The Times, It Is A-Short-Changin'

Democrats, that is, with apologies to Bob Dylan.

All you need to know about the latest New York Times/CBS news Poll released on Friday is on the first page of the poll document. I'll save you some time and recreate it below:

September 10-15, 2011
Total N= 1,452
Registered N=1,356
Republican N= 781
Republican Primary Voter N=747
That's right, according to these numbers, 53.8% of the polls' respondents identified themselves as Republicans. Yet, when you go to page 27 of the poll, only 26 % say they are Republicans and 32% identify themselves as Independents. What gives?

In any case,  the news coming out of the poll will be the decline of President Obama's approval rating, 43% and that 72% of the respondents say that the country is headed in the wrong direction. Rick Perry and Mitt Romney seem to be putting some distance between themselves and the rest of the Republican pack, which is not surprising, and Congress continues to have abysmally low approval ratings (12%). Oh, the economy and jobs ranked 1 and 2 as the most critical issues facing the United States. At least the voters have recognized this.

And yet.

According to this Rasmussen Poll of likely voters, Obama is leading every Republican challenger except Mitt Romney. The breakdown:

Rasmussen
9/14-15/11; 1000 Likely Voters, 3.0% Margin of Error
Mode: Automated
Rasmussen Results
National
2012 President: General Election
Obama: 46%, Perry: 39%
Obama: 46%, Bachmann: 33%
Romney: 43%, Obama: 40%
Obama: 39%, Paul: 38%
Obama: 50%, Palin: 33%
Obama: 42%, Cain: 35%
Obama: 48%, Gingrich: 30%
Obama: 44%, Huntsman: 28%
Obama: 45%, Santorum: 31%

These results are surprising for a couple of reasons. First, Rasmussen polls are generally seen as having a slight Republican bias, so a poll that shows Obama ahead of most of the field given the results of other poll results, shows that the media narrative and the reality are somewhat divergent.

Second, the New York Times poll is of adults and registered voters, which means that the Times didn't ask the respondents questions about their voting patterns or delve into their voting history. It asked only if they were registered to vote. Rasmussen, on the other hand, screens voters by asking them how likely it is that they will vote in the presidential election at various times in their poll. They also ask about the respondent's voting history in order to gauge the likelihood that they will vote in 2012. The assumption is that likely voters will give a more accurate picture of the race than if you just asked those who were registered.

To compare further, the Gallup Daily Tracking Poll, from the other major polling firm cited in the media, doesn't screen for either registered or likely voters in their daily polling. They only ask if an adult is home to answer the question. That makes their daily tracking poll the least reflective of what's actually transpiring in the presidential race.

So if we look at today's numbers for all 3 polls that measure Obama's approval/disapproval numbers, here's what we get:

NY Times/CBS News (registered voters) 43/50
Gallup (adults) 39/52
Rasmussen (likely voters) 45/55

Even though the spread is greater in the Rasmussen poll, the approval numbers are higher among those likely to vote, which provides a more accurate account of the race at the moment.

National polls can give us a snapshot of what the electorate is thinking, and 14 months before the election that picture can be rather blurry. Remember, too, that our electoral system is based on results in the individual states. For that, take a look at this Public Opinion Strategies analysis of  the race so far. The entire election could come down to the vote in 9 states.

Numbers can be funny things: They don't lie, except when they do.