So Just Who Did Vote For The Bailout?

I've been engaged in a little argument with the guys over at Open Left over the nature of the various bailout bills that have come before Congress over the past several months.

My position is essentially as follows: Legislators whom we usually think of as liberal or progressive have been more, rather than less, likely to support bailout legislation. It is therefore difficult to argue, as David Sirota does, that a vote against the bailout represents "a major victory for the progressive movement [...] against kleptocracy".

Chris Bowers has responded by suggesting that, while progressive congressmen may have been more likely to support the most recent bailout held over the past couple of weeks, this was not true of the original bailout bills last autumn, before Barack Obama took over the presidency, and was particularly not true of the "original" bailout bill last September 29th, which failed 205-228 in the House and infamously made the stock market very unhappy.

I find Chris's comments thoughtful, as I usually do. Undoubtedly, the bailout involved some incredibly complex political considerations. Imagine some of the issues that a Congressman must have been wrestling with when deciding how to vote on the bailout last September:

- The impact of the passage or failure of the bailout bill on the presidential campaign;
- The impact of the passage or failure of the bill on his re-election campaign;
- Pressure from leadership to vote for the bailout;
- Pressure from industry to vote for the bailout;
- Pressure from the blogs and talk radio to vote against the bailout;
- The long-run impact of creating hundreds of billions of dollars in additional debt;
- The long-run impact of creating (arguably/probably) a moral hazard for banks;
- The near-term impact of a credit market crash caused by the failure of the bailout;
- The near-term impact of a stock market crash caused by the failure of the bailout;
- The possibility that by voting against the bailout, a superior bailout bill could be facilitated;
- The particular impact of the bailout on the economy of his Congressional District;
- The Bush Administration's support for the bailout;
- The (then-tepid) support of Obama and McCain for the bailout.

That's a lot to chew on, and it's not surprising that whole sets of Congressmen who usually vote alike wound up splitting their votes on this issue.

Nevertheless, perhaps we can find some useful evidence in the fossil record. I went back to the bailout vote, compiled a whole bunch of relevant variables, and tried to find which of these variables best predicted voting behavior. These might be characteristics having to do with the Congressman himself (such as his race or his party), or characteristics having to do with his congressional district. After many permutations, here is the most coherent model that I came up with:



All right -- so, running raw regression output probably isn't going to endear me to very many of you. But basically, what we have is the following set of variables:

- 'voteshare06' represents the share of the two-party vote that the Congressman received in 2006; it is a measure of how electorally vulnerable the Congressman was. The higher the Congressman's vote share had been in 2006 -- meaning, the less vulnerable he probably was in 2008 -- the more likely he was to vote for the bailout.

- 'retire' is a flag indicating whether the Congressman had announced his intention to retire at the end of his term (but was not, at that time, still engaged in a campaign for higher office). Retiring congressmen were considerably more likely to vote for the bailout.

- 'banklobby' is the amount of political contributions that Congressman received from the banking industry. Congressmen receiving more of these contributions were more likely to vote for the bailout.

- 'demleader' and 'gopleader' are indicators of whether the Congressmen held a leadership position in the Democratic and Republican parties, respectively, either as a member of his party's floor leadership or as a chairman or ranking member of a permanent House committee. Leadership members of both parties were more likely to vote for the bailout.

- 'is_hisp06' and 'is_black06' indicate whether the Congressman is Hispanic or black, respectively. Hispanic and black Congressmen were less likely to vote for the bailout, all else being equal.

- 'libcons', finally, represents the Congressman's liberal-conservative score as provided by National Journal. The more liberal a congressman's voting record, the higher his score. More liberal members of congress were more likely to vote for the bailout, holding other variables constant. Interestingly, party affiliation did not appear to be a determinant of voting on this particular bill, except insofar as it was reflected in ideology.

To get some idea of how all of this works, here are the Congressmen predicted by the model to be the most likely to vote for the bailout along with their actual votes:

98.9%  Frank, Barney       D    MA-4    Aye
98.1%  Bachus, Spencer     R    AL-6    Aye
98.0%  Emanuel, Rahm       D    IL-5    Aye
97.5%  Rangel, Charlie     D    NY-15   Aye
97.5%  Hoyer, Steny        D    MD-5    Aye
97.3%  McNulty, Michael    D    NY-21   Aye
96.8%  Crowley, Joseph     D    NY-7    Aye
96.3%  Davis, Tom          R    VA-11   Aye
95.9%  Hooley, Darlene     D    OR-5    Aye
95.9%  Pelosi, Nancy       D    CA-8    Aye

Basically, this is the Democratic leadership in the House, plus a handful of Congressmen from both parties who received especially large contributions from the banking industry. And here, conversely, are the Congressmen deemed least likely to vote for the bailout:

6.1%   Diaz-Balart, Mario  R    FL-25   Nay
6.5%   Diaz-Balart, L.     R    FL-21   Nay
8.0%   Lamborn, Doug       R    CO-5    Nay
8.5%   Ortiz, Solomon      D    TX-27   Nay
9.5%   Franks, Trent       R    AZ-2    Nay
9.7%   Jordan, Jim         R    OH-4    Nay
10.3%  Westmoreland, Lynn  R    GA-3    Nay
10.7%  Pitts, Joseph       R    PA-16   Nay
11.0%  Souder, Mark        R    IN-3    Aye
11.2%  Radanovich, George  R    CA-19   Aye

There are a lot of cross-currents here. On the one hand, Congressmen clearly understood that the optics of the bailout bill were poor. Those most vulnerable to being voted out of office were inclined to vote against the bill, and those least vulnerable to electoral pressure -- especially those that were retiring -- were inclined to vote for it.

The banking industry -- Sirtoa's kleptocracy -- appears to have had its influence; members receiving a lot of contributions from the banking sector were significantly more likely to vote for the bill. Then again, we may be confusing cause and effect. Does Spencer Bachus tend to vote for bills that help the banking industry because he gets a lot of financial support from it? Or does he get a lot of support from the industry because he tends to vote in its favor? Or, might there be some characteristics of his Congressional District -- such as high employment in the banking sector -- that motivate both the contributions and Bachus' votes? This is a tricky problem to solve. I did, however, look at the number of employees in each district in the financial sector, as that data is available from the Census Bureau, and it was not a very strong predictor of voting. So money/lobbying from the industry a factor in some of these votes -- it was not simply a matter of Congressmen protecting local jobs in those industries.

Race was a factor; the Hispanic and black caucuses really did not get behind the initial bailout bills. The black caucus (though not the Hispanic caucus) did get behind the second bailout bill in October -- the one that passed -- perhaps as a result of Barack Obama's firmer support for the measure.

And finally, ideology was a factor. Holding all else constant, we still show more liberal/progressive members as being more likely to support the bill.

What I see, frankly, are some "bad" votes on both sides of the issue. On the one hand, the banking industry may have had some influence on certain of the 'yes' votes. On the other hand, I see demagoguery in certain of the 'no' votes, particularly as retiring congressmen -- those who are presumably freest to evaluate the bill on its merits -- were overwhelmingly inclined to support the bailout.

What I still don't see is much evidence that voting against the bailout tended to be the "progressive" position; in fact, the evidence seems to point in the opposite direction, although really the votes on the bailout seem to be more nonideological than anything else.

I am well aware, certainly, that there are other ways to define "progressivism" besides through the voting tendencies of the Congress. At the same time, certain of those definitions may be self-serving. I could make an argument, for instance, that "progressive" values should (as I happen to believe) include fairly strong support for Second Amendment rights, but that is not the generally accepted position. Terms like "progressive" lose their meaning if they tend to be defined as "whatever the particular blogger thinks about the world".