However, you can color me (tentatively) skeptical. Now I'm no social scientist. But I'm not convinced convictions per capita is the appropriate measure. If we're looking at federal corruption prosecutions, wouldn't convictions per federal employee be a better yardstick? Using Glaeser & Saks's figure for convictions 1990-2000, and the U.S. Census report on paid civilian employment in the Federal government for 1998, I came up with the following conviction rates, which make Alaska appear a much less formidable contender in the corruption contest:
- Alaska: 0.0030
- Mississippi: 0.0045
- Louisiana: 0.0109
- South Dakota: 0.0050
- Tennessee: 0.0052
The Northwest retains remarkably clean hands:
- Washington: 0.0017
- Oregon: 0.0012
- California: 0.0044
- New York: 0.0072
Link via Marginal Revolution
UPDATE:
Professor Tabarrok points out the conviction figures are for "government officals," not just federal employees. So I've recalculated my figures, based on federal and state employees combined (and in parentheses, just state employees). The decimal places started to get a little long, so I've moved them over four places:
- Mississippi: 12.94 (14.80)
- Louisiana: 12.82 (14.51)
- Illinois: 12.26 (14.26)
- North Dakota: 11.88 (14.47)
- Montana: 10.33 (12.56)
- South Dakota: 10.03 (12.55)
- Tennessee: 8.38 (9.96)
- New York: 8.09 (9.11)
- Alaska: 7.06 (9.23)
- California: 6.58 (7.73)
- Oklahoma: 4.70 (5.71)
- Washington: 3.08 (3.75)
- Oregon: 1.78 (2.10)
These figures still don't match Glaeser & Saks's results. They report the following results for the "Top 10" convictions per capita:
- Alaska 0.643
- Mississippi 0.612
- Louisiana 0.513
- South Dakota 0.472
- Tennessee 0.464
- Illinois 0.458
- New York 0.439
- Oklahoma 0.415
- Montana 0.414
- North Dakota 0.398
Obviously, using the number of government officials rather than population makes a difference. Apparently, Alaska and Oklahoma have a high ratio of government officials, deceptively inflating their score.
1 comment:
Nice post! I can definitely see Mississippi being a top dog in this contest. The city government in Jacktown can't seem to make up its mind whether graft or incompetence is its first priority. Makes you think twice about all that federalism snake oil they feed you in law school (or not). Of course, then you think of the alternative...
I wonder whether there're other factors that distort the numbers, such as differences in the opportunity cost of investigating and prosecuting corruption cases. Other matters (drugs?) might hog federal resources in some states.
Just curious, 'cause I'm too lazy to read the report: Which statutes were counted? Is there generally one statute (or a small group) that the vast and representative majority of cases are brought under? (I know wire/mail fraud has an honest-services component that reaches public officials.)
Post a Comment