Can't Get It Right When He Tries

Last week, Paul Krugman mentioned in passing that various recount studies showed that Gore "won" Florida in 2000, prompting howls in the blogosphere.

Yesterday, due to overwhelming reader demand, Krugman dedicated a column to justifying his intital assertion.  And he still can't get it right, as Patterico points out.

Let's clip the highlights.  From Krugman:

The 2000 election is still an open sore on the body politic. That was clear from the outraged reaction to my mention last week of what would have happened with a full statewide manual recount of Florida.

This reaction seems to confuse three questions. One is what would have happened if the U.S. Supreme Court hadn't intervened; the answer is that unless the judge overseeing the recount had revised his order (which is a possibility), George W. Bush would still have been declared the winner.

The second is what would have happened if there had been a full, statewide manual recount - as there should have been. The probable answer is that Al Gore would have won, by a tiny margin.

The third is what would have happened if the intentions of the voters hadn't been frustrated by butterfly ballots, felon purges and more; the answer is that Mr. Gore would have won by a much larger margin.

About the evidence regarding a manual recount: in April 2001 a media consortium led by The Miami Herald assessed how various recounts of "undervotes," which did not register at all, would have affected the outcome. Two out of three hypothetical statewide counts would have given the election to Mr. Gore. The third involved a standard that would have discarded some ballots on which the intended vote was clear. Since Florida law seemed to require counting such ballots, this standard almost certainly wouldn't have been used in a statewide recount.

The Herald group later did an analysis of "overvotes," in which more than one choice was recorded, but this wasn't a true recount, because some of it was based on computer records rather than the ballots themselves.

All clear?  Krugman refers quite specifically to the "undervote" standard when he says that "Two out of three hypothetical statewide counts would have given the election to Mr. Gore".

Now let's flip to the consortium:

George W. Bush would have won a hand count of Florida's   disputed ballots if the standard advocated by Al Gore had been used, the first   full study of the ballots reveals. Bush would have won by 1,665 votes —   more than triple his official 537-vote margin — if every dimple, hanging   chad and mark on the ballots had been counted as votes, a USA TODAY/Miami   Herald/Knight Ridder study shows. The study is the first comprehensive review   of the 61,195 "undervote" ballots that were at the center of Florida's disputed   presidential election.

...   USA TODAY, The Miami Herald and Knight Ridder newspapers hired the national   accounting firm BDO Seidman to examine undervote ballots in Florida's 67 counties.   The accountants provided a report on what they found on each of the ballots.

The newspapers then applied the accounting firm's findings   to four standards used in Florida and elsewhere to determine when an undervote   ballot becomes a legal vote. By three of the standards, Bush holds the lead.   The fourth standard gives Gore a razor-thin win.

I hope that one of Krugman's acolytes can clarify this for us, if only to spare us a third column rehashing the 2000 election.

As to "overvotes", here is what the consortium said, but nothing there seems to tie in to Krugman's notion that two out of three ain't bad for Gore.

I wonder what Krugman was reading as he researched his book review.  Given the tone of his column, he actually does seem to be trying to set the record straight, and he does acknowledge in his intro that the Supreme Court did not halt a process that was headed towards a Gore victory.

However, given his evident confusion about the basic facts reported by the Miami Herald consortium, this bit of his latest column seems especially poignant:

So why do so many people believe the Bush win was rock solid?

One answer is that many editorials and op-ed articles have claimed that no possible recount would have changed the outcome. Let's be charitable and assume that those who write such things are victims of the echo chamber, and believe that what everyone they talk to says must be true.

The other answer is that many though not all reports of the results of the ballot reviews conveyed a false impression about what those reviews said.

I don't know how many people would describe the Bush win as "rock solid", but perhaps Krugman could turn his attention to sorting out why so many think that Scalia gave the election to Bush.

Comments

[Bob]

What would have happened if the phony felon list didn't knock tens of thousands of black voters off the rolls? Or if the Republicans weren't allowed to threaten potential black voters to further suppress the Demo vote?

[cathyf]

What would have happened if CNN hadn't called Florida for Gore while the (heavily-Bush) central-time-zone counties in the panhandle were still open?

Oh, and the name confusion with the felons list excluded people who had the same name as a felon but were not felons, while allowing the felons to vote. Some knowledge of polling data and some elementary mathematics shows that that if a) convicted felons were more likely to vote for Gore, and b) people who were not felons were equally likely to vote for Bush or Gore, then screwing up the felon lists helped Gore, not Bush.

Oh, and cantcha get the conspiracy theories right? I thought it was O'Connor who handed the presidency to Bush...

cathy :-)

[kim]

Bob, that is so much woulda coulda shoulda. Krugman speaks to the consortium findings and misinterprets them. There are any number of ways to claim victory for either candidate there. The fact of the matter is that the democratic will was so evenly split on that day in Florida that our processes for determining that will could not distinguish the majority. That would have demanded accuracy greater than the machine could produce.

And yes, you'll find that Republicans support measures to improve the voting process, from paper trails to verifiable voters, to enfranchising everyone legal.

And the Suprreme Court vote was 7-2.
=========================================

[Dwilkers]

I am still amazed that 7 million people could vote in Florida and it come out essentially dead even. I wonder what the odds are on that?

[kim]

It was a bizarre and unique event in social history. And of course, it's importance was magnified by it being the essential tiebreaker nationwide, and it will enter myth because of all the ambiguity in it.

But my first thought D, was, 50-50, even-up.
================================================

[James]

Cathy is correct. For starters there were only about 1,000 people improperly removed from the list, not tens of thousands. There were over 40,000 felons who were correctly on the list, but twenty counties just ignored the list, allowing thousands of pro-Gore felons to vote. This was a net loss for Bush, not Gore.

And those reports of black voter intimidation were specious. They were complaints such as a police officer voting in uniform, like police should not be allowed to vote, or blacks will start running from polling places screaming in fright if they see a cop.

[Patrick R. Sullivan]

'Given the tone of his column, he actually does seem to be trying to set the record straight...'

No, he's just trying to weasel out of admitting he was wrong. As usual. He stated flatly that the book documented the simple truth that Gore won in 2000. And he concluded the first column with:

'The Republicans will be strongly tempted to make sure that they win those elections by any means necessary. And everything we've seen suggests that they will give in to that temptation.'

Which is pretty silly when you realize that all the election day problems come from counties controlled by Democrats. And Krugman conveniently ignored what happened in Washington state last fall. Democrat Christine Gregoire won a 129 vote victory only after a re-re-count. A judge ruled in the ensuing lawsuit that King County (78% Gregoire) had allowed almost 1,700 illegal votes to be cast. However, he also ruled that he couldn't tell who those 1700 would have voted for, and refused to invalidate the election of the Democrat.

[Patrick R. Sullivan]

'Given the tone of his column, he actually does seem to be trying to set the record straight...'

No, he's just trying to weasel out of admitting he was wrong. As usual. He stated flatly that the book documented the simple truth that Gore won in 2000. And he concluded the first column with:

'The Republicans will be strongly tempted to make sure that they win those elections by any means necessary. And everything we've seen suggests that they will give in to that temptation.'

Which is pretty silly when you realize that all the election day problems come from counties controlled by Democrats. And Krugman conveniently ignored what happened in Washington state last fall. Democrat Christine Gregoire won a 129 vote victory only after a re-re-count. A judge ruled in the ensuing lawsuit that King County (78% Gregoire) had allowed almost 1,700 illegal votes to be cast. However, he also ruled that he couldn't tell who those 1700 would have voted for, and refused to invalidate the election of the Democrat.

[oceanguy]

Most likely, the deciding factor in the 2000 election was NOT undervotes, but OVER-votes. In Jacksonville, over 20,000 ballots were thrown out due to over-voting.

The problem was the number of presidential candidates. In south Florida, they tackled the problem with the infamous Butterfly Ballot, here in northeast Florida they simply used two pages. The majority of those 20,000 over votes came from three precincts, almost entirely African-American and likely Gore voters. Poll workers were cheerfully instructing voters to "Make sure you vote for someone on each page!" Apparently many did.

The only way Gore could have had a chance to win a Florida re-count was to have a do-over election. Every recount of actual ballots showed Bush winning. Only projections and assumptions gave the nod to Gore. Those who complain about the results are probably correct in thinking that the “intention” of the majority of Florida’s voters was to vote for Gore… it’s just that thousands of likely Gore voters were not sharp enough to vote for only one presidential candidate.

[James]

Maybe Democrats should get an election handicap to account for the competence of their voters? Let's see if Bush can beat the spread!

[Greg D]

And what would have happened if the three networks hadn't repeatedly misinformed people, claiming that the polls were closed in Florida while they were still open in the (Bush friendly) panhandle? This dispite the fact that Kathleen Haris had sent them all notes teh previous week reminding them that polls in the Panhandle would be open for an hour later than those in the rest of the state?

What would have happened if Clinton had not sent thousands of troops based in Florida off on "exercises" that were called at the last minute, and made them unable to vote?

For that matter, how DID Gore "close the gap" at the end? Here's the numbers I saved from the AP at the time:

• 95% reporting
• George W. Bush, GOP 2,736,318 - 49%
• Al Gore, Dem 2,672,793 - 48%
• Ralph Nader, Grn 92,202 - 2%
• Pat Buchanan, RP 19,026 - 0%
• Harry Browne, Lib 18,162 - 0%

That's Bush up 63,625

• 99% reporting
• George W. Bush, GOP 2,906,997 - 49%
• Al Gore, Dem 2,904,519 - 49%
• Ralph Nader, Grn 96,843 - 2%
• Pat Buchanan, RP 20,284 - 0%
• Harry Browne, Lib 18,886 - 0%

Which is to say that Bush got 170,679 votes in the final pickup, while Gore got 231,726.

To put it another way, of the votes counted in the final submissions (and recall that the polling booths closed an hour later in the pro-Bush Panhandle) 58% of the votes that that went to one of the two major party candidates went to Gore.

In a state that supposedly went 50-50.

If you want to talk about "vote fraud", the place to start is with the impressive number of "votes" that Gore collected at the last minute.

That is, that's the place to start if you're honest, instead of a partisan hack.

[Greg D]

Oh, and if you're going to talk about an honest revote, let's not forget all the snowbirds who voted twice, once in Florida and once in their other state. None of them should be allowed to vote in that revote.

[Leon]

Just to complicate matters further. What happened to the military absentee ballots that were thrown out due to the (documented) effort by the Democratic party?

[Dr. Weevil]

Leon: This is from memory, but as I recall, the Gore campaign used hundreds of lawyers to sue in every county to keep all military absentee ballots from being counted. They lost nearly all of those suits, but there were still hundreds of military ballots in litigation when the election was decided in Bush's favor, and the remaining suits were then declared moot. If "recounting" had continued, those lawsuits would have continued, and Bush would almost certainly have picked up more votes from them, offsetting any votes gained by Gore from "undervotes" or "overvotes" or chadly divination.

[DennisThePeasant]

The 2000 election is an open sore for the body politic.

Right.

But only for those whiling away the hours waiting for their visa application to be processed and the results of their Canadian citizenship test to be mailed to them.

[richard mcenroe]

Has anybody actually seen an American parlor pink set up house in Canada yet? Or New Zealand, the other sexy spot?