Here are the ideology and leadership indices for the House and Senate for 2011-2016:
As you can see, the ideology score seems to do a good job of separating the two major parties. Looking at the 2014 scores specifically, it's even clearer. For example Senators Jim Inhofe, John Cornyn, and Pat Roberts all score near 1 on the ideology index (far right on the chart above); Senators Elizabeth Warren, Barbara Boxer, and Bernie Sanders score near 0 (far left on the chart); and Senators Joe Manchin, Lisa Murkowski, and Mark Pryor all score near the middle.
I took the data for the 2013-2014 legislative period and split it into 5 groups: far-left wing, left wing, center, right wing, and far-right wing which allows for non-linearity in the effect of ideology on the probability a legislator voted for the 2014 Farm Bill. I plan to use a few different versions of this breakdown in the paper.
In the model I controlled for the legislators' party, sex, whether they were Senators or Representatives, and whether or not they served on the Ag Committees in their respective houses.
As you can see below, 68% of Senators and about 60% of Representatives voted in favor of the bill (i.e. voted "yea") and only 8 of the 65 members of both Ag Committees voted against it.
Looking at the correlation between a "yea" vote and the raw ideology index from GovTrack shows a weak but statistically significant positive correlation. The interpretation of this is that more right wing legislators were more likely to vote in favor of the bill.
 |
| (click the image to enlarge) |
So estimating a probit model with controls for party, house, sex, and Ag Committee membership gives us the following marginal effects:
Senators and members of the Ag Committee were, as expected, more likely to vote in favor of the 2014 Farm Bill. Compared to far-right wing legislators (the reference group), far-left wing legislators were 29% less likely to vote "yea" on the bill and right-wing legislators were 19.6% more likely to vote for the bill. There was no statistically significant difference in the probability of voting "yea" for left wing and centrist legislators relative to far-right wing legislators.
Looking again at the cross-tabs above, I think the regression analysis shows that there was, at least in the case of the most recent Farm Bill, a coalition of primarily right-wing and centrist legislators banding together with some far-right wing and left-wing legislators to pass the bill. While I have other variables to add to the model, I suspect that these relationships will hold up pretty well given the importance of ideology in other studies of legislative voting behavior. As always, I'm interested in readers' thoughts, suggestions, and questions.
*11:45 PM 4-24-2016 - I had previously stated that SNAP spending increased. Thanks to Keith Coble for the correction.
Levi-The ideology scores look a little weird to me. I assume that what it means to be a conservative is something self identified conservative groups would be best equipped to determine. GovTrack ranks the 10 most conservative Senators as Inhofe, Risch, Enzi, Roberts, Lee, Coats, Barrasso, Cornyn, Vitter, and Sessions. They are graded by the conservative group Conservative Review as C (74%), C (78%), D (62%), F (55%), A (100%), F (45%), F (54%), F (46%), C (71%), and B (80%). I'll grant you this is a group that counts it as part of the legislator's score their position on the bill in question, but it's striking to me how different how self styled conservatives rank senators in terms of their conservatism, and how these scores rank them.
ReplyDeleteI imagine there are self styled progressive groups that grade senators in a similar way. It might be worth looking into alternative ideology scoring methods to see how it influences the results (problem: finding scoring methods that don't use the farm bill vote as part of the score, or where they do, backing that out of the score. Hm.)
(The leadership scores are a little wonky, too. Harry Reid, who is, you know, minority leader, is described and ranked as a "rank-and-file Democrat." Maybe that just means he's a terrible minority leader in his party and the real control lies elsewhere. But it's very counter-intuitive.
On an unrelated note, I have a bit of data analysis I've done on the business cycle and labor allocation I think you might find interesting (it's what I mentioned a bit back that I was working on-well, one of the things I was working on). Actually it's something I've been sitting on for a while but I've updated it recently. I can email you about it if you want.
Andrew_FL
Andrew,
DeleteThanks for the suggestions! I'll certainly look deeper at the literature on ideology to be sure I have plenty of robustness checks in the analysis.
If you check out the GovTrack methodology section, they determine the Leadership and Ideology indices using co-sponsorship data. So, for Leadership, Senator A is considered more a leader than Senator B if Senator B co-sponsors more of Senator A's legislation than A does of B. Here's how they explain the calculation of the Ideology score:
Here’s how it works: Form a matrix (a grid of numbers) with columns representing Members of Congress and rows also representing Members of Congress. Do this for the House and Senate separately. We include (co)sponsorship from the current and previous two Congresses, so between four and six years of data. For the Senate, you have a 100x100 table. In each cell of the table, put the number of times the senator for the row cosponsored a bill introduced by the senator for the column. Or if it's the same senator in the row and column, put in the number of bills he or she introduced. Then compute the singular value decomposition of the matrix (which is how Principal Components Analysis is often done).
Every square matrix has a singular value decomposition. The magic is in how you interpret it. The singular value decomposition takes one matrix and gives you back three: called u, s, and v-transpose. V-transpose can be interpreted as a set of scores for each Member of Congress on a new set of dimensions. The dimensions are ranked in order by how much of the original data they explain. We have found that the second dimension best corresponds with ideology. We use the scores from that dimension in our charts.
Each score is a number. It’s entirely arbitrary whether liberal or conservative is positive or negative — the original matrix is blind to actual information like that. In fact, there’s no guarantee that these numbers even have anything to do with liberal- and conversative-ness. All it tells us is how to separate Members of Congress into two groups, or more precisely how to spread them out along a spectrum in a way that explains their record of cosponsorship. But in practice it captures ideology very well.
https://www.govtrack.us/about/analysis#ideology
Please do send your analysis, I'd love to read it!