« It's Official: Kerry's Ahead | Main | Republican Analysts Lower the Bar »

All Trial Heats Are Not Created Equal

There's been considerable confusion about which trial heat results to pay the most attention to at this point in the race. Here's my take, which should help clarify why I choose to focus on certain results over others in this blog.

One issue is likely voters (LVs) vs. registered voters (RVs). At this point, most polls are surveying only RVs and I believe that's appropriate and, in fact, preferable. It is way too early to put much faith in likely voter screens/models as representing very accurately the voters who will actually show up on election day. There is reasonable evidence that careful likely voter methodologies work well close to the election and do fairly accurately capture that pool of voters. But there is no such evidence for LV samples drawn this far out.

Indeed, my understanding is that Gallup does LVs this early not so much because they believe they are capturing election day voters this early, but more so that they can avoid having to explain sudden shifts in the horse race question as LV data replaces RV data in the fall (the traditional time to switch from RVs to LVs). There have apparently been some problems with this in the past, so reporting both from the very beginning of the campaign eliminates any potential embarrassments along these lines. But that doesn't mean the LV data is any better at this point in time--it merely means they're providing it.

In fact, since the sample size for LVs is smaller and since the composition of the LV sample will shift depending on how political developments are affecting interest and intensity levels among different groups of voters, additional volatility is built into the LV samples that is not there with the RV samples.

And then there are the comparability problems. LV samples are difficult even to compare to one another, since methodologies differ, and clearly can't be compared very well to RV samples, which are the bulk of polls at this time. That's another strike against paying much attention to LV results this early.

Another issue is how much attention to pay to the Rasmussen tracking poll. My view is that this poll is worth looking at but with caution. It is the only tracking poll out there at this point and therefore on any given day it may have the "freshest" results. But it is only one poll and there is no other tracking poll to check it against. Moreover, it is a poll of LVs, which, as outlined above, should be viewed with some skepticism at this point. Therefore, I am inclined to pay more attention to standard national polls of RVs and the trends these polls imply than to this one tracking poll.

Let me sum up my current approach, both as outlined above and in previous posts (for example, on the Bush-Kerry vs. Bush-Kerry-Nader question) with the following:

Ruy's Rules

1. Pay more attention to Bush-Kerry results than to Bush-Kerry-Nader results.

2. Pay more attention to RV results than LV results.

3. Pay more attention to recent national polls of RVs and the trends they imply than to the Rasmussen poll.

These rules don't mean ignore the rest of the data. On the contrary, immerse yourself in as much of it as you can stand. But these rules provide a way to sift through the available data to get the best sense of how the contest between Bush and Kerry is developing.

Comments

Thank you. You answered a lot of questions and sustained your robust credibility.

I'm no expert, but only six months ago new Iraq trouble would probably have gotten Bush a 15 point bounce, what with continuous "support the president" and "we must win" from the media. To get only this much (if even this) means that Bush has permanently lost a lot of people.

Excellent analysis of the wisdom of surveying RVs v. LVs this far from the election.

You might want to go back and examine Gallup's surveys of LVs in the 2000 election for a casebook example of the foolishness of polling LVs months before the election.

Back in 2000 Gallup's surveys of likely voters in September and October had huge swings from Gore to Bush and back to Gore. I believe these swings represented the different interest levels (one of the factors used to identify LVs) of Democrats and Republicans to events in the campaign -- and not massive swings of voters from one candidate to the other.

Since the final weeks of the campaign will have a big effect on which likely voters become actual voters, polls of LVs at this point are bound to be less reliable than polls of RVs.

One thing that's remarkable to me is the stability of Bush's approval numbers in the face of some very negative news, both on his handling of 9/11 beforehand, and the situation in Iraq. While it's true that he got some counterbalancing good news on jobs, it WAS just one month's numbers, and it certainly didn't affect his numbers on handling the economy in any case (so far as I know).

It seems pretty obvious that the low hanging fruit in pulling down Bush's numbers have already been picked, and that turning even another 5% away from Bush -- enough likely to put the election completely out of his grasp -- is going to be a major undertaking. On the other hand, it may also be that the final dip in Bush's numbers is now being prepared, and it may take just one final episode of bad news to turn a substantial number of voters against him.

The Plame affair looks to me like it could be that event. Would it be wrong of me to pray to God that Bush might be so disgraced?

frankly0, I think you make some very good points. Bush is getting around 43 percent now in two-way matchups with Kerry, which is probably getting close to his core support. But I agree the Plame affair could drop him further and be a major blow. Does anyone know when indictments might be forthcoming?

Rasmussen uses "automated polling technology". Is there any evidence that this produces a systematic bias of any sort? For example, I instantly hang up on any computerized voice I hear. Also, I never cooperate with market researchers. But I might answer questions from a live pollster who identifies themselves as being from a major polling or news organization and is interested in political opinions.

frankly0, I presume you're referring to Bush's propensity for holding onto 48-49% approval in most polls (Zogby/Pew excluded) -- a number which would suggest a narrow loss, at worst. I, too, have found this puzzling, given that 1) his internals in many areas are well below that -- as are the right track-wrong track numbers -- and 2) his share in trial match-ups with Kerry never go above 45-47.

I wonder if this continues to reflect a waning halo effect from September 11th. Bush got a huge jolt in approval from that, a jolt that has since faded steadily but quite slowly. It may be that we just have to be patient, that the internal and re-elect numbers are Bush's true "approval", but the main poll number still has a bit of fog attached to it that will burn off at its own pace.

Numbers are obviously helpful in determining electoral outcome, and we should be rigorous about viewing them realistically. However, if I can be permitted an emotional evaluation: I have over 30 years experience watching political races, and have found some presidential years clearly giving off a re-election vibe ('84 and ''96), and others ('80 and '92) clearly not. This one isn't quite in the latter category yet, but it's miles away from the former -- and, in fact, feels less favorable than '76, when Ford lost by only a 3% margin.

I guess what I'm saying is, the way things feel out there, it's hard to foresee re-election, and my bet would be on that approval rate dropping rather than rising.

I'm pretty sure there's never been a president re-elected during a recession, or when his approval ratings were below 50%. Economic indicators still look pretty mediocre.

In response to frankly0's comment about Bush's approval ratings not budging much even when his other numbers seem to be steadily dropping, I wonder about the role of cognitive dissonance.

I've been noticing (along with Ruy and others) that the polls have been reporting general approval of the U.S. war in Iraq for some time now, with around 60% of the people responding that it was "the right thing to do" to invade Iraq. Meanwhile, the question, "are things going well for the U.S. in Iraq" has seen erosion of approval from around 60% down to around 40% now. Cognitive dissonace would explain this -- people realize at one level that invading Iraq was a mistake, but to say so out loud implies that all the American blood that has been shed in Iraq was a waste. People aren't comfortable admiting that, so they report to pollsters that they agree with the U.S. decision to invade Iraq, but also must accept that the invasion isn't going very well.

I wonder if something similar isn't happening with the president's approval. People are beginning to recognize he's not competent in the job (which shows up in the specific questions about his job performance), but people aren't willing to admit they disapprove of the Commander in Chief during a time of war -- because disapproval sounds too much like not supporting the troops or being unpatriotic or whatever the reason. Even if cognitive dissonance prevents them from answering a pollster by saying they disapprove of the president, it won't take much to tip them to pulling the lever for Kerry in November either.

On this question of why Bush's approval ratings haven't gotten as low as the public's assessment of many of his actions might suggest...

Bush has gotten consistently among his highest marks for being a strong, decisive leader. A portion of the survey respondents appear willing to give credit to, and possibly stick by for a long time, a President who they think is making some bad decisions. Not sure how to interpret this. Is it a grudging sort of respect that, verbalized, might be expressed in something like the following?: "Sure, he's made some decisions that haven't worked out well. That's true of anyone at the top. But at least he's firm, he knows his own mind, and he seems determined to stick to his guns."

(To my way of thinking if a President wants to lead us off a cliff I want him to be an ineffective leader although not a complete, ahem, idiot lest our adversaries exploit the situation. "Strong leader" and "making all the wrong decisions" just don't square in my view of the world. But that's just me. I'm not a Republican.)

Another interpretation may result from surmise--and that's all it is--that Republicans tend to stick up for, and stick with, someone they see as one of "their own" (maybe Bush I didn't fall into that category) under more dire circumstances than a large share of Democrats do. Arguably this is consistent with B Clinton's take that Democrats want to fall in love with their candidate while Republicans want to fall in line behind their candidate. Could it be that the Republican mindset is less questioning of authority and more tolerant of hierarchy and hierarchical thinking? I wonder if anyone has done a Ph.D. thesis on this subject? The old joke "I'm not a member of any organized political party; I'm a Democrat" comes to mind in this context.

Never in our life time has this country had such a horrible president as bush. Every corner he has turned comes more harm to the citizens of this country and the world. We must do everything in our being to remove this person from office. Our country cannot survive as we know it if bush remains in office another 4 years.
Get active with everyone you know and get the word out and convince everyone that this must be done now. Ignore the nay sayers and march ahead with the message. It is up to all of us to save our country. Yes it is that serious.
The poles are just beginning to show what will turn into the largest victory for the democrats in the history of this country. Keep pounding and don't give up. I have faith in my fellow Americans to do the right thing. Even republicans know that this man is bad and that he must go. A surprising number of them will overwhelmingly vote for Mr. Kerry in this election. Tell every republican you know that it's OK to vote for Mr. Kerry and they are doing what is best for our country. bush has bertayed everyone. For every reason that a republican can give you for why they are going to vote for bush you can give them 10 reasons not to vote for him. Your reasons are based in facts theirs are based in lies and deceptions and false faith.
Lets just hope and pray that between now and election day bush doesn't do something else that will not be reversible for our country and the world.

Don't focus on the poles just get out there and do what it takes to turn every vote in for our next President, John Kerry.

I may be speculating here but while Bush's poll numbers have only dropped by a point or two while support for his Iraq Policy has dropped more considerably, the drop in support for his Iraq Policy may-just may, foreshadow a big drop ( maybe 5 points) in his head to head poll numbers against Kerry. I remember in 1980 that Carter was running ahead or even with Reagan most of the year when I believe failure after failure with the hostage drama finally caused the bottom to drop out of his support during the last week of the campaign. Am I grasping for straws or is this a reasonable take on current poll numders? Help me out here.

Probably. I'm not sure...rt?

In response to BKW, it sounds sound, but in the 1984 election, a lot of people voted for Reagan who didn't really like him, but they assumed nobody else didn't like him, and that they were weird, so they either stayed at home or DID vote for him, just as in the '80 election where a considerable number of people who voted for Reagan didn't agree w/ him, they just wanted Carter out, which may do wonders for Kerry. Right? Somebody help me out here....rt?

Oops. Reignman, I am guessing, perhaps incorrectly, that the questions you posed to "rt" in the previous post in this thread are intended for Ruy, whose initials are the same as mine. When I first chose my handle for this site it did not occur to me that the choice of "rt" might generate confusion on this point. Not long after I began posting here I "introduced" myself to Ruy via email--we had met FTF a couple of months before. Part of his friendly reply was words to the effect of "Oh, so *you're* rt." I guess if I'd been more alert it might have occurred to me to change my handle then. But I wasn't.

I apologize for any confusion this might have caused--and especially if anyone thinking "rt" was Ruy formed a lesser opinion about his work as a result of reading my (decidedly less data-driven!)posts. Beginning with my next post I will use the handle "bt" instead.

Again, my sincere apologies to you, Ruy, and any of you other good people who come to this site to get the benefit of Ruy's terrific analyses who were confused as a result of my actions.

May I ask why it's better to discount the Nader effect in those polls? Is it the assumption that his support will wane as time goes by and the election nears? Thanks.


In response to David de la Fuente, my recollection (without going back to check for sure) was that Ruy was arguing that without Green Party backing, Nader will have serious trouble getting his name on the ballot in any more than a couple states. Oregon (my home state) has historically given Nader a significant percentage of the vote, but last I heard he was nowhere close to getting enough signatures to make the ballot. Without his name listed, he's reduced to a write-in candidate, which will result in a tiny fraction of whatever he's polling. That said, you're absolutely correct that he won't be polling more than 2% by November.

In response to reignman, you may be right about people who aren't comfortable with Bush voting for him anyway. By my thinking, though, a challenger needs two things to happen. 1) people need to be dissatisfied with the incumbent (or else nobody even thinks twice about replacing him), and 2) people need to prefer the challenger. My point was simply that cognitive dissonance might mean there's a larger dissatisfaction with Bush than the polls sometimes indicate. Which means that if Kerry presents a favorable alternative, the polls will begin reflecting a significant movement of voters to the Kerry camp. Then again, it could just be my wishful thinking!

Alan Snipes: You are not grasping at straws. A few weeks ago, I checked some old poll figures on Gallup's Web site and was surprised to see that Reagan did trail Carter for most of 1980. In fact, the race was very close until the end, when undecideds swung heavily toward Reagan, giving him his landslide victory. I'm not saying 2004 is 1980 all over again, but I believe Kerry is in very good shape if you look at it from a historical perspective.