Trend or Fluke?

I only get to celebrate special days like my birthday, wedding anniversary, Thanksgiving, and Christmas once a year, but I can rejoice every month or so when an NBC News/Wall Street Journal poll comes out (yes, this is sadly true). The survey, conducted by Democrat Peter Hart and Republican Bill McInturff, two of the most experienced and respected pollsters in the business, contains a treasure trove of data, most of which never appears on the air or in print.

Luckily for psephologists (yes, this is a word, meaning "students of elections") and political junkies, both NBC News and The Wall Street Journal release the full questionnaires online. This allows mere mortals to peer over the shoulders of top political pollsters and peruse data not dissimilar to what campaigns see. (Although, I am an NBC News political consultant, this isn't just sucking up; it's true.)

One of my favorite questions tests public attitudes toward government's role. The version that Hart and McInturff use gives respondents a choice between "Government should do more to solve problems and meet the needs of people" or "Government is doing too many things better left to businesses and individuals." The order is alternated to prevent bias.

Back in 2007 and mid-2008, the government-should-do-more camp was a slight majority, in the 52-55 percent range; the government-doing-too-much position was in the 38-42 percent range. Starting a month after Lehmann Brothers collapsed in September 2008 and when credit markets seized up, the results tightened up. The more skeptical view of government pulled ahead in the September 2009 poll, 49 percent to 45 percent. In the national exit poll taken by various news organizations on Election Day 2010, the government-should-do-more response dropped to 38 percent, and the more antigovernment attitude soared to 56 percent.

However, in the latest NBC/WSJ poll of 1,000 adults (including 200 by cellphone; overall margin of error plus or minus 3.1 points), conducted from February 24-28, 51 percent of respondents said the government-should do more and 46 percent said the government was doing too much. One could conclude that the antigovernment bandwagon certainly isn't picking up speed.

More important-and I have to give NBC Political Director Chuck Todd credit for pointing this out to me-independents shifted significantly. In the February survey, 47 percent of independents said the government was doing too much, compared with 60 percent who said so last October. Independents who said the government should do more jumped 13 points, from 38 percent to 51 percent.

Why is this important? Because independent voters are the ones who matter most in American politics. More than 90 percent of Democratic voters can be expected to vote Democratic, just as more than 90 percent of Republicans reliably vote Republican. In a bad year for Republicans, such as 2006, voters who call themselves Republican voted for GOP candidates over Democratic candidates by 91 percent to 8 percent. Last year, a great one for the GOP, Republican voters stuck with the party by 95 percent to 4 percent. In 2006, a great year for Democrats, party members voters cast their ballots for Democrats by 93 percent to 7 percent; last year, the numbers were 92 percent to 7 percent.

It's not about defections, and it isn't so much about turnout either. In 2006, 38 percent of all voters called themselves Democrats and 36 percent called themselves Republicans. In 2010, it was 36 percent for each party. The big difference was that independents in 2006 swung from backing Democrats over Republicans (by 57 percent to 39 percent), to preferring Republicans last November (by 56 percent to 38 percent). The swing in both elections was 18 points.

Democrats can be expected to support bigger government, as they did in this survey, siding 75 percent to 21 percent on the do-more side, and Republicans can be reliably expected to be on the antigovernment side, as they were, 75 percent to 22 percent. It's when independents make a big swing that election results shift. The point is, politically speaking, we need to focus specifically on how independents are moving as policy fights play out over the next two years. We know how partisans are going to see things. The question is whether independents will buy more into the Republicans' message or the Democrats'.

One poll isn't a trend, but the responses to this key question will be worth watching closely.