(Click chart for full size)
Here is my overview of the GOP Field vs Obama as of mid-September 2011. GOP Candidates are in their order of strength for the nomination among GOP voters, not strength against Obama.
While Perry has surged among Republican voters for the nomination, Romney remains the most competitive against Obama among all voters. As for the nomination, we’ve seen a recent drop-off in Perry support, suggesting his debate performances and mostly negative media coverage of several issue positions has cost Gov. Perry among independents and Democrats at least.
Romney has been quite competitive with Obama for the past 2 months, with a tiny hint of a shift recently. While Perry has surged ahead of Romney in the GOP nomination race, it appears as of now that Romney has more appeal for the general election.
Michele Bachmann has seen a very sharp drop in support since winning the Iowa straw poll in August and is now far from competitive.
Non-candidate Sarah Palin trails Obama by a dozen points, though she has gained about 5 points over the last two months while Obama has lost about the same against her.
Of the other GOP candidates, Cain and Paul run closer to Obama than any others, though we have fewer polls for them, especially for Cain.
The second tier candidates all have significantly fewer polls than do the top two. This makes the estimates less stable and exaggerates the potential for pollster “house effects” to obscure or exaggerate candidate support. For example, if polls that lean Republican are more likely to include questions about Cain, then part of his support is due to the pollster’s “house effect” rather than to Cain’s personal support.
One should be careful of the trend estimates for any of the candidates with few polls, mostly obvious due to the much more ragged estimated trends in those cases.
An alternative way of looking at the candidate’s strengths is to see each candidate and the entire GOP field against Obama. The chart below shows all GOP candidates as gray dots and highlights each in turn with red dots. Points above the diagonal are polls with the GOP candidate ahead of Obama while points below the diagonal have Obama ahead. While Obama has clearly led the field in the substantial majority of polls, there are some polls with GOP candidates ahead.
While no GOP candidate has established a consistent lead, Romney and Perry show the greatest number of leads. Romney’s results are in a more compact region, showing some consistency as a serious challenger to Obama. Perry’s data is much more spread out, though this could reflect both his recent entry and relative unfamiliarity with voters as well as his recent downturn seen in the first chart above. We need more time to get a good fix on the dynamics of the Perry candidacy.
Finally, we can look at how the GOP field has changed against Obama in the last six months. The chart below plots all GOP contenders against Obama by month. As the cloud of points moves up and to the left, the GOP field as a group is stronger and Obama weaker. As the cloud of points moves down and to the right, Obama is stronger and the field weaker. The goal here is to see the strength of the field as a whole, rather than individual candidates, and to assess Obama’s relative strength against that field.
Overall, the GOP field is a bit stronger, or Obama a bit weaker, over the past six months, though the vast majority of polls clearly show Obama with a substantial lead. With September only half done, the number of polls is smaller than previous months.
One important overall note is that GOP candidates are still unfamiliar to many voters, even Republican voters. Obama’s advantage in these polls is partly real but it is also partly due to unfamiliarity with the challengers. There will be a substantial change in candidate familiarity over the next six months or so as the primaries begin.
Methodological notes: The gray dots include every GOP candidate ever tested against Obama since April 2011, including some (e.g. Pawlenty) who have dropped out and some (Trump, Giuliani) who never entered. I retain them because the plots against the field are intended to give an overall sense of Obama versus all potential challengers. The highlighted candidates, however, are all currently active or the exceptional Sarah Palin. The trend estimates in the first chart are based on local regressions that use all the polling data, rather than average only the most recent polls. This trend estimate works very well when there are a substantial number of polls but becomes jagged and irregular with few polls. It should be obvious that candidates with few polls provide too little data to be confident of trends. I go ahead and plot these trends but be aware that with less than 15 polls or so they may be unstable.