A poll from the Catholic Association was taken August 19-21, and found that President Obama has retaken the lead for the Catholic vote. The survey, which had a sample of 2,629 "likely voter" Catholics, asked the question, “If the election for President were being held today for whom would you vote if the candidates were Barack Obama, Democrat, or Mitt Romney, Republican?” 49% of all surveyed went with Obama, while 41% went with Romney.
Then, a recent Gallup Poll found that Romney’s lead among Catholics dissipated through August from 46-45% over Obama in the first week of August, to an Obama lead over Romney, 47-45% on September 9th. These two polls have elicited some concern from conservative Catholics.
However, let’s take a closer look at the data and the method of collecting the data.
The Catholic Association poll’s sample was 48% Democratic, 33% Republican, and 19% Independent. Even if we assume a two or three point Democratic bias in Catholics, +15% is an extremely large Democratic edge.
In addition, the poll was taken with autodial methodology, which experts believe can affect accuracy. Standard telephone polls use live interviewers, while autodial surveys use automated voice questioning. A “plus” for autodial interviewing is that people may tend to exaggerate their likelihood to vote when interviewed by a live person but respond more honestly when prompted by an automated voice. The recorded voice may allow survey responders to feel less pressured and give the sense of a higher level of privacy than a live person. So, with regard to “likelihood to vote,” the autodial method may provide more valid data.
On the other hand, autodial methodology has its negatives. First, while about 30% of people telephoned by live interviewers actually participate in the survey (called the “response rate” in “poll-speak”), fewer people are inclined to respond to an automated call. A lower response rate is often associated with a higher possibility of bias in the sample.
A second concern of autodial polls relates to the randomness of the sample. In standard, “live-person” polls, the interviewer may ask the initial responder if he or she can speak to another household member, thereby adding another tier of randomness to the survey. For example, a poll interviewer may ask to speak to the person in the household whose birthday is closest to the polling date. Automated polls aren’t able to do this, leading some polling experts to assert that by removing the additional opportunities for randomness, autodial pollsters make it necessary to heavily weight their samples to obtain demographic diversity. But such weighting then affects the validity of the poll itself.
There are two more interesting features, however, in the Catholic Association poll results. First, responses to the survey’s questions were provided for all respondents, and then for respondents divided into groups depending on the regularity of their Mass attendance. When Catholics were divided in this way, 47% of “regular” Mass-goers would vote for Romney, while 45% would vote for Obama. For those who do not attend Mass regularly, 53% would vote for Obama, while 36% would vote for Romney. True, in the “regular Mass” category, Romney is only up 2%, but we need to remember the D+15 advantage.
However, even when Catholics are not divided by Mass attendance, the Catholic Association poll overwhelmingly demonstrates that Catholic likely voters believe that opinions associated with President Obama and his administration’s policies have been unfavorable. 78% of all surveyed said our rights come from nature and God, not from government. In reference to the HHS mandate, 66% said that religious charitable institutions should not be forced by the government to provide or pay for goods and services to which they morally object. Regarding First Amendment freedom of religion rights, 57% agreed that the Obama administration has gone too far in placing restrictions on religious freedom when implementing their programs and policies.
The Gallup poll used a sample of 9,000 registered, not likely voters, 2,000 of whom say they are Catholic. The likely voter screen is more accurate than that of registered voters because registered voters are not as committed to vote as likely voters. Registered voters tend not to be as interested in politics and often say they are Democrats because the messaging of the Democratic Party has been more effective.
It goes without saying that, if conservative Catholics want Romney to be the next president, they need to get out the vote. However, neither the Gallup nor Catholic Association polls may accurately reflect a Catholic nod to Obama.