Here's the link to the paper and an associated FAQ. Note that the paper is unpublished (meaning it hasn't undergone peer review) yet.
http://www.eco.utexas.edu/faculty/Ha...istpapers.html
I find the raw data of the paper to be interesting. For example, there are 25% more white batters than pitchers, 76% more hispanic batters than pitchers, and 570% more black batters than pitchers. I wonder what can drive such a disparity? (And I'm pretty sure it isn't umpires!)
I can't comment on the utility or applicability of their econometric models, but the graphs which show the differences in pitcher performance (strikeouts, homeruns, hits, walks, runs, Bill James' Game Score, wins, ERA) depending on whether pitcher/umpire ethnicity matches or is different are provocative. These are just raw data; you can form your own opinion of the statistical validity.