Wednesday, February 25, 2004

DEBUNKING THE MYTH: Ralph Nader didn't cost Al Gore the presidency in 2000

On Jan. 31, I posted an opinion piece about whether or not Ralph Nader should consider a run in 2004: ["Let Ralph decide"]. In the analysis, I made the case, that through first hand experience on the ground in New Hampshire running the field organization and extensive exit polling, Nader did not cost Gore the state. I also posted some data from Florida challenging the assertion that Nader's candidacy may have cost Gore that state.
Since the Jan. 31 post, Nader has announced that he will run for president again and the media and Democrats are attacking him. Smartly, Nader has used some of the information on this site to defend himself and his effort in 2000. And there is a debate raging on the Web as to whether or not this information is accurate or not.
Earlier today, Alex Beam wrote an interesting column in the Boston Globe challenging the mantra and pointing out that Pat Buchanan's candidacy may have weakened George W. Bush's position in some states: ["Ralph Nader: the odd man out again?"]. Beam only mentions Florida - with Buchanan's 17,000 votes that may have gone to Bush had "Pitchfork" Pat not been in the race. Beam doesn't mention, however, that Buchanan's numbers may have cost Bush the states of Iowa, New Mexico, Oregon, and Wisconsin.
In turn, the Boston Phoenix's Dan Kennedy, on his MediaLog, posted the article and then was critical of the data: ["Measuring the Buchanan effect"]. Kennedy does some "quickie analysis" and says Beam gets it wrong, and points to some NYT "news analysis" by David Rosenbaum: ["Relax, Nader Tells Democrats, but the Math Says Otherwise"].
In his "analysis," Rosenbaum uses national exit poll data to make his assumptions:
But based on who voted for him four years ago, [Nader's] analysis looks shaky. Voters leaving polling places in 2000 were asked by Voter News Service, a consortium of television networks and The Associated Press, how they would have voted if George W. Bush and Al Gore had been the only candidates on the ballot. Among Nader voters, 45 percent said they would have voted for Mr. Gore, 27 percent said they would have voted for Mr. Bush, and the rest said they would not have voted.
Well, Nader's analysis - which is also mine - isn't "shaky" at all, as I will show in this post.
Rosenbaum continues:
In Florida, Mr. Nader received 97,488 votes, 1.6 percent of the total, and Mr. Bush carried the state by 537 votes. In New Hampshire, Mr. Nader won 22,198 votes, 3.9 percent of the total, and Mr. Bush carried the state by 7,211 votes. Had Mr. Gore won in either state, he would have become president. Mr. Nader said at the Press Club that surveys of voters leaving the polls showed he had received more Republican votes than Democratic votes in New Hampshire in 2000. That is true. New Hampshire has 30 percent more registered Republicans than registered Democrats.
Rosenbaum is right on the New Hampshire registration, which in 2000 broke down this way: Republican - 265,679, Democrats - 197,816, and Undeclared - 274,927. But that has nothing to do with the point about whether Nader being in the race cost Gore. Yes, there are more Republicans in New Hampshire, and yeah, independents trend conservative, so why would they vote for Gore? They wouldn't and Rosenbaum is making my case without knowing it. Voter registration in the state has nothing to do with whether Nader cost Gore the state.
However, if reporters are going to figure with figures, they should use the correct figures, right? Well, Rosenbaum isn't using the correct figures. He is comparing national exit poll data and cross-referencing them with state ballot returns. He should have used individual state exit polls from both New Hampshire and Florida and cross-referenced them with returns from New Hampshire and Florida. Had Rosenbaum done this, he would see that Nader didn't cost Gore these states and his "news analysis" wouldn't achieve its outcome - to attack Nader.
If Rosenbaum had looked at the exit polling from both of these states, here is what he would have found:

In New Hampshire, CNN's exit polling showed Nader taking more votes from Republicans than Democrats by a two-to-one margin. Nader received 2 percent of the Republican vote and 1 percent of the Democrat vote. Nader also took 7 percent of the independent vote.
At the same time, 6 percent of registered Democrats voted for Bush! This is the real reason why Gore lost: He couldn't hold his own base! The Democrats never want to talk about this. They never want to talk about their negativity or the lousy campaign Gore ran. It's all, 'It's Nader's fault, it's Nader's ego,' ad nauseum. But back to the exit polling.
On the ideological front, 7 percent of Nader's vote identified themselves as "liberal," while 4 percent called themselves "moderate" and 2 percent "conservatives." Again, 7 to 6, pretty even and the votes could have gone either way.
When asked who they voted for in 1996, 3 percent of Nader's voters said they voted for Bill Clinton, 1 percent said they voted for Bob Dole, and 9 percent said they voted for Ross Perot. Seven percent said they did not vote in 1996. Here are how the numbers were charted:

No vote---12------50-------44---------0-------------7

As everyone knows, in study after study, Perot voters trend conservative. Here in this exit poll, by a more than 2 to 1 margin, Perot in New Hampshire voters went with Bush. So it is safe to say that at least 66 percent - if not more - of Nader's 7 percent of Perot voters, would have gone to Bush. Combine that with the 1 percent from Dole and it is an even split with the previous Clinton voters which most certainly would have gone to Gore. With Bush beating Gore by more than 7,000 votes though, and then splitting the Nader vote, Gore would still lose. Also note: 17 percent of Clinton's vote went to Bush! Gore couldn't hold the previous administration's support.
However, in a two-way race, CNN showed 3 percent of Naders voters said they would have voted for Gore, with 2 percent going to Bush, the same as Rosenbaum stated in his article. But even if the Nader vote was split 60-40, there is still no gain for Gore:

Here are the official New Hampshire results:
Bush: 274,290 - 48 percent
Gore: 266,121 - 47 percent
Nader: 22,188 - 4 percent

Results had the split been 60-40:
Bush: 283,165
Gore: 279,433

Again, even Rosenbaum's figuring with the figures doesn't add up to a Gore victory. As well, in a hypothetical two-way race, there would be no Buchanan candidacy, awarding thousands more votes to Bush.
The same chart, showing a two-way race, the results would have been the same - Bush 48 percent, Gore 47 percent, with 4 percent not voting. Somehow, Rosenbaum forgot to mention that! Again, a Bush win.
Now, this is just CNN and we know that exit polls aren't always accurate. So what do the polling experts say about the New Hampshire vote in 2000?
On Nov. 9, 2000 in a Nashua Telegraph article by David Brooks entitled "Nader probably didn't rain on Gore's N.H. parade," three different political experts challenged the notion that Nader cost Gore the election.
"You can't say, 'Oh, darn, (Gore) would have won otherwise,'" said Dick Bennett of American Research Group, a Manchester-based polling firm. Bennett said the evidence indicates Nader's vote was a solid core that wouldn't have moved to the Democrats. Dartmouth College government professor Dean Spiliotes agreed that no blame should be parceled out.
"I think it's really hard to know, without some kind of systematic means of testing, how people would have voted in a different set of circumstances," he said. "A lot of these people might have not voted at all, if not for Nader."
Spiliotes also doubts the stereotype that Nader voters would naturally turn to Gore as their second choice.
"I know a number of people in this community who voted for Nader, and ideologically these folks were all over the place. Folks I would have pegged for Bush voted for Nader," he said.
Robin Marra, a professor of political science who runs the polling institute at Franklin Pierce College, pointed to two polls done by his group before the election. They showed that about half of Nader supporters had voted for U.S. Sen. John McCain, R-Ariz., in the presidential primary.
"The conventional wisdom is that a vote for Nader helped Bush, but that may not be true in New Hampshire," he said. "Maybe some of them were just anti-establishment ... and if they saw Gore as establishment, they wouldn't have backed him."
Bennett said he thinks Nader's tally of 22,156 represented a core group of support that would have been unlikely to support another candidate - an opinion he reached partly because it appears a lot of Nader backers did switch to Gore at the last minute.
"When we polled in September we learned that the Nader voters were generally strategic voters - they would vote for Nader if it wouldn't hurt Gore. The minute they thought it would (hurt him), they changed," he said.
Three experts, some switching to Gore late in the race, but no blame for Nader.
But here is another example: On Nov. 10, 2000 in an Associated Press article published in the Foster's Daily Democrat entitled "Exit polls show Nader wasn't a spoiler in N.H.," the same figures were discussed:
In a New Hampshire exit poll, neither Gore nor Bush got a boost when voters were asked whom they would've chosen in a two-way race. Four percent said they wouldn't have voted at all. While there were not enough Nader voters interviewed in New Hampshire to draw firm conclusions about their effect, national exit polls show 46 percent of Nader voters would have chosen Gore in a two-way race. Twenty-one percent said they would have voted for Bush, but 30 percent said they would not have voted at all. If New Hampshire's Nader supporters matched the national trend, it appears unlikely Gore would have won. When those who would have stayed home or voted to Bush are taken into account, the net gain for Gore would have been only one in four votes.
Again, the state polls show no boost for either candidate. But again, using the national exit polls, with 46 percent of the Nader vote going to Gore and 21 going to Bush, here is what the returns would have looked like:

Bush: 278,949
Gore: 276,327

Here are the figures based on Rosenbaum's 45/27 split:

Bush: 274,290 + 5,991 = 280,281
Gore: 266,121 + 9,976 = 276,097

In other words, Gore would still lose New Hampshire with either of the mathematical formulas.

But back to the AP story:
Clark Hubbard, a political science professor at the University of New Hampshire, said Nader almost certainly hurt Gore, but how much remains unclear.
"The perceived closeness of the race pushed a lot of partisans back to the fold," he said. "All of the true partisans ended up voting for their party, it was that last-minute voting booth guilt."
Hubbard says the same thing as Bennett in the other article and his comments are a pretty safe bet. The Nader vote literally collapsed on Election Day, dropping for highs of 5 to 6 percent nationally, to the eventually 2.8 percent. In New Hampshire, Nader was as high as 8 percent at one point and ended up with a little more than 4 percent.

I was also quoted in the same article:
"We really drew on the McCain factor. We had a lot of progressives who've stayed out of the process. The majority of Democrats I talked to were never going to vote for Gore ... The Democrats can be angry at us all they want. But they ran a filthy, disgusting campaign. They spent millions and millions of dollars scaring and stealing votes from us. Gore lost by slim margins all over the place. What that says is we didn't cost them the vote, this is a divided nation."
Kathy Sullivan, the chairman of the New Hampshire Demoratic Party, said this in the story:
"All I can say to people who voted for Ralph Nader is that if George Bush is sworn in, the day they start drilling in the Alaskan wildlife refuge, it's your fault. And when the Supreme Court overrules Roe v. Wade ... remember you voted for Ralph Nader."
Well Kathy, yes, some of us voted for Ralph Nader in 2000 and it is almost four years later and ANWR is just fine [with senate Democrats voting for the drilling] and abortion is still legal.
In fact, John Kerry, the presumptive Democratic nominee, is telling James Hoffa over at the Teamsters that he is ready to drill, drill, drill! If Kerry wins, and the drilling starts, we'll remember that you voted for John Kerry.

Now, on to Florida.

As I said in the Jan. 31 post, this is a bit trickier because of the large number of votes Nader received [more than 97,000] and the slim margin of votes which resulted in the state being "stolen" from Gore [under 600 votes]. It is a safe guess - using the national exit poll model - that Gore would have received more than enough votes to have kept the election from being sent to the Supreme Court, who selected Bush.

However, that's the national exit polling. The Florida exit poll data shows completely different results:

In Florida, CNN's exit polling showed Nader taking the same amount of votes from both Republicans and Democrats: 1 percent. Nader also took 4 percent of the independent vote. At the same time, 13 percent of registered Democrats voted for Bush! Again, Gore couldn't hold his own base and because of this, he lost. The Democrats don't say one word about the fact that 13 percent of their own party members voted for Bush.
On the ideological front, 3 percent of Nader's vote identified themselves as "liberal," while 2 percent called themselves "moderate" and 1 percent called themselves "conservatives." An even split: 6 to 6.
When asked who they voted for in 1996, 1 percent of Nader's voters said they voted for Bill Clinton, 1 percent said they voted for Bob Dole, and 10 percent said they voted for Ross Perot. Here is the chart:

No vote---12------50-------44---------0-------------7

Again, Perot voters trend conservative. In fact, by a 3 to 1 margin, Perot voters in Florida went with Bush. So, with Nader taking equally from voters who cast votes for Clinton as they did from Dole, and then 10 percent previously voting for Perot being split on a 3 to 1 margin to Bush, that shows that if Nader had not been in the race, the majority of those voters would have gone to Bush, by a 7 to 4 margin. Also note: 16 percent of Clinton's vote went to Bush! Again, Gore couldn't hold the previous administration's support.
In a two-way race, CNN showed the results as Bush 49 percent, Gore 47 percent, with 2 percent not voting. And again, with no Buchanan, Bush gains thousands of votes. So, an even split of Nader voters offers no resolve to the matter of the state being thrown to the Supreme Court, with or without Nader.
Now, if you use the national exit polling data, yes, Gore wins handily. But it just doesn't work that way. Each state is different. Both New Hampshire and Florida have large sections of independents who are conservative but not Bush conservatives. Many of them might vote for Nader but wouldn't vote for Gore. Both New Hampshire and Florida are not like Oregon and Wisconsin. From personal experience, having lived in this state off and on for most of my life, I can tell you that independent voters in New Hampshire swing conservative. They might throw votes to Clinton; they might throw votes to Nader; but that doesn't mean they would go to Gore.
On the flip side, Florida is different. There is no doubt in my mind that 600 or more Nader voters from Monroe or Gainesville counties could have gone to Gore. These two counties have pockets of liberals - environmentalists in Monroe County, around the Everglades, and thousands of college students at University of Florida in Gainesville. According to the New York Times, which had a green map of Florida right after the election, Nader's strongest support was in those two counties. But the numbers from CNN don't show this at all so ...
Then again, as talked about in this great article by Jim DeFede, then with the Miami New Times, there was a lot of animosity for centrist Democrats and Gore in Monroe County over the Homestead Air Force base conversion to a multi-use airport: ["Collision Course"]. Again, when the Democrats abandon their base - in this case, liberals and environmentalists - they lose, especially when voters have other choices.

I contacted Kennedy earlier today and gave him some of the data which he posted on his site: ["Buchanan and Nader: a contrarian view"] which he later pooh-poohed, suggesting that because I was a "devoted Nader supporter" he's "not buying his overall thesis." Kennedy then again points to Rosenbaum's flawed analysis.
Now, I like Kennedy a lot. I try to read his Media Log a couple of times a week. While he is a great media critic, he falls a little short in the political analysis department, ignoring the overwhelming exit poll evidence showing that Nader didn't cost Gore the election - and registered Democrats hurt their own cause.