Left-right presidential swings since Truman

[cthulu2016]

This is interesting. One need not take it as gospel, but you can find more about the assumptions and methodolgy at the bottom link. The fact that it purs Obama to the right of other Dems is not (in my view) about him personally as much as the fact that the national center has been moving rightward (certainly where money is concerned) since at least 1980.

It is interesting to see someone try to represent, in some way, how very large the difference between parties is versus the small differences within the parties.

Obama the Moderate

...

I’ve long been a great admirer of the work done by Poole and his collaborators. What they do is use roll-call votes to map politicians’ positions into an abstract issue space. You can think of this as a sort of iterative process: start with a guess about how to rank bills from left to right, use that ranking to place politicians along the same spectrum, revise the ranking of bills based on the politicians, and repeat until convergence. What they actually do is more complicated and flexible, and allows for multiple dimensions; but that sort of gets at the general idea.

And it turns out that US politics really is one-dimensional, that once you know where politicians stand on a scale that clearly has to do with taxation and the size of the welfare state, you can predict their votes very well. There used to be a second dimension, clearly corresponding to race; but once the Dixiecrats became Republicans, that dimension collapsed into the first.

The new result comes from identifying cases where presidents clearly endorsed or opposed legislation, and using those cases to place presidents on the left-right scale. Here’s what they find, with up meaning moves to the right, down moves to the left:

...

I’m not bashing Obama, by the way; I wish he took stronger stands, but I think he’s moving in that direction; also, even if his health reform was devised by Heritage and implemented by Mitt Romney, it’s a lot better than nothing.

http://krugman.blogs.nytimes.com/2012/02/05/obama-the-moderate/

http://voteview.com/blog/?p=317

[MFrohike]

1. Hmm

Can't seem to find a list of the votes that make up the basis of this graph. I find it kind of unbelievable that Carter and Clinton are somehow to the left to LBJ. I can find a bunch of gobbledegook about the use of roll call votes when I click through Krugman's cited source, but somehow the exact votes and the methodology as to what constitutes "clearly supports/opposes" seems to be missing. Hell, what constitutes left/right in this graph is missing. Is free trade left or right? What about voting rights? Are tax incentives/expenditures generally one or the other or does it break down depending on the exact tax under consideration?

Additionally, I find this whole exercise suspect because it doesn't take non-legislative actions into account. How do you measure the use of executive orders by presidents who did not do well with Congress, like Kennedy and Truman?

I don't think this graph is terribly useful. Its results are pretty suspect from the start. How is Carter, as president, to the left of every Democratic president since Truman? Carter and Clinton were fans of deregulation, yet they're on the left? What?

[cthulu2016]

2. The work is imperfect, as any

attempt to quantify ill-defined terms will be. There's not really a "right" answer to liberal and conservative.

And I think almost anyone would rank the republicans that way - Ike a moderate, Nixon surprisingly moderate domestically, GW Bush to Reagan's right, Papa Bush less extreme than Reagan... if a method gets that right (which it appears to) then it is interesting to see how it ranks other things, even if one disagrees with the conclusion.

It's interesting stuff though. For instance, this (below) is intriguing:



http://voteview.com/blog/?p=284

[MFrohike]

9. And that's the underlying problem

It's fantasy to try to quantify this sort of thing. The use of statistics is valid when you're dealing with data that is readily quantifiable. This data doesn't lend itself to being quantified, nor is the methodology clear. I just don't care for trying to graft statistical analysis to every subject. It's been pretty damn disastrous in economics and I don't expanding the reach of the "statistics uber alles" mindset is a wise move.

[joshcryer]

11. Here's their methodolgy, on first glance I consider it sound:

http://en.wikipedia.org/wiki/NOMINATE_(scaling_method)

[MFrohike]

15. Thanks

That was interesting. It seems like it may be valuable for single bills and maybe issues. I don't think it scales up so well when you have to weight social vs. economic issues, state vs. individual, and the like. That being said, in a simple form, it may be useful.

[joshcryer]

18. I think it fails at nuance, but it helps show the divergence in parties.

It's an abstract model to be sure.

[ProSense]

3. That

chart is absurd. Let's see Clinton was pro-deregulation, he signed welfare reform, DADT, DOMA, NAFTA, the repeal of Glass-Steagall, the Enron loophole, declared the era of big government is over.

<...>

Mr. Obama’s third State of the Union address is widely seen in parallel with the one delivered in 1996 by President Bill Clinton. Mr. Clinton likewise was seeking re-election, after voters in the midterm elections had put Republicans in power in Congress as a rebuke to his perceived big-government liberalism.

Yet Mr. Obama is charting an opposite response. While Mr. Clinton sought to co-opt Republicans’ small-government message — his State of the Union line “the era of big government is over” is among the most memorable of his presidency — Mr. Obama is confronting it, and framing the election-year debate in a way that aides say will challenge Republicans’ view of the role of government in a time of economic transition.

http://www.nytimes.com/2012/01/22/us/politics/obama-to-draw-an-economic-line-in-state-of-union.html

This chart is suppose to show Clinton is more liberal than LBJ and Truman? Even Carter was pro-deregulation. He was, in fact, a conservative Democrat.

It's rather interesting that Obama ends up on the same line as LBJ, father of the Great Society. The chart uses an odd model to basically rewrite history in favor of Clinton as the great liberal (it seems an ongoing project). What hogwash.

[joshcryer]

4. Said this from the start, but still, he's been consistently moderate.

And consistency is something I admire in politicians, since typically they're two faced.

And if he makes a move to the left, with a new populism as it appears he is doing...

...then you can expect his graph to actually change.

[cthulu2016]

6. The R-D gap and ranges is far more interesting than

the small differences between parties.

Ike was quite moderate, but it's interesting to see a measure whereby even an okay republican is well to the right of the D-range.

And the general move rightward of both parties after the 1970s.

[joshcryer]

7. I agree did you go to that site's blog? Look at this:

Utterly fascinating.

[cthulu2016]

5. They found Obama to the left of H Clinton

close, but to the left.

http://www.voteview.com/recentpolitics.asp

http://voteview.com/Clinton_and_Obama.htm

And a ranking of everyone who served in congress 1937-2002, which is cool. (Ron Paul is the most conservative of the 3320 people on the list by their reckoning of conservative vs. liberal. Russ Feingold is remarkably high on the liberal list -- it's typically harder for senators since they represent a whole state, but he's well up there.)

http://www.voteview.com/Is_John_Kerry_A_Liberal.htm

[joshcryer]

8. This site is awesome and there's a serious shift to the left in the country...

...since the 100th or so Congress. I am quite fascinated by these results.

[ProSense]

10. More

absurdity:

253 USA D CLINTON, WILLIAM J. -0.440
307 TENNESS D FORD, HAROLD -0.421
478 MASSACH D KERRY,JOHN F. -0.366

So Clinton and Harold Ford are more liberal than Kerry?

Looks like they came up with a model to put people whereever the hell they want to on their scale.

[joshcryer]

12. "presidential ideal points are somewhat biased towards the ideological extremes"

You cannot compare a President to a Congressperson using this method.

They are compared against one another. So basically Presidents get a kick to the left or right because they have a smaller sample.

Below we plot the estimated positions of presidents between 1945 and 2011 along the liberal-conservative scale, which produces a pattern we call the “presidential square wave”. Because we use first dimension (ideological) Common Space DW-NOMINATE scores, presidential locations are directly comparable across time. However, because presidential estimates are based on a limited number of “presidential support” votes– roll calls on which the president clearly indicates his support or opposition to a particular (often contentious) measure, presidential ideal points are somewhat biased towards the ideological extremes (however, this effect is roughly constant for all presidents, so it is unlikely than any particular estimate would be affected more than others).

Our findings here echo those discussed in a prior post that Republicans have moved further to the right than Democrats to the left in the contemporary period. Indeed, as seen below, President Obama is the most moderate Democratic president since the end of World War II, while President George W. Bush was the most conservative president in the post-war era.

http://voteview.com/blog/?p=317

[ProSense]

13. What

"presidential ideal points are somewhat biased towards the ideological extremes You cannot compare a President to a Congressperson using this method."

...does that mean? The methodology is bizarre.

Below is a plot of the first (Liberal-Conservative) dimension of the Common Space Scores. (Below the plot I show, in order, from left to right all 3,320 unique individuals that served in the House and Senate from 1937 - 2002.) The histograms are for Democrats and Republicans in the two Chambers. Senator Kerry (D-MA) is located at -.366 which is to the left of the mean of the Senate Democrats, -.243 (standard deviation, .187). Senator Edwards (D-NC) is located at -.239 which is almost exactly on the mean. Senator Kennedy (D-MA) is located at -.476. President Bush is located at .408 which is to the right of the means of both the Senate Republicans, .271 (standard deviation, .187), and the House Republicans, .300 (standard deviation, .158). Vice President Cheney is located at .517 and former Senate Helms (R-NC) is located at .658.

And they end up with:

98 SOUTH D D MCGOVERN, GEORGE STANLEY -0.516
99 NEW YOR D KOCH, EDWARD I. -0.516
177 VERMONT I SANDERS, BERNARD -0.472

[joshcryer]

14. It's not a left right scale.

[ProSense]

16. Well,

" It's not a left right scale."

...they seem to disagree with you:

Below is a listing from left to right of all 3,320 individuals who served in either or both the House and the Senate from 1937 to 2002. Senator Kerry is number 478 and President Bush is number 2,971.

[joshcryer]

17. You are correct, I thought you were linking the actual data, but that is an analysis page.

It would've been nice if you had provided the link.

[ProSense]

19. It was

"It would've been nice if you had provided the link."

...in the comment I responded to.

[cthulu2016]

20. ...