Is John Bohlinger a Democrat? Show me the data!

Two weeks ago, the Montana Democratic Party saw fit to endorse Senator John Walsh in the Democratic primary. Bryan Watt, spokesman for the Democrats, said in announcing the party’s support for Walsh, that neither Adams nor Bohlinger were actual “Democrats.”  Of course, John Walsh’s opponents—Wilsall rancher Dirk Adams and former Lt. Governor John Bohlinger—were extremely upset by the decision and took umbrage at the party’s decision to declare them party non gratis.

In this post, I’d like to assess the claim that John Bohlinger is not a Democrat. The former Lt. Governor points to a slew of progressive legislation that he carried while serving in the state legislature, and of course, claims his work in the Schweitzer administration as demonstrating his fidelity to the principles of the Democratic Party. Others, notably liberal blogger Don Pogreba over at Intelligent Discontent, have pointed to other evidence suggesting otherwise. They include advertisements aired by Schweitzer’s campaign during the 2008 reelection where Bohlinger notes he’s a Republican, Bohlinger’s support for (and willingness to chair the state campaign committee for) John McCain during the 2008 presidential election, and Bohlinger’s record on abortion—which some claim is not consistent or liberal enough for a Democrat. (And, see here, here, and here for the tit for tat between Pogreba and Dirk Adams over Dirk’s record—fun stuff).

Like most things, I prefer to look at solid empirical data to sort out these “he said, she said” type claims. How can we get closer to understanding John Bohlinger’s claim to be a Democrat?

Fortunately, we can look at public positions in the aggregate and individual level. Since John Bohlinger served in the Montana House and Senate, we can compare his record there to other legislators to see how he stacks up to other Republicans and Democrats. Unfortunately, since Dirk Adams has not served in public office, I can't assess his record using this method.

In a previous blogs, I utilized DW-NOMINATE scores to examine the voting records of Montana’s congressional delegation in the post-World War II era (here and here) Two political scientists have taken the NOMINATE estimation model developed by Keith Poole and Howard Rosenthal and, using the Project Vote Smart National Political Awareness Test (a survey of federal and state legislators), generated common space NOMINATE type scores of state legislators serving in all fifty states between 1993 and 2011. Grab these data here. I encourage you to read their APSR paper on the process and to go to their website where they data are available for download and analysis in the comfort of your home. It’s available here.

I simply downloaded the Schorr and Rosenthal data and pulled out the Montana legislators for analysis. Like NOMINATE, higher positive scores indicate more conservatism. High negative scores, more liberalism. Unlike NOMINATE, there is only one score for each member per chamber rather than a score per session. This means there is one score for the member—that is, unless they change parties. Then a new score is computed. Again, let me refer you to Schorr and Rosenthal’s FAQ section of their website here.

Before getting into the analysis, let’s do something called a face validity check. Do the scores assigned legislators make intuitive sense given what we generally know about Montana legislators? Here are the five most conservative and liberal legislators for the whole period listed in Table 1. Anyone who follows the state legislature is probably not surprised by this list of the most liberal and conservative members. And, as a result, this measure of ideology would seem to exhibit a certain degree of face validity.

Table 1: Ideology Rankings of Montana State Legislators, 1993-2011
Most Conservative			Most Liberal	Ideology Score
Jore	2.486		Toole	-1.531
Sales	2.219		Ellingson	-1.349
Koopman	2.047		Ellinson	-1.328
Everett	2.038		Bixby	-1.286
Adams	1.963		Buzzas	-1.275
Hawk	1.907		Doherty	-1.265

The mean value of the Republican Party is .976 for the entire period, and for the Democrats, it’s -.654. Jon Tester, who has established a voting career in the U.S. Congress to the right of Democrats in the U.S. Senate, had a very similar voting record in the Montana State legislature. Again, his score of -.431 is to the right of the Democrats serving in the legislature. In fact, it is about half a standard deviation to the right of the mean. One thing that is fairly well-established among those studying roll call behavior: members rarely change their ideology during their careers. Tester, by this measure, has been consistent—as we would expect from the literature.

Where does John Bohlinger sit? For the entire period, the average score for Republicans is .976. Bohlinger’s ideology value is .322. Essentially, this indicates that out of the 247 legislators serving in the Montana legislature and coded in these data, John Bohlinger compiled the 14^th most liberal voting record. That is two and a half standard deviations to the left of the Republican mean. Bohlinger was a pretty liberal Republican during his time in Helena.

But….

Had John Bohlinger served as a Democrat, it would have made him the fourth most conservative Democrat to serve during the period. In fact, John Bohlinger would be about two and a half standard deviations to the RIGHT of the mean Democrat. As liberal as Bohlinger was as a Republican, he’d still be a pretty conservative Democrat in the state legislature.

Relative to Republicans, John Bohlinger is quite liberal based upon his NOMINATE vote score accumulated during his time in the legislature. Relative to other Democrats, he’s pretty darned conservative. Maybe he’s become more liberal since then as he’s served with Governor Schweitzer. But, many political scientists suggest that such dramatic ideological conversions are relatively rare. I’d peg Bohlinger, from these data, as a moderate who sits ideologically in the broad center between the two parties—two parties that have, at least in the national legislature, have polarized over the past three decades. And, by the way, a center that has become essentially abandoned in American politics.

Is John Bohlinger a Democrat? Frankly, that’s the beauty of primary elections: it is up to the electorate participating in the Democratic primary to decide—notwithstanding the party’s endorsement of Lt. Governor John Walsh. The voters, ultimately, get to make the call.

Of Campaign Ads and Campaign Arcs

Today, Congressman Steve Daines announced that the first TV ad of his Senate campaign will drop on Montana's stations tomorrow. Watch "Interview" below:



Let's compare it to Jon Tester's first ad from 2012, "Combine":




And Denny Rehberg's first ad, "Honest":





Note how Tester and Daines start with positive biographical narratives. This is the traditional campaign arc followed by most campaigns. Rehberg's Senate race, alternatively, chose to use a contrast ad instead of starting with the positive, biographical narrative.

Also note that in 2012, Tester aired his first spot in March of 2012 after organizations had been airing ads for over a year. Rehberg dropped his first spot in May. Daines is launching his ad in February. Why? First, Daines has less name recognition than either Tester or Rehberg. He needs to build that up--the earlier, the better. Second, I suspect Daines wants to build up his favorability ratings as well before the hits start coming from outside groups (that could happen at any time). Third, Daines has raised more money than the Democratic candidates in the race and has more on hand. He might hope that this early expenditure puts pressure on the Democrats to start their own ads, too--which would put some pressure on their financial resources earlier than they would like.

Political science scholarship is mixed on the effectiveness of early ads. Most recent work suggests that ad effects decay quickly (see for example the discussion in Sides and Vavreck's book The Gamble), thereby suggesting that early advertising is not terribly useful or beneficial. I think, however, that the story is not so clear, especially when an early advertising buy occurs when the airwaves are essentially free of other messages (at the very least, Ridout and Franz find some positive effects of early advertising in their 2008 book). At the very least, I conclude that Tester's March-April buy in 2012, which represented the single largest advertising advantage in terms of spots aired during the campaign (when you include the outside group CSS that had an ad up attacking Rehberg on congressional pay raises in March), helped the Tester campaign over the long haul because it helped remind some Montanans that they genuinely liked the guy even if they disagreed with him politically. Political scientists generally under-appreciate the power of positive advertising and the likeability of a candidate as a factor in voting decisions.

Will the Daines ad buy yield a similar result? Time will tell. If anyone has information on the size of the buy and on what stations, I'd love to know.