I expressed frustration about this via Twitter, but I may as well take to the blog to point out that by far the most ridiculous thing about the impulse to point toward VA/NJ/NY-23 as indicators of the national mood is not so much the historical unreliability of these indicators, it’s the fact that we have statistically valid surveys of national public opinion available for our perusal.
For example, how do people feel about Barack Obama? Well:
And what are people’s intentions with regard to congressional voting? Well:
By the same token, you don’t actually need to know the outcome of the NJ gubernatorial election to know what New Jersey residents think—the polls are very clear that most people don’t like Corzine and most people don’t like Christie. The question of who wins the election in the end (which will come down to turnout issues, tactical voting, and second-choice preferences) is very important to the future of public policy in New Jersey but it doesn’t tell us anything we don’t already know about underlying sentiments.
Also worth noting, Obama job approval:
Of course this is national data. But again if we want to know whether Obama has become unpopular in Virginia, the way to do that would be to do a poll of Virginia and ask questions about Obama.