I'm beginning to think that National Journal's Josh Kraushaar is just baiting political scientist bloggers because he believes that emulating Matt Bai is a clever career strategy (well, I might, if I thought that getting picked on by us was even plausibly helpful). Anyway, his column about special elections this week is, well, awful.
Did Scott Brown's win early last year "[foreshadow] significant gains"? Was the GOP loss in a NY special earlier in which conservatives bailed on the Republican nominee and wound up electing a Democrat a "harbinger" of Republican losses in Nevada and Delaware Senate races when they nominated Tea Party candidates? Well, sort of -- but unfortunately, the same is true of the other way around, right? How do we know that Scott Brown didn't falsely predict that the GOP would unite behind moderates and sweep those Senate races -- or the NY special falsely predict that Republicans would fail to take advantage of their opportunities? Note that Democrats won several other House specials in 2009. What that reminds us is that we tend to remember the contests that correctly "predicted' what later happened, and forget those that don't.
I wrote recently on real reasons that special elections might be (slightly) worth paying attention to, other than the quite real, if limited, importance of single House seats. But actually believing the spin...no, that's not a good idea.