Voter Fraud and DOJ Firings

Statisticians will tell you the most common error in reading statistics is to believe that because two things have a correlation must mean there is a causation. For example, suppose that all red-heads were left-handed, a fact we know to be false, but assume it is true. That is a correlation, but being red-headed does not cause left-handedness, nor the other way around. It is important to always ask the question, just because two things are related does not mean one is caused by the other.

The media, in particular, routinely messes this up. The Washington Post is carrying a front page story today about the DOJ firings, discussing that voter fraud is at the heart of the matter.

Of the 12 U.S. attorneys known to have been dismissed or considered for removal last year, five were identified by Rove or other administration officials as working in districts that were trouble spots for voter fraud -- Kansas City, Mo.; Milwaukee; New Mexico; Nevada; and Washington state. Four of the five prosecutors in those districts were dismissed.

It has been clear for months that the administration's eagerness to launch voter-fraud prosecutions played a role in some of the firings, but recent testimony, documents and interviews show the issue was more central than previously known. The new details include the names of additional prosecutors who were targeted and other districts that were of concern, as well as previously unknown information about the White House's role.

This is typical of the kind of spin the Post is putting out about the DOJ firings. Were the firings political in nature? Of that I have no doubt, but no real proof to substantiate my belief.

Interestingly, that is what makes me and the Post very much alike. The story stretches hard to prove causation where only a correlation exists. Yes, five U.S. attorneys who were fired worked in areas that the Administration considered trouble spots for voter fraud. Yes, the attorneys were dismissed. But does one necessarily implicate the other? Professor Dan Tokaji, whose expertise in voting rights and voter fraud is perhaps second to none notes

"There is reason for worry and suspicion at this point as to whether voting fraud played an inappropriate role in personnel decisions by the department,"

but that is far different from causation.

How much of this "scandal" is political? All of it. Did failure to prosecute cases of voter fraud have a role in the dcision? Probably, but isn't that political too?

Admittedly, the GOP has got voter fraud on the brain. But on the opposite side, the Democrats have voter intimidation on the brain. In reality, can one find instances, anecdotally, that each crime has happened. Of course, but that doesn't mean either is widespread. Yet, the media's obsession with this "scandal" is only made worse when charges of causation are generated out of mere correlation.

It may be possible that voter fraud issues did cause the firings, but we don't have any proof of that, merely guesses and suspicions.