Not terribly long ago we linked to an Art Devany article where he claimed that the steroid era had no statistically discernible effects on home run production. Eric Gould and Todd Kaplan look carefully at the numbers and determine that Jose Canseco had a big influence on his peers’ performance numbers.
From the abstract:
[W]e estimate whether Jose Canseco, one of the best baseball players in the last few decades, affected the performance of his teammates. In his autobiography, Canseco claims that he improved the productivity of his teammates by introducing them to steroids. Using panel data on baseball players, we show that a player’s performance increases significantly after they played with Jose Canseco. After checking 30 comparable players from the same era, we find that no other baseball player produced a similar effect. Clearly, Jose Canseco had an unusual influence on the productivity of his peers.
If you are a baseball fan, this is a nice research paper to take a look at. The problem identification is clear, the statistical analysis is straightforward, and the interpretation of the coefficients is central to the analysis. In other words, it isn’t enough to be statistically significant, it also must be “economically” meaningful.
Still, I am going to talk to Prof. Finkler, because I don’t think the numbers are quite consistent.
Here’s a nice summary at Slate.
And here’s Mr. Canseco’s tell-all, Juiced. When this came out, it was scandalous and the denials were ubiquitous. But, as time marches on, several allegations have come to pass, and few have been discarded. By the way, that’s the famous “ball bouncing off Canseco’s head and over the fence for a home run” picture, run ad nauseum on sports bloopers back in the day.