Thursday, June 5, 2014

Battle for the Big Sky: The Director's Cut

In order to celebrate and build interest in my book on the Tester-Rehberg Race (Battle for the Big Sky, to be published this fall with CQ Press), I'm going to release material that I just couldn't get into the final version. Please check out my post at the blog for the book, Battle for the Big Sky. You can access the blog here.

Stay tuned for more in the weeks and months ahead!

Tuesday, May 27, 2014

Who is the “most” conservative House candidate? The data--not so much.



Sometimes data illuminates and makes patterns easier to see. Other times, data do not illuminate and instead confuse. When it comes to understanding the ideologies of the Republicans running for the House here in Montana, it seems to obfuscate more than clarify.

Mike Brown over at the Western Word opines thusly about the upcoming Republican House primary:
“Republicans across Montana can be heard saying, “Candidate, candidate on the ballot, who is the purest conservative of them all.”

One of the problems with primary elections is they are, by their very nature, low information environments. This makes it hard for voters to distinguish clearly the ideologies of the candidates. I said in a recent interview with Chuck Johnson that endorsements matter a lot in this type of environment because they provide easy cognitive shortcuts for voters to make decisions.

One possible way to help sort this out is by looking at empirical data. I propose using endorsements and the state NOMINATE values generated by Boris Schorr and Nolan McCarty to put three of the House Republican candidates (Else Arntzen, Matt Rosendale, and Corey Stapleton) on a left-right spectrum to decide who is the “purest” conservative of them all. In an earlier blog, I used these same data to conclude that John Bohlinger’s voting record in the legislature suggests he would be quite a conservative Democrat if elected to the United States Senate.

Corey Stapleton and Elsie Arntzen served in the state legislature before 2011, so they actually have NOMINATE scores based upon their votes. Stapleton had a score of .934 as a representative and 1.032 as a senator—averaging the two scores gives us a .983 for Stapleton. Arntzen had a score of .795. 

Rosendale is not included in the Schorr and McCarty data. I have, however, generated a score from him based upon the 11 legislators who have endorsed him and have NOMINATE scores in these data. Note that Rosendale has more than 50 current and former state legislators who have endorsed him, so this is only a slice of his total endorsement haul. The endorsements were culled from Rosendale’s campaign website. Averaging the scores of these 11 gives us a value of .872.

Again, the higher the value, the more conservative. The mean value of Republican state legislators for the entire period is .976. According to this analysis, that would place Elsie farthest to the left, followed by Rosendale, and then Stapleton—with only Stapleton to the right of the average state legislator.

Finally, there is the case of Ryan Zinke. As far as I can determine, he has no public endorsements from state legislators, current or former. Chuck Johnson of Lee Newspaper reported none, I found none listed on Zinke’s website, and an e-mail to Zinke’s campaign manager requesting a list has received no response. 

So, using this method, I can’t evaluate Zinke’s conservatism relative to Arntzen, Rosendale, and Stapleton.

But I CAN use Common Space scores calculated by Poole andRosenthal to help us pin Zinke down. Zinke received the endorsement of former U.S. Senator Conrad Burns. Rosendale received former Congressman Ron Marlenee’s blessing, and Artnzen landed former Congressman Rick Hill’s support. Stapleton, unfortunately, has not received the support of a former federal officeholder, so again, I’m left without complete information.

Using this metric, Arntzen comes out the most conservative as Rick Hill’s score of .448 is the most conservative, followed by Marlenee’s .397, and Burns’ .364. This suggests a ranking from least to most conservative of Zinke, Rosendale, and Arntzen with Stapleton as indeterminate.

One method finds Arntzen the most conservative, and another Stapleton. What do we make of this?
I’d put a bit more stock in the method based upon legislative voting records if only because there are more data points available for analysis compared to the voting records of only three federal officeholders. It is clear to me that Zinke is the hardest to pin down ideologically—and while that may provide him with a clear advantage in the general election, it may serve to give some Republican voters pause in the Republican primary.

Unfortunately, the mirror on the wall doesn’t give us a clear answer as to who is the “purest” conservative of them all. Instead, it leaves us with a blurry image of the voting records and beliefs of all three candidates.  

Monday, April 14, 2014

2014 Montana House Race: How the Fundraising Numbers Stack up

*UPDATE: I originally posted this blog last week, but there were errors in the numbers that were my fault. I have corrected those mistakes and now actually have all the data from the candidates filing their April quarterly reports from the FEC. My apologies, but the good news is the substantive interpretation does not change.

Candidates for the open Montana House have released their first quarter fundraising numbers for 2014. It is important to put these numbers into some broader context instead of merely focusing on who is doing better relative to others in each quarter.

Fortunately, the 2012 House seat was an open seat race--as the seat is for this cycle. That helps us make apples to apples comparisons.

First, let's look at fundraising in the 2012 cycle compared to the 2014 cycle, quarter by quarter to date. Note that we still have yet to hear from one of the House candidates (Driscoll) as to what their April numbers look like:


Click the picture above to zoom in and check out the numbers.

Four points. Look at Zinke and Lewis' fundraising totals relative to the 2012 nominees. Zinke has already raised more money in the cycle YTD than Daines did in 2012. And, more impressively, he's done it much less time. John Lewis, in the three quarters he's been raising money, has more than doubled Kim Gillan's take and raised more than the top two vote getters in the 2012 Democratic Primary, Gillan and Wilmer. Finally, note the steep decline in Stapleton's numbers. It certainly suggests to me he's fading quickly. I felt that the emergence of Ryan Zinke really put a dent in his candidacy--at least, I'm not sure what Stapleton's niche among Republican primary voters is. Finally, Rosendale's campaign is largely self-funded. This presents a double-edged sword. On the one hand, they can concentrate on campaigning instead of raising money. On the other, it is important to connect with the activist-donor base of the party--and if you aren't, that's potentially an issue. Even Stapleton outraised Senator Rosendale in individual contributions for the quarter.

John Lewis, relative to the Democratic House candidates in 2012, is performing much stronger. He's raised almost as much as all 2012 Democratic candidates, and in less time. And, financially, Ryan Zinke is clearly the strongest candidate as evidenced by his performance relative to the Republican nominee in 2012, Steve Daines.

Let's look at the numbers in a slightly different way.




Here I totaled the amounts by party. Note that in 2012, Steve Daines raised more than all the Democratic House candidates combined through the first quarter of the election year--but the difference between the parties was a relatively modest $130,000 or so.

In this cycle while John Lewis has raised about $700,000 (note John Driscoll's numbers aren't available yet, but I'm not expecting much at all from him in terms of contributions), it is only 34 percent the combined total of the Republican candidates! That suggests to me, at least, an enthusiasm gap among Republican and Democratic donors--an enthusiasm gap that seems to favor the Republicans. Midterm elections tend to favor the party opposing the president--and perhaps the fundraising numbers are indicative of that fact. At least, donors tend to be politically sophisticated and certainly understand this.

Republicans will need to spend much of their money fighting for the primary nomination, but conducting an expensive primary election doesn't necessarily mean you will be ill-prepared for the general election. In fact, Marquette Political Science Professor Amber Wichowsky (a fellow graduate of the University of Wisconsin) finds little evidence that a tough primary hurts candidates in general elections. See the NPR story here or her research paper here.

I tell my students this repeatedly: Having the most money does not mean you win an election. You only need enough. But in an open seat contest, having the most helps considerable because it allows one to build much needed name recognition both in the primary and in the general election.


































































































Others 2012Zinke 2014Lewis 2014Rosendale 2014Stapleton 2014Arntzen 2014Driscoll 2014




140151

26435


91881

32692
18957128643767751

2634944761020489932089930796

61995460760300000???
14747190837050489932089930796??








































































Monday, April 7, 2014

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 14th 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.