The Higher the Vote Turnout, the Fewer Votes for Kerry!
Go figure!
By Brad Friedman on 11/10/2004, 4:54pm PT  

There's been much talk about the 97,489 more votes than voters cast in Cuyahoga County, Ohio. A plausible explanation for the huge difference between votes and actual voters, as reported via the Cuyahoga County website), is being attributed by County Officials to the strange method they have for including absentee ballots into the vote counts separately from the Election Day voters. Or something.

So setting that oddity aside for the moment, take a look at this interesting graph. I have not been able to confirm the specific numbers it shows, so I can't testify to it's accuracy, but I'm posting it here because, if accurate, this seems worth getting out to the world for further scrutiny.

Though it's difficult to read this small, the trend-lines are all that really matter here (The full sized graph is here).

If correct, this graph would seem to show that precinct-by-precinct in Cuyahoga County, Ohio --- where Cleveland is, and thus it's a heavily Democratic county --- the larger the turnout in each precinct, the smaller the percentage of the vote for Kerry!

The GREEN trend line is the size of voter turnout in each precinct, and the BLUE trend line is the percentage of votes for Kerry.

While conventional wisdom is often wrong, if the above numbers are accurate, it does seem a bit odd to say the least that the higher the turnout in a strongly Democratic county (Kerry won Cuyahoga 63%-33% if I recall correctly), would result in a smaller percentage of votes for the Democrat!

I don't know what to make of it, or how to explain it, but it seems worth putting out there for the armchair investigators.

UPDATE: I was asked what the balloting/counting methods are in Cuyahoga County. Apparently punch card, as counted by machines made by our old friends at ...Election Systems and Software, Inc. (ES&S). Source