What Does Evan Bayh's Retirement Say About Centrism and the Modern Congress??? Or, Why Are Progressives So Glad He's Leaving???

The recent announcement of the retirement of Indiana Senator Evan Bayh has set off more of a conversation among commentators than that of any of the other recent congressional retirees. The debate, to be sure, has largely been taking place on the left. Much of this discussion, from the standpoint of more progressive liberals, stems from a long running frustration, if not exasperation, with Bayh. To those who have favored a robust health care reform bill, a progressive approach to energy policy, a fiscal and tax regime that rolled back many of the Bush era policies, and a less belligerent foreign policy, Bayh was someone who could never be counted upon. Though a Democrat, Bayh either sided with Republicans (an original supporter of the Bush tax cuts and the Iraq War) or required numerous concessions from Democrats before delivering his vote. Even worse, many have argued that Bayh’s approach to policy was rooted not so much in a centrist set of positions, but rather his own political agenda. In the past two presidential cycles, Bayh was on the short list for Vice Presidential consideration. He was a moderate from a historically red state, someone who would be attractive to independents and more centrist Republicans—seemingly gold for a party trying to cobble together 270 electoral votes after two narrow losses. But for those on the left, he had no core set of beliefs, but rather positions that seemed to shift with the prevailing political winds. This frustration with Bayh came to a head in the 60 vote Democratic Senate of the past year. In this environment, with no Republican support for any of the President’s major initiatives, each Democratic Senator has come to exercise maximum leverage. Thus, we’ve gotten to know Bayh, Ben Nelson, Blanch Lincoln, Joe Lieberman, and Mary Landrieu quite well over the past year. For many liberals, this has not been time well spent.

Bayh’s announcement, and his critique of Congress on the way out the door, got me thinking some more about some of the topics I’ve blogged about in the past as it relates to Congress, legislating, and elections: partisanship; centrism; and the relationship between members and their constituencies. I’d like to hit on these topics while putting them within the context of some of the political science literature. As I noted above, one of the criticisms of Bayh was that he was a political centrist, not a policy centrist. Over at Ta-Nehisi Coates’ blog at the Atlantic Monthly (one of my all time favorite blogs), a commentator brought this distinction to my attention. Its worth hashing out a bit, I think. A policy centrist would seem to be someone who both holds a consistent set of policy beliefs and locates those beliefs somewhere near the middle of the ideological spectrum. What might this look like? On fiscal matters—something that Bayh prided himself on—this might include a recognition that balanced budgets and fiscal responsibility require both spending cuts and revenue increases. How these are arranged may vary depending upon the circumstance but the point would be that someone is willing to accept both of them and that they do so from a reasoned set of arguments and assumptions and, hopefully, over time. Another example could be abortion. A centrist position could entail accepting abortion under certain circumstances (rape, incest, etc.) or under certain time constraints (during first trimester). Policy centrism doesn’t need to be the result of indecisiveness or uncertainty. Rather, in the best sense, it is the result of someone recognizing that the more ideologically extreme positions aren’t likely to prove workable or successful. Political centrism is a slightly different animal. Here, a politician could hold positions that are actually to one end of the ideological spectrum but could, because of how they fit into a broader political community (say a legislature), position themselves as someone who could provide concessions and as a result, be a dealmaker (or deal-killer). Also, this person’s centrism could result from policy positions that aren’t tethered to a core set of beliefs. These stances could be unpredictable and in fact could also be contradictory (pro-tax cut and pro-spending increases while emphasizing fiscal responsibility for example). During the health care debate, Joe Lieberman embodied this for many liberals. They noted, for example, that during his own presidential run Lieberman had campaigned on the idea of a public option and health care insurance exchanges. When he later disavowed support for these, critics charged that he was less interested in policy than in his own ability to shape the debate given that his vote was essential to the Senate moving forward.

Bayh’s announcement got me thinking again about some of the themes I’ve blogged about in regard to Congress: polarization and partisanship; centrism; and the electoral connection between members and their districts. I’d like to hit on each of these and try to put some of the furor around Bayh into a broader—largely political science based—context. The criticism that Bayh was more interested in positioning himself as a “political centrist” and that he deep down had no true policy convictions brought me back to an influential piece of work from the early 1950’s that acted as the theoretical underpinning for what has come to be known as the “responsible party” model of government. “Toward a More Responsible Two-Party System” was generated by the American Political Science Association, through the collaboration of many scholars, and presented in 1950 as a way to create a more accountable governing system. At the time, many scholars (known as pluralists) had been describing and even advocating a more group or civil society based organization of politics. Parties, to them, were a remnant of a past, more corrupt, and less responsive mode of governance. To those responsible for the new report, though, group-ism was even less democratic in that it dispersed accountability too much, especially among those not electable. Only parties could provide true accountability and an organized set of choices for voters. To quote a bit of the report:

“An effective party system requires, first, that parties are able to bring forth programs to which they commit themselves and, second, that the parties possess sufficient internal cohesion to carry out these programs. Such a degree of unity with the parties cannot be brought about without party procedures that give a large body of people an opportunity to share in the development of the party program…The fundamental requirement of accountability is a two-party system in which the opposition party acts as the critic of the party in power, developing, defining and presenting the policy alternatives which are necessary for a true choice in reaching public decisions. The opposition most conducive to responsible government is an organized party opposition.”

In other words, parties produce policy proposals that give voters a clear choice of how the government will move forward. A winning party will enact that set of policies and do so as a united bloc. This movement will be opposed by a unified opposition—with their own set of policies—that frame the choice in the next election. Voters will evaluate the performance of the majority party either positively or negatively, setting the course for the next move forward.

Let’s bring this into the current environment. For those on either end of the ideological spectrum “responsible party government” is extremely attractive and desirable. Their party is unified around a common set of objectives and pledges itself to work as a bloc. Any internal dissension works against this. This is why people like Bayh, Landrieu, Lincoln, Nelson, and Lieberman are so maddening to liberals. They refuse to be part of a unified Democratic Senate caucus. Rather than being able to enact a health care, cap and trade, or jobs bill—which can then be put to the voters for their consideration—these members produce (again in the eyes of their liberal critics) gridlock. While Senate Republicans are acting “responsibly” by presenting a unified opposition (although some would argue without their own set of policy prescriptions) the result works against the Democrats in two ways. Not only do they not have a set of results to point to, they also can be credibly blamed for not producing them. If only they acted like the Republicans and stuck together, they could get things done. Notice how few elegies to Bayh have been given by Senate liberals and other progressives over the past week.

A critique of the Responsible Party model that one could give is that it’s impossible for parties to 1) elect members from a diverse electorate and 2) stay unified afterwards. Or, in the case of Bayh, one can argue that he had no choice but to act the way that he did given where he came from. Indiana is not California, which is not New York, which is not Pennsylvania, etc. Constituencies and districts matter. They have different electorates with different interests and thus elect different types of politicians. This is the stuff on which we’ve dwelled so much on this site.


What I decided to do was come up with a way of seeing how close Senators are to their states based on the ideologies and partisanship of the two. The table above is the result of this attempt. To be clear up front, this is going to be very, very inexact. Methodology is crucial, yet tricky. What I decided to do was come up with a way to rank states based on their level of liberalism. The only way—or perhaps the only way that matters in the end—to do this is to look at voting. Because states change and because support for one party can vary based on who the candidates are and other variables, I decided to look at the past three elections. I took the Democratic percentage of the vote for each election and then added the three to give an overall score. Thus, a state that gave the Democrats 50%, 52%, and 55% would get 157. The result is a continuum of states running from the least liberal (Utah) to the most (Massachusetts). Running through the list I don’t see anything that’s too out of whack with what we would predict coming in.

From this, we want to try and predict how that state’s Senator should behave. If Senators work hard to represent their constituencies, there should be some correlation between the Senator’s behavior/ideology and that of their state. We should, in theory, see a continuum of Senator ideology that conforms with the state voting continuum. So what I did was create a column that suggests an “expected” order of Senators based on their ideology. Because each state has two Senators, each state is listed twice as you go down the column. Thus, Utah should have the country’s least liberal Senators, Massachusetts the most. While this is what we would expect, what we’re ultimately interested in is what we’ve actually seen—and whether it confirms our expectations.

To measure the ideology of individual Senators, I’ve relied on the most recently released set of Ideology Ratings by National Journal. Each year, the magazine selects a representative sample of Economic, Social, and Foreign Policy votes. Each member’s vote is recorded across all of these votes in each category. Members are then compared to each other. The result is a percentile score that places each member in relation to their colleagues in each of these policy categories. The higher the score, the higher the level of liberalism. National Journal then combines the Economic, Social, and Foreign Policy dimension to give each member a single score. Columns 7, 8, and 9 of the table list the actual order of the Senators along with their respective state and liberalism scores. Each Senator is color coded based on party as well.

I should note a couple of things right here. First, the most recent data available is for the year 2008 so we aren’t looking at the current Congress. We might expect behavior of Senators to change a bit given a new President, somewhat different agenda, and other factors. Overall, though, I doubt that individual behavior would change dramatically although I could be wrong in certain cases. Based on when this data came out last year, I would expect the 2009 numbers to be coming out shortly so I’ll try and update this ranking to include the cast of Senators elected in 2008 (and exclude those no longer in office). A second quick note relates to three Senators whose data was not available for 2008. First, Senator Ted Kennedy, due to his illness, did not cast enough votes to produce a meaningful score. Thus, as a proxy I used his 2007 score. Given that National Journal compiles a broadly representative sample of votes from year to year, I didn’t expect there to be much difference between his behavior in 2007 and what it would have been the following year. Similarly, Barack Obama did not cast enough votes in 2008 due to his presidential campaign. I also used his 2007 score. Like Obama, John McCain didn’t have a 2008 score. He also didn’t have enough recorded votes in 2007 (???!!!) so I had to use his 2006 score as a proxy.

Anyhow, you can see the continuum of Senators that we have as a result. The least liberal Senator is John Barrasso of Wyoming while the most liberal was Obama. One point we need to make clear is that although I have produced an ideological continuum of Senators, the ideological distance between each is not the same. Rather, as we have talked about a lot here, there is considerable polarization of the modern Congress. Thus, most Republican members are clustered on the right of the ideological spectrum whereas most Democrats are clustered together on the left. A better way to represent and visualize this would be to produce a bell-curve of this distribution of scores. Here is one that was generated based on the 2007 rankings. Going back to an earlier point, this clustering and polarization is what we might expect in a “Responsible Two Party System.” Also worth noting is that Senators from the same state are not necessarily located close to each other on the continuum but in many circumstances are quite distant.

What we’re interested in with people like Bayh is what’s going on in the middle of the continuum. How liberal/conservative are these Senators, where are they from, and how does this conform with their states? With National Journal’s rankings, we actually see a bit of overlap between Democratic and Republican moderates. Senator Bayh is actually more conservative than Maine Republicans Olympia Snowe and Susan Collins while Democrats Ben Nelson and Mary Landrieu score more conservative than Snowe. Its no surprise, then, that these members have been the ones getting the most attention over the last year—and driving their respective party ideologues crazy in the process.

How do they compare with their states??? What I did next (Column 10) is calculate the distance between where they are based on their Senate behavior and where they should be based on their state’s behavior (Column 6). Thus, is a Senator more liberal or conservative than their state and by how much? What we see is a range of distances as well as a small handful of Senators who are exactly where we would expect them to be (Enzi, Feinstein, Schumer, and Reed). The Senators who are far away are interesting to speculate about and also show us how elections and survival in Congress are oftentimes based on more than just a member’s voting. We should note, though, that because of the clustering around the ideological poles (that bell curve above) the distances I'm calculating here might look much further than they really are--i.e. moving 20 spots might mean traversing over 20 Senators who are only marginally different than you. Because we’ve had an election since these numbers were produced, I highlighted the Senators who lost re-election (Yellow) and the seats that changed party after a retirement (Green). While most of the flips took place in states where the incumbents had a bit of distance from their expected slot, two switches (Dole and Warner) did not. Overall, though, there is a tremendous amount of variance among the members. To further demonstrate how states and members can differ in their behavior, the final column is color coded to show the degree to which each state splits its vote among its Senate delegation. Red coincides with both Senators being Republican; Blue for both being Democratic; and Purple connotes a split delegation. Here, too, one sees some pretty good variation. As I noted above, we often see Senators from the same state behaving quite differently. National Journal has an interesting discussion of this phenomenon here.

To come back to those “centrist” members including Bayh, we see that there is a range of variance between their actual and expected behavior. Bayh, it turns out, is actually pretty closely aligned with Indiana (12 spots more liberal than expected) whereas Republicans Snowe and Collins are 25 and 28 spots respectively more conservative than we would expect for a Maine Senator. With Ben Nelson, his reluctance to sign on to the Democratic agenda seems understandable given his distance from Nebraska’s expected ideological location (he’s 38 spots more liberal).

So where does this leave us??? What this discussion, I think, shows is that (no surprise here) legislating is difficult, especially in the modern Congress. “Responsible Party Government” is attractive in that it is clean and efficient. For those who find themselves in wide agreement with a party’s ideological positions, there will be a demand that politicians act accordingly. Unfortunately for these people, we don’t seem to have a country—geographically, demographically, etc.—that is conducive to producing this type of politics. This wasn’t the case in 1950 and it is even less the case today. What you get instead is a politics that is much more messy and slow. When you have polarization coupled with this demographically and geographically produced diversity, things slow down even more and frustration mounts. In the past week, the frustration felt by those on the left has been aimed at Evan Bayh. Similarly, in past months Ben Nelson and Joe Lieberman have occupied the position of “Democrats’ Most Hated Democrat.” I’d expect this type of frustration to continue.