Looking Back at 2004 and 1992: Solidifying Support Two Weeks Out

So we hear it everyday now. The wheels are coming off of the McCain campaign. With poll numbers at a seemingly insurmountable level, many pundits are preemptively calling the election for Senator Obama. Though it definitely makes me nervous to agree with them, history tends to agree. I looked back at the 2004 and 1992 elections for reference, and both instances made me feel more confident that polling one to two weeks out is a strong harbinger of the final point spread on election day.

On Thursday (10/23), the New York Times posted the results of their latest NY Times/CBS poll. Obama is up 52-39 in the poll among likely voters. This spread is quite incredible and when you look back at the Times's past polls at roughly the same time in the 2004 and 1992 campaigns, the results seem to provide an early window to the eventual spread. Take this from a similar October 25, 1992 article.

The upshot is a tighter race. Among the probable electorate, which reflects the likelihood of each respondent's voting on Nov. 3, Mr. Clinton now has the support of 40 percent, Mr. Bush of 35 percent and Mr. Perot of 15 percent, the poll shows. In early October, before the debates, Mr. Clinton stood at 46 percent, Mr. Bush at 38 percent and Mr. Perot at 7 percent. Elation in Bush Camp

Bush campaign officials, elated over signs in this and other polls that Mr. Clinton's substantial lead was eroding, predicted that voters were entering a new phase of decision-making that would put a premium on trust. "It's moving our way," said Charles Black, senior adviser to the Bush campaign.

But Clinton strategists disputed the Times/CBS poll, noting that other independent surveys, while showing a closer race, did not show the gap as small as five percentage points. "I never question somebody's polls," said James Carville, senior strategist for Clinton. "But this is the first private or public poll with a margin like this."

The quote from the Bush campaign sounds similar to some of the soundbites coming out of the McCain campaign this weekend. In the end Clinton beat George H.W. Bush 43.0% to 37.7%, a difference of 5.3%. Ross Perot received 18.9%. If according to the poll that 10% were undecided nine days before the election, it seems that each candidate got roughly equal shares of that group. In 2008 this group is standing at 9%. Say they break 5% and 4% for each candidate. The final result should stand at Obama 57% and McCain 43%. My feeling is that the undecideds could break the opposite way just as easily, but really any changes are marginal.

In 2004 the same poll showed President Bush with 47% support and John Kerry with 46%. Again, very much in line with the eventual result, Bush 50.7% and Kerry 48.3%. This is a 3.7%-2.3% break. Again, the undecideds split with each candidate getting a sizeable percentage.

So barring something unforeseen, Senator Obama seems to be solidifying support similar to the winners of the 1992 and 2004 elections in the closing days of this election.