The "Bradley Effect" in New Jersey?

The breadth of opinion among 68 people was surprising to me, that and the amount of “not either” that materialized in such a small sample in the liberal state of New Jersey. Is it the case that New Jerseyans would rather vote for no one at all before they vote for Obama? Or is it that they would rather vote for Ron Paul than Bob Barr?

In the beginning, Hillary led the polls against McCain in New Jersey. Now, only half of her former followers are voting for Obama. In our sample, at this stage of the game, the other half of that group must be our “not eithers.” If we believe what Quinnipac told us in early September, those are men.  I think this split is tenable.

I wanted to talk about Walter Lippmann and the flawed dilemma of two party systems but these results are probably just an example of what people are calling the “Bradley effect.” That’s an interesting term that makes racism seem like a respectable phenomenon of political science and makes racists seem like partisans, rather than racists. The popularizing term, many now know, describes a California gubernatorial race in 1982, and then again in 1986, wherein an African American man lost to a white man despite being ahead in the opinion polls. We “discovered” then that when people were asked their opinion publicly, being averse to sounding racist, they voted for Bradley. Their votes turned white once they got into the voting closet. In this poll, which is absolutely anonymous, public scrutiny is absent so Obama loses half of the sample. 

Hopefully this disparity between public and private opinion will not show up this coming week. That would be unfair and, well, racist. Obama is doing well, at this point in the election, about as well as Gore did against Bush in New Jersey in 2000: 49-35 percent around this time. It was close in New Jersey for a while, but, according to Quinnipiac Obama has McCain 59-36 percent.  McCain sure is feeling the Palin!