The New Hampshire Primary Expecto-Meter

By Major Garrett

January 10, 2012 | 5:41 PM

Leave a Comment

New Hampshire is all about expectations. I have my own (not that anyone asked or particularly cares). But I have them. I think they are valid.

I measure them on my Expecto-Meter. I use three sets of data. 1). Polling since the Iowa caucuses 2). How candidates finished in Iowa this year compared to polls taken before the caucuses 3). How candidates who ran in 2008 and are running now performed in Iowa and New Hampshire relative to polling data before that caucus and primary.

Yes, every race is different, but polling data in 2008 and 2012 reflected the final results with more than a modest degree of precision. These sets of data, I believe, ought to tell us something about tonight - before the spin cycles begin.

Tonight's Expecto-Meter

Mitt Romney: Total vote percentage = 41.2 percent

Spread over 2nd place = 19.7 percentage points

Jon Huntsman: Total vote percentage = 23.8 percent

Ron Paul: Total vote percentage = 18.6 percent

Rick Santorum: Total vote percentage = 9.5 percent

Newt Gingrich: Total vote percentage = 8.8 percent

Rick Perry: Total vote percentage = 1 percent

Here's how I arrived at each calculation for the Expecto-Meter. (Note: All polling data complied from RealClearPolitics.com and its running tabulation of state polls in 2012 and 2008).

ROMNEY

Ten polls have been taken in New Hampshire since the Iowa caucuses. All polling experts advise taking a long-average of polling data to smooth out differences in methodology, voter availability (weekends versus non weekends), sample size and other factors. I've taken the 10-poll average for all the candidates and made other calculations to reach today's final score on the Expecto-Meter.

Romney's 10-poll New Hampshire average is 38.1 percent. He's also the dominant front-runner. His spread over his closest polling rival, Rep. Ron Paul, is 19.7 percentage points over those same 10 polls. Romney finished first in Iowa. His 10-poll average before the Iowa caucuses was 21.7 percent. He finished with 24.6 percent of the vote, a 2.9 percentage point improvement. This shows in Iowa, where the sledding has historically been tougher for Romney, he knew how to exceed his polling position.

In 2008, of course, Romney ran in New Hampshire and finished second to Sen. John McCain with 31.5 percent (McCain won with 37 percent). In 2008, Romney finished 3.1 percentage points ahead of his 10-poll average of 28.4 percent before that primary. Romney fell short of his 10-poll average in Iowa in 2008 - he polled at 26.6 percent and finished with 25.2 percent, a 1.4 percent deficit. Romney has proven this year he can run ahead of his polling average in Iowa. He should run ahead of his polling average in New Hampshire because he's done it before, he's run before and is well-known, has organized heavily there and is practically a native-son (and certainly a summer resident). Big finish. Romney's 10-poll average in New Hampshire is 38.1 percent. If you add the 3.1 percent by which he beat his 10-poll average in 2008, you get 41.2 percent.

But let's say Romney does only as well as he did in Iowa, where he beat his 10-poll average by 2.9 percentage points. That would put Romney's Expecto-Meter reading in New Hampshire at 41 percent. One last point. When McCain won New Hampshire in 2008, he beat his 10-poll average "spread" (the difference between his lead over Romney) by 1.4 percentage points. McCain's 10-poll spread was 4.1 percentage points and he won by 5.5 percentage points. If you take Romney's current spread of 19.7 percentage points and hold him to the McCain front-runner standard of 2008 (which I am not doing), Romney's spread should be 21.1 percentage points.

HUNTSMAN

Huntsman wants to be the Santorum of the New Hampshire primary. Considering he didn't compete in Iowa and has staked everything on New Hampshire, it's a fair standard to set. Whether he can meet it is a separate question. Huntsman's current 10-poll average in New Hampshire is 11.7 percent. That puts him third behind Paul. Santorum also was behind Paul in Iowa. To be the Santorum of the Granite State, Huntsman has to catapult past Paul with a surge of late-breaking voters he's courted consistently for months - just as Santorum did in Iowa.

Is there a way of setting a suitable expectations standard? I believe there is. Again, Huntsman's 10-poll New Hampshire average is 11.7 percent. In Iowa, Santorum's 10-poll average before the caucuses was 12.4 percent. He finished with 24.5 percent. Since a primary in good weather is much easier than a caucus in good weather (which Iowa had this year), it seems Hunstman ought to be able to hit the same catapult factor as Santorum. If you add Hunstman's N.H. poll average of 11.7 percent to Santorum's Iowa "catapult" (12.1 percentage points), you arrive at 23.8 percent. (Incidentally, McCain in 2008 finished with 13 percent in Iowa and 37 percent in New Hampshire - an even bigger gain than the Expecto-Meter sets for Huntsman because McCain ran and won in 2000 in New Hampshire and was much better known).

Huntsman is claiming a "surge" in New Hampshire. This is, I believe, a credible way to place his vote outcome tonight in a surge-relevance context. If Huntsman doesn't meet the Expecto-Meter standard, it won't mean he's failed. But it will suggest he surged too late and his all-in New Hampshire strategy was less effective than Santorum's. That will make it very difficult for him to campaign as a credible alternative to Romney in South Carolina or Florida.

PAUL

Paul's Expecto-Meter is based on his 10-poll average in New Hampshire, which is 17.9 percent. It's also based on Paul's history of finishing very closely to where he polls (people who say they are voting for Paul, vote for Paul). In Iowa, Paul's 10-poll average was 21.4 percent. He finished third to Romney and Santorum with 21.4 percent. In Iowa in 2008, Paul's 10-poll average was 6.5 percent and he finished with 9.9 percent. In New Hampshire in 2008, Paul's 10-poll average was 7 percent and he finished with 7.7 percent. Owing to Paul's larger prominence and stronger organization than in 2008, he should easily hit his 10-poll average in New Hampshire now. If he exceeds it, that will allow him to claim momentum and intensify his desire to extract policy concessions if he does not -- as many expect - win the nomination.

SANTORUM

Santorum's 10-poll average going into tonight is 9.9 percent. He has campaigned in New Hampshire and was better organized there than some appreciated before he finished a close second in Iowa. He should meet this 10-poll average. But even if he doesn't, there is this historical context to consider. Since Santorum is most often compared in ideology and Iowa-centric strategy to the 2008 campaign of former Arkansas Gov. Mike Huckabee, it's worth looking back at Huckabee's Iowa and New Hampshire experiences. Huckabee finished third in New Hampshire with 11.2 percent after winning in Iowa with 34.4 percent. That was four-tenths (0.4) of a percentage point percent less than his 10-poll New Hampshire average of 11.6 percent. So, if Santorum is the Huckabee of 2012, he can finish with 9.5 percent and still fall within the Expecto-Meter range.

GINGRICH

Gingrich's 10-poll average is 9.7 percent in New Hampshire. He's fighting desperately and angrily not to finish fifth behind Santorum. In the five polls taken before today, Gingrich's average is 10.4 percent, suggesting a slight uptick. But Gingrich underperformed his 10-poll average in Iowa. He polled at 14.2 percent and finished with 13.3 percent for a deficit of nine-tenths (0.9) of one percent - making him the only top five finisher in Iowa whose final results were lower than his 10-poll average going into the caucuses. If Gingrich is blighted by the same under-performance, he could finish with 8.8 percent. That's the low end of the expectations range. If he exceeds his five-poll average going into the primary (10.4 percent) he could claim some momentum heading to South Carolina. But if Gingrich falls below the low end on the Expecto-Meter, he will have distinctly and measurably under-performed in the first two contests of this nominating process. You can do that once and bounce back if you're near the top of the heap. You can never bounce back from this if you under-perform twice at the bottom of the heap.

PERRY

In ways the Texas Governor may not know and may not appreciate, he is the true "one percent" in New Hampshire. Perry's registered 1 percent in New Hampshire in the last 14 polls. Talk about consistency.

View All Decoded Posts by Major Garrett