Can We Trust the Polling?

The Hotline's Reid Wilson has an interesting thought piece today entitled "Why Democrats Will Keep The House," listing a number of reasons why the Republicans may not be able to gain control of the lower chamber of Congress: money, turnout, opposition research and modeling. The last reason stood out to me in particular.

Hours after the polls closed in a special election in Pennsylvania's 12th District earlier this year, Boehner was still predicting that businessman Tim Burns (R) would beat former congressional aide Mark Critz (D). Republicans' polling showed Burns would win. Their polling was wrong, and Critz is now a congressman.

Democrats who kept a close eye on the Pennsylvania polls detected the same excitement gap that Republicans saw -- Republican voters were much more enthusiastic than their Democratic counterparts. But Democratic polls also saw their voters come home in the race's waning days, and adjusted their targets accordingly. Republicans did not, resulting in a blown call and a self-deprecating examination of the party's polling techniques.

Wilson makes two important points here. One, polling has been extremely erratic this year. Pollsters have been having a doozy of a time getting proper samples, and polling data have often not comported with actual results. Wilson points to the special congressional election in Pennsylvania's 12th district, where the Democrats pulled off a victory few outside the Democratic Congressional Campaign Committee thought they could, or think most recently the Florida Republican Governor primary, where most of the final polling showed Bill McCollum winning a race he eventually lost to Rick Scott.

Why does this matter? The answer is simple: There has been a lot of polling showing Republicans trouncing the Democrats around the country. But polling an election does not necessarily make. If it did, Tim Burns would be a Congressman, and McCollum would be his party's nominee. They're not, and it's possible that predictions of the Democrats' demise are hasty.

The other important, and related, point that Wilson makes is that although Republican voters were more enthusiastic than Democratic voters leading up to the election, Democrats did turn out. Enthusiasm is important -- but to a point. In the end, if Democrats show up to the polls, enthusiastic or not, they are still voting. It may be harder for Democratic campaigns to get out their vote, no doubt; but just because Democratic poll respondents are telling pollsters they aren't thrilled about voting doesn't mean that they won't, in the end, vote.