COMMENT ON A TWEET #21: NATE SILVER ON JOHN ROBERTS’ USE OF STATISTICS

COMMENT ON A TWEET #21: NATE SILVER ON JOHN ROBERTS’ USE OF STATISTICS

Matt HealeyThursday,7 March 2013

The Snap:

Tweet: “@fivethirtyeight: Here’s my evaluation/critique of John Roberts’ use of statistics in the Voting Rights Act case. http://t.co/ZJ39OJ5s1E

The Download:

This article was a very good and fair evaluation of the statistical arguments John Roberts used in the Voting Rights Act case. As usual, it is never a good idea to argue statistics with Nate. It is the intellectual equivalent of picking a fight with a grizzly bear without the benefit of a weapon. Nate does object to some aspects of Roberts use of statistics. Roberts was asking that since the ratio of black to white votes in Mississippi, which is covered by the voter rights act, is better than Massachusetts, which is not, then perhaps section 5 of the Voting Rights Act is no longer necessary. The problem with this is exactly the same problem that the Romney campaign had – they are cherry picking data. A better data point would have been to compare all states covered by the act to all states that are not covered by the act. If you use that metric, surprisingly, the point is the same. There is little difference between the two types of states. However, this does not mean that the act should be overturned. The court needs to determine if this is true because of the act, or because the attitudes have changed. This data point alone can be used to argue both sides – there is little difference so we should over turn or there is little difference because it is working so it should not be overturned.

Hat Tips:

Fivethirtyeight, NYT review of Roberts, Image Credit: Wikimedia Commons

Take Action!



Subscribe to get updates delivered to your inbox