A puzzle about the polls

by Harry on November 3, 2008

No doubt if I had more time I’d find the right answer to this question, but I’m lazy and/or pressed for time, so here goes. For quite a while now the polls which include Nader and Barr show a significantly bigger distance between Obama and McCain than those which exclude Nader and Barr. I thought I understood why, but when my 12 year old asked me to explain my explanation was so bad that not only did it collapse as I tried to articulate it, but it also disappeared (I no longer remember what my supposed understanding was). Why is it? Barr seems to get about 1%, and Nader 2-3% when they are included (both numbers seem remarkably high to me, but what do I know?). Could it be that Nader, as well as Barr, is drawing mainly from McCain (protectionist Republicans who are either too racist to vote for Obama or too sophisticated to believe he’ll be a protectionist?). Or is there some technical explanation that I don’t understand?

{ 45 comments }

1

Cryptic Ned 11.03.08 at 3:30 pm

About seven years after Nader first claimed that his presence on the ballot actually helped Democrats, I finally figured out how this might be true.

Freakonomists have recently harped on the notion that stores, restaurants, etc. include some extremely expensive things on the menu which they don’t expect anyone to buy, with the purpose of making the other things on the menu look cheaper. Suddenly the $12 burger looks like a lot more plausible option if the alternatives include the $16 burger and the $20 burger.

Now, say you are a voter who has no idea what is going on or why you should vote for anyone in particular, but you want to be in the mainstream of national politics so you can say with a weary sigh that both sides are full of liars and fools and extremists whereas you are steadfastly in the middle. In some countries, people who don’t like any of the candidates are encouraged to stay home, but here voting for a random person is seen as a civic good, better than staying home to express your true preference. So if there were only two candidates, the unmotivated voters would be equally split up between them. But then if there’s an extra, super-leftwing candidate, the regular ordinary sort of leftwing candidate looks like an acceptable middle ground, because at least you’re not voting for the most leftwing person on the ballot.

2

Dave 11.03.08 at 3:33 pm

Hi Harry. I think out of the crooked timber of humanity, no perfect poll was ever taken. Ahem.

3

Gregorus 11.03.08 at 3:58 pm

It takes a certain sort of contrarian to vote for a third party candidate; you essentially have to be excited enough about losing to show up at the polls. These contrarians, if not given a choice to vote for a candidate that will *certainly* lose will be willing to vote for the candidate who will *probably* lose. They also obviously don’t think that there is much of a difference between McCain and Obama (which there isn’t on the issues they care most about: election reform, ending the drug war, ending corporate welfare like farm and oil subisdies). Hence, they’ll just vote for McCain to be contrary.

There are also people who want to vote for a losing candidate because then they can claim it’s not their fault when things go haywire (as they invariably do with politicians).

Also, McCain is the more libertarian candidate between McCain and Obama. Neither is libertarian, but McCain is moreso. I’m a libertarian and will almost certainly be a Nader voter, and I know many other libertarians who are planning to be as well. He’s not that ultra-left-wing. Nader, Barr, McKinney, Baldwin, and Ron Paul all tend to make common cause on essentially libertarian issues like the ones I listed above, which are also among the most important to libertarians.

4

Rich B. 11.03.08 at 3:59 pm

My guess would be Republicans who don’t like McCain, but like Obama even less. They will pick McCain over Obama, but anyone else over McCain.

As for going for Nader over barr, Nader probably has more name recognition than Bob Barr. If I’m going to pick a third-party guy just because I want to protest my major-party candidates, though, I want to make sure I don’t accidentally vote for the Nazi-Commie-Pedophile Party guy, and if I haven’t head of Bob Barr, I can’t be sure he’s not one of those, so I go with the guy who will definitely lose, but who I have actually heard of.

5

Alex Gregory 11.03.08 at 4:04 pm

I’m a total non-expert on this, but here’s a guess: Ignoring people who support Nader/Barr means that you’re looking at a smaller group of voters, and working with a smaller group thereby exaggerates any difference in popularity between the main two candidates.

An extreme example: Imagine there are 100 voters. Imagine Obama has 80 votes, McCain has 10, and Nader/Barr have 10 between them. So expressed as a percentage, Obama is 70% ahead of McCain. Now ignore Nader/Barr. So we’re now considering 90 voters, of which Obama has 80. So Obama has 89% of the vote, and McCain 11% (rounded).

But as I say, this is pure guesswork.

6

Slocum 11.03.08 at 4:28 pm

My guess would be Republicans who don’t like McCain, but like Obama even less. They will pick McCain over Obama, but anyone else over McCain.

Substitute goods? McCain, Nader, and Barr are all protest votes against a presumably inevitable Obama victory. It would be interesting to see if the effect is stronger in states where the outcome is not in doubt than in battleground states. I assume that would be the case.

7

David Weman 11.03.08 at 4:31 pm

I’m pretty sure the explanation is that pollsters that screening choices etc, favor Obama, mainly I think dailykos/research 2000, are the only ones who include Nader and Barr.

8

Ed 11.03.08 at 5:53 pm

The simplest explanation is the best. Nader and Barr are drawing from McCain, n0t Obama.

The mistake liberals keep making as to Nader is paying too much attention to the candidates ideologies. The assumption is that if Kerry/ Gore are kind of left, and Nader is further left, Nader must be drawing exclusively from Kerry/ Gore. I really don’t think many voters use ideology when casting their ballots, the mental process is more tribal. Fringe party candidates are identified foremost as fringe party candidates and I think only a few obsessives care about their platforms.

For example there is plenty of anecdotal and voting evidence that the Perot phenomenon in 1992 was fueled by Republicans angry over GHW Bush agreeing to raise taxes. Bush agreed to raise taxes to balance the budget. Perot’s signature issue was balancing the budget. Anyone who took a few minutes to look up his stances on issues would know he was not a taxcutter. But for a certain type of Republican, Perot was a useful stick to beat Bush with.

I think minor party candidates tend to draw protest votes from the party in power, these people can’t vote for the main opposition party because they have demonized the latter all their life but they disagree with something the government had done. If I’m correct, the anti-Nader tactics of the Democrats backfired in 2004.

9

geo 11.03.08 at 6:02 pm

the anti-Nader tactics of the Democrats backfired in 2004

And not just in 2004. If Gore had agreed to debate Nader, had made the “lesser-evil” case directly to potential Nader voters, and had agreed to support electoral reform, he would have won easily.

10

MattF 11.03.08 at 6:32 pm

I suppose it’s barely possible that the discerning would-be Nader or Barr voter has decided not to be stupid, leaving the undiscerning voter to make a more-or-less random choice. Since Obama is ahead in the polls, this means McCain gets a benefit from the ‘WTF’ voter.

11

politicalfootball 11.03.08 at 6:36 pm

In fact (contra Ed’s 8), Nader voters in 2000, when asked about their second choice, expressed a strong preference for Gore over Bush. The idea that Nader has some special appeal to Republicans that Buchanan, say, lacks, is just the sort of fantasy that sometimes leads one to be a Nader voter in the first place, I’m guessing.

The idea (per geo in 9) that Gore could have persuaded Nader voters had he included Nader in a debate seems even more insulting to Nader voters than I am inclined to be. Nader voters certainly understood Gore’s position that he was different from Bush. They just disagreed with that position.

12

politicalfootball 11.03.08 at 6:49 pm

I like MattF’s explanation, but if it weren’t listed among the choices available, I would choose Dave’s explanation.

13

c.l. ball 11.03.08 at 7:07 pm

Alex, Harry is saying that the split goes the other way — the poll which has Nader and Barr shows a greater gap than polls that don’t offer them as a choice. We would expect it would go the other way, as you outline, that the gap widens if you calculate it as a percentage of the two-party vote.

I’m not sure what polls Harry is referring to, however. Pew and Ipsos/McClatchey, which include Barr and Nader by name, show only a 6 and 5 point gap, below the RCP average.

14

christian h. 11.03.08 at 9:27 pm

I can think of various reasons why this effect might occur, some already given here. But the easiest explanation is: we are talking about different polls. Comparing them is an iffy proposition at best. You might as well ask why outfit A’s poll shows candidate X with more of a lead than outfit B’s poll.

15

PersonFromPorlock 11.03.08 at 9:28 pm

What’s the problem? Nader/Barr voters are saying “neither of the above” to McCain/Obama but when forced, pick McCain for whatever reasons. I suspect that all this shows is that Republicans are less enthused about McCain than Democrats are about Obama (huge revelation, eh?), and some are much less enthused.

16

geo 11.03.08 at 9:37 pm

My dear football: Of course Nader voters knew Gore was different from Bush. That’s not what they would have learned from a Gore-Nader debate. What they might have learned is that Gore was an honorable man who understood the undemocratic nature of the American electoral system and could perhaps be trusted to help reform it. Unfortunately, he wasn’t, he didn’t, and he couldn’t.

17

harry b 11.03.08 at 9:40 pm

I’ve been comparing the average of polls at RCP every couple of days (comparing the with Barr and Nader average to the without Barr and Nader average) and have noted that specific polling companies get bigger gaps for their with Barr and Nader polls than for their without Barr and Nader (despite, presumably, using the same basic methodology for both).

political football — geo’s idea wasn’t that Gore would have convinced Nader voters (highly unlikely!) but that the contrast would have made him seem more moderate and acceptable to swing voters who either sat it out or went for Bush. That’s entirely plausible. It wasn’t however, a risk-free strategy — its entirely possible that Nader would have made Gore look like a fool, and/or forced him to alienate left-ish voters who he had in the bag. Given that he ended up with a populist strategy to appeal to left-ish voters, I imagine that is what his strategists were worried about.

Finally, left-wing Nader voters had very complex reasons for voting for Nader. In some States they believed that Gore had the state sewn up so voting for Nader was costless. Attention to the polls led them to believe (wrongly) that there might be a real pay-off for the Green Party (if N reached 5%) which would have made it worth losing to Bush in the long run. And, to be honest, even I, who have very little regard for the Democrats, didn’t believe they could screw up opposing Bush. I even thought that 9/11, when it happened, would help them (as it would have helped a political party with any level of competence). If Nader supporters shared my unreflective assumption that the Dems were not completely inept they might well have thought that there was a not a large long-term cost to Gore losing. I’ll just add that I know a lot of people who voted Nader in 2000, and not one who thought that Bush and Gore were the same.

18

harry b 11.03.08 at 9:44 pm

I misread geo. Still, my comment is worth thinking about.

19

MSS 11.03.08 at 10:12 pm

Ed’s succinct statement really says it best: “I think minor party candidates tend to draw protest votes from the party in power.”

I would also agree with Harry B’s comment about some of the calculations behind the larger vote for Nader in 2000. Of course, the goal of getting Nader to 5% for the sake of the Green party can’t apply this year, because he is not the Green nominee. That would be Cynthia McKinney. She is not present in most of these polls, but the few I have seen do not change the picture much; she does not even seem to cut into Nader’s vote, despite nearly indistinguishable policy prescriptions. And then there is Chuck Baldwin, social-authoritarian, endorsed, over “Libertarian” Bob Barr, by that other alleged libertarian, Ron Paul.

Yeah, basically, other than the tiny committed ideological core, the 1% or so that is probably each minor candidate’s ceiling this year comes from anti-incumbent-party voters. But Ed said it more succinctly than I just did.

20

Barbar 11.04.08 at 1:18 am

And, to be honest, even I, who have very little regard for the Democrats, didn’t believe they could screw up opposing Bush. I even thought that 9/11, when it happened, would help them (as it would have helped a political party with any level of competence).

“Screw up opposing Bush”? The party in power has more power than the party out of power. If power is important, how will that not lead to negative consequences? I guess power is not important.

If the Republicans were a competent political party, they could have used 9/11 to wipe the Democrats off the map, instead of focusing on a 50%+1 strategy while creating disaster after disaster.

21

harry b 11.04.08 at 2:08 am

In democracies, opposition parties can use the incompetence of ruling parties to generate widespread support. One might have expected, in a democracy, that the party responsible for failing to prevent a disastrous terrorist attack, and then launching a pointless and distastrous war, would be punished by an electorate, especially if the opposition party had all but won the previous election. Failure to make real political gains in the circumstances they faced reveals a lot about the capabilities of the Democratic Party.

22

Mike Jenkins 11.04.08 at 2:23 am

Good time to mention my all-time favorite CrookedTimber post (4 years ago)

http://crookedtimber.org/2004/11/02/kerry-317/

23

Charles S 11.04.08 at 2:30 am

That they failed to do it in 1 and a half years, but succeeded in spades in 3.5 years, and are now expanding upon that in spades this year tells us what, exactly about the ineffectuality of the Democratic Party?

I suppose one can reasonably claim that it shows the difference between the pre-Dean Democratic Party and the current Democratic Party, but it seems equally likely that it shows that one and a half years just wasn’t long enough to get Americans to turn completely against the war.

24

Barbar 11.04.08 at 2:44 am

Yeah, it took a little while, but the Democrats have brilliantly capitalized on Republican incompetence to gain widespread support. 58-60 Senate seats, large House majorities, a black guy whose middle name is Hussein as President — all this would have been inconceivable in 2000. So why would anyone think that there was a downside to Bush becoming President back in 2000? It all looks like gravy to me!

25

vivian 11.04.08 at 2:56 am

specific polling companies get bigger gaps for their with Barr and Nader polls than for their without Barr and Nader (despite, presumably, using the same basic methodology for both).

What happens when they include Morgenbesser?

26

roy belmont 11.04.08 at 3:15 am

#23:
“one and a half years just wasn’t long enough to get Americans to turn completely against the war.”

I can’t stand it.
The 2006 elections were about Americans being against the “war”. What that really meant was being against the occupation, and what that really really meant was the sacrifice of American soldiers in the occupation of Iraq.
The “war” in Iraq is over.
American soldiers are still dying in Iraq, because they’re enforcing an occupation against a great deal of native resistance.
The Iraq War, meaning the invasion of Iraq and the complete dismantling of its military and the destruction of its economy, was a success.
Now it’s over.
Whoever was waging it, whoever wanted it fought, they won. It’s an occupation now, and has been for some time.
It will remain an occupation for a very long time, because that was the whole idea all along.
It’s over, the war part. Got it? Over, done.
Being against the Iraq War is like being against the Korean War.
This group orgasm of Democratic triumph is as two-dimensional as a car commercial.

27

LFC 11.04.08 at 4:36 am

harry b@17: “left-wing Nader voters [in 2000] had very complex reasons for voting for Nader. In some states they believed that Gore had the state sewn up so voting for Nader was costless.”
In states that Gore had sewn up, voting for Nader *was* costless. There was no cost to voting for Nader in, say, Massachusetts or Maryland. Voting for Nader in Florida was quite different.

28

bad Jim 11.04.08 at 6:17 am

To hell with Florida, the Nader vote in New Hampshire cost Gore the election.

29

Matthew Kuzma 11.04.08 at 6:23 am

As much as it may be fun to guess about what motivation drives the trend, the only statement that can be solidly made from the numbers is also the blatantly obvious one you already made: that when not given the option of picking a third-party candidate, more of their supporters pick McCain than Obama and the gap is narrower. I’m sure there are quite a lot of reasons for this. There are, after all, an infinite number of wrong answers to any question.

30

geo 11.04.08 at 6:29 am

But Jim (#27), since Nader was an immeasurably superior candidate, why not say that the Gore vote cost Nader the election?

31

J Thomas 11.04.08 at 6:35 am

The Iraq War, meaning the invasion of Iraq and the complete dismantling of its military and the destruction of its economy, was a success.
Now it’s over.
Whoever was waging it, whoever wanted it fought, they won. It’s an occupation now, and has been for some time.
It will remain an occupation for a very long time, because that was the whole idea all along.

Well, we won the far fast, and it didn’t cost all that much. The occupation has cost us a whole lot more casualties and a whole lot more money and there’s no end in sight, no victory to look forward to.

The occupation is a whole lot more like a war than the war was. You can say it isn’t a war if you want to, but for most of us it’s a war we want ended. Officially you could stress that it isn’t an occupation either — we had to stop calling it an occupation because some of the things we were doing were officially war crimes during an occupation, but the Geneva conventions don’t apply to non-war/non-occupations. But these niceties are pretty much irrelevant.

I say it’s broccoli and I say the hell with it.

32

Barbar 11.04.08 at 6:40 am

Well, the Greens weren’t able to capture the Presidency and the Executive Branch but at least Nader’s run allowed them to win a few House seats and take over some local governments, and now they’re in much better shape than they were in 2000.

33

whomever1 11.04.08 at 7:05 am

Don’t forget the Party for Socialism and Liberation is running candidates too:
Gloria La Riva and Eugene Puryear.

34

peter 11.04.08 at 8:51 am

The phenom is yet more evidence for the empirical invalidity of the assumption of the Irrelevance of Independent Alternatives.

35

Alex Gregory 11.04.08 at 1:47 pm

c.l.ball (#13): Ah, thanks. I misread the post; I should have known it couldn’t be anything so simple.

36

KAS 11.04.08 at 2:46 pm

I know someone whom is voting for Nadar; in their answer as to why, they said it was because they liked him the most out of all the candidate’s; when I pointed out that by voting for a predetermined loser instead of the best of the worst (if you don’t like either that is) that the affect is not simply irrelevant, it’s detrimental- they said that it was just ‘something’ about Obama they didn’t like and McCain/Palin scared them. I took it as racism on the Obama end, whether she recognized it or not and later learned that was the case among her parents…

I think many, many people are simplistic and though they portray reason as it is far more sociologically accepted; feelings end up ruling their decision making process…

KAS

37

Harry 11.04.08 at 3:40 pm

LFC — yes, I didn’t mean to imply that the belief was false (but see how it might have seemed that way).

38

geo 11.04.08 at 5:11 pm

Dear KAS,

It seems to me that by not voting for the candidate you know (or ought to know) is much the best one, ie, Nader, you are helping to “predetermine” the continuance of our irrational and corrupt two-party oligarchy. Perhaps we Nader voters are not altogether simplistic.

39

harry b 11.04.08 at 5:35 pm

Why are so many Dems so invested in blaming Nader and his voters for the 2000 loss?

40

Cryptec Nid 11.04.08 at 6:34 pm

Because they wish to avoid losing in the future.

41

Barbar 11.04.08 at 6:54 pm

I’m not so interested in blaming Nader voters as I am in pointing out in ridiculous errors in logic and fact (for example, that all the negative consequences of Bush’s presidency can be blamed on the ‘incompetence’ of the Democratic Party, or that Nader contributed anything to the Green Party’s nonexistent long-term success).

42

harry b 11.04.08 at 6:57 pm

Neither of your examples constitute errors of logic, and nor did anyone make either of the claims you cite. So, we’re all in agreement!

43

Barbar 11.04.08 at 7:03 pm

If Nader supporters shared my unreflective assumption that the Dems were not completely inept they might well have thought that there was a not a large long-term cost to Gore losing.

Fantastic!

44

Righteous Bubba 11.04.08 at 7:13 pm

Why are so many Dems so invested in blaming Nader and his voters for the 2000 loss?

An Al Gore loss couldn’t have happened. It didn’t happen.

45

geo 11.04.08 at 10:31 pm

Because they wish to avoid losing in the future

Perhaps Democrats also wish to deflect Nader’s devastating criticisms of their miserable (though not, of course, as bad as the Republicans’) policy positions.

Comments on this entry are closed.