Tag Archives: New Hampshire primary

Who Really Won New Hampshire, and Why

The on-site comments* to my post yesterday for the U.S. News “debate club” remind me that perhaps the biggest story coming out of the New Hampshire primary was not Romney’s decisive win – it was Ron Paul’s unexpectedly strong second-place showing.  You will recall – and undoubtedly remind me in months to come – that I had Paul coming in a very close third, just behind Huntsman, with about 18% of the vote.  Although I nailed Huntsman’s vote totals, Paul did better than I projected, winning 23% of the vote to finish a strong second.

And yet despite his strong finish, very few if any commentators bothered analyzing why Paul did so well in New Hampshire.  That oversight is consistent with the more general media view that Paul is no threat to win this nomination, and that he has committed but relatively small support consisting of a core group of “Paulistas” who contribute to Paul’s moneybombs and lurk on every website, but who don’t constitute much more than 10-15% of likely Republican voters.

I think that characterization, while not completely inaccurate, fit Paul better in 2008 than it does this time around.  In fact, the New Hampshire exit polls suggest Paul has expanded his base of support beyond his libertarian core by attracting conservatives and Tea Partiers who are worried about the deficit and who want to reign in government spending.   Let’s take a closer look.

By now, it is clear that Paul does well among younger voters.  Interestingly, that support is not just among the very young; he edged Romney, 11%-10%, among all voters 44 years or younger who voted in New Hampshire’s Republican primary (based on exit polls).  Romney racked up his winning margin by relying heavily on the much larger 45 year and older vote.

But the most intriguing findings come when we look at the breakdown of the vote by income.  As the table below shows, Paul’s share of support in each income category drops in linear fashion as you climb up the income scale, while Romney’s is the mirror opposite – his support climbs as you go up the income ladder:

Income Paul 2012 Romney 2012
Under $30k (11%) 35% 31%
$30-50k (15%) 28% 31%
$50-100k (37%) 22% 35%
$100-200k (27%) 20% 47%
$200k or more (10%) 12% 52%

In the aggregate, Paul ties Romney at 31% support among wage earners below $50,000, but Romney trounces him, 41-20%, among those earning over $50,000.   If we look back at 2008, however, we see a slightly different pattern of support by income for Paul:

Income Paul 2008 Change In Share of Income Category from 2008 to 2012
Under $30k (10%) 8.4% +26.6%
$30-50k (14%) 8% +20%
$50-100k (39%) 6.5% +15.5%
$100-200k (29%) 6.4% +13.6%
$200k or more (7%) 7% +5%


Although Paul shows gains across the board from 2008 to 2012 in each income category, the gains are larger in the lower income brackets.  His coalition now has a distinct economic skew in a way that it did not four years ago. Much of Paul’s New Hampshire support in 2008 came among voters who strongly disapproved of the Iraq War, who thought the economy was doing very poorly and who were self-identified liberals.  This year he again drew heavily on self-identified liberals and moderates who formed a slightly higher (47% to 45%) proportion of the voting pool than they did in 2008.  But he increased his support among independents, winning 31% of that category, compared to in 2008 when he won only 13%, even as the percentage of independents voting in the Republican primary climbed 10% from 37% to 47% in four years.

Note that Paul’s overall support does not correlate very well with opinions regarding the Tea Party movement; among New Hampshire Tea Party sympathizers; he wins 22% of those who support the movement, 27% who are neutral, and 19% who oppose.  (Interestingly, Huntsman won the most – 41% – among the 17% of voters who opposed the Tea Party.)  What this suggests then, is that Paul was able to draw on a subset of Tea Party voters: those most concerned – particular lower and middle-income voters – with government spending and the deficit.  These are the economic populist voters that Santorum appeared to make some inroads with in Iowa.  My guess is that Santorum’s social views, particularly toward gay marriage, made him anathema to New Hampshire voters who might otherwise have backed him for his economic policies.

What does this mean as we head into South Carolina?  I think Paul is unlikely to fully replicate the success he had in New Hampshire, with its favorable mix of libertarian, middle and low-income fiscally-minded voters. But he could still do well – if the trio of Gingrich, Perry and Santorum let him.  There is an opening, I think, for someone who can reclaim the economic populist portion of the Tea Party vote and still appeal to social conservatives.  So far, Paul doesn’t seem to attracting much support among social conservatives, and his foreign policy views may play less well among South Carolina’s more traditional Republican voters.

The degree to which Paul can replicate his New Hampshire success, then, depends in part on whether someone contests his claim on the economic populist vote.  At this point, I think Gingrich is best positioned to do so. If he is to win back these voters, however, he needs to have a much better answer to the Fannie Mae/Freddie Mac lobbying connection than he has given so far.  If he can provide one, and if he hones his Romney-as-job-destroyer campaign theme without appearing overly negative, he is going to give Mitt a run for his money in South Carolina.  The key to attracting the grass-roots Tea Partiers is for Gingrich to successfully paint Romney as part of the Wall St. banking crowd that benefitted from government bailouts and engaged in “predatory capitalism”.  Even then, however, Gingrich won’t beat Romney if Perry, Santorum and even Paul effectively appeal to the same Tea Party faction.

*The Paulistas were none too happy with my U.S. News piece because I dismissed Paul’s chances of winning the nomination.  Here’s a representative comment:

“Stopped reading this article after bracketed sentence in first paragraph. Author, you are an unpatriotic American, someone who will say anything as long as you can collect money for doing so. You have the integrity of Newt Gingrich.”

Why Mitt Romney Is A Weak Candidate

There is a growing consensus among the talking heads that with his victory in New Hampshire, Mitt Romney is on the fast track to win the Republican nomination.  That sentiment, judging by comments from Tom, Paul and others last night, is one that many of you share.  In response, I want to develop an argument I started in this U.S. News and World Report opinion piece suggesting why it is too early to anoint Romney as the Republican nominee.  In fact, I will go further here to explain why New Hampshire confirms that Romney may be one of the weaker candidates Republicans could nominate.

It is true that he won New Hampshire decisively, with his 39% share of the vote almost exactly matching the average winner’s share in contested primaries dating back to 1988.  That represents a gain of 7% over his performance here four years ago.  But that decisive win masks a second important point: Romney did not broaden his coalition at all in the intervening four years. Consider this graph put together by MIT Professor Charles Stewart that compares Romney’s support in New Hampshire towns yesterday (the left-hand vertical axis) to his support in those towns four years ago (the horizontal axis).  (Courtesy of the Monkey Cage website.)  Each circle is a town (bigger circles=bigger proportion of overall turnout).   What does it show?

Essentially, geographically speaking, Romney drew on the same voting coalition as he did four years earlier, but  his vote total was boosted about 5-7% in those areas.  (If he had exactly the same vote total in a town in both elections, it would be situated on the diagonal line running just below most of the circles.)  What explains the boost in his support?  I suspect it is mostly because Jon Huntsman, a much weaker candidate, was running in the place of John McCain and Rudy Giuliani this time around.  In short, there’s no evidence that he expanded his voting coalition geographically or widened his support to new voting blocs.

And, as Stewart shows, this is almost identical to what happened in Iowa; there Romney also largely drew on the same voting coalition in 2012 as he did in 2008.  However, his support actually declined in most areas of Iowa in the intervening four years, but that decline was almost offset by a substantial boost he received from voters in Polk County.

Pundits have seized on the fact that Romney is the first non-incumbent Republican presidential candidate to win both the Iowa caucus and New Hampshire’s primary.  Given the states’ dramatically different demographic pool, this is evidence, they suggest, that Romney has put together a broad Republican Party coalition capable of carrying him to the nomination.  But this is nonsense.  Never mind that he likely didn’t win Iowa (the votes there have not yet been certified but there is credible evidence that Santorum’s vote was undercounted.)  The voting results indicate he hasn’t broadened his support at all beyond the Republican banking/country club set that backed him four years ago.  In Stewart’s words, “Thus far at least, Romney has almost nothing to show from five years of presidential campaigning.  Romney stands at the top of the heap right now because he has the traditional Wall Street/Country Club wing of the party to himself.”

My point is simple: the reason Romney is “winning” this race has almost nothing to do with his gaining strength as a candidate, and everything to do with the fracturing of the non-Romney vote among several candidates.

In addition to overestimating Romney’s support in this election cycle, pundits have made a second mistake: underestimating the strength of the Tea Party movement. Ezra Klein, among others, suggests that Romney’s success indicates that a desire for moderation is driving Republican voters during the current election cycle. In fact, the center of the Republican Party has moved Right, driven by the Tea Party influence in the last four years, and not to the Center where the moderates reside.  Rather than losing influence, the Tea Party remains as potent an electoral force as it was in 2010.  The problem is that no single candidate has been able to unify the economic populists with the social conservatives, two voting blocs that live uneasily together under the Tea Party label.

As Stewart persuasively argues, some of the Tea Party strength has been siphoned off by Paul, which explains how he has gone from a fringe candidate supported by the small libertarian wing of the party to a second-place finisher in New Hampshire. Republican voters may be turned off by his isolationist/non-interventionist foreign policy views, but his argument against spending and corporate bailouts are exactly the issues that motivated the Tea Party movement in the first place, even if they don’t buy his more extreme views on ending the Fed or monetary reform.

Paul’s influence has expanded to fill the vacuum left by the implosion of the other Republican candidates who might have been expected to represent the Tea Party.  The most logical candidate to unite the anti-Mitt forces is Gingrich, but his candidacy wilted almost entirely because of his link to Fannie Mae and Freddie Mac – the government-backed mortgage giants that most Tea Partiers see as the Prime Movers behind the housing collapse.  Gingrich never came up with a satisfactory answer to the negative ads touting his lobbying for these firms, and it likely fatally wounded his candidacy.  The other possibility was Rick Perry, but he was simply unprepared when the spotlight turned on him.  Santorum’s success in Iowa had much to do with Gingrich’s and Perry’s collapse.  But he also tapped into the economic populism that fueled much of the Tea Party movement.  It’s not clear, however, that he is poised to build on his Iowa success.

My point is that Romney would be losing this race if any of the three – Gingrich, Perry or Santorum – had not stumbled out of the starting gate.  And he is still vulnerable if any of them can regain their footing – or if a new candidate with Palinesque stature enters the race. In this vein, the Republican establishment has been busy trashing Perry’s and Gingrich’s “attack on capitalism”.  But in fact the Bain-as-job-destroyer theme is precisely the one that should have been used against Romney from day one because it capitalizes on the original sentiment – opposition to crony capitalism – that fueled the Tea Party movement in the first place, and which may yet resonate in South Carolina, where unemployment is much higher than in Iowa or New Hampshire.

Romney is ahead not because he is a strong candidate. It is because his opponents and their supporters face a classic collective action problem: each would prefer any of the other non-Mitts as the nominee, but none are willing to sacrifice their own candidacy to make it happen.  There are only two ways Republicans are going to solve this problem.  One is to negotiate an agreement, one brokered by Tea Party activists and social conservatives, to back one of the three non-Mitts (perhaps in return for promises to be placed on the ticket or to have a place in the administration).  The second is to let the process play out and hope that two of the remaining three are winnowed by the voters in time to stop the media-induced rush to coronate Mitt.  The problem is that as long as the media buys into the Republican establishment’s mantra that Mitt is the One, the harder it will be for the non-Mitt’s to raise money.  And when the money goes, so does the candidate.

Mitt may yet win this nomination by default. If he does, I have no doubt many Tea Partiers and social conservatives will hold their nose and vote for him in the general election. But we ought not to overlook the reason why – it’s not because he’s the strongest Republican candidate. It’s because the majority of the Republican Party cannot make their mind up between three candidates – any one of which might beat Romney in a head-to-head matchup.

That’s my argument because I think that’s what the data shows. Now let’s hear your objections.

Meanwhile, Bert Johnson and I are up with our latest prognostications on the race.

What To Look For Tonight In New Hampshire

Polls close in the remaining New Hampshire towns within an hour, at which point we will begin seeing results begin coming in.  Here’s how to interpret the results by location.  Romney, as you might expect, must do well in the southern tier counties – particularly Rockingham and Hillsborough.  If he gets over 40% in these counties, he will have a good night.  In terms of towns, this means doing well in Derry, Bedford and Salem.

Ron Paul, meanwhile, will hope to draw well along the Connecticut River towns on the western mid-part of the state, such as Lebanon and Hanover, where there’s lots of college students and a strong Democratic presence.   Huntsman, however, will be competing for some of these college students as well as with voters in Romney’s strongholds.  Generally speaking, he wants to reprise the McCain coalition from 2008.  That means doing well in Hanover, Concord, and Keene.

Gingrich spent a lot of time focusing on the more rural areas, particularly in northern New Hampshire (Coos County).  These areas don’t have a lot of voters, but he’s hoping to piece together a strong turnout based on many smaller vote totals.  Santorum will be competing for some of these voters, but he also wants to draw heavily in highly populated blue-collar areas down south like Manchester, Merrimack and Rochester, which have a strong Catholic vote and Tea party support.

If there’s one bellwether town in New Hampshire, let it be Nashua – it has a relatively diverse population.

So, how is the vote likely to go?  As always, I’m relying not on any sophisticated voting model. Instead, as I did with Iowa, I’m going to rely on the most recent polling data.  This time, however, I’m going to correct the mistake I made in Iowa by ignoring late trends.  In New Hampshire, recent polls suggest both Huntsman and, to a lesser extent, Gingrich, are trending up.   So…..drum roll please… here’s what we should see when the dust clears:

Romney 35%

Huntsman 18.5%

Paul 18%

Gingrich 13%

Santorum 11%

Remember, I’m a professional.  You shouldn’t do this at home. No wagering please.

I’ll be back on in about 15 minutes for live blogging.  Please join!

By How Much Must Romney Win To Really Win?

There are two contests on tap tonight. The first is the race to win votes in the New Hampshire primary, and with those votes earn all-important delegates.  To be sure, there are only 12 delegates at stake (New Hampshire’s total was reduced as a party penalty for moving its primary earlier in the calendar), and they are awarded proportionally, based on the candidates’ popular vote totals.  A candidate must clear the 10% threshold to win a delegate.  (By the way, if anyone is counting the New Hampshire popular vote, at this moment Mitt is leading 7 votes to 5 over Ron Paul, with Huntsman close behind at 4 votes. But I expect these totals to change as the night goes on.)

The second, and arguably more important contest, is to beat the media expectations game.   Keep in mind that there is no journalist’s manual for determining just what the benchmarks that each candidate must reach are; instead, they are apparently divined through a complicated process of consultation, group think, astrology and rooting through pig’s entrails.

It would help, therefore, if we could establish some real benchmarks by which to evaluate the media’s benchmarks.  Consider Mitt’s candidacy – what would be the proper benchmark by which to evaluate his vote totals tonight?  I’m going to use the same benchmark I proposed for Iowa: how does his total compare to past New Hampshire Republican primary winners?  If we look only at contested primaries back to 1988, the winner – on average – has received about 40% of the vote (standard deviation of about 9.3%)  Of course, one can quibble that this percentage depends in part on how many candidates there are, but in fact the number running today isn’t that different than in most past elections. So, based on past winners, I declare Romney’s success benchmark to be 40% or greater.  If he clears half a standard deviation above the average winning percentage – that is, 45% – he has had a very good night and the pundits are free to claim the race is over.  If he comes in below 45%, however, but is above 40%, he has still had a very good night – perhaps not a knockout, but pretty darn close.  Anything between 35% and 40% is fine, but it won’t convince the diehards that he’s the New Mitt.

So much for his benchmark of success. What about failure?  That’s easy.  Four years ago he won 32% of the vote.  He needs to clear that total tonight, or pundits have the right to begin writing “if Mitt can’t close the deal in his backyard” stories.

There you have it: clear benchmarks by which to evaluate the pundits’ benchmarks, and not a single pig killed in the process. Isn’t science the best?

It is traditional that I make my predictions regarding vote totals and the order in which the candidates will finish.  I will post that in a moment.

Meanwhile, keep in mind that the polls close in many New Hampshire towns at 7:30, but many others remain open until 8 p.m.  If the networks do not call this election for Romney precisely at 8 p.m., then he is in big doo-doo and you can screw the benchmarks.  Of course, the real excitement – beyond seeing if Mitt meets his benchmarks – is who finishes second – and who does not.  I’ll be live blogging tonight beginning shortly before 8.  As always, feel free to join in.

And journalists – you can thank me later for establishing your benchmark.

Huntsman in Dead Heat With Romney in New Hampshire!

That’s based on the early voting from Dixville Notch.  I always wanted to do that. In fact, Romney is going to win this. But maybe by not enough.

The last three polls before tomorrow’s (actually – today’s!)  New Hampshire primary are in, and they have both good news and bad news for Mitt Romney.  The bad news is that his support is definitely eroding.  He’s fallen 3%, from 40% to 37% in two days according to this ARG poll.  The rolling Suffolk tracking poll shows him dropping 7% in three days, from 40% to 33%, and WMUR poll has him dropping 3% in 5 days, from 44% to 41%.  And all this was before his widely quoted (but usually out of context) remark today that he likes to fire people.

He was actually making the point that under a free market system we should be able to fire insurance companies that don’t provide adequate health care service,  but why let the facts get in the way of a good campaign moment?

The good news for Romney is that he retains a substantial lead, and that it doesn’t appear that any single candidate is benefitting from his slide.  As  this RealClear politics graph shows, he retains a substantial lead over Ron Paul, one that will be hard to overcome in the last 48 hours (Mitt=purple, Paul=Yellow, Gingrich=green, Huntsman=purple, Santorum=brown):

As the graph shows, and as I noted earlier, there’s some evidence that Huntsman is the primary beneficiary of Romney’s slide, but the Newtster also seems have reversed his polling decline and Paul – although slipping – remains a good bet to finish second, particularly given the evidence that he has a strong organizational presence in New Hampshire.

In short,  I’m sticking by my earlier assessment that the belated decision by Mitt’s opponents to focus on his Bain record won’t be enough to change the outcome in New Hampshire.  But it may be a factor down the road.  As I predicted during my debate coverage, portions of Newt’s exchange with Mitt are already the centerpiece of one of Newt’s campaign videos:

And Newt’s SuperPac is poised to air an almost half-hour long infomercial that takes dead aim at Romney’s  record while CEO at Bain.  Progressives, of course, are chortling that these attack ads are simply making the Democratic case against Romney even easier, but this is nonsense.  Democrats are not stupid – they would have brought Bain up on their own.  Instead, the fact that this is coming out now, in the Republican nomination contest, is good news for Romney – assuming he wins the nomination – because it gives him an opportunity to formulate a response.

New Hampshire, then, is significant not for who wins here – it’s almost certain to be Romney – but for its implications for South Carolina.  That means that the real issue in New Hampshire is who exceeds media expectations. The media frame of the New Hampshire results has important implications for South Carolina. Gingrich has already gone on record that he has to win, or finish a very close second there, to stay in the race.  This is no sure thing.  Keep in mind that while the media tends to portray South Carolina as a “southern Red State”, it actually has a substantial population of Midwest retirees that are more moderate in outlook.  In fact, there’s really three somewhat distinct voting areas in that state.  The bottom line is that Romney could do quite well there.   I’ll develop that point in more detail in a later post.  Later today, however, Bert Johnson will be up with a guest post describing his view of New Hampshire from the ground. We’ll begin our primary-day coverage with Bert’s analysis, and I’ll be on periodically during the day to update results as they come in.