Religiosity Remains Huge Predictor of Presidential Vote Choice

Americans’ religiosity remains a powerful predictor of their vote choice. Take a look at these data among non-Hispanic whites from our tracking for the period May 1-July 18.

These data are among whites only. Why? Because black voters form a special case in this regard. Blacks are the most religious racial and ethnic group we track. Blacks are also are the most Democratic. Black voters have such a high probability of voting for Barack Obama, across the board, that none of their demographic characteristics make a difference. So, by looking at whites, we get a more highly focused portrait of the impact of religion.

The differences among highly, moderately, and nonreligious white voters are startling. (These three categories of religiosity are based on responses to two questions about the importance of religion in their daily lives and also religious service attendance).

Nonreligious voters are about 30% of the registered voter population. Over the period of time May 1-July 18, they swing to Obama over Mitt Romney by a 56% to 35% margin (a 21-percentage-point Obama gap). The highly religious white voters -- about 42% of registered voters -- swing to Romney by an overwhelming 46-point gap, 23% for Obama, 69% for Romney. Those who are moderately religious (28%) go for Romney by 16 points.

In other words, the mere knowledge of a white voter's response to two seemingly simple questions: "Is religion important in your daily life?" and "How often do you attend religious services?" continue to be extremely effective predictors on that white voter's voting intentions.

Of course, the state of being nonreligious does not occur in isolation. Americans who are nonreligious also tend to be unmarried and younger, characteristics which, in and of themselves, are associated with voting for Obama.

But these demographic correlates are not the total explanation.

I looked at the relationship between religiosity and vote choice (again, using the big sample of interviews conducted between May 1 and July 18) among each of four age groups.

Age in and of itself, of course, certainly makes a difference in the projected vote choice of white voters. Note that 18- to 29-year-olds as a total group are virtually tied in their support, while 65 years and older go 21 points toward Romney.

But, and this is the key factor here, within each of these age groups, the very religious are skewed much, much more toward Romney than are the nonreligious.

Look at 30- to 49-year-old whites, for example. Voters in this group who are very religious go for Romney over Obama by 73% to 20%. Voters in this age group who are nonreligious go for Obama over Romney by 56% to 35%.

I also looked at the relationship between religiosity and vote choice within marital status groups.

As you will note, taken as a group, single/never married white voters and white voters living in domestic partnerships skew significantly toward Obama. Single white voters go for Obama by a 49% to 42% margin; those in a domestic partnership are even stronger in their support for Obama, giving him a 52% to 39% edge over Romney. On the other hand, married white voters skew significantly toward Romney, giving him a 59% to 34% margin over Obama.

But, again, within each of these marital groups, there is a big difference between those who are very religious and those who are nonreligious.

The biggest difference comes within the ranks of those who are married. Very religious married white voters go for Romney by a whopping 52-point margin. Nonreligious married white voters go for Obama over Romney by a 15-point margin.

We find the same pattern within each of the other groups based on marital status.

Bottom line:   How religious one is in America today continues to be a major predictor of one's vote for president.


This religious divide is not new or unique to this Democratic president. It has been apparent for decades in American politics.

Barack Obama is, by all accounts, a religious man. In fact, his remarkable prayer breakfast speech this past February gave evidence that he has been thinking about the relationship between religiousness and politics.  But at this point in American politics, as a Democratic candidate for president, Obama continues to attract those voters who are the least religious, and has significantly lower appeal to those who are the most religious.

The Central Issue of the Campaign

Americans' responses to our recent set of questions about the impact of the Affordable Care Act help us understand the major overarching difference point in this presidential campaign: the appropriate role of government.

The results, as I reported here on Monday, basically show that Americans are aware of the benefits and the costs of the ACA legislation:

Americans clearly perceive that the healthcare law would benefit those without insurance and those who get sick. Americans agree, it appears, that government programs can in fact help citizens. As we have seen in other countries around the world, government can provide health insurance, provide income to those who don't have much money, provide retirement, provide healthcare for all, guarantee job security, and do almost anything it wants.

At the same time, Americans clearly perceive that the ACA would hurt taxpayers and businesses. Taxpayers would be “hurt” presumably because Americans perceive that someone will end up paying for a government program that provides insurance to those who currently do not have it. And businesses presumably are perceived as being hurt for the same reason.

Thus, Americans' reactions to the ACA underscore the central trade-off of our times -- the value of the government stepping in to help those who are in need or discriminated against versus the costs in terms of money, efficiency, and involvement in people's personal lives. Barack Obama, and the groups of Americans who tend to support him and the Democratic Party, clearly view the government as an entity that should use its force and might to fix social problems and help the unfortunate and those who cannot fend for themselves. Mitt Romney, and the groups who support him and the Republican Party, view the government as an inefficient mechanism for fixing problems, and one that takes away individual freedom and initiatives.

Note that even Republicans -- who overwhelmingly oppose the ACA -- are as likely to say the ACA will benefit those without health insurance as to say it will hurt them. Republicans presumably still oppose the healthcare law because they think there are other ways to help those without health insurance.

All of this goes along with the fact that the public remains roughly divided in its choice between the two presidential candidates, who represent in broad ways the two approaches to the appropriate role of government.

This continuing divide occurs even though consumer confidence as we measure it here at Gallup is as low as we have seen it since January. About six in 10 Americans say the U.S. economy is getting worse. And two-thirds of Americans mention some aspect of the economy as the most important problem facing the country.

Ordinarily this would spell trouble for an incumbent president.

Of course here we can get into a discussion of what “trouble” is. As Susan Page and I discussed in our Election Matters webcast this week, one can argue whether the incumbency glass is half empty or half full. Some say Obama has managed to stay competitive in the race despite the dour economic perceptions of the average American out there across the land. Others can argue that Romney has managed to stay competitive in the race despite the substantial and focused criticisms of Romney from the Obama campaign team.

What we know for certain is that we simply have not seen major shifts in this race. As I discussed here, we have seen Obama move marginally ahead in our massive, large-sample tracking. The latest updated three-week average through July 15 shows Obama maintaining a two-percentage-point lead, 47% to 45%, over Romney. But this is not a huge margin. Registered voters nationally still appear to be generally split in their vote preferences.

Watching Obama at the U.S. Olympic basketball game in Washington on Monday night – as broadcast on ESPN – brought home one significant Obama strength. He comes across well. Our latest measurements showed Obama with a significant “likable” edge over Romney. Eighty-one percent of Americans say Obama is likable, compared with 64% who say that about Romney. Obama also does better as the candidate who understands Americans’ problems. And cares about people like “you.”

How much this is helping Obama is unknown at this point. But in general it certainly can’t hurt to be personally likable. This likability edge for Obama adds to Romney's challenges.

Romney has opportunities ahead to improve his likability quotient. First will be the hoopla surrounding his announcement of his vice president, most likely coming in August after the Olympics are over -- if he doesn't surprise everyone and announce it this week. Then it’s his convention performances, and his debate performances. Both of these are more important for Romney than for Obama. The public knows Obama after four years. Romney, despite being around and being heavily involved in the GOP primary process earlier this year, is less well known to Americans. His goal for his high-visibility moments during the VP announcement, the GOP convention, and the debates is presumably to a) become more likable and b) convince voters that as president he can do a better job than Obama on the jobs front.

I focused on “jobs” in b) above because that is what Americans were most likely to tell us is needed when we asked them last week what they would recommend to fix the economy. Twenty-eight percent mentioned jobs outright, while another 9% said quit “outsourcing” jobs. We’ll have more on that here on gallup.com on Thursday morning.

Obama's Slight Edge Over Last Three Weeks

The generalized narrative of the presidential election contest so far this cycle is that not much has changed...which is true in the broad sense....but we are seeing a little movement toward the Obama side of the ledger.


In the four, separate, non-overlapping three-week periods since Gallup began tracking on April 11, Barack Obama’s percentage of the vote has been 46-47%.... Mitt Romney’s has been 45%-46%....each of these three-week periods is based on about 9,000 interviews with registered voters.

Still, the most recent three-week period shows Obama at his upper range, 47%, while Romney is at his lower range, 45%...these results suggest that within the broad framework of broad stability, Obama has gained a slight edge...also evident in our short-term seven-day rolling average, where, as of the July 12 report, Obama has a 47% to 44% lead over Romney.


These are data from registered voters....Romney most likely will get some benefit from turnout, as the Republican candidate nearly always does...meaning that Obama needs a cushion within the registered voter totals to have a high probability of breaking even or winning the popular vote on Nov. 6.

The structure of the support for each candidate remains basically static....Obama’s highest support comes from any registered voter who identifies him or herself as something other than white....along with those groups of voters who are divorced from what we might call the traditional structures of marriage and religion...and young voters and voters with postgraduate educations…and low-income voters….and women.

Romney’s base of support is whites -- along with those who are religious, married, middle aged, who don’t have a postgraduate degree, and who are at a middle-income level or higher.

Plus, my colleague Jeff Jones’ analysis shows that Romney wins in the South, Obama in the East, and the two are very close in the Midwest and West…all of these data and much more are now available in Gallup’s amazing new U.S. Presidential Election Center.

The dynamics of each campaign’s strategic efforts are becoming clear...Obama’s team is taking a page from the playbooks of previous incumbents and attempting to paint Romney as an unacceptable alternative to four more years of Obama....which is basically built off the assumption that an incumbent election is all about a choice between what you know and what you don’t know. Obama’s relatively weak position, based on his sub-50% job approval ratings, means he can’t afford to sit back ala Ronald Reagan and talk about the new "Morning in America" under his regime....so the strategy is to focus on the assumption that whatever the current situation….the alternative is worse.

The main target of the Obama campaign is Romney’s wealth, how he acquired it, and where he has invested it. New Gallup data show that most Americans don’t care about Romney’s wealth…75% say it makes no difference to their vote...but 19% of independent voters say it makes them less likely to vote for Romney, which could be a factor.

This whole push has been given an accelerant with the Obama team's re-emphasis on a call for higher-income families to pay more in taxes....this idea is poll-tested and does well…when Gallup last asked about it, 59% supported the idea of higher taxes on those making $250,000 or more.

Still, Americans like having a rich class and want to be rich themselves, and reducing the income gap is a low priority for Americans…but the Obama team’s idea is to raise doubts that Romney’s policies would help the average American....The latest ABC/Washington Post poll, however, shows that Romney does slightly better than Obama in terms of Americans' views of who would do the better job handling the economy.

Romney has three time periods in which to fill in the blanks on his image...and to counter the Obama campaign's efforts to define his image for him...his VP pick, the convention, and the debates. There is some thought that Romney may spring his VP pick before the Olympics get underway on July 27...but the historical precedent would be for him to announce just before the Aug. 27 beginning of the Republican convention.

Romney’s campaign team made the decision to have their candidate speak to the the NAACP convention in Houston on Wednesday...by no means the world’s friendliest audience for a Republican candidate...at the moment, black voters are going for Obama over Romney by 87% to 5%. It is highly unlikely that Romney will make any significant headway into this monolithically Obama-supporting voting group...the assumption is that Romney’s team made the decision that his appearance might serve to soften his image among other key voting groups.

No sign yet of major shifts in any of our key Gallup tracking indicators after last Friday’s announcement of the June jobs report...economic confidence is beginning to slide a little, however. We will see if that continues with our weekly report due out next Tuesday at Gallup.com.

What Do Obama, George W. Bush, and Gerald Ford Have in Common?

The Obama-Romney race continues to cut down the middle of the (limited) historical information we have on previous presidential races. This underscores the conclusion that the winner of this year's race cannot be predicted with any certitude from existing data at this point.

There have been 10 presidential elections since World War II in which an incumbent ran for re-election. In seven of these, the incumbent won. Therefore, everything else being equal, this very limited sample suggests that the incumbent should have an advantage -- a home-field advantage of sorts.

The first president since World War II to seek re-election was Harry Truman in 1948. The year 1948 represents a major anomaly as far as polling is concerned. For one thing, Gallup quit measuring presidential job approval in June 1948, when Truman had a 40% approval rating and a 44% disapproval. We don’t know for sure, but we think Dr. George Gallup and his associates did that because they were measuring the trial heat ballot and felt it wasn’t necessary to continue to track presidential job approval. Still, one thing we learned from the 1948 race is the fact that a 40% job approval rating for an incumbent in June of the election year can still result in a victory for that incumbent. Second, as is well known, Gallup and other polling firms’ trial heat ballots predicted a clear victory for Thomas Dewey, governor of New York. Gallup’s final poll had Dewey with 45% of the vote, Truman with 41%, with the rest going to Henry Wallace, Strom Thurmond, or to “other/undecided.” Of course, that was wrong. Truman won by a 50% to 45% margin.

There have been five incumbent elections since World War II that produced “easy” victories for the incumbent. All were clearly evident in the data ahead of time. These incumbents were Dwight Eisenhower in 1956, Lyndon Johnson in 1964, Richard Nixon in 1972, Ronald Reagan in 1984, and Bill Clinton in 1996. All five of these presidents had job approval ratings above 50% in the summer and fall before their elections and were winning handily in the trial heat ballots prior to their elections.

Obama’s job approval ratings and trial heat position at this point do not emulate these strong positions. He is not in an “easy” election position.

There were two incumbent presidents whose eventual losses were clearly evident by virtue of their job approval ratings at this point (the summer) of their election years -- Jimmy Carter and George H.W. Bush. Both had job approval ratings in the 30% range. The trail heat ballot for these two presidents was not as clear, however. Carter from time to time through September of 1980 actually tied with or beat Ronald Reagan, who surged at the last minute and went on to win convincingly. Bush was leading Clinton and Ross Perot in early July 1992, but when Perot dropped out of the race (he came back in later), Clinton surged ahead and stayed ahead for the rest of the race.

Obama’s job approval ratings at this point do not emulate the weak positions of Carter and Bush, Sr.. He is not a clear “loser” at this point.

That leaves two races: Gerald Ford in 1976 and George W. Bush in 2004.

Ford, a congressman from Michigan, had been appointed vice president by Richard Nixon after Spiro Agnew had to resign in disgrace in 1973 because of corruption charges. Ford ascended to the presidency in August 1974 after Nixon resigned in disgrace because of Watergate.  Becoming president as part of a scandal was certainly not an auspicious precursor to a successful re-election bid two years later.  Plus, Ford pardoned Nixon in September 1974, a decision with which the American public soundly disagreed. This also didn’t help his case for re-election and is thought by some to have cost it for him. Still, in June 1976, Ford had a 45% approval rating and a 40% disapproval rating. That was it. Like 1948, Gallup quit asking approval for Ford after that point (until after the election).

In the trial heat ballot from July 16-19, 1976, Jimmy Carter was massively ahead of Ford, 62% to 29% among registered voters. Carter maintained a lead -- but by September the race got a lot closer. In fact, when the dust settled, Carter only won by two percentage points in the popular vote, 50% to 48%. These results underscore how much the trial heat ballot can change over time.

George W. Bush was not a highly popular president, in many ways similar to where Obama is today. Bush’s average job approval rating in July 2004 was 48%. Last week, the first week in July, Obama’s was 45%.  Democratic nominee John Kerry was leading Bush among both registered voters and among likely voters in July of 2004. Bush moved ahead in the fall, but ended up winning by three points, 51% to 48%.  So, like Ford, and like Obama today, Bush was in a mid-range historical position and ended up in a close race in the popular vote. 

So, we can say that Obama’s job approval rating at this point puts him in the same broad range as Gerald Ford and George W. Bush, further suggesting that the outcome of the popular vote this fall may be quite close.  Obama's trial heat positioning against Romney -- statistically tied at this point -- also suggests a close race, although as we have seen, the trial heat is not a solid predictor this far out.

Public Opinion After the SCOTUS Decision

The reaction to the Supreme Court Affordable Care Act decision last Thursday is divided, as we would predict.

I’ve seen five polls conducted after the decision -- our one-night USA Today/Gallup poll on June 28, and those conducted by the Kaiser Family Foundation, Pew Research, CNN/ORC, and ABC News, Washington Post.  All found that the public’s reaction to the ruling was split. In our one-night poll, amazingly enough, not only did we have a 46% to 46% split on agreement/disagreement with the ruling, but we had exactly -- I mean exactly -- the same number of respondents who said they agreed as said they disagreed.

The divide in reaction is predictably partisan. In our poll, 79% of Democrats agreed with the Supreme Court’s ruling, while 83% of Republicans disagreed. Independents tilted toward agreeing by a three-percentage-point margin. The other polls found basically the same type of political divide.

Why would we expect this type of political divide? Well, it’s because the underlying Affordable Care Act (ACA) is politically divisive -- a piece of legislation that has come to be extraordinarily partisan. The ACA was initiated and championed by President Barack Obama, and the informal sobriquet most often used to describe it incorporates his name.  Republicans have generally opposed the ACA and focused on it as a main political target this year. The ACA has taken on symbolic gravity for partisans on both sides, as legislation sometimes does. Ergo, basic views of the Affordable Healthcare Act have been and continue to be highly divided along party lines. And so was the reaction to the Supreme Court decision.

So, in a way, asking why the reactions to the ACA, and the Supreme Court decision which affirmed its constitutionality, are divided is like asking why opinions of President Obama are divided. That’s the reality of American politics today.

To be sure, it’s clear that the fundamentals of the ACA resonate with two disparate themes in American politics today. Democrats on average favor more government action to improve the lives of Americans, while Republicans favor letting individuals and businesses handle more of this responsibility. Since the ACA is demonstrably the government attempting to make things better for individuals when it comes to healthcare, it fits with a Democratic view of social and political life. It does not fit with a Republican view.

People ask: How important will the Supreme Court's decision be in this fall's election? The answer: It will probably be highly important for some, but not that important for the majority. When we recently asked Americans why they were voting for Obama or for Mitt Romney, healthcare was relatively low on the list of categorized responses. Just 10% of Obama voters mentioned healthcare and just 4% of Romney voters mentioned healthcare. Further, as my colleague Jeff Jones pointed out in his recent analysis, and I as reviewed here, few Americans at this juncture (at least through early June) mention healthcare as the most important problem facing the country. And, in our June 28 poll, only about one-fifth of Americans said that the ACA was going to be the determining factor in their voting decision for this presidential election, although another 59% said that it would be one of many factors they would take into account.

As I noted here, it may be possible that the decision, perhaps like President Obama’s public announcement that he favors legalized same-sex marriage, will galvanize conservative voters to get out and vote this fall. On the other hand, there is some indication that the decision may have helped make Democrats more enthusiastic.

What does the American public want to happen now? Several polls have asked about that, with different results.

Our USA Today/Gallup poll gave Americans four choices:

• Expand government’s role 25%
• Keep the law in place 13%
• Repeal parts of the law 21%
• Repeal the law entirely 31%

The CNN/ORC poll gave them two choices:

• Repeal all of the provisions in the healthcare law 51%
• Keep all of the provisions in place 47%

The Kaiser Family Foundation gave these four alternatives:

• Expand law 28%
• Keep law as is 25%
• Repeal and replace with Republican alternative 18%
• Repeal and not replace 20% 

Obviously there is no one general conclusion here.  This is one of those instances in surveys when the options the researcher give to respondents make a difference.  There is clearly an appetite in some segments for repealing all or part of the ACA, up to about half of the public in our USA Today/Gallup and the CNN polls.  These people are highly likely to be Republicans.  When the Kaiser poll asked respondents if opponents of the law should "continue trying to block the law from being implemented" or "stop their efforts to block the law and move on to other national problems", the latter won with 56% of the choices.   This suggests a tilt towards not attempting to repeal the law, although the question wording here, it seems to me, needs to be looked at carefully in interpreting the results.

The real impact of the SCOTUS decision won’t be known for a while now. In modern elections, events tend to reinforce existing perceptions rather than change people's minds about the candidates they support. So one key result of the decision will be its long-term impact on the Holy Grail of elections -- voter motivation and turnout -- that campaign consultants so covet.