April 12, 2006

The many faces of John McCain

Posted by Arcane Gazebo at April 12, 2006 2:14 PM

I'm not convinced by Jacob Weisberg's argument that John McCain's not really a conservative. The idea seems to be that, sure, if you look at his voting record he's an ultraconservative, but if you look at his offhand comments you see that in his heart he's really a progressive. He just has to throw a little red meat to the base every once in a while, consisting of almost every vote he's ever made in his Senate career.

Even if he does have more liberal personal views than his voting record indicates, it's the votes that actually matter. The fact that he'll make moderate statements about abortion or gay rights doesn't do a lot of good when he's voting the other way. The only way this argument could work is if there's reason to expect him to be more moderate as president. But the political pressure from the Republican base doesn't go away the day after the election—just look at the trouble Bush got into with Harriet Myers. In fact, there's no reason whatsoever to think McCain would make policy differently as President than he does as Senator.

Let's consider the reverse hypothesis: McCain is actually an ultraconservative, but is trying to play to the moderate center in order to boost his presidential prospects. This actually explains the data better—the meaningless public statements placate moderates, while the ultimate policy decisions are still very right-wing. Now, I'm sure there is still some calculation going on in his recent appeals to the religious conservatives; no one as intelligent as McCain is going to view Jerry Falwell as anything other than loathsome. But in fact Falwell is so loathsome that cozying up to him goes beyond just political calculation: it's just plain unprincipled. I'll pass, thanks.

Tags: Politics, Republicans, Senate

Back when there were such things, Democratic Congressmen from Southern (and pretty conservative) districts commonly held their seats with this strategy: vote left, press release right.

McCain just borrowed their formula and converted it for his own use.

Posted by: JSpur | April 12, 2006 2:21 PM

I'm looking at the SVD-produced rank-ordering of the 107th Senate (2001-2002) right now. (This is the one I have handy because we put it in a paper. I don't have time this second to look at the other years.) This actually puts McCain far to the left of most Republicans---based only on one way of looking at the data and only votes taken during those two years. Also, the only one "less predictable" than him during those two years was Jeffords. Less predictable essentially means that taking only the first two SVD coordinates reconstructs a smaller % of his vote outcomes than others (roughly speaking, you can interpret that as meaning the error bar is higher in the first statement above).

Of course, we were using these measures of partisanships only loosely because it wasn't our main point. One would have to go through it in more detail (which can be done, and some poly sci people have done this using some methods) to make more specific statements about an individual. (These are collective methods after all, so one has to be careful with conclusions concerning things like rank-orderings.)

Posted by: Mason | April 12, 2006 2:44 PM

Weisberg didn't write this article - Marshall Witman did

Posted by: Gari N. Corp | April 12, 2006 3:11 PM

Bleh, crap like this is why I rarely visit Slate anymore... Granted it would be nice if it's right and there are still a few sane people left in the Republican party, but I'll believe that precisely one election cycle after I see it.

McCain was attractive in '00; I registered Republican for the specific purpose of voting for him over Bush in the primaries (hey, IIRC McCain did win Santa Cruz county!), and I would at least have thought about voting for him in the general election. His outspoken attacks on Falwell and the rest of the American Taliban were quite compelling... But now he's just another Bush-bot (see http://shakespearessister.blogspot.com/2006/04/ask-and-ye-shall-receive.html for a more graphic expression).

Mason, that's interesting to know that he's less wingnutty than his recent remarks would suggest. Though I remember seeing those vague "most liberal/conservative" lists that sometimes come out and McCain was way down, like 6th most conservative or something like that. It would be interesting to see the results of the more detailed PS analysis, if it has indeed been done (and can be understood by non-PS folks).

Posted by: Justin | April 12, 2006 3:44 PM

It has been done and one basically has to go through several methods and try to get collective insights. Saying somebody is 6th most conservative means a lot less when a slightly different metric ranks them as, say, 25th most. The key is that there are certain names that always show up somewhere between say, ranks 1 and 40 depending on the algorithm. Ordering between them becomes a lot trickier and is tied very fundamentally to the question one is trying to answer (which will typically have a lot to say about the specific schemes one uses).

By the way, it was necessary for me to check a couple things for my current grant proposal (in which such issues are showing up, given the topic involved) and the poly sci consensus is basically that McCain is a maverick among the Republican party in that he reasonably often votes completely against the party line. The data definitely supports this and it is what our work saw as well, but by no means does that mean that he's not a @!#?@! nut. Nevertheless, this is a fair designation given the actual votes.

If you want to see a ton of detail for a set of methods to look at data mining of roll call votes (including a discussion of whether and to what extent Lieberman and McCain changed their voting practices for their campaigns), see Keith Poole's VoteView website. I had this guy as a prof at Caltech when he visited for one term in spring 05 and his pioneering work on looking at the roll call this way was a very important influence in that component of my own research on Congressional networks.

Posted by: Mason | April 12, 2006 4:07 PM

I messed up the link above. Here's the url: http://www.voteview.com/

Posted by: Mason | April 12, 2006 4:09 PM

Very interesting data, thanks!

If I'm reading those lists right, McCain comes out of the analysis as extremely conservative in the last four years, but with the "maverick" effect of being less predictable by whatever algorithm these guys use. And apparently he's an extreme outlier on the "social issues" axis. I can't tell what that axis means, though. It's clearly not the abortion and gay axis the name might suggest. How Bush can be tightly sandwiched between Kennedy and Feingold on any axis is utterly baffling!

I'm pleased to see that '97-'02 McCain was in fact a fairly reasonable character according to these ideological rank orderings, so apparently I wasn't wrong to support him then.

I really wish I had the time to dig into this stuff! I wonder how (or if) these analyses take into account symbolic votes. For example, Chafee makes occasional meaningless "show" votes to try to maintain his liberal credibility for Rhode Island, but these always happen when his vote is not decisive. Lieberman's votes for cloture and against the nomination of Alito are another example. And back to McCain, the torture amendment kabuki comes immediately to mind...

Is there a good way to embed links into comments, btw?

Posted by: Justin | April 12, 2006 6:51 PM

Just use html; my goof was a typo in the url itself.

To address things along the line you mention, one can do the same analysis on appropriate subsets of the votes. One has an n x m matrix for n Congressmen and m votes, so one just uses appropriate submatrices using whatever data mining technique one wishes. By the way, not even close to all interesting possibilities have been studied in the past (which is a big reason why people are listening to the stuff I've done recently), especially with respect to trying to find correlations between the voting data and other data (such as committee assignments and legislation cosponsorship).

Posted by: Mason | April 12, 2006 7:11 PM

(Josh's inner monologue)

"Oh good! Politics! Oh good! Statistics! Two things I know very little about! Let me look at this webpage!" *clicks URL, head explodes* "Back to memorizing iambs, then."

Posted by: Josh | April 12, 2006 9:35 PM

My deed is done here.

Posted by: Mason | April 12, 2006 10:20 PM

My issue is I dont see any way this country politics will be improving any time soon (next 6 years). As of today there are no democratic presdential candites who could run and win. Sorry Hillary your a carpet bager and tend to inpsire outright hate form large percentages of the population (including many democrats). And the republican party has rolled over to its ulta conservative wing no treu moderates (if they exist) could pull the primary.

Posted by: shellock | April 13, 2006 5:05 AM

ps I want a viable thrud party to vote for is that to much to ask? (a fiscially conservate socially liberal party and i dont even mind small government)

Posted by: shellock | April 13, 2006 5:09 AM

Shellock, if you'd like to go ahead and form the third party you're describing, I'll be the first one to make a campaign contribution to it.

Posted by: JSpur | April 13, 2006 7:33 AM

One of the things that defines McCain is his loyalty to the GOP, not to any particular figure within. It might seem like a strange distinction, but I think it follows from his voting record. He toes the party line in most respects, but will go out on his own on issues he feels is important (think Gang of 13 which prevented the nuclear option and his attempts at campaign finance reform). If McCain runs in '08, I'm still voting for him, but I'm uncertain how good his chances are in the conservative dominated primary system.

Posted by: Chris L-S | April 13, 2006 9:04 AM

This country has historically been unable to support third parties. However, when one of them becomes strong enough, that has sometimes led to a reorganization among the two major parties (lines getting changed to absorb the new challenger and all that), which is probably our best hope.

That reminds me... is anybody submutting code or pseudocode to that political experiment I advertised?

Posted by: Mason | April 13, 2006 9:09 AM

Shellock, why don't you think Warner, Clark, or Richardson could win in '08? Is it just the hurdle of getting past Hillary in the primaries? Personally I have a hard time seeing how the Republicans can field a viable candidate in '08. McCain is demolishing his "straight talk maverick" image to get past the base (though the Weisbergs of the world may still believe it for a while); no freakin' way is Giuliani going to get many primary votes; and another far-right theocrat (Brownback, Allen, Santorum, Huckabee) will go down hard in the general election to anti-Bush "throw the bums out/time for a change" voters. If Powell (even with his tattered shreds of credibility, post-'03) or possibly Rice were in the mix, things might be different, but both seem quite emphatic about not running.

Ever since taking various PS classes as an undergrad, I've wished I lived in a parliamentary democracy structurally capable of supporting many parties (IIRC Germany was my personal ideal, but UK and Canada are both steps up from the US in terms of electoral design). Sadly, that ain't gonna happen in this country...

Posted by: Justin | April 13, 2006 10:31 AM

Mason: Sadly, no. I suppose there are a few days left, but it's only as of yesterday that I'm FINALLY F-ING IN THE CLEAR. (sorry, that was just directed at the sky) I did, in fact, sleep last night.

At the very least, thinking about their simplified model has given me an interesting perspective on 3rd parties.

Posted by: Lemming | April 13, 2006 11:06 AM

Shellock: I'm a bit more optimistic than that, like Justin I think that the Democrats have some decent candidates if the Hillary juggernaut can be overcome. Warner in particular seems like he'd be a good choice. (I personally like Russ Feingold but I don't know if he has much moderate appeal.) Also, "fiscally conservative, socially liberal" describes many of the leading Democrats these days including the DNC chair.

On the other hand the "buffoon with an accent" strategy has worked so well for the Republicans, it may well work again if they nominate George Allen. A lot may depend on whether voters finally stop trusting Republicans on national security after three more years of disaster in Iraq. And the effects of a war in Iran... I'm hoping the administration isn't crazy enough to invade Iran, but all evidence indicates that they are. (That's a different post I have yet to write.)

Posted by: Arcane Gazebo | April 13, 2006 11:20 AM

Question: if there is a controlling majority of right-wing nuts in the legislative, judicial, and executive branches, as well as active political gerrymandering going on to keep it that way, can we slap the Republican party with Antitrust laws? That would be a funny day in my book.

Posted by: Josh | April 13, 2006 11:27 AM

Josh, I think RICO might be a more appropriate statute than the Sherman Act. After all, it was adopted specifically with organized crime in mind.

Posted by: JSpur | April 13, 2006 11:45 AM

Organized crime is so much more despicable than organized Hime.

Posted by: Josh | April 13, 2006 11:50 AM

I actually saw a very amusing post recently (can't remember where, unfortunately) about using RICO or something similar on Bush and the Republican government. We're now seeing clear, beyond reasonable doubt proof that the Iraq war was indeed based on deliberate lies to the American public and not merely incompetence plus cherry-picked "intelligence". The (tongue-in-cheek) argument is to use these racketeering laws to have a judge impound the ill-gotten gains from the crime, namely authorization to use military force in Iraq, thereby immediately getting the troops home where they belong. :-) Cute, but not quite practical...

More seriously, it's unfortunately very hard to prosecute people with near-absolute power... Arguably that's why Stevens defected to deny Supreme Court review of Padilla, since if Justice Kennedy was leaning pro-Bush on that case a Supreme Court decision could easily have nullified the Fourth Amendment. I'm very impressed with Fitzgerald's success (so far) with the Plame leak investigation, though.

AG: my take would be that Feingold has by far the best appeal of any Senator; I'm personally ambivalent about him only because of the history of Senators running for President. One of the big spins in the Bush campaigns was that Bush stood on principles rather than catering to public opinion like most other politicians, and nobody can use that argument better than Feingold! :-) I think the "stands for what he believes, no matter the polls" factor should far outweigh any disagreements moderates might have with specific positions. And do moderates actually like the Patriot Act or support Bush anymore? Censure (and impeachment) had quite broad support last time I looked.

Posted by: Justin | April 13, 2006 1:32 PM

It is pointless to attempt to locate McCain on a simple left-right spectrum. It is also not very illuminating to say that McCain speaks one way and votes the other. He has his own opinions, many of which are scattered across the spectrum. Like any realistic politician, however, he understands that he will lose his influence if he loses the support of his party. He works to maintain his influence, keep his powder dry, and then inserts himself fully into the calculated issues where he knows he can make a difference.

Of course he's going to campaign for Bush, give some tacit support to the social conservatives, and so on. It's simple math--he cant win the Republican primary without at least a significant minority of social conservatives. That doesn't mean he's one of them. Look at where he's made his big stands...campaign finance reform (opposition to republican big donor base), banning torture (absolute opposition to bush), Gang of 14 (absolute opposition to Frist).

Here are a few other of his positions you may have overlooked...he supports stem cell research, has said he has no confidence in Rumsfeld, criticized republicans for keeping the tax cuts during war, opposed some of the Bush tax cuts in favor of less cuts for the rich and more relief for the poor, and opposes the Bush administration's environmental record. And of course you could have a similar list of his conservative votes and positions, but none of the above were mere meaningless public statements--they were all significant fights he picked with either Bush or his party.

If you're going to force him onto a spectrum, at least seperate his social, economic, and foreign policy views. After some research, you'll probably find him to be socially moderate (but only after balancing some left and some right views on abortion and the environment), slightly right of center on economics (but what's the center anyway? he wants to reduce the tax cuts, but also cut entitlement and pork barrel spending--something neither republicans nor democrats want to do), as for foreign policy he's conservative (but don't even attempt to put him left or right of neocons--neocons were once liberals that came to believe in the hard power pursuit of Wilsonian ideals. McCain is a classic conservative and follower of realism that believes in the pursuit of a calculated national interest.)

It is easy enough to paint someone like Bush in black and white. I would only suggest more caution when attempting describe a maverick like McCain--his views and motivations don't lend themselves to convenient classifications, gratifying and reassuring though they may be.

Posted by: Anonymous | April 15, 2006 6:12 PM

Torture - bullshit. That was plain and obvious kabuki theater to polish McCain's "maverick" image. The Bush administration does not consider itself bound by these cute little "law" thingies Congress passes, so what do they care if McCain organizes an anti-torture amendment? Torture has always been illegal under US law, one more amendment (which Bush promised to obey only if and when he felt like it in his signing statement) doesn't make any difference. All it does is help McCain with his image, and to the extent that McCain has been a good boy recently - especially the '04 campaign - Bush's people are happy to go along by "opposing" what to them is a legally meaningless PR gesture. That's not to say that the torture kabuki is completely meaningless; it does show that McCain is measurably less evil than Bush and (apparently) most Republican Senators. But it would be a hell of a lot more meaningful if McCain were willing to do something to hold Bush accountable - supporting Feingold's censure motion, introducing his own specifically on the torture issue, putting together a budget amendment holding Defense funding contingent on (say) International Red Cross inspections of Guantanamo, Abu Ghraib, Bagram, etc.

Not winning the primaries without the fundies - sadly true, most likely. Though he did pretty well in '00 until the Rove smear machine got into gear in South Carolina. And the natives are getting restless since they're not getting nearly as much of a reactionary theocracy as they voted for. If they get disappointed enough, '08 might be susceptible to a reality-based Republican campaign, contradictory though the concept would appear. :-) One can hope, at least...

I find the inclusion of the environment under a "social" category to be quite bizarre, btw. If you have to label it, "economic" would seem to be the logical place.

Campaign finance reform is not exactly an issue McCain wants noticed right now, since he's busily trying to rewrite the legislation to severely limit 527 spending while permitting the National Committees to spend freely again. No way anyone can cast that as fair and bipartisan reform!

Posted by: Justin | April 15, 2006 8:45 PM

Wow, Anonymous's post reads like a campaign advertisement.

If you take a look at the research that was mentioned in this discussion (see the url way above), the spectra being used are not one-dimensional. Parts of the first dimension (which has been shown to carry the largest amount of variance) were mentioned explicitly for the purposes of discussion, but the actual issue is taking a high-dimensional situation and trying to see how many dimensions one needs to keep the information (that is, be correct some high percentage of the time). In fact, this analysis agrees with you about McCain's "maverick" status, and keeping only the first two dimensions picks this up (the fact that all these analyses bring out this feature is also mentioned above). Because he's a maverick, it means that keeping those two dimensions reconstructs a somewhat lower percentage (maybe it was 80 %? I don't remember the precise number; this is why he shows up as the "least predictable" Senator in 01-02; that's exactly picking out who the method captures the least with two dimensions; this was discussed above, as I mentioned) of his individual votes instead of the 98 % or so gets for people like Jessie Helms. One can argue that for McCain, one needs to keep one more dimension to get a similar reconstruction, but keeping the first 2 dimensions at the level of an entire body of Congress for a term reconstructs nearly all the roll call outcomes successully--one gets 984/990 results correct for the 107th House, for example, and the number of errors is basically always comparable to that in every single Congressional term. One gets about 93% of individual votes with two dimensions, so looking at collective phenomena, this really is telling you that the data taken as a whole is well modeled by a two-dimensional spectrum (one picks up very small details at higher dimensions except for a couple specific Congressional terms, where one needs a third dimension that is purported to represent a North/South split); these tend to correspond to ones in which the parties are reorganizing, and such analysis has in fact allowed people to find such reorganizations in the first place!). It's true that for individuals that one can't just stop there (or one must state one's conclusions accordingly so as not to overstate one's case; that's why I made it a point to specify that it's not fair to give a specific rank ordering), but these sorts of analyses have repeatedly shown their worth (both academically and otherwise). The money that many lobbyists dole out is actually based in part on such spectra, although they usually just deal with the subset of votes about which they care rather than taking the data set as a whole. The 'political action groups' (I think that's not the right term) that go after various politicians they don't like and/or support ones they do also make use of this type of information. (And they also eventually give me the data they promised me that they can acquire, I hope...)

Posted by: Mason | April 15, 2006 11:36 PM
Post a comment