eligible voters can have a considerable impact on the outcome, especially since almost all the subgroups that are more likely to lack ID currently show a preference for the President.

The survey shows that, in passing this voter suppression law, Pennsylvania's Elephants have succeeded beyond their wettest dreams. When House Majority Leader Mike Turzai told his fellow Elephants that the law would “allow” a Romney victory in November, this was no idle boast. A better word might have been “ensure.”

Prior to this month's hearing in Commonwealth Court, the Corbett administration agreed in a court document that it knows of no cases of in-person voter fraud in Pennsylvania, and does not anticipate any in November. However, they argue that evidence of fraud is unnecessary to justify the law. It is sufficient that legislators can “imagine” the possibility of fraud.

If the court allows this outrageously bad law to stand, it's time for Pennsylvania's citizens to grab their torches and pitchforks and begin the long march to Harrisburg.

Addendum

I'm glad to see that the plaintiffs are thinking about one of the less obvious but more important impacts of the voter ID law—its tendency to slow down lines at the polls. Long waiting lines at the polls are already a problem in many areas of Pennsylvania, especially in low income areas and near college campuses. Unfortunately, if people give up and go home rather than face long queues, there will be no public record of the fact that they have been discouraged from voting by the law.