Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.


root
Posts:
53
Registered:
8/26/09


Statistically, is the Supreme Court constitutional?
Posted:
Feb 8, 2012 9:46 PM


Maybe I should word the question: Is the Supreme Court consistent with the ideas of the founding fathers?
The role of the USSC is to decide the constitutionality of laws passed by Congress and the States (primarily). If all decisions rendered by the court came out 90 then we might expect that the mandate of the court was clear cut. On the other hand, if the justices voted as if each tossed a coin for every decision then their function would be more harmful than anything else.
Shortly after Sandra Day OConnor retired, I took all the decisions rendered by each justice over a period of several years. I assigned a 1 to a yes vote (concur) and a 0 to a no vote. I then had 9 time series, one for each justice, consisting of a sequence of 1s and 0s.
I calculated the covariance matrix of these 9 time series and the eigenvectors/eigenvalues of the matrix. By looking at the eigenvalues I can determine the effective number of justices on the court. If a single eigenvalue represented all the variance, then the court would be voting as a single mind. On the other hand, if the nine eigenvalues were roughly equal then that would mean no coherence to the voting.
It turned out that just a little over 2 eigenvalues accounted for almost all of the variance. I think that means that philosophical differences account for the decisions. It may not be partisanship, but it certainly represents factionalism.
The Founding Fathers wanted to avoid factionalism.



