Why Nader Won’t Matter: A Political Scientist’s Perspective

[JDPhD]

First, a disclaimer: I am a political scientist. This means that, although I am a lifelong Democrat, I probably view the ongoing political struggle with less passion than the rest of you. So, if you read something you don’t like in what follows, don’t assume that I am somehow supporting Bush. I am just calling things as I see them. Now, about the impact of Ralph Nader….

Last year I sat in on a dissertation defense for a Political Science PhD candidate at Stanford University. (That candidate now, PhD in hand, teaches at Harvard.) The candidate’s dissertation dealt with voters’ changing choices over the course of a political campaign. In particular, she was interested in what she called “mismatched voters”—liberals who were supporting Bush, or conservatives supporting Gore, as examples. Her central conclusion was that mismatched voters almost always find there way home to the party where they belong. In the 2000 campaign, virtually every voter who started out mismatched figured out their mistake by election day and voted for their “right” candidate. I’ll share a couple other interesting things about her study after I get to the point about Ralph Nader.

In her study, the PhD candidate tracked over 20,000 voters for more than a 6 month period. She discovered that virtually every single voter who “should” have been voting for Gore—based on their approval of the Clinton/Gore administration and/or their agreement with Gore’s issue postions—who had initially supported Nader, ended up voting for Gore on election day. I’ll say it again—virtually every single voter who “should” have been voting for Gore, but who initially supported Nader, ended up voting for Gore on election day. The people who did vote for Nader, almost without exception, would not—I repeat, would NOT—have voted for Gore! Nader voters rejected both major political parties. Only a very tiny percentage of eventual Nader voters were actually stolen from Gore. This was a stunning discovery. In the big picture sense, this means that Nader did not cost Gore any votes—or at least very few votes. Now, Nader still could have made a significant difference in the 2000 election—but only because the margin in Florida was so excruciatingly small.

Based on the results of this PhD candidate’s dissertation study, it is likely that Nader will steal almost no votes from the Democratic nominee in the general election. Thus, unless some state is again decided by only a couple hundred votes, Nader’s presence will have no substantive impact on the results of the 2004 election. Nader will attract some voters who are disaffected with both major parties, but he will not steal Democratic votes. Even Democrats who initially support Nader will find their way home by election day.

Now, a couple other interesting things from her study.

First, it was not Nader who cost Gore the most votes in 2000. It was Clinton! There was only one group of “mismatched” voters in 2000 who failed to vote for the “right” candidate on election day. Liberals who were disgusted by Clinton’s personal behavior voted for Bush, though they agreed with Gore on most issues. During her dissertation defense, the PhD candidate described this as the “Clinton anchor”, because its impact on Gore’s candidacy was so dramatic. If these voters had correctly matched up with Gore, he would have won the election easily. From her study, it appears that Gore was right to do all that he could to run away from Clinton.

Second, the source of data for this study is fascinating. This PhD candidate got her data from “Knowledge Networks”. Knowledge Networks is a company that was started by a group of Stanford professors, essentially as a really slick way to get social science data. After securing financial backing from the private sector, these professors gave away thousands of TiVos to a sample of people from across America. In exchange for the free TiVo, each of these people had to agree to immediately and honestly answer any questionnaires sent to them over the interactive system. Most of the time these surveys deal with products—like “Do you use Palmolive dish soap?” This is how Knowledge Networks pays for the system. But, every now and then, these professors throw in one of their social science surveys. As a result, they not only know the political leanings of the people in their sample, but they can track their attitudes over time. So, by asking policy preference questions, the PhD candidate was able to determine whether each person in the study “should” have been a Democratic voter, Republican voter, or Green voter, etc. She could then compare who these people should have been supporting with who they did support. And she could track changes in their preferences as the campaigns progressed. This is how she could say that almost all mismatched voters found their way back to their correct party by election day—except for those towing the “Clinton anchor”. This is also how she could discover that those who ended up voting for Nader would not have voted for Gore anyway.

In case any of you are wondering where these results have been published—they haven’t, yet. She is working on putting them out as part of a book.

I am now reminded of a couple of other interesting bits of information from the last election that can help us evaluate the potential impact of Nader in 2004—but they are NOT from the study I discuss above.

First, remember the forgotten impact of Pat Buchanon. Many people focus on the two states that Bush won that were within the margin of the Nader vote—Florida and New Hampshire. They assume that if Nader had not been in the race, Nader votes would have gone to Gore and Gore would have won those two states, and thus won the election. The study discussed above makes this line of thought questionable, but even if it is true, we should also consider what may have happened if Buchanon had not been in the race—just to be fair. Most people don’t even think about Buchanon’s impact, but they should. If we assume that votes for Buchanon would have gone to Bush without Buchanon in the race, then it appears that Bush would have also won Iowa, New Mexico, Oregon, and Wisconsin in 2000, instead of Gore (assuming Nader still ran, but not Buchanon). Gore’s margin of victory in each of these states was less than the number of votes cast for Buchanon. And, without Buchanon in the race, Bush may have won Florida by thousands of votes. Now, I don’t personally believe that Buchanon voters would have automatically gone for Bush—anymore than I believe that Nader voters would automatically go for Gore. I am just trying to point out that if we are going to get all upset about the impact of third party candidates, we should admit the impact that they have an impact on both major parties—not just our own. I think it is better to not get upset about them at all, or over estimate their impact—either way.

Second, remember that Joe Lieberman will not be on the ticket this time around. Lieberman practically lived in the Jewish retirement communities in Florida for months prior to the 2000 vote. Gore handled the rest of the country while Lieberman worked Florida, day in, day out. At the time it was joked that Lieberman would have to run for reelection to the Senate from Florida—because he now lived there. Without Lieberman’s efforts, Florida would have probably been won handily by Bush, and Nader’s impact—whatever it was—would not have mattered. Just look at the margin in the Florida legislature, and at Jeb Bush’s election numbers. Florida is clearly a majority Republican state. Unless Graham is the VP nominee, I doubt Florida will be in play in 2004—and the Nader factor won’t be factor, even there.

So, in sum, everybody just calm down. In all likelihood, Nader will make no difference in 2004. The Democratic nominee will win or lose for entirely different reasons.

Finally, before anyone accuses me of abandoning this thread, I must now head off to class. I’ll check back later today to possibly reply to any comments.

[Jackpine Radical]

1. Thank you for a fascinating report.

The thing that strikes me with the most impact is that Knowledge Networks seems to be a potentially very effective workaround for the sort of sampling problems that are increasingly vitiating the effectiveness of conventional polling methods.

[billhos]

2. Very good

This is something that really makes this forum interesting. I hope too see more postings from you.

[newyawker99]

23. Hi billhos!!

Welcome to DU!! :toast:

[Florida_Geek]

3. I take an exception to that

in Florida I was emailed from a group of famales called Nader Raiders, and one my neighbor after the election said, "I did not know my vote really counted". She would have vote Gore over Bush and said most of her friends would have too.

How many of Florida's Naders Raiders would have voted Gore just to stop Bush from getting elected?????

[Nicholas D Wolfwood]

4. Forgive me for saying this,

But it's total and complete bullshit to suggest Nader didn't cost Gore the election when only a hundred votes' difference in Florida would've put him in the White House. There is no way in hell statistically that out of the some 100,000 votes Nader got in Florida that a couple of hundred of them would absolutely not have voted for Gore instead.

Edit: I'll also note this coming election is expected to be just as close as 2000. A few hundred votes WILL make a substantial difference.

[mdmc]

14. Nader voters would have stayed home

Even the one down in Florida.

These people chose to vote for reckless Nader over Gore. They made that choice.

If Nader didn't run, they would have made the reckless choice of a write in or just stayed home.

Fear may have "straightened" some lefties out, but the only remedy for Naderitis is to reform the Dem party.

[AP]

26. Many people might have tuned out the entire election if Nader had not

been running.

There's no way to predict who would have showed up to vote. Many might have been bored to death by the canididates.

Wandering to Nader and back to Gore might have been because of Nader keeping things interesting. Without Nader they might never have had a doorway into caring about 2000.

[Nicholas D Wolfwood]

30. It's very, very easy to predict

That a couple of hundred people out of 100,000 would have voted Gore. Seriously, if 1.78% of the people who voted for Nader in Florida switched to Gore, Gore wins. You can't convince me at all that such a small percentage of obviously left leaning people wouldn't have voted for Gore if Nader wasn't there. It's not possible.

[AP]

31. 25% of all voters preferred Nader to Gore and Bush. 90% voted for Gore.

If They hadn't had Nader to listen to and to keep their attention focussed on the election, turnout might have been as low as ti was in 96. Instead, if memory serves, more people voted in 2000 than in any election ever, and trunout was up as a percentage of registered voters too.

Gore and Bush were trying to bore us to death. Without Nader, they might have succeeded.

[Iverson]

5. Gee, what a thoughtful contribution.

I'm glad you posted it. It was ... refreshing.

[Mattforclark]

6. About the Buchanan effect

I think you are assuming that Buchanan's votes would go to Bush, but are still counting Nader's run?

In Oregon, for example, Nader had about 5% and Buchanan had 1%. If Buchanan hadn't run and all his votes went to Bush, he would have barely beaten Gore, but if even 20-40% of Naders votes went to Gore, he would have won the state.

And in Florida, which is the whopper, if all of Nader's votes went to Gore and all of Buchanan's votes went to Bush, Gore would have won with a 1% margin.

Good post.

[Florida_Geek]

8. We will not go into all the Jewish WPB voters that voted

for Buchanan

[Mattforclark]

10. That's very true as well

[HFishbine]

7. Exit Poll Results Support Findings

2000 exit polls support the findings you report.

- If it had been a two-man race, 2% of voters would not have voted. Of that 2% who would not have voted if it had been a two man race, 31% voted for Nader, 27% for Bush, 23% for Gore and 9% for Buchanan.

This is the most telling statistic to me, and one that exonerates Nader:

- If it had been a two-man race, 2% of those who voted for Gore would have voted for Bush. (1% of Bush voters would have voted for Gore if it had been a two-man race.) This suggests that the presence of third party candidates actually inspired 2% of Gore voters to vote for Gore instead of Bush.

http://www.udel.edu/poscir/road/course/exitpollsindex.html

[AP]

33. This is an important point. Read this post.

[displacedtexan]

9. All dissertations are 'published' and available from the school library.

Yet you wrote:

"In case any of you are wondering where these results have been published—they haven’t, yet. She is working on putting them out as part of a book."

Anyone who has completed a doctorate knows this.

Give us the name of this Harvard instructor and the title of her dissertation at Stanford.

[snippy]

22. Good point.

[Jack Rabbit]

11. Response

(BA, Political Science, San Francisco State University, 1975)

This is a different election with different issues. It is probably fair to say that Nader will not be as much of an impact on this election as he had on the last one. One person who doesn't believe that he will have any great impact on the chances of the Democratic nominee is Nader himself. Here he responds to an open letter from the editors of The Nation urging that he not run:

The Nation's open letter does not go far enough in predicting where my votes would come from, beyond correctly inferring that there would be few liberal Democratic supporters. The out-of-power party always returns to the fold, while the in-power party sees its edges looking for alternatives. Much more than New Hampshire in 2000, where I received more Republican than Democratic votes, any candidacy would be directed toward Independents, Greens, third-party supporters, true progressives, and conservative and liberal Republicans, who are becoming furious with George W. Bush's policies, such as massive deficits, publicized corporate crimes, subsidies and pornography, civil liberties encroachments, sovereignty-suppressing trade agreements and outsourcing. And, of course, any candidacy would seek to do what we all must strive for-getting out more nonvoters who are now almost the majority of eligible voters.

Personally, I will feel more comfortable if Nader not run at all. The Democratic nominee is going to need every vote he can get and that means every vote Nader would peel off from a disaffected progressive.

However, there will be fewer disaffected progressives voting for Nader in 2004 than in 2000. I can offer myself as anecdotal evidence of such a trend. I cast my vote for Nader in 2000, but I will support the Democratic nominee in 2004. Among the factors leading to that vote was that I live and vote in California, where there never doubt that Gore would carry the state easily, and that I perceived Bush to be simply a dimmer version of the kind of obnoxious conservative that Ronald Reagan and the elder Bush were; we survived those administrations and would survive another like it. Today, I view Bush as a genuine threat to the health of American democratic institutions and as one whose ideology is alien to traditional American political thought.

While Senator Kerry or Senator Edwards may not be all a progressive could ask for in a president, defeating Bush is the overriding focus in 2004. As the leftists who made up the French Resistance were willing to take direction from General de Gaulle, politically a sober conservative, to rid their country of Nazi occupation, so should American progressives in 2004 be willing to follow a Democratic nominee whom they perceive to be less than perfect in order to rid America of Bush. President Kerry or President Edwards may disappoint us and get the country into a colonial war or negotiate a NAFTA-like trade pact to the detriment of working people at home and abroad. If so, we can hit the streets again, just as we have to protest Bush and his colonial wars. However, we would have less fear of having dissent met by being put on no-fly lists, or having federal agents visit our public librarian to ask what we've been reading and threaten to throw her in jail if she lets us know about it. We won't be reading about justice department officials drafting legislation that would allow the president or the attorney general to strip Americans of their citizenship.

Otherwise, your PhD candidate is correct in that the Nader factor in 2000 is overrated. Nader didn't cost Al Gore Florida in 2000; Katherine Harris did by purging the voter rolls before a single chad was hung. Other factors were Gore himself (he could have run a much better campaign) and, as your student points out, many people visited Clinton's personal sins on Gore; that may be unjust, but it happened.

Clinton's personal life will not be a factor this year; those who raise the issue are probably voting for Bush anyway. Al Gore isn't the candidate; the eventual nominee will rise or fall on his own strengths and weaknesses. And, most importantly, we know George W. Bush a lot better and we now know we cannot afford the luxury of casting a protest vote.

We can go back to bickering about the issues that divide progressives from centrists after this election. For now, we will do best to make common cause with each other.

[Florida_Geek]

12. ALSO

And in Florida Ralph Nader got 97, 431 votes. If only 1 % of the Nader voters had voted for Gore, it would be President Gore today.

[Dhalgren]

13. And if 1% of Republicans had stayed at home

Gore would have won Florida! Damn those damnedly mobile Republicans and their damned cars!

[Nadienne]

29. Florida

The recount was stopped, correct?

Is it possible that if all of Nader's Florida votes instead went to Gore, Gore still wouldn't've won Florida due to tossed-out votes and other dirty tricks?

The blame-Nader sentiment prevents us from focusing on what can be done to prevent more voting scandals...

[DemocratSinceBirth]

15. As Someone Who Has Done Post Grad Work In Political Science

I question the efficacy of the "matching " theory of voter choice because it doesn't allow for second choices. Surely, Al Gore was the second choice of more Nader voters than George Bush and if Nader wasn't in the race he would have gotten the lion's share of those votes.... I will allow for the possiblity or probability that many Nader voters would have stayed home....


Let's bring political science to a level that the average Joe can appreciate.. ...

Let's say I'm at a bar.... At nine o'clock I think I match up perfectly with a woman who looks like Anna Kournikova.... At 11 o'clock I'm lowering my expectations and I think I match up perfectly with a woman who looks like Helen Hunt..... By 2:00 o'clock I think I match up perfectly with anything with a heartbeat...


Peace

Brian

[littlejoe]

38. So, which one is Anna?

[TreasonousBastard]

16. Very interesting...

and goes along with the not often mentioned theorizing from some of the better observors that Nader voters would likely have stayed home and not voted at all should he not have run. Their disgust was with the system, and neither candidate appealed to them.

I suspected that Buchanan voters would have been more likely to vote anyway, and would have voted the Republican line.

I wasn't aware of the Lieberman effect in Florida, and that now explains how Gore got as much as he did down there. This does not bode well for November unless dissatisfaction with Bush exceeds expectations.

[YNGW]

17. Nader Forces Dem Nominees To Stake Their Claim NOW!!!!

Neither Kerry nor Edwards can afford to move further to the center or they alienate the left and drive them into the Nader camp. If Kerry or Edwards move further left to appease them, then they allienate the center. By Nader entering the race, he forces them to to stay put where they are. There's a certain element of voters who have already been lost to the Nader camp. There's a whole center up for grabs and two Dem candidates who can't shift closer to it without further alienating the left and appearing "political".

[mdmc]

18. Dems have to BUILD

they can't rely on Greens being scared of Bush.

This study reported on "lost voters" - this group believes in one thing but supports a candidate that doesn't agree with their issues. The conclusion that they made their way home should be obvious.

[bearfartinthewoods]

19. why assume buchannan voters would have voted pubbie in numbers

greater than nader voters for gore? if the nader voters never belonged to gore why assume the buchanan voters belonged to bush?

[IndianaGreen]

20. Nader is a fringe candidate

Nader's name will be in the ballot of a handful of states, at best.

Perhaps the hysterics about Nader have more to do with Kerry's own weaknesses as a nominee than with Nader himself.

[Nazgul35]

21. Dear God...

As someone who holds a PhD as well in political science, the last group of people who should be asked for advice is us....especially Ivy league schools!!!

The answer to every single question in political science is "it depends." As the science of human behavior is not an exact one. As a prominent teacher of mine says..."God gave the physics department all the easy questions."

As to this being a startling discovery, any one who has ever read Maurice Duverger and Gary Cox knows that it is possible for individuals to violate strategic voting if the honestly see no difference between the other two alternative candidates.

Rational choice has much to say about this subject. A key one would be that in 2000, many voters lacked information about the policies that Bush would enact (his Compassionate Conservatism), but as of 2004, those same voters would possess a considerable amount of data about this regime.

There are many levels of voters with regards to the amount of information they possess. The electorate is certainly more informed about Bush this time around.

There are much more important factors to consider as to why Nader will not be a threat this time around.

1) Nader does not have the organizational base (such that it is) of the Greens to get on most ballots...thereby removing him from consideration in most states.

2) Historically, third party runs always suffer a significant drop off in the follow up election...the same will happen here...

3) George Bush has a record that is clear and understandable to the electorate, should the eventual Dem candidate be able to exploit it. The signs suggest that as the electorate has been exposed to Bush's record, his support has eroded.

4) The incumbency factor is more important than anything, as the election will revolve around the performance of the President. Two questions will be asked by a significant portion of the voters:

a) Has George Bush done a good or bad job as president?
If Yes --> vote for Bush
If No --> see next question
b) Would the challenger be a better candidate?
If Yes --> Vote for challenger
If No --> anything is possible (don't vote, vote for a third party candidate, flip a coin, etc.)

People need to keep in mind that the vast majority of political science is so narrow and specific that it is useless to use it as any kind of predictor...this is not the "fault" of political science, but rather the subject matter that we cover...

Another point, as any poli-sci person would need to see is

1)what was the model used to test the hypotheses offered by this student?
2) How was it specified?
3) Was the correct sampling distribution used or was it the standard classic linear regression approach?
4) How did it reject alternative theories of voter behavior?

It should also be noted in closing that the area where we do the absolute worst as a science is in voting behavior....most of the science is based upon inferences that are heroic and polling data that contains multiple problems....it is hard to get at the real reason why voters behave the way they do and any study that tries should be looked at with a suspicious eye...

[snippy]

24. This raises several questions but they and the entire post are irrelevant.

This is not 2000 and Bush is an incumbent. Nearly every election involving an incumbent is a referendum on the incumbent. This election will be no different. It definitely will be a referendum on Bush. The three main issues will be the economy, Iraq, and Bush's incessant and compulsive lying. These issues likely will determine 10 to 15 million votes. Those 10 to 15 million votes will preclude another close election. Nader's vote totals will be much lower than in 2000 and insignificant. The people who will vote for Nader are people who would not vote for the candidate of either major party.

Since 9/11 Bush has been in a steady down trend with the occasional bump. Bush's decline in the polls is nothing other than the American people looking at and listening to Bush and realizing that he is an incompetent liar. Day by day and voter by voter, Bush himself is convincing the American people that he should not be president. If the trend since 9/11 continues, Bush will lose badly. If the trend reverses soon enough, Bush will win easily. Nader will not be a factor in any voters' perceptions of Bush.

The economy and Iraq must improve between now and November for Bush to win. Between now and November Bush will continue to lie at every opportunity about every subject on which he speaks. Between now and November, Bush will continue to demonstrate his incompetence. Between now and November Bush will continue to demonstrate his Fecal Midas Touch. Bush can benefit only from what happens to the economy and Iraq between now and November. Those issues, not Nader, may or may not change the trend for Bush and will decide the election. If there is no positive change on those issues, Bush will lose.

In 2004 Nader voters are not people who care about improving this country or the world. They care about trying to appear that they are more concerned than anyone else because of their enlightenment. Their vote is solely about self-gratification. They differ from Nader himself only in that Nader also is concerned with using his position to amass personal wealth.

[AP]

25. Another study confirmed the same thing:

There's another study by a professor at Harvard which said that 90% of the roughly 25% of voters who preferred Nader over the other three candidates ended up voting for someone else (and, for 95% of them, that was Gore). The data he relied on, however, didn't break down those national numbers by state, and I'm sure that the 10% who stayed with him were in NY and Cal.

[WiseMen]

27. What are the Validity concerns regarding conclusions from this study?

Isn't it premised on assumptions regarding the political dynamics
without Nader. How do we know how this would change Gore's or
Buchanan's campaign, for example.

[Meldread]

28. This is *EXACTLY* what I've been saying all along.

I found the "Clinton Anchor" part interesting, but the bit about Nader -- I've been saying that all along. I've said that the majority of people who voted for Nader would *NEVER* have voted for Gore, and in the end Gore lost very few votes to Nader.

I am interested though to know how many people would have voted for Bush had Nader not been in the race. I know that people voting Nader favored Gore more 2 to 1, but there is still room there for Gore to STILL have lost Florida regardless of Nader depending on how many voters would have voted for Bush. (It probably would have been by a smaller margin, but it's completely possible. The Clinton Anchor makes it even more possible.)

[AP]

32. The Clinton anchor: by avoiding Clinton, Gore affirmed the perception that

there was something about Clinton that was justifiably despicable.

I think there was a middle-ground. I think Gore could have embraced him in a way that made people who despised Clinton still appreciate that personal problems notwhithstanding, hatred for Clinton wasn't a good reason to vote for someone who would not look after your best interests.

Another thing worth noting: this study proves that the Republicans impeachment strategy worked. It won them the election in 2000. That's all they needed to do. They didn't need the impeachment. They just needed to wait two years to win an election there was no reason they should have won.

[NNN0LHI]

34. kick to bookmark for later reading n/t

[JPJones]

35. Sweeping conclusions based on 16 Florida voters

20,000 Tivos

Florida has about 5% of the nations population. About 1000 Tivos in Florida.

1.6% of people in Florida voted for Nader. So, about 16 Nader voters had the Tivo system.

How well do these 16 people represent the 97,000 people who voted for Nader in Florida?

To truly make any conclusions here, you would have to study many more Nader voters in Florida. That is the problem with using this system for essentially state politics. I have no doubt that it is valuable for soap and nacho chips. But political "scientists" should know better. The margin of error would be about 25% in this case.

[JPJones]

36. Sorry, 8 voters

I forgot that only about 1/2 of people vote.

[bullimiami]

37. Lieberman 2K

Lieberman was a big mistake in 2000.
The jewish communities in N Dade, Broward and Palm Beach are already extremely liberal and would have overwhelmingly voted democrat anyway.
Lieberman was preaching to the choir.
Also I think he was a negative on the ticket in the rest of the south. Prejiduce in play.
What was needed was a VP on the ticket that could have swayed some actual swing voters.

[GreenPartyVoter]

39. Thanks. Nice to hear a rational voice when it comes to Nader. *lol*

[GodHelpUsAll2]

40. Great post JD

I though it was great and well stated. But I somehow don't see the think tank accepting it.