Lyin' and Cheatin' Midwest'ners

So, Rod Blogojovichmuskanuwazzisnameagin' has been accused of a corruption. This proves the Illinois political machine is the worst oiled int eh country, yes? Well, sort of. Chicago is historically a hotbed of trouble. It's hard to get somewhere in IL without involving Chicago, and to do that means rubbing shoulders with some freaky business. But I was a little surprised at this article. So I had to read it. I mean, seriously, those cattle traders need to be watched, yes?

It turns out that the situation is probably more flash and sparkle from USA Today, who did the analysis, than anything actually newsworthy. North Dakota is not, as the headline implies, the most corrupt state. It does top the particular analysis of corruption, however, so no lie has been made. The analysis that it tops? Take the number of federal convictions for corruption during ten years and divide by the state population. Um... okay? First off, even though some factor is being divided by a population, it's not being divided by the right population. What percentage of the people in the state could be involved in corruption? Theoretically everyone, but in practice no state is that pure of a democracy. Second off, federal laws are not the only laws that could be enacted against corruption.

A better measure for the first point, why not consider what percentage of the population is involved in government? Here are a few states

WY 11,810
SD 12,150
VT 13,682
ND 14,597
MT 17,013
RI 18,511
DE 22,810
AK 23,647
LA 77,597
IL 105,471
MI 118,667
OH 119,147
PA 141,478
FL 171,342
NY 236,719
TX 259,578
CA 334,432

Now, what is the ratio of state government employees to population in those states?

WY 2.259%
SD 1.526%
VT 2.202%
ND 2.282%
MT 1.776%
RI 1.750%
DE 2.638%
AK 3.460%
LA 1.801%
IL 0.821%
MI 1.178%
OH 1.039%
PA 1.138%
FL 0.939%
NY 1.227%
TX 1.086%
CA 0.915%

(Note that I cheated a little. I used November 2008 employment with 2007 populations. But my main point lies in the first digit or two of the percentage, which won't change much.)

Notice anything about those states? The smallest states by population (1,000,000 or so and under) have about twice as many state employees as the largest states by population (about 10,000,000 and over). This is an estimate for the amount of government in the state, and is probably related to the amount of people likely to be involved in corruption. True, it misses local and federal employees. Those should to be factored in as well, along with convictions on non-federal charges.

But seeing that small states have bigger governments, loosely speaking, raises a question. If we divide convictions on corruption by relative size of government, would states like North Dakota still be on top of the heap? In other words, if we took the number of corruption cases divided by the number of people in position to be corrupt, would the numbers line up the same way as when convictions are divided by population? My guess is no. States like Illinois would probably be higher. Disagree? Then why is a state like Louisiana, which the article mentions as pretty darn corrupt, conveniently located with a government to population ratio close to the smaller states?

If anyone has time to kill, a more detailed analysis (including all government employees, not just state; doign all states; etc.) would be fun to see. Of course states like Illinois and New York would probably end up on top, so in the end it woudln't be news.