Sunday, March 7, 2004
previous entry | main | next entry | TrackBack (12)
How conservative is Bush? How liberal is Kerry?
Northwestern University political scientists Jeffrey A. Jenkins has an interesting essay in today's Chicago Tribune on where George W. Bush and John Kerry stand in the political spectrum, using standard methods in the study of American political science:
For an introduction to the methodology Jenkins used for this op-ed, click here.
UPDATE: James Joyner provides a cogent critique of the Poole-Rosenthal method for determining ideological position:
ANOTHER UPDATE: Chris Lawrence has further thoughts on methodology. And Jenkins responds by posting comments here, here, and here. One point is particularly interesting -- Poole and Rosenthal used the early 1990's Kerry as an example of their methodology in their 1997 book Congress: A Political Economic History of Roll Call Voting. Kerry came out as quite liberal. What happened?:
posted by Dan on 03.07.04 at 12:08 PM
I'm just curious: how valid is this methodology, or any methodology that looks at voting records? An awful lot of votes are based simply on supporting the party, and don't really say much about the person himself. Do these models take this into account?posted by: Kevin Drum on 03.07.04 at 12:08 PM [permalink]
Thank you for the interesting perspective on the main presidential candidates. This raises the question of how well a politician's behavior in congress predicts his or her behavior once in the executive branch of government.
President Bush certainly is not a far-right conservative, but many of the things he has done in office *have* been examples of extremely conservative thought.
In Congress, the behavior of the indiidula must be understood in the context of cooperation with a large group. This is less of an issue in the executive branch, where the executive has greater freedom to act unilaterally.
I think that, in addition to understanding the voting record, we need to get a view of how much a candidate makes an effort to cooperate with others. Ultimately, the willingness to compromise could be a better measure of suitability for office than one's position on the left-right political spectrum.posted by: Joseph on 03.07.04 at 12:08 PM [permalink]
Liberal/conservative seems pretty meaningless to me, if you're just looking at voting records. What about people who undercut the political support of liberal groups while voting for liberal policies, and vice versa?posted by: MattS on 03.07.04 at 12:08 PM [permalink]
The answer to Kevin Drum's question is that the methodology Jenkins uses can provide a sound though incomplete comparison of legislative voting records.
It cannot provide an reliable comparison between one man's legislative voting record and another man's action as President, because while the legislative and executive functions do overlap they do not completely overlap. In particular, a President is often in the position to do things that a legislator can only respond to later, indicating perhaps his approval or disapproval but not whether he would, or could, have done what the President did had he occupied the Oval Office.
Take, for example, Nixon's China initiative in 1971-72, an opening toward one Communist dictatorship that aimed to isolate another, the Soviet Union. It's fair to ask whether Nixon was acting as a conservative (making difficulties for Moscow) or a liberal (as some Democrats had supported talks with Beijing), but my point is that, as the Vulcan saying goes "only Nixon could go to China" -- because only Nixon was President. Whether a legislator placed himself on record in a roll call vote favoring or opposing Nixon's step is essentially irrelevant.
On the other side of the coin, not all Presidents have voting records: only one of the last four Presidents (the elder Bush) ever served as a legislator. Their announced positions on or even their actions in response to legislation are dubious points of comparison, because while Presidents can influence legislation their level of influence varies widely depending on what the Congress looks like. Gerald Ford's Presidency looks more liberal than George W. Bush's, but Ford faced a Congress after the Watergate landslide of 1974 gave Democrats near veto-proof majorities in both Houses. This meant that while Bush could advance his own legislative initiatives, Ford mostly had to respond to Democratic bills.
There is finally the consideration that voting records do not fully summarize what a legislator does. John Kerry in legislative terms is a show horse, a lightweight. He has written little legislation that became law in almost two decades in the Senate; he has never held a position in the Senate leadership; he has never even chaired a major committee, which is unusual in an era of small (55 or less) Senate party majorities that afford most senior members the opportunity to secure a chairmanship if they really want it. But Kerry, like past Presidents Truman and Nixon, has been unusually active in leading Congressional investigations -- for example into the BCCI scandal and the Vietnam POW issue -- devoting tremendous amounts of his time and energy to efforts that by their nature produce mostly information for the public. I am not arguing that such effort are without value (I believe the opposite, in fact), but they do not appear to be reflected in Jenkins' methodology. Considering the nature of Kerry's Senate career, his voting record may not be a wholly reliable guide to how conservative or liberal his Presidency would be.
As an aside, and because trade is discussed so much on this board, I'd point out that the two Republicans Jenkins considers on the far right of the party usually took opposite positions on major trade issues throughout their respective careers in Congress, Phil Gramm as a doctrinaire free trader and Jesse Helms as a defender of North Carolina's declining textile industry and tobacco farmers. Whether or not one thinks that important from the standpoint of ideology depends on how important one thinks trade issues are.posted by: Zathras on 03.07.04 at 12:08 PM [permalink]
Didn't I just read that Kerry was rated the most liberal senator? Who's more liberal? Noam Chomsky? Kwawzi Mfume? Jimmy Carter?posted by: erp on 03.07.04 at 12:08 PM [permalink]
The Poole-Rosenthal technology is certainly not perfect, and inducing ideal points for Presidents adds another layer of uncertainty. For one thing, Presidents (when treated as legislators), as Zathras rightly points out, can select on those issues that they support/oppose. This introduces some definite measurement error into the estimation.
These limitations are most problematic for the comparison to Kerry. My intent in the article was to focus on Bush -- and, there, the limitations are less problematic, as I'm comparing Presidential behavior to Presidential behavior. In a sense, the measurement issues related to the estimation should be consistent across the various Presidents. Of course, I could not avoid the temptation to discuss Kerry as well...
Re: the last point. Yes, a recent National Journal article reported Kerry as the most liberal Senator. The National Journal scores suffer from the same types of problems inherent in most interest group scores -- they incorporate a fairly small set of votes, which typically leads to findings of "artificial extremism." That is, moderates (in each party) are pulled toward the extremes. Why? The votes that are culled for these various interest group scores are often highly-visible votes... such votes typically provide very distinct choices (i.e., the location of the policy output from the bill and versus the policy output under the status quo policy are very far apart). Moderate legislators therefore are left in a bind... neither policy position is what they really want, but they are left with a dichotomous choice -- vote for one or the other. Thus, they appear much more extreme in these small-vote scores than they really are. Kerry is a case in point. The Poole-Rosenthal scores (or NOMINATE scores) incorporate ALL votes -- which is in the order of 1,000 per Congress, versus less than 100 in most interest-group scores) -- so they provide more fine-grained estimates.posted by: Jeff Jenkins on 03.07.04 at 12:08 PM [permalink]
I just took a look at Chris Lawrence's website... some good stuff there.
Re: his thoughts on my piece. First, I used both W-Nominate and DW-Nominate in my analysis. W-Nominate was used for my quartile-based analysis, i.e., pinning Bush down in the 107th Congress, for example. (For more analysis on Kerry, see my February 15th op-ed in the Chicago Tribune). DW-Nominate was used for across-time comparisons, such as the Presidential comparisons that I reported.
Interesting point -- I tried to really lay out the "science" behind the ideal-point estimation in my op-ed piece. But it didn't make the Tribune cut. The editor didn't think the average reader would understand or care ... and he's probably right. Still, it probably makes it harder (and more frustrating) for the reader who really DOES want to figure this stuff out. I worry that the analysis will appear as "smoke and mirrors" to those who don't know Poole and Rosenthal's work.
Chris also makes a good point regarding Kerry's National Journal score -- that is, Kerry's abstention rate is quite high in the 108th, given that he's been on the campaign trail. Thus, the number of "active" votes that make up his score is smaller than the already-small National Journal set.posted by: Jeff Jenkins on 03.07.04 at 12:08 PM [permalink]
Last point -- for those who want to learn more about the Poole-Rosenthal technique, pick up their book, "Congress: A Political Economic History of Roll Call Voting" (Oxford University Press, 1997).
Note -- in their first chapter, they give a very nice example of a simple spatial model using 5 well-known legislators... and John Kerry is one of them! AND, at the time of their analysis, which was the early 1990s, John Kerry actually WAS an extremist within his party. What's interesting is that over time Kerry has remained remarkably consistent, BUT the party has continued to move left. So, because of leftward movement of carry-over members along with liberal replacement of moderates, Kerry now looks left-of-center, but not extremist.posted by: Jeff Jenkins on 03.07.04 at 12:08 PM [permalink]
Keith Poole frequently post current policy analysis on his website: http://voteview.uh.edu/default_recpol.htm
Quick responses to two other questions:
1. Liberal-conserative is by far the dominating dimension in roll call voting in the current Congress.
2. Kerry for the longest time was around 10th-most liberal Senator. Picking the 100th Senate at random he was to the right of: Simon (IL), Cranston (CA), Kennedy (MA), Matsunaga (HI), Pell (RI), Mikulski (MD), Adams (WA), Leahy (VT), and Sarbanes (MD). He has moved to the center recently as someone else noted, but imho not only because of turnover, but also because he prepared himself for a presidential bid.posted by: ogmb on 03.07.04 at 12:08 PM [permalink]
One note regarding the large number of absentions possibly skewing Kerry's score leftward. The flip side is that these are the votes that he felt the need to actually show up to work and vote for, so these are more likely to reflect strongly held positions. For example, the recent vote on gun control.posted by: Jim Thomason on 03.07.04 at 12:08 PM [permalink]
A few comments on methodology:
1) If there is such a thing as the gold standard in political science research on estimates of ideology, the Poole-Rosenthal methodology is it. For a partial bibliography of academic and non-academic work that uses these scores see here: http://voteworld.berkeley.edu/bibliography.html
2) The National Journal qualified its earlier statement that kerry is "the most liberal" and showed that even using its methodology this was not so for earlier years when more votes were used: http://nationaljournal.com/0305nj2.htm
3) The over-time comparisons with earlier presidents are possible because there is overlap in Senate membership and because liberal-conservative has been such a dmoninant dimension of conflict for such a long time. The conclusion that Kerry is close to the center of his part does not depend on the dynamic model though.
4) Correlation between NOMINATE scores and interest group scores is generally very, very high.posted by: Erik Voeten on 03.07.04 at 12:08 PM [permalink]
John Kerry does not seem to possess any well thought out ideas. Core beliefs really don’t matter in his corner of the world. He is a postmodern Democrat---and therefore intellectually lazy. Indeed, it is fair to describe him as anti-intellectual. The Massachusetts senator merely puts his wet finger into the air and flip flops accordingly. President Bush may not be a scholar, but he has spent far more time contemplating on these matters. Not even Senator Kerry knows what he will stand for in the next five minutes.posted by: David Thomson on 03.07.04 at 12:08 PM [permalink]
Kerry has a 20 year record of political opportnism. Which, of course, is why the 'hate Bush, nothing else matters' platform is such a perfect fit for him.posted by: Bithead on 03.07.04 at 12:08 PM [permalink]
Yeah, and another lousy opportunist was FDR. That guy campaigned in 1932 on reducing the deficit and then went and engaged in deficit spending! He said he'd keep our boys out of war in 1940 and then took us into WWII.
Boy, I hope we never get another lousy politician who keeps changing his mind over time!posted by: TedL on 03.07.04 at 12:08 PM [permalink]
To get closer to an apples-to-apples comparison with Kerry's voting record, Jeff Jenkins might consider comparing it with the records of Truman, Kennedy, Johnson, Nixon, Ford and the elder Bush -- not as Presidents, but as members of Congress.
I have no idea what such a comparison would show as to how liberal Kerry is compared to past legislators who went on to the White House, but it would be an interesting exercise.posted by: Zathras on 03.07.04 at 12:08 PM [permalink]
David Thomson wrote:
posted by: Jay Dien on 03.07.04 at 12:08 PM [permalink]
"Take, for example, Nixon's China initiative in 1971-72, an opening toward one Communist dictatorship that aimed to isolate another, the Soviet Union. It's fair to ask whether Nixon was acting as a conservative (making difficulties for Moscow) or a liberal (as some Democrats had supported talks with Beijing), but my point is that, as the Vulcan saying goes "only Nixon could go to China" -- because only Nixon was President. "
"Only Nixon Could go to China" because he had an almost unblemished record as a staunch anti-communist for over 20 years, not because he "was president". For instance, George McGovern could never have "gone to China" without appearing to condone and even sympathize with Communism. In fact I'm pretty sure it wouldn't just have been an appearance.
In the same vein unfortunately, only a Democrat can reform Social Security, which means it will be a cold day in Hades.posted by: DSpears on 03.07.04 at 12:08 PM [permalink]
This complicated voting-record analysis is MEGO stuff, i.e., "my eyes glaze over." And then there are the predictably partisan responses. This is sad. How do we deal with the flourishing of unthinking partisanship. With identifications that we hold without regard to others' ideas and feelings. This is way big a question for me to solve, but I may have two suggestions:
One, read the delightful book "Night Comes to the Cretaceous," 1998, by James Lawrence Powell. Oh, by the way, it's subtitled "Dinosaur Extinction and the Transformation of Modern Geology." The extinctions include a school of geologists who clung stubbornly to views that were increasingly proved to be erroneous. It's a well-told tale of how some very intelligent people who have accomplished much in their lives can be so stubborn as to march right into the shadows. (Do I have a political view? Yes, that, with exceptions, but by and large, those who are economically liberal and socially conservative are, likewise, headed into the ideological dustbins of history.)
Two, try to think of the political center as, in one perspective, a solid anchor by which to moor one's political understandings. That center is a fairly wide swath in which criticism and self-esteem both have a comfortable share. If one is only critical and antagonistic or even despising of one's sources of identity, one is a partisan and does not live in this swath. Similarly, one who finds only one's own kind (not beliefs) always to be right and desirable, likewise. Self-hatred and self-love are good, but only combined and in moderate doses. These two ingredients are the overlooked emotional roots of our political beliefs.
In toting up votes or positions, try measuring how often the candidate criticizes his country and fellow citizens, and how often he finds his land and people in esteem.posted by: John Hall on 03.07.04 at 12:08 PM [permalink]
"and John F. Kennedy, to whom Kerry is often compared, appears considerably more liberal than the Massachusetts senator trying to follow in his footsteps"
I must be getting old. but I don't recall Jack Kennedy as being particullarly liberal. I recall him running against Nixon on a plege to close the Missle Gap. I recall his massive tax cuts for the rich. I recall his anti-communist actions in Laos and Viet-nam. I recall some unhappiness with him from black people for his failure to push harder for civil rights legislation. But I do not recall anything particularlly liberal.posted by: Robert Schwartz on 03.07.04 at 12:08 PM [permalink]
Someone mailed me aquote this monring:
When Teresa Heinz-Kerry arrived, she handed me a pin that read in the center: “Asses of Evil” with “Bush”, “Cheney”, “Rumsfeld” and “Ashcroft” surrounding it.
Yep. Kerry's sounding more moderate by the day, isn't he?
I have read a lot of your comments especially on the side of what look to be republicans, I think that neither of the candidates are well equiped to guide the country, especially dick chaney, I mean lets put our feet on the ground for a moment here, if we want to be called patriots lets start by accepting the truth, and that is, that DICK is the real president here, he has hes agenda and he needs 4 more years to close it. As to who would be a better choice of president for this country now a days, hands down Kerry, why? I'LL tell you why, kerry is more flexible (that gives him a advantage to get the 25 million jobs that Bush has loss because he is stuck in one place) yes he votes for one thing now and tomorrow for the opposite, (people if that is what the country needs , whats the problem, bush like in iraq, he wont even accept hes wrong let alone change course, so there Bush SUCKS)He is a way by far far better choice to be comander in chief that Bush, were was Bush in vietnam, hdidng behind daddy and oh yeah getting high on COCAINE and getting drunk daily and speeding all the time, YOU KNOW THAT WOULD BE A GREAT LEADER, THAT TELLS A LOT OF HIM, by the way that was sarcasm.
Post a Comment: