Scott Walker performed as well as a typical Wisconsin Republican. Explanation below.
Chances are, if you read an article about Wisconsin Governor Scott Walker in the national press, it will mention he's a conservative who's won an election more than once in a blue state. But what it might not mention is that his margin of victory isn't unusual for the midterm electorate he faced, and this makes a big difference.
Nate Cohn has supported this argument using results from governors' elections across the nation in 2010 and 2014. The Wisconsin contest sits nicely in the middle of the pack, smack on a regression line when results are plotted against Romney's percentage.
But wait! Scott Walker is no ordinary Republican! He is especially conservative—the second-most conservative governor to run for re-election in 2014, according to CFscores presented by Harry Enten. So achieving an ordinary win in a blue state for an extraordinarily conservative governor was an extraordinary feat!
In fact, on the topic of Walker's conservatism, Jason McDaniel, as quoted by Kevin Drum, says "it is not a stretch to argue that if nominated, Walker would be the most conservative Republican nominee since Barry Goldwater."
But Sean Trende and David Byler took exception to that argument, and with good reason. The CFscore is, after all, a campaign finance score, based not on the politician's words or actions but on who is donating to his or her campaigns. These numbers can change based on who is "hot" among nationwide conservative donors, among other reasons to be careful with them. Trende and Byler give the example of Mitt Romney, who had a CFscore of 0.88 before winning the GOP nomination, and 1.18 after. (More positive is more conservative.) Adam Bonica, who created the metric, has a similar example for the infamous Joe Wilson, whose numbers made a similar leap after his outburst during President Obama's speech to Congress.
The arguments are truly dizzying. So what are we to do?
The results are summarized in the graph above (Walker is smack in the middle of the pack), with details and explanations below the fold.
First, eliminate the question of degree of conservatism. In this post, we're going to compare Scott Walker's performance to other Wisconsin Republicans—who, in most cases, passed the same bills he signed, and are therefore as a whole just as conservative if not more so. Second, we will compare the gubernatorial election to races that were held the same year in the same state: US House, state Senate, and state assembly races.
Thanks to the Daily Kos Elections presidential results by CD and legislative district, we can treat Wisconsin as we did the entire US House in my previous post. Only contests with 99 percent of the votes going to Democrats and Republicans are included, with the exception of the governor's race, which was only 98.9 percent. Here's the graphs, showing how Republicans in Wisconsin races stacked up in 2014 compared to Romney in 2012:
Click to enlarge.
On the left, the vast majority of points are above the diagonal line, showing that most Republicans did better than Romney. On the right, points on the positive upper half of the graph show where Republicans did better than Romney. This is the same measure that is plotted in the first graph of this post. On both the left and the right, the point for Walker is visually in the middle of the cluster of points. Walker's vote share was 6.4 points better than Romney's, the same as the median of all the points shown.
Really, though, we should try to make comparisons within the same geographic unit. Below, I show comparisons between the Republican US House candidate's vote share in a given county and Walker's vote share.
Click to enlarge.
If Walker were a Superstar Republican, he should be doing better than other Wisconsin Republicans, and the data points on the left-hand graph should be shifted to the right. This would correspond to a downward shift on the right-hand graph. In fact, the median difference between Walker and the Republican House candidate shows Walker doing 0.3 points worse. It's obvious from the graphs that Walker is performing about the same as the other Republicans.
Is this last analysis a good way to judge Walker's performance? I'll go through the imperfections in a minute, but it should be good enough to notice a big difference if Walker were indeed a much better candidate than other Republicans. And, overall, this type of analysis showed, nationwide, that US House candidates, on average, typically have just about the same vote share as their presidential candidate in presidential election years.
The caveats for the Wisconsin analysis are, first, that we're essentially looking at only eight different elections to compare to the governor's race, just split up and mixed and matched. And second, not all of these counties had at least 99 percent of votes going to Republicans or Democrats, although the vast majority were 98 percent+. (You can see what effect that may have here—probably not much.)
So, in the end, Scott Walker is not Stupendous SuperRepublican, who can win blue states in a single bound. He's just a typical Wisconsin Republican, performing as well as other Wisconsin Republicans in a Republican-friendly year.
And perhaps the final nail in this particular coffin: PPP, which showed Walker winning over Burke in every poll of the 2014 governor's contest, now shows Walker losing to not just Hillary Clinton in the 2016 presidential race, but Joe Biden and Elizabeth Warren as well. And, to add a dash of insult, he does worse against Clinton than Mike Huckabee, Rand Paul, Rick Perry, and Marco Rubio.