The last polling update before the debates... Gallup's weekly demographics are out without any other companion polls reporting Catholic vote estimates recently. With a final fielding date of Sept. 30 (the most recent available), Catholic registered voters nationally favor President Obama 50% to 45% over Gov. Romney. This is the second consecutive week in which the Catholic vote narrowed. The current estimate is very similar to the first estimate of the race a year ago where Obama led Romney 50% to 47%. On average, for the series, Obama has led Romney 48% to 45%. The percentage of registered Catholics indicating they are "definitely voting" reached a series high this week of 84%.
The voting preferences of Catholics remain a bellwether mirroring national estimates. In 22 demographic breakdowns provided by Gallup since April, President Obama has led Gov. Romney among Catholic voters 59% of the time with an average advantage of +1 percentage point (Gallup estimates differ from those above as CARA's tracking includes estimates from any available survey reporting data for Catholic registered voters; aggregated through averaging). In addition to being a bellwether the Catholic vote could also be considered a "swing" sub-group that has not consistently polled in one direction in this election season. The table below puts the Catholic vote in this context compared to other sub-groups of voters in Gallup's surveys.
The top block of the table, in blue, are groups where Gov. Romney has never led, which President Obama is likely to win. The bottom block, in red, are groups where President Obama has never led, which are likely to be won by Gov. Romney.
There are a few groups that may be a bit surprising to some. The first is seniors, those ages 65 and older. Democratic Party candidates have not won this group since 2000 (i.e., in national exit polls) and President Obama does not look likely to win it in 2012 even with his campaign focusing heavily on Medicare, Social Security, and Rep. Paul Ryan's budget plan in an attempt to bring this group back into the Democratic Party coalition.
The second interesting group are those in households earning $36,000 to $89,999. This is the closest approximation one can arguably make in the breakdowns that reflects the "middle class." Despite this group being a major focus of the Democrat's campaign and party platform it is leaning toward Gov. Romney. President Obama has only led Gov. Romney once in Gallup's sub-group tracking summaries among this income group (i.e., Sept. 3-23). The President is currently down a percentage point (technically tied) with this group but the overall trend gives an advantage to Gov. Romney.
Finally, although many envision ideological moderates as being in the middle of the political spectrum, they are not in the middle of the electorate in terms of voting preferences. President Obama is leading strongly among this group by +23 percentage points and has led here in all of Gallup's estimates this year.
Of course no one wins electoral votes for winning a demographic sub-group of the electorate! In the end, the swing states are what really matter now. However, within those states demographics will be key. In reviewing recent swing state polls CARA estimates that there are more than 5 million Catholic registered voters combined in Colorado, Iowa, Florida, New Hampshire, Ohio, and Virginia representing about 24% of key "swing" electorate for 2012.
Note: A more recent update on the Catholic vote has been posted here.
A Note about the Polls
There has been a lot of recent discussion about "bias" in election polls. There was a similar discussion in 2004 that even extended to exit polling (...also a bit of this in 2008). Supporters of the party on the "wrong end" of the polls are often more critical (it was Democrats in 2004, Republicans now). Presidential elections are the Super Bowl and Oscars of polling. Reputable pollsters want to be as close to the eventual outcome as possible. So what is happening? Why is there so much volatility between polls? In my opinion this is not really about partisan or candidate bias and more about methods. As Nate Silver has pointed out, surveys that are under-sampling cellphone-only households are skewing more toward Republicans and Romney. How would one under-sample? It's a story of methods and regulations. Cellphone interviews are supposed to only use human interviewers and there are several firms that rely on automated surveys that fail to reach cellphone-only respondents. What Nate Silver doesn't note is that at the other end of the spectrum firms can under-sample landline-only households (...based on their assumptions about the size of this population... there is also a third population to consider in those who have a landline and a cellphone). I have a hunch that some of the polls landing furthest left are under-sampling landline-only households (I also think declining response rates are introducing more "noise" into the polls, which could also be leading to more volatility). But we won't really know until the election is over...
Nineteen Sixty-four is a research blog for the Center for Applied Research in the Apostolate (CARA) at Georgetown University edited by Mark M. Gray. CARA is a non-profit research center that conducts social scientific studies about the Catholic Church. Founded in 1964, CARA has three major dimensions to its mission: to increase the Catholic Church's self understanding; to serve the applied research needs of Church decision-makers; and to advance scholarly research on religion, particularly Catholicism. Follow CARA on Twitter at: caracatholic.
- ► 2016 (14)
- ► 2015 (21)
- ► 2014 (19)
- ► 2013 (30)
- Were U.S. Catholics Raptured? ... Again?
- A Final Pre-Election Peek at the Catholic Voter
- Catholic Vote Before 2nd Debate: Obama on the 16th...
- Author Meets (Online) Critics: CARA Vocations Stud...
- Catholic Vote Update Before the All-Catholic VP De...
- Millions of Never-Married Catholics Have Considere...
- Twenty Years of Change and Stability: Lay Ecclesia...
- Catholic Vote 2012: 'Before the Debates' Edition
- ▼ October (8)
- ► 2011 (33)
- ► 2010 (25)