Tag Archives: polarization

A Transparent Failure

In the week since it officially folded, there has been a lot of discussion of why the super committee (hereafter SC) failed, whether it was ever intended to “succeed”, and what the future ramifications are. So in the spirit of better-late-than-never, I will share a few of my thoughts on these topics.

Let me begin first as to whether the SC process was designed to succeed. From the perspective of a student of legislative bargaining, I can think of at least three mechanisms that the SC could exploit to succeed under circumstances where the normal legislative process would fail.

  1. The Agenda Control Mechanism: The SC would have the privilege of bringing a package of spending reductions and tax increases to the floor of each chamber that would be subject to up-or-down votes. Consequently, any agreed upon bargains could not be undone by floor amendments.
  2. The Composition Mechanism: The SC might have a composition that is more conducive to reaching an agreement than Congress as a whole.
  3. The Procedural Mechanism: The SC might adopt internal procedures that were more conducive to reaching an agreement than the procedures used in standing committees.

Continue reading

CNN on Gerrymandering

I did not get a chance to see the CNN Special Report on gerrymandering over the wekend. Maybe it is all for the best as Jamelle Bouie at American Prospect suggests that it was entirely typical of journalistic treatments on the topic. When it comes to the effects of gerrymandering on polarization, with few exceptions, journalists have focused on the talking points of reform activists without really engaging the now substantial social science literature on the topic.

Turnout and Polarization

The Monkey Cage has an interesting discussion about whether boosting turnout would reduce levels of polarization and partisanship among our elected officials. While none of my work directly addresses the question of the effect of general election turnout on polarization, several of its strands would place me clearly in John Side’s camp of skeptics.

  • Chapter 4 of Polarized America analyzes trends in turnout among different income groups during the period of expanding polarization (1972-2004). There were several important take aways. The first is that our evidence is consistent with earlier findings of McDonald and Popkin that there is no trend in turnout among the voting eligible population (VEP) (roughly citizens over 18 who are not disqualified due to felony convictions). The downward trend in participation by the voting age population (VAP) (all residents over 18) is caused by increased numbers of non-citizen residents and (to a lesser extent) disenfranchised felons. Moreover we find that the relative incomes of voters and non-voters are roughly constant over the time period. So it would seem that trends in turnout could hardly explain the increase in polarization. Indeed, reversing the trends in VEP would mean reversing disenfranchisement laws and allowing non-citizens to vote. It would be reasonable to expect that those reforms would shift the spectrum to the left but not de-polarize it.
  • My other relevant work concerns the effects of participation in primary elections. In this rejoinder to Richard Pildes, I report some estimates of the effect of primary election turnout on the extremity of a House member’s NOMINATE score. A key problem for such analysis is that turnout may itself be influenced by a candidate’s position. So I use the coincidence of competitive presidential, gubernatorial, and Senate primaries as exogenous variation in turnout. The resulting IV estimates show no relationship between turnout and the winner’s NOMINATE score. I concede, however, that the effect of the instruments (the competitive primaries for higher office) is small. The second piece of evidence is reported in my working paper with Eric McGhee, Seth Masket, Boris Shor, and Steve Rogers. Here we find little evidence that the institution form of primary election effects polarization of state legislators. This undercuts the common argument that open primaries reduce polarization by boosting participation.

But perhaps the most relevant piece of information about Bill Galston’s original contention that mandatory voting would reduce polarization is a new paper by Anthony Fowler at Harvard. In a careful study, he looks at the effects of the adoption of compulsory voting in Australia. He estimates that the causal effect of compulsory voting was a significant boost the left-wing Labour Party. So the effect of compulsory voting appears to have shifted Australian politics to the left rather than shore up the center.

A Sign of the Times

It is a sign of the polarized times in which we live that partisans of different stripes can look at the same objective indicators – persistent and increasing unemployment—and come to radically different conclusions about what is to be done. The Republicans see proof in the unemployment numbers that the original stimulus package was ineffective and misguided. The Democrats see it as a sign that the original package was not enough and that we need another one.

Because I like to cast myself as a moderate, let me split the difference. I think the original stimulus package probably was beneficial. It was always unrealistic to think that it would prevent unemployment from going above 8%, but it is equally unrealistic to believe things would be better now without it.

Yet I do not believe another package is what we need for two reasons. The first is that the problem with the original stimulus plan is that the money has not been spent fast enough (mainly because too much of it went for pent up Democratic spending priorities rather than “shovel-ready” projects). So a much better plan would be to reprogram some of the original package toward projects that will spend money more quickly. Helping state governments avoid layoffs might be a good place to start. The second reason that I’d oppose another stimulus bill is that I think unemployment numbers may not be a good indicator of where the economy is going or whether the first package worked. As we observed in last recession (for reasons we do not quite understand), employment and wages only started improving after the economy had recovered. So throwing another batch of stimulus at an economy that may be already recovering would not be prudent.

Senate Retirements

These are indeed gloomy days for the Republican Party. Just two months after a disastrous election, ten percent of their Senate contingency is calling it quits for 2010 (twenty percent of those up for reelection). Those retiring are Kit Bond (MO), Sam Brownback (KS), Mel Martinez (FL), and George Voinovich (OH). So far the only Democratic “retirement” is that of seat warmer Ted Kauffman who plans to make way for Biden fils.

What is perhaps more troubling for the party is the difficulty the party will have in retaining all but the Kansas seat. Obama won Florida and Ohio and barely lost in Missouri. But the longer term damage may be even greater. All of the retirements are members from the most moderate half of the Republican party (Brownback is indeed socially conservative, but significantly more moderate on other issues). Thus, these retirements will further erode moderate wing of the Republican party and further polarize the Senate.

Update: Chris Bowers makes the point that the four retiring senators may be more likely to support President Obama’s economic iniatives than the non-retiring senators. He bases this prediction in part on behavior of retirees on the TARP and Auto Bailout votes. I also found evidence for such an effect here and here.

Catching Up

My earlier post on the continued high levels of polarization in Congress generated a few comments to which I thought I might respond.

First, both the left and right took some umbrage at how I presented the data. R Weeks protests that

Using the link to read the entire article, I find the information presented fails to support the headline used. It is apparent increased polarization is due to the continued drift right by Republican members of congress…Perhaps the headline should read “Republican Polarization?”

while Liberty responds that

I always have to smile when I hear that Republicans must moderate their positions and show “restraint” in order for everyone to get along. And if one is to believe this article, then Democrats are considered to moderates and reasonable and exhibit a non -polarizing effect in Congress. Ya, right. You only have to look at the number of Republicans representing the Northeast in the U.S. House of Representatives, ZERO, to observe an example of polarization. Good Grief!

It is generally a good sign when you offend both sides, because it means that you’ve also treated both fairly. But let me make a few additional points.

  • R Weeks is correct in that the preponderance of the evidence indicates that contemporary polarization is a reflection of the increasing conservatism of the Republican party. The position of the Democratic party has been relatively stable, pulled to the left only because its southern wing is now heavily African-American rather than white conservative. This is consistent with almost all of the political science research on the topic. In assessing blame for polarization, however, what we would really like to know is how the parties shifted relative to voters rather than to each other. If, for example, the electorate has moved sharply to the right over the past 30 years, the Democrats might be blamed for not going along. But comparing the conservatism of Congress with the electorate is a very methodologically-tricky thing to do, and there is much less consensus on this question. My personal view is that the electorate moved modestly to the right over the last generation (at least until the last 3-4 years) but not nearly as much as the Republican party did. So I do assign the Republicans a very large share, but not all, of the blame.
  • Conservatives and libertarians should not be too upset when they are blamed for polarization. After all, many on the right take great pride in transforming the Republican party into a conservative party. Of course that transformation led to polarization, and many conservatives think that is a good thing.

Second, King Politics adds the following to my analysis:

the breathless reporting by the MSM usually neglects to include how redistricting (generically) makes for less competitive general elections where the primary winner has to run to the wing of his/her party and not to the middle.

I’m going to dissent here. The MSM actually greatly exagerates the effects of redistricting. Keith, Howard, and I have the following paper under review:

Does Gerrymandering Cause Polarization?


Both pundits and scholars have blamed increasing levels of partisan conflict and polarization in Congress on the effects of partisan gerrymandering. We assess whether there is a strong causal relationship between congressional districting and polarization. We find very little evidence for such a link. First, we show that congressional polarization is primarily a function of the differences in how Democrats and Republicans represent the same districts rather than a function of which districts each party represents or the distribution of constituency preferences. Second, we conduct simulations to gauge the level of polarization under various “neutral” districting procedures. We find that the actual levels of polarization are not much higher than those produced by the simulations. We do find that gerrymandering has increased the Republican seat share in the House; however, this increase is not an important source of polarization.

Here’s a copy gerrymander38.pdf.


In the aftermath of the 2006 election, the era of polarization was declared over in such astute analyses as this one:


The impetus behind such conclusions was the extraordinary success of “Red State” Democrats such as Jon Tester and Heath Shuler. But few pundits took note of the fact the these Red Democrats were only moderate or conservative on a few social issues, but quite populist on economics and trade. Even fewer considered the consequences of the extinction of “Blue State” Republicans for polarization in Congress.

But now that Congress has adjourned sine die, Keith Poole has fired up the NOMINATE machine, and we can look at what impact, in any, the 2006 elections had on the level of party polarization in the House and Senate.

This first figure is an update of the data presented in our book with Howard Rosenthal showing the average difference between Democrats and Republicans on the DW-NOMINATE conservatism scale.


Do you see the dramatic collapse of polarization in congressional term beginning in 2007? Me neither. In fact, polarization rose in the 110th Congress just as it has almost every term since 1975. The House had set a record for polarization in the 109th, but the 110th broke it. The Senate broke its own record set in 1867.

So what might the future bring? To get a prediction that is hopefully at least as accurate as Joe Klein’s, I have forecast the average conservatism of Democrats and Republicans for the next congressional term in the following way:

  1. Assigned all returning members their DW-NOMINATE score from the preceding term.
  2. Assigned all new members the average DW-NOMINATE score for their party and region. In other words, a new Democrat from the Midwest gets the average of all midwestern Democrats and a new Republican from the South is assigned the average of all southern Republicans.

Essentially, this procedure captures the effects of the regional distribution of partisan seat shifts. A seat shifted from Republicans to Democrats in the Northeast increases polarization whereas a Democratic pickup in the South decreases it. So here is what the House and Senate may look like next term.


senate_fc.jpgIn each of the figures, the red line is the average conservatism of Republicans, and the blue line is the average conservatism of the Democrats for each term since 1969. The triangles are my prediction for the next term. The Democratic average is expected to change very little, but the Republicans will be considerably more conservative. This, of course, is due to their continued hemorrhaging of seats outside the South. The net effect is again an increase in polarization.

All of this is predicated on the assumption that there will not be any major deviations from recent historical patterns. Of course, things could change. In the conclusion of our book (written in January 2005), Keith, Howard, and I speculate that a financial crisis triggered by a housing bubble might lead to a swing in the public’s partisanship and ideology that might cause the Republicans to moderate. So we have the crisis, a modest swing in public attitudes, but if the congressional votes on the bailouts are any indication, the Republicans haven’t take that last step.

Polarized America

In 2006, I published Polarized America: The Dance of Ideology and Unequal Riches with my colleagues Keith Poole of UC San Diego and Howard Rosenthal of New York University.

Here’s the cover blurb:

The idea of America as politically polarized–that there is an unbridgeable divide between right and left, red and blue states–has become a cliché. What commentators miss, however, is that increasing polarization in recent decades has been closely accompanied by fundamental social and economic changes–most notably, a parallel rise in income inequality. In Polarized America, Nolan McCarty, Keith Poole, and Howard Rosenthal examine the relationships of polarization, wealth disparity, immigration, and other forces, characterizing it as a dance of give and take and back and forth causality.

The authors find that polarization and income inequality fell in tandem from 1913 to 1957 and rose together dramatically from 1977 on; they trace a parallel rise in immigration beginning in the 1970s. They show that Republicans have moved right, away from redistributive policies that would reduce income inequality. Immigration, meanwhile, has facilitated the move to the right: non-citizens, a larger share of the population and disproportionately poor, cannot vote; thus there is less political pressure from the bottom for redistribution than there is from the top against it. In “the choreography of American politics” inequality feeds directly into political polarization, and polarization in turn creates policies that further increase inequality.

Links to several reviews and discussions of the book follow the jump.

Continue reading