Close doesn’t count in elections — just ask Al Gore or Norm Coleman.
It appeared as late as midnight on Tuesday that Democrats had a chance to win the three seats necessary to give them control of Wisconsin’s State Senate in an unusual and expensive recall election. In the end, however, the Republican incumbent, Alberta Darling, surged passed the Democratic challenger, Sandy Pasch, as votes from suburban Milwaukee were counted. Democrats picked up two seats — but Republicans maintained a 17-to-16 majority.
Where margins of victory do matter, however, is when you’re making inferences about future elections from the results. Here’s what I wrote yesterday about Wisconsin, and the implications for Gov. Scott Walker, a Republican:
The one thing it would be safe to say is that, if Democrats have a strong night, Mr. Walker will be in some jeopardy, as he could face his own recall election next year.
If you are going to read into the results, it is probably best to compare them to Mr. Walker’s performance in 2010. […]
Mr. Walker carried the six districts on Tuesday’s recall ballot by an average of 13 percentage points in 2010 — better than his statewide margin of 6 percentage points. If Democrats were to split the vote across these districts about evenly, that would be a reasonably troubling sign for Mr. Walker, however many of the seats Democrats actually win.
One Republican incumbent, Robert Cowles, bettered Mr. Walker’s performance in his district, winning 60.4 percent of the vote as compared to the 57.4 percent that Mr. Walker achieved there in winning the governor’s race last November. The other five lagged Mr. Walker, some modestly and some significantly.
Combined across the six districts, the Democratic candidates received 47.3 percent of the vote, and the Republican incumbents 52.7 percent.
Each of these districts are swing districts — but, nevertheless, five of the six are slightly more Republican than Wisconsin as a whole. In total, they gave Mr. Walker 55.6 percent of their votes in 2010, as compared to the 52.3 percent he achieved statewide — a Republican lean of 3.3 percentage points. (The difference mostly reflects the fact that no portions of Democratic-leaning Madison, and few portions of Democratic-leaning Milwaukee, participated in the elections last night.)
What that means is that if you wanted to use Tuesday’s results as a proxy for a recall vote on Mr. Walker, you’d want to subtract 3.3 points from the Republican total and add 3.3 points to the Democrats’ total to get a sense of how the whole state might have voted. That would point toward a recall vote that would be too close to call: Republicans would have gotten 49.4 percent of the vote, and Democrats 50.6 percent, according to this method.
But, we should be careful. One thing that was apparent Tuesday is that the individual candidates mattered quite a lot. Compare the Second District to the 18th District: they voted almost identically in the governor’s race in 2010 as well as the presidential election of 2008. In the Second District, however, the Republican incumbent, Mr. Cowles, won in a landslide, beating his Democratic opponent by 21 percentage points, whereas the Republican incumbent in the 18th District, Randy Hopper, lost his seat to the Democratic challenger, Jessica King, by 2 percentage points.
So, there’s at least some uncertainty about how these results might roll up at the state level (never mind the national level). What evidence we do have, however, suggests that a recall election against Mr. Walker would be quite close. There are some risks to Democrats in pursuing one, like inducing voter fatigue and draining resources and candidates from other races (like Russ Feingold, the former Democratic senator who could run for either governor or the Senate next year).
But in theory, their calculus shouldn’t have changed much based on Tuesday’s results.
In practice, however, a recall effort against Mr. Walker would require collecting 500,000 voter signatures, something that requires not just money but enthusiasm. Tuesday’s results won’t affect the math much, but they will negatively affect Democrats’ morale.