FiveThirtyEight: Mormon Voters Raise Romney's Advantage in Arizona

Arizona might have less symbolic importance than Michigan, but it is probably of more practical significance. That is because it’s one of the few Republican states to award its delegates on a truly winner-take-all basis, without any qualifications or complication. Get one more vote than your rivals in Arizona, and you take all 29 of its delegates.

The most tangible advantage in Arizona belongs to Mitt Romney, and it is because the state has a reasonably high Mormon population. In the 2008 primary there, Mormon voters constituted 11 percent of the electorate — and Mr. Romney won 88 percent of their votes, versus 8 percent for John McCain.

If Mr. Romney posts similar numbers among Mormon voters this year in Tuesday’s primary — and there’s no reason to think that he won’t — that works out a nine-point built-in advantage in the state.

Without that Mormon edge, in fact, the state would essentially be a toss-up. Mr. Romney is now projected to win the state by 11 points over all, according to a FiveThirtyEight forecast model, which works out to an 89 percent chance of winning given the uncertainty inherent in the forecast.

The 11-point advantage that the FiveThirtyEight model gives Mr. Romney is larger than his lead in all but one recent poll of the state. But the model is designed to be aggressive about identifying a trend, and there is perhaps the slightest hint of one toward Mr. Romney in Arizona. The two polls that have been in the field most recently — from NBC/Marist and We Ask America — gave him his largest advantage there.

With that said, there has been quite a bit of variance between different polls of the state, with Mr. Romney’s lead varying by margins of 3 to 16 points. I don’t know that there is anything special about Arizona that is causing this; we are also seeing a fairly wide spread in the polls in Michigan, as well as in national surveys. As there now seem to be differences in the level of voter enthusiasm for the candidates — more for Rick Santorum, less for Mr. Romney — the assumptions that pollsters make about turnout will tend to be more important.

Mr. Santorum’s deficit in Arizona is just large enough, however, that it would qualify as something of a surprise if he won the state. It might require either that he performs strongly at Wednesday’s debate in Mesa, Ariz.,  or that he simply beats the pollsters’ expectations on primary night itself, something he has done in several states so far.

Because there is no delegate value in merely keeping things close in Arizona, however, it is also worth watching to see whether Mr. Santorum essentially abandons the state and downplays expectations there, instead concentrating his efforts on Michigan, which also votes on Tuesday.

If Mr. Santorum wins Michigan narrowly but loses Arizona by a clear margin, he will probably fall further behind Mr. Romney in the delegate count given Michigan’s complex delegate-selection rules. Such an outcome, however, would very probably allow him to put a strong spin on the outcome — and a credible enough one, since if Mr. Romney loses Michigan, his native state, it is hard to know where he might win elsewhere in the Midwest.

But if Mr. Santorum loses both Arizona and Michigan, however narrowly, he would fall behind, according to both math and momentum.

View the original article here

FiveThirtyEight: Reads and Reactions: Santorum's Rise

Reactions

Our last Reads and Reactions came just two weeks ago, but that was when Mitt Romney was the clear front-runner in the polls. He was about to notch a commanding win in Nevada, Rick Santorum was polling in the midteens nationally and Newt Gingrich was the main anti-Romney candidate.

But, as FiveThirtyEight pointed out, Mr. Romney had yet to finish in front in one area of the country: the Midwest.

Nate Silver wrote on Feb. 5: “Imagine … that Mr. Santorum wins both Minnesota and Missouri. That could revive his campaign, especially given that he also took Iowa. Mr. Santorum is, in many ways, a more dangerous opponent for Mr. Romney than Mr. Gingrich at this point.”

FiveThirtyEight gets plenty wrong, but not this one. Mr. Santorum won not only those two states, but he upset Mr. Romney in Colorado as well. At the very least, Mr. Santorum has tacked on a few extra miles onto Mr. Romney’s race.

Mr. Romney is still the most probable nominee, but he seems to have lost his air of inevitability, which may change his campaign’s calculus. At The American Prospect, Patrick Caldwell highlighted Mr. Romney’s somewhat lacking field operation, something FiveThirtyEight has noticed, too. Mr. Caldwell wrote:

Romney has built a campaign that is just good enough to dispatch a lackluster field of opponents. Santorum’s sudden rise in the polls threatens that. A meager field staff would have been sufficient to dispatch Romney’s competitors when he was the runaway front-runner, yet might not be enough to right his path now that he’s vulnerable.

Not only does Mr. Romney seem to have lost the “inevitable” tag, his “most electable” tag is also in dispute. Nathaniel Rakich, at his Baseballot blog, wrote that Mr. Romney’s inability to connect with white, working-class voters would at least partially negate one of President Obama’s chief weaknesses — an inability to connect with white, working-class voters. “Much of Romney’s advantage derives from the fact that he is perceived to be strongest in a matchup with Obama — in other words, the most electable Republican,” Mr. Rakich wrote. “But given this new information, the candidate with the most credible claim to that mantle may be Rick Santorum.”

Mr. Romney’s rocky stretch has even renewed talk of a brokered convention. The Week magazine had a great overview of how a brokered convention would work, how it might play out and why it’s still unlikely.

Next up are Michigan and Arizona on Feb. 28. FiveThirtyEight’s Michigan forecast has Mr. Santorum as a 72 percent favorite to win the state. Jonathan Bernstein, writing at The Washington Post, thinks those odds are too kind to Mr. Santorum — and Mr. Silver raised questions about them earlier today. Still, Mr. Bernstein laid out the significance of a Romney loss in Michigan, saying it would — at a minimum — prolong the Republican nominating process well beyond Super Tuesday on March 6.

Reads

The Washington Post’s Lori Montgomery looks at fears that the end of the payroll tax cut will help sink the economy. Check out the graphic, too.

At Talking Points Memo, Kyle Leighton argued that low G.O.P. turnout in the primary may hurt the party in November.

Our colleagues put together an interactive graphic on the distribution of government benefits.

At the Columbia Journalism Review, Curtis Brainard looked at what influences public opinion on climate change.

View the original article here

FiveThirtyEight: Santorum and the Emphasis on Social Issues

Last week, I made the case for why the “electability gap” between Rick Santorum and Mitt Romney might not be all that large. But there’s a contrary point to consider, one that would not necessarily work in Mr. Santorum’s favor.

One of the factors that our forecasting model uses in conjunction with presidential approval ratings and economic forecasts is a score to indicate the ideological disposition of the opposition candidate. Historically, candidates whose issue positions are closer to the center of the electorate tend to do a little bit better than the fundamentals would dictate; those who are closer to the extremes do somewhat worse.

There are various objective ways to measure candidate ideology, and the different systems tend to be in fairly strong agreement for Mr. Romney. He is neither a moderate nor is he extremely conservative, according to these ratings, but somewhere in between.

The ratings for Mr. Santorum, however, are all over the map. Without going into too much detail into how each system is calculated, there are some that deliberately place more emphasis on economic policy and others that weight economic policy and social policy more equally.

In the economically driven systems, Mr. Santorum rates as essentially no more conservative than Mr. Romney. But he is substantially more conservative in the systems that also consider his social views. This is particularly true for the system that directly measures voter perceptions about the candidates from surveys. For the time being, Mr. Santorum is seen by voters as being quite conservative indeed, and Mr. Romney less so.

So if the general election is primarily about economic policy, Mr. Santorum may not be at much of a disadvantage to Mr. Romney. In fact, a case can be made that he’d be a little stronger than Mr. Romney based on Electoral College considerations. However, the more that social policy is emphasized, the more the electability gap could grow.

The ideology scores that the model uses, as I’ve mentioned, are somewhat deliberately slanted toward economic policy. There are several good reasons for this. First, the economy remains in pretty bad shape, although it is improving. Second, voters routinely indicate in surveys that they are much more concerned with economic issues than social ones. Third, the president probably has less influence on social policy than on economic policy. Social policy is often determined more through the courts and at the state level than through the executive branch.

To some extent, the emphasis that voters and the news media will place on these issues is likely to be dictated by economic conditions. Frankly, if issues like gay marriage and contraception are making headlines — rather than unemployment or gas prices or the European debt crisis — that will probably be a bullish sign for President Obama, since there is more bandwidth available to cover these issues when the economy is improving.

It is also true, though, that the amount of emphasis placed on economic issues versus social policy may be influenced by the Republicans’ choice of a nominee. If Mr. Santorum is picked, there will probably be relatively more coverage of social policy — both because he tends to place more emphasis on social issues himself, and also because the news media tends to focus on his social policy positions.

There are already some indications of this. When Mr. Santorum’s name is mentioned in news coverage, it tends to be associated with social issues more often than economic ones. The opposite is true for Mr. Romney.

I used Lexis-Nexis, a database of news articles, to look up how often each candidate’s name was mentioned in conjunction with various keywords related to economic and social policy. Examples of social policy keywords include “abortion,” “gay marriage,” “values” and “evangelical.” Examples of economic policy keywords are “jobs,” “taxes,” “Medicare” and “deficit.”

In order to ensure that the usage of these terms was in context, my search required that they were used within 15 words of the candidate’s last name. I then compared the number of “hits” to how many articles used the candidate’s last name in any context from Jan. 1 through Feb. 19.

For instance, the terms “birth control” or “contraception” were used in conjunction with Mr. Santorum’s name in 1,302 articles. There were a total of 31,548 articles that used Mr. Santorum’s name in any context during this period, which means that about 4 percent the articles about Mr. Santorum covered these issues.

By contrast, the same terms were used 1,049 times in conjunction with Mr. Romney’s name out of 55,906 articles about Mr. Romney. That means that about 2 percent of the articles about Mr. Romney covered his position on birth control — about half the frequency as for Mr. Santorum.

Indeed, this pattern was typical. Mr. Santorum’s name was used more commonly in conjunction with 12 of the 15 social policy keywords; the exceptions were “guns” (or “Second Amendment”), “immigration” and “Mormon.”

However, Mr. Romney name was used more commonly in conjunction with all 15 of the economic keywords.

One of the biggest discrepancies was for the keywords “sex” and “sexuality”; those terms were used about three times more often in conjunction with Mr. Santorum’s name than Mr. Romney’s, adjusting for the overall amount of news coverage that each candidate received.

At the other end of the spectrum were a set of three keywords that I grouped together in the search: “wealth,” “wealthy” and “rich”. These were used about three and a half times more often in conjunction with Mr. Romney’s name than Mr. Santorum’s.

So if you want a somewhat jaundiced view of the associations that the news media tends to make disproportionately often with each candidate, it is this: Mr. Romney and the term “rich,” Mr. Santorum and the term “sex”.

If you group all of the terms together, social policy keywords were used about 80 percent more often than economic policy keywords in discussions of Mr. Santorum. By contrast, economic policy keywords were used 50 percent more often than social policy keywords in articles about Mr. Romney.

Does this reflect news media bias of some kind — or is it a fair reflection of the emphasis that the candidates tend to place on these issues themselves?

That is not an easy question to address, but here is one approach. I looked up how often each of the candidates used the keywords on the “issues” sections of their respective campaign Web sites.

Both candidates used the economic keywords more often than the social keywords on their Web sites, but the disparity was greater in Mr. Romney’s case.

Mr. Romney’s Web site used one of the economic policy keywords in 217 instances and one of the social policy keywords 60 times, a ratio of about 3.6 to 1. By contrast, Mr. Santorum’s Web site used the economic keywords 190 times but the social keywords 128 times, a ratio of 1.5 to 1.

So the answer here is somewhat complicated. For both candidates, the news media coverage placed more emphasis on social issues than the candidates did themselves. I would need to think more carefully about the research design before conclusively labeling this a form of bias, but that is certainly one interpretation: the news media tend to place more emphasis on social issues than the candidates do — and probably also than most voters profess to, given how much economic issues dominate among voters across the political spectrum when they are asked to identify which issues are most important to them.

However, to the extent there is bias in the news media coverage, it is at least somewhat proportional to the respective stances of the candidates. Mr. Santorum does tend to place more emphasis on social issues than Mr. Romney — if not as much as the news media seem to think he does.

Still, this could put Mr. Santorum at a disadvantage. If he is the nominee, the coverage might tend to place more emphasis on social policy, where his views are further removed from those of the median voter.

It should be mentioned, however, that Mr. Romney is not as much of a social moderate as he is sometimes described — at least based on his current issue positions. In fact, if you look at public statements that Mr. Romney has made recently, they are nearly as conservative as Mr. Santorum’s on social issues: slightly less so on issues like gay rights, but slightly more so on others like immigration.

Some of this disconnect, of course, is because Mr. Romney’s positions on some social issues were formerly more moderate when he was running for office in Massachusetts. It may also be because of Mr. Romney’s tone: he is somewhat less strident than Mr. Santorum on social issues, even though they don’t differ very much when you examine their actual policy positions.

Nevertheless, Mr. Romney will probably have an easier time maintaining a focus on economic policy if he gets to the general election. Although the electability question between Mr. Romney and Mr. Santorum is debatable, that is a point in Mr. Romney’s favor.

View the original article here

FiveThirtyEight: Could Ron Paul Still Win Maine?

Ron Paul’s campaign is claiming that it could still win the presidential preference poll in the Maine caucus because of a county that postponed its vote and will hold its caucus next Saturday, Feb. 18.

On Saturday, the Maine Republican Party declared Mitt Romney the winner of the presidential preference vote, which he led by 194 ballots based on the caucuses that have been held so far.

State Republicans said they considered the results of the straw poll final. However, Washington County, in the easternmost part of the state, postponed its caucus after a snowstorm was forecast there. The Washington County G.O.P. Chair, Chris Gardner, said his county would conduct the straw poll at its caucuses and will report the results to the state.

All if this will be moot unless Mr. Paul is able to make up 194 votes in the county.

Based on how the county voted in 2008, that seems unlikely. Just 113 votes total were cast in the county in 2008, and only 8 of those were for Mr. Paul. John McCain, instead, won the plurality.

In addition, Mr. Romney narrowly won the two counties, Hancock and Penobscot, that border Washington County to the west and which are probably the best demographic match for it — although Mr. Paul won sparsely-populated Aroostook County, which borders it to the north, where he took 81 votes to Mr. Romney’s 26.

However, Washington County might theoretically have some untapped potential for Mr. Paul. It is rural and relatively poor — demographics that tend to suit him more than Mr. Romney. And it is relatively conservative, having split its vote about evenly between Barack Obama and Mr. McCain in 2008 when Mr. Obama won Maine as a whole fairly easily.

What such an outcome would require is for Mr. Paul’s campaign to make a concerted effort to turn out any supporters it has in the area. There are 6,907 registered Republicans in Washington County, and another 8,247 unaffiliated registered voters, who are eligible to participate by changing their registration to Republican at the caucus site. Unregistered voters, for that matter, are also free to participate provided that they register at the caucus site.

Imagine, for instance, that voters turned out in the county at a rate comparable the Iowa caucuses, where Mr. Paul had a strong turnout operation. In Iowa, 122,255 Republicans participated in the caucuses as compared to a total of 644,220 voters who were registered as Republican prior to caucus night.

Were turnout in Washington County to occur at the Iowa rate, it would produce about 1,300 participants at the caucuses, enough to swing the outcome if Mr. Paul received about 15 percent more of their votes than Mr. Romney.

A more likely scenario, perhaps, is that Mr. Paul would work to turn out his supporters while the other campaigns would not. Mr. Romney’s campaign, in particular, might seem to give credence to the notion that the straw poll was an unresolved issue if it made extra effort to turn out voters there.

If Mr. Paul were somehow to secure the 194-vote margin, that would create a messy scenario for the Maine G.O.P. It would either need to backtrack on its previous statements that the straw poll results are final, or might risk appearing as though it was disenfranchising voters.

Such a result would also create problems of interpretation. Under this scenario, turnout would be much higher in Washington County than in other parts of Maine — and it would be higher in part because the results from the rest of the state were already known and Mr. Paul had doubled-down in the county as a result.

Put differently, it’s possible that Mr. Paul would receive enough votes there to swing the outcome, but would not have done so if the caucus had taken place as originally scheduled.

A handful of Maine towns outside of Washington County have also not yet held caucuses. In those cases, however, the caucuses were scheduled for after Feb. 11 all along. Only 45 votes were cast in these towns in 2008, so any impact is likely to be small.

The results of the presidential preference poll are nonbinding and serve mostly for vanity — delegates are selected through a separate vote at the caucus sites. Although this is also true in most other Republican caucus states, Maine’s delegate selection process is especially cumbersome and can potentially reward candidates whose supporters are more enthusiastic and who sit through the entire process, which can be hours long.

Mr. Paul’s campaign has predicted that it will win the most delegates from Maine regardless of the result of the straw poll.

View the original article here