Wednesday, February 29, 2012

The Montana Senate Race: Who are these guys and what are their records?

Watching this Montana Senate race closely has proved to be immensely interesting, enjoyable, and fascinating. Readers of this blog will note that I have repeatedly stressed the notion of “one of us”—campaigns work hard to show how their candidate is connected to voters and how their opponent is not. This really gets at the heart of representative democracy: voters are more likely to trust decisions made by representatives who look, act, and think like them. To use the language of postmodernism, what voters do not want and fear is the “other”: the representative who is beholden to someone else and who will make decisions not in their best interest.

Part of proving or demonstrating “one of us” is to discuss legislative records: What have you done for me lately and how is it evidence that you are serving the best interests of the state? Senator Tester casts himself as an independent voice for an independent-minded state, while calling Congressman Rehberg a yes man for Republican Party bosses. Congressman Rehberg, alternatively, says he’s the true voice of the Montana voter and Senator Tester is merely another vote for President Obama and his agenda. Each side will use a vote or a series of votes to support their position: Senator Tester voted with President Obama X number of times, and Congressman Rehberg is too conservative for Montana because he voted for House Republican proposal X, Y, or Z.



How can the average voter make any sense of these claims? As with all things, it depends on how you look at the records of both candidates. I am going to present three ways of thinking about Senator Tester’s and Congressman Rehberg’s records. I am going to present some additional data that will help you contextualize those records. Using this information, I provide some brief commentary but will largely leave it to Montanans to draw their own conclusions about the claims of both campaigns about their own records and the records of their opponents.



In general, the claims made by the campaigns about themselves and their opponent revolve around three distinct concepts:



Party Unity: How often does the member vote with their party in Congress?



Ideology: How liberal or conservative is the member?



Presidential Support: How often does the member support the president’s legislative agenda?



There are a number of ways these concepts can be empirically measured. Here are some standard measures utilized by political scientists that we’ll examine:



Party Unity: There are many ways to measure party unity, but the most common measure is produced annually by Congressional Quarterly since the late 1950s. CQ defines a party unity vote as a majority of one party opposing the majority of the other party on a given vote. To be specific, if 51% of Republicans vote yea on a motion and 51% of Democrats vote Nay, that vote meets the minimum standard to be included as a party unity vote. A party unity score is the percentage of times a member votes with their party on all party unity votes cast in a given year. The percentage of party unity votes as a percentage of all votes cast in Congress has ebbed and flowed throughout the years—the mid-1970s were a low point, while the last twenty years has seen a rebound in the number of votes clearly dividing the parties. Party unity scores have also increased over the late 20th and early 21st centuries, meaning members of the House and the Senate have become more loyal to their parties and less likely to break partisan ranks. Averages today are typically in the high 80s to low 90s in both chambers, with the House usually exhibiting more party loyalty than the Senate.



Ideology: One way to measure how liberal or conservative a member is to use DW-NOMINATE scores. DW-NOMINATE is a measure of ideology created by Poole and Rosenthal in their seminal work, Congress: A Political-Economic History of Roll Call Voting. Essentially, DW-NOMINATE takes the all the votes cast in Congress over time and develops factors which explain the variation found in those votes. The first dimension score Poole and Rosenthal call ideology, and this measure places members of Congress on a left-right dimension. Positive scores approaching 1 are more conservative, while negative scores approaching -1 are more liberal. It is important to note that both parties have become increasingly polarized after 1975, which means the average Democrat has become more liberal and the average Republican more conservative in both the Senate and the House. The other important thing to note is that both parties have become more ideologically homogenous—meaning moderates have become increasingly rare. More information on NOMINATE and NOMINATE scores can be found at Keith Poole’s website here.



Presidential Support: Again, Congressional Quarterly has set the standard for developing a measure of Presidential Support. CQ determines whether the presidential administration has taken a clear and public position on legislation. If the administration has, the vote on taken on that legislation is included in the universe of votes utilized to calculate an individual legislator’s presidential support. Simply put, a presidential support score is the percentage of times a member of the House or the Senate has voted to support legislation on which the administration has taken a public position. Higher scores indicate more support for the administration, lower scores less support. Traditionally, administrations receive the bulk of their support from their parties in Congress—with support of the administration’s positions from the opposition party declining in recent decades. Congressional Democrats often support Democratic administrations, and Congressional Republicans support Republican administrations.



Below, I have created a series of graphs depicting the voting records of the Montana congressional delegation between 2001 and 2011. In each instance, I plot the party unity, ideology, and presidential support scores for Senators Tester, Baucus, Burns, and Congressman Rehberg. I also provide contextual information, namely the average scores for Democrats and Republicans in both chambers. Finally, I wrap up the discussion with some concluding thoughts and by examining key congressional votes to see if that information tells a story that is any different from these more general measures of party unity, ideology, and presidential support.



Party Unity



The first chart plots the party unity scores of Montana’s Senate delegation and compares it to the party means in the chamber. Again, a party unity score is the percentage of time the member votes with their party when the majority of each party is in opposition. The second chart plots Congressman Rehberg’s party unity score relative to his party’s average scores.



Figure 1: Party Unity in Montana's Senate Delegation, 107th-111th Congresses





Figure 2: Congressman Rehberg's Party Unity, 107th-111th Congresses





NOMINATE



How liberal or conservative is Montana’s Congressional Delegation? NOMINATE scores help sort this out. The first chart plots the first dimension DW-NOMINATE values of Montana’s Senate delegation, the second plots Congressman Rehberg relative to his peers in the House.

Figure 3: Montana's Senate Delegation and Ideology, 107th-111th Congresses



Figure 4: Congressman Rehberg's Ideology, 107th-111th Congresses




Presidential Support



How often did Presidents Bush and Obama receive support from Montana’s Congressional delegation? The final two charts show us. Again, the first chart is the CQ Presidential Support score plotted for Senators Burns, Baucus, and Tester and the second chart is Congressman Rehberg relative to his Republican peers.



Figure 5: Montana's Senate Delegation and Presidential Support, 107th-111th Congresses





Figure 6: Congressman Rehberg's Presidential Support, 107th-111th Congresses



Summary

In general, we learn the following from these measures of party unity, ideology, and presidential support:

1. Members of Congress generally support their parties on party unity votes.

2. Members of Congress generally support the administration when the administration shares their party affiliation.

3. Ideologically, Montana’s Democratic Senate delegation votes to the right of the average Senate Democrat. It is difficult to call either Senator Baucus or Tester a liberal—they are moderate or—perhaps more generously—conservative Democrats. Congressman Rehberg faithfully exists at about his party’s mean in the House—and quite possibly a tad to the left of the mean GOP member in the last Congress. Congressman Rehberg is conservative, but his conservatism is not extreme relative to his peers. He’s no Ron Paul or Randy Neugebauer, both representing Texas and two of the most conservative members of the Republican delegation in the House.

Three different ways to consider the legislative records of Montana’s congressional delegation. Overall, the picture is far more complex and nuanced than we might expect given the campaign rhetoric we hear on Montana’s airwaves, read in our inboxes, and a view on Facebook.



Thursday, February 23, 2012

New poll--but same story in the Montana Senate Race

Montana Republicans and the Rehberg campaign are touting a poll showing their man leading in the Montana Senate race. The poll was done by Rasmussen and shows Congressman Rehberg with 47% of the vote and Senator Tester with 44%. The poll was of 500 likely Montana voters conducted on February 22, 2012. The margin of error is 4.5 percent, so really, the race is too close to call. I generally do not like to say a candidate is in the lead unless the candidate has a lead outside the margin of error. You can read about the particulars here.

I should note that in 2010, Rasmussen's polling consistently showed a considerable Republican bias as noted by Nate Silver who blogs about survey methodology and statistics over at The New York Times. Read his analysis here. That said, I have less of a reason to believe that any bias--should it remain in Rasmussen's methodology during this election cycle--is reflective in THESE results given that they show us pretty much what we've seen since March of 2011: the race is...too close to call.

Monday, February 20, 2012

Does House Service Make it Harder to Move to the Senate? Remember Fenno's Paradox!

The New York Times recently published a piece on the difficulties House members face when running for the Senate. Read it here. The article discusses Congressman Rehberg, and the chief point it makes is given that Americans hate Congress, so why would they select a House member to move to the Senate? The author notes that the dismal rankings Congress receives overall will make it much harder for individual House members to make the move up to the Senate.

As a political scientist, I had three problems with the article.


First, the serving in the House is the single best path to the United States Senate. Looking at the current membership 112th Congress, 46 senators served as a House member as their last job prior to moving to the Senate. No other career path comes close. 13 served in some statewide elected office other than Governor, and another 10 served as their state’s Governor. 10 had no elected office experience at all, 9 served in the state legislature, and 12 had some elected office at the local level. The New York Times underestimates the skills House members obtain and hone during their service which make them strong Senate candidates.


Second, the article presumes that because Americans hate Congress, they hate their own member of Congress. Anyone who has taken my introduction to American politics class should be familiar with the concept of Fenno’s Paradox. Fenno notes that many Americans hate the institution of Congress, but like their member of Congress. The following graph—which I used in a lecture just last week—illustrates the point clearly. Here in Montana, both Senator Tester and Congressman Rehberg have job approval ratings well to the north of the Congress.


Fenno's Paradox: Ratings of Congress and Incumbents in the American National Election Studies, 1990-2008.



Finally, serving in a state as the lone House member is very different from serving in the House from a state with multiple House members. Generally speaking, the key difficulty facing a House member moving from a congressional district to a statewide office is moving from a more politically homogenous political entity to one that is more heterogenous, representing a more diverse constituency. The same rules do not apply, obviously, to House members representing an entire state. Sure, the House is a more partisan political institution and the name of the representative game is slightly different in how Washington Work is approached. But the task of representing a state—in a single-member House delegation—is not entirely dissimilar from serving in the U.S. Senate (but see Richard Fenno's book, When Incumbency Fails, on North Dakota Senator Mark Andrews who, after having served many years in the House, was elected to the Senate in 1980 and lost reelection in 1986).



This is not to say that Congressman Rehberg or Congressman Berg in North Dakota don't have a tough path ahead of them. But that path is not necessarily more difficult because of their service in the House. The conventional political science wisdom would suggest otherwise.


Friday, February 17, 2012

Two articles of note have appeared in the media over the last week. In my opinion, both should be read by each and every student of politics.

The first is James Fallow’s analysis of the Obama administration – Obama, Explained. It is the cover story of this month’s edition of The Atlantic. He has three themes in the piece. First, the behavioral and psychological Obama. Second, how do we view the administrative successes and failures of the last 3+ years. Third, a temporal look. For the last, he references a memo written by James H. Rowe Jr., at the time a young official at the Bureau of the Budget (what is now the Office of Management and Budget) to President Harry Truman. Rowe’s memo could be written today and still be mostly accurate about the political setting in which Obama finds himself today. The lesson is twofold – the political strategy of opposition (by Republicans in both cases) is totally predictable and, given the political strategy, political stalemate is inevitable. If Obama is going to enjoy a second term, we can pretty much predict how the campaign will unfold no matter who the Republican nominee is.

The second piece appeared in the New York Times last week. The Geography of Entitlements interactive map is illustrative of the hypocrisy of much of the political debate in the current primary season. Those areas that most depend on federal funding for social programs, retirement benefits, Medicare and Medicaid and other federal benefits are, in large part, the same parts of the country that support candidate that would reject government in general. The map does not show military spending but trends would be the same.

In his NYT column today (3.17.12), Paul Krugman cites work by Aaron Carroll of Indiana University. Carroll points out that “in 2010, residents of the 10 states Gallup ranks as ‘most conservative’ received 21.2 percent of their income in government transfers, while the number for the 10 most liberal states was only 17.1 percent.” For those economists that assume rational economic man, they need to explain why large parts of country are willing to vote against their economic self-interest.Others have written about this seemingly illogical political stance - notably Ben Barber, Walt Whitman Professor of Political Science Emeritus, Rutgers University.

Psychologists tell us that when cognitive dissonance gets bad enough, we typically seek to resolve it. I’m waiting for the collective dope slap from supporters of Rick Santorum, Gingrich, and Romney but I’m not holding my breath.