From WNYC in New York, this is NPR's On the Media.

I'm Bob Garfield.

And I'm Brooke Gladstone.

If you were wondering how the Justice Department thought it could get away with firing eight federal prosecutors on the thinnest of pretexts, now you have an answer.

Eight is a lot fewer than 93.

And that, we now know, is how many the White House

believed should be canned. Emails documenting that recommendation surfaced this week,

resulting in the resignation of a top aide to Attorney General Alberto Gonzalez. The email

suggests that the so-called Gonzalez-Aid may have been fired for going too hard on Republicans in their districts or too soft on the Democrats.

Now, a few years ago, two communications professors, Donald Shields and John Cragan, had a hunch that something odd was afoot in justice, so they set out to catalog federal investigations

and indictments of some 375 elected officials. Cragan was struck by what he found.

What we found is that about seven to eight out of ten times they investigated a Democrat over a Republican.

And the statistical analysis says that the chances of that happening by random about 1 in 10,000. There are roughly 50% Democrats and 41%

Republicans and 9% independence nationally. So if they were doing their job, we should have found

an investigation rate of that same ratio.

Instead, we found 79% Democratic, about 17% Republican and the rest of the few percent's independent.

Now, just to be clear, all the entries in your database are investigations or indictments that were previously reported in the media, right?

Yes. What we looked at were posted news stories by journalists.

So it's possible that the number of investigations could be a lot higher since a lot of the original

stories were launched on leaks.

Exactly. I mean, that's our point, is that it's a sad state of affairs in this country when a couple unfunded university professors are having to gather this data.

And we can't even get data of who was convicted, you know, who is currently indicted,

...