As Romney sunk in the polls, my Republican friends began dissing the methodology used by the pollsters. But even if the methodology is flawed, the polling can still be relevant. Let's say, for example, that a Gallup poll shows Romney even with Obama and then take another poll that, post-47% comment, shows him down 6%. Pollsters tend to use the exact same methodology from poll to poll, so two polls from the same pollster, employing the same methodology, would still be relevant in tracking movement. Movement was definitely shown to be in President Obama's favor following Romney's 47% comment.
I should note that if the polls started showing a Romney lead, Democrats would be challenging the methodology of those polls. Neither side much believes in the inexact science of polling when the polls don't come down in their favor.
I, on the other hand, believe strongly in the science of polls. Most of these pollsters take their craft seriously. While a flawed methodology might affect the result of the poll, generally you're only talking a few points. Each pollster has to make adjustments to adapt to changing circumstances. For example, the fact is many people no longer have land lines with a published name and address that can be cross-referenced against voter registration lists. When you're making those land line calls, you have to call people who are specifically on the list in order to get a representative sample of the electorate. If a husband;s name is on the list, you can't instead talk to his wife who answers the phone. However, when pollsters start calling cell phones they are dialing into the dark.
In the last two days, four polls have been released showing Obama with leads of 3, 3, 4 and 2 points. The most recent poll released late yesterday was CNN's which showed President Obama ahead of Romney 50-47%. Looking at the cross tabs on the poll, 756 interviews were conducted of people with landlines and 257 interviews were with cell phone users.
The CNN poll interviewed 305 people who identify themselves as Democrats and 107 who call themselves as independents who lean Democratic, for 412 Registered Democrats in total. People who identified themselves as Republicans in the poll was 270 with 143 identifying themselves as independents who lean Republican, for a total of 413 Registered Republicans. Apparently the rest are independent voters who refuse to identify with either party.
What head-to-head poll results don't do, however, is capture enthusiasm which drives turnout to the polls. You get elections where the polls are close, but once side wins by a landslide because of the enthusiasm factor. The presidential election of 1980 is the classic example. The polls were very close, but Republicans were enthusiastic about Ronald Reagan while the Democrats were not thrilled about another four years of President Carter. Republican voters flocked to the polls while Democratic voters stayed home. Reagan won an enormous landslide win that year that was not at all reflected in the polls..
I sensed that enthusiasm for the Romney had been waning. The CNN cross tabs showed that. When asked about how enthusiastic they were about voting in the presidential election, here are the results:
Registered Democrats v. Republicans (first number is Democrats)
Extremely enthusiastic: 39 - 38
Very enthusiastic: 25 - 27
Somewhat enthusiastic: 21 - 17
Not too enthusiastic: 8 - 11
Not at all enthusiastic: 7 - 7
The CNN poll shows that Obama has more than closed the enthusiasm gap. Romney needs to get Republicans and Republican-leaning independents re-excited about his candidacy. If he fails to do that, he won't be able to close the gap on Election Day with better turnout.