« Are Law Schools Hiring More PhDs? | Main | Law & Psychology Conference »

02 March 2006

Comments

Stephen Jessee

We are actually in the process of updating the site to give a better explanation of the numerical scale on which the justices'ideal points are estimated.

Basically, the scale is defined by our modeling restrictions. We have fixed O'Connor to be at zero and Rehnquist to be at one. Therefore, all the other justices are estimated relative to these two points. We could have easily fixed two other justices or fixed Rehnquist at 100. This would change the scale, sort of like measuring temperature in ferenheit vs. celsius, but wouldn't change the relative distances and rankings of the justices.

We chose to fix O'Connor and Rehnquist because they are no longer casting votes, and hence they will not be moving at all. Plus, it's nice to be able to look at the graphs and see whether Roberts or Alito is more conservative than the justice he replaced.

As for interpretation of the scale, it really can't be fully interpreted apart from the statistical model. If we were to assume a different error distribution, for example, then this would effectively change the distances between the justices (but almost certainly not the rankings). On the whole, though, the scale goes from liberal to conservative, with very liberal values being more negative and very conservative values being more positive.

We believe that some of the most interesting results, however, deal with rankings rather than actual values on the scale. Saying, for example that there's about a 50% probability of Roberts being more conservative than Rehnquist, or that there's a very small chance Roberts is the most conservative justice on the current court, seems like the big payoff. See our "Trend Lines" section for these things.

Thanks for the interest, and I hope this clarifies things a bit.

Geoff Urland

This looks interesting, but it would be a lot more useful if somewhere on their website they explained what the numbers on the scale mean. (or did I just miss it?).

The comments to this entry are closed.

Conferences

November 2014

Sun Mon Tue Wed Thu Fri Sat
            1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30            

Site Meter


Creative Commons License


  • Creative Commons License
Blog powered by Typepad