Trying to follow the polls during this election season can make one dizzy. Are we headed for an Obama landslide and a 10- or 11-percent margin of victory? Or is this race still close, with Obama ahead by about 5 percent? The answer depends not only on which polls you read, but — I would argue — on how those polls are reported, especially with regard to the partisan breakdown of survey sample and the seeming dearth of undecided voters.
PARTISAN BREAKDOWN
Let me start with the Obama landslide polls — the most recent examples being a Newsweek poll released on October 11 showing an 11-point lead for Senator Obama and an ABC News/Washington Post Poll released on October 12 showing a similar 10-point lead. No doubt such polls cause great concern among McCain supporters — but should they? As with all polls, the best data comes from what we political diehards call “the internals” or the results below the line that measure candidate support among key demographic groups or public support for candidates on certain issues (the economy, the war, etc).
Consider the
Newsweek poll. Although the poll reports that Obama leads McCain 52 percent to 41 percent overall, the poll also shows that McCain is winning Independent voters 45 percent to 43 percent. Obama is winning Democrats 91 percent to 5 percent and McCain wins Republican by 89 percent to 7 percent. From these numbers, it’s difficult to understand Obama’s 11-point advantage — until you look further into the polls internals and see that the results are based on a respondent sample that was 40 percent Democrat (D), 27 percent Republican (R), and 30 percent Independent (I). Three percent expressed no preference. To see how that affects overall support one need only multiply each candidate’s level of partisan support (D, R, or I) by each group’s percentage of the overall sample. As shown in the table below, Obama’s 52 percent is based on 91 percent support among Democrats multiplied by Democrats 40 percent proportion in the poll, his 7 percent support among Republicans (27 percent of the respondents) and his 43 percent support among Independents (33 percent of respondents — add the “no preference” folks to this group). Tally Obama’s support across the three groups and you arrive at 52 percent. If you follow the same calculations for McCain, you’ll arrive at 41 percent.
| Poll Results Based on Newsweek's Partisan Breakdown |
| |
Respondent
Sample (RS) |
Level of Support by Party (LSP) |
Resultant Level of Support
(RS x LSP) |
| |
|
Obama |
McCain |
Obama |
McCain |
| Democrat |
40% |
91% |
5% |
36% |
2% |
| Republican |
27% |
7% |
89% |
2% |
24% |
| Independent |
33% |
43% |
45% |
14% |
15% |
| Overall Level of Support |
52% |
41% |
The problem is: In no recent election has the Democratic Party (or any party) enjoyed such an advantage among the American electorate. In 2004, exit polling data found the electorate to be 37 percent D, 37 percent R and 26 percent I. In the 2006 midterm elections for the House of Representatives the electorate was 38 percent D, 36 percent R, and 26 percent I. In 1996 and 2000, Democrats enjoyed a 4-point edge over Republicans. Given this history, it seems hard to believe that the Democrats have suddenly leapt to a 13-point partisan advantage. If one takes the Newsweek results and re-weights them to reflect a more realistic 4 point partisan advantage for Democrats the results change significantly. As can be seen in the following table, Obama’s 11-point lead shrinks to 4.
| Poll Results Based on Historic Partisan Breakdown |
| |
Respondent Sample (RS)
Historically Weighted |
Level of Support by Party (LSP) |
Resultant Level of Support
(RS x LSP) |
| |
|
Obama |
McCain |
Obama |
McCain |
| Democrat |
40% |
91% |
5% |
36% |
2% |
| Republican |
36% |
7% |
89% |
3% |
32% |
| Independent |
24% |
43% |
45% |
10% |
11% |
| Overall Level of Support |
49% |
45% |
CONTINUED 1 2 Next >
