Having read the "Authoritarians" pdf book, and seen some other work along related lines, I'd guess that this phenomenon is much more heavily a Republican phenomenon than a Democrat phenomenon, although certainly members of both parties are subject to it to some extent.

Note both these examples are ones where reality has a liberal bias.

You assert "the lines almost certainly would have reversed," but do you have some good examples of self-declared Democrats having a worse objective grasp on reality than self-declared Republicans? It seems like a reasonable hypothesis that members of both parties are prone to self-delusion, but it's nonetheless a hypothesis which must be supported by evidence, and not merely asserted.

How do you reverse a distance?

I don't think this graph proves what they say it proves.

First off, there's a weird convergence of opinions between 10 and 40 percent. That suggests to me something very odd is happening among casual news viewers.

Also, the sudden crash of R's at 100 percent is really bizarre. Suggests to me there's a sampling error somewhere.

And yes, you'd need to do this experiment the other way to see if Dems did the same thing.

They're also trying to draw an objective conclusion from subjective answers. "Gotten Much Worse - Gotten Much Better." We'd need to know more about the people involved and their socioeconomic status to make meaningful correlations with this kind of data.

Way too many variables to draw any firm conclusions on this one. Looks like good dinner table conversation fodder but not exactly adding anything to understand people and their behavior.

If if George W. Bush had been in office during that period, the lines almost certainly would have reversed,

What exactly is the basis for that claim? Reading through the Bartels pdf I didn’t see any examples are striking as the 1996 economic conditions data. It’s a lot different think a politician agrees with you more then they do, than just totally flub a simple well reported trend question.

I don’t think the paper even tries to judge the relative rationalization of different voters. Saying that both sides rationalize is very different then saying both sides rationalize equally.

The reagan years would be a good example to throw in, to balance things out.

It seems like partisan lefties over 30 have conniption fits over any praise heaped on Reagan.

Reason four:

Right wing types are seeking to dial down concerns because they are so philosophically opposed to the type of fixes that higher level concerns may lead to.

Think of this as a mirror image of how they used heightened concerns over terrorism and safety.

"the lines almost certainly would have reversed,"

So here we have a study [&blogger] that/who confirm the He Said, She Said mind set.

I'm not buying any of it. "Educated" is a loaded word and it means whatever the aurthor needs it to mean in order for the aurthors study to be correct.

zzzzzzzzzz....stupid post.

Ezra, I try to read as much right-of-center stuff as I can stomach. I'd be very interested in your list of sources (blogs, columnists, etc.) that you find compelling / informative / well-written. I'm frustrated with the all-over-the-place-ness of the Corner, RedState is pretty decent, I don't know what else is really out there.

Thanks.

Good post. I have long seen bias everywhere, including in myself.

I once told a scientist that I thought that scientists were bias like everyone else, perhaps less than most, but still biased and he responded badly.

Try telling a Doctor or teacher that they are biased, they will not like it (I think that they see themselves in the caring professions and thus above bias). I remember talking to doctor and asking him questions like I ask an auto mechanic, assuming possible bias in his recommendations and he started to insult me.

But what do I know I am biased like everyone else? (smile)

As for: "...lefties over 30 have conniption fits over any praise heaped on Reagan."

There reasons for this, most importantly, is that they may have information that precedes the Haplographies of the recent past.

http://www.fair.org/index.php?page=1832

"Ronald Reagan was the most popular president ever to leave office," explained ABC anchor Elizabeth Vargas (6/6/04). "His approval ratings were higher than any other at the end of his second term." Though the claim was repeated by many news outlets, it is not true; Bill Clinton's approval ratings when he left office were actually higher than Reagan's, at 66 percent versus Reagan's 63 percent (Gallup, 1/10-14-01). Franklin Delano Roosevelt also topped Reagan with a 66 percent approval rating at the time of his death in office after three and a half terms.

In general, Reagan's popularity during his two terms tends to be overstated. The Washington Post 's lead article on June 6 began by declaring him "one of the most popular presidents of the 20th Century," while ABC 's Sam Donaldson announced, "Through travesty, triumph and tragedy, the president enjoyed unprecedented popularity." The Chicago Tribune (6/6/04) wrote that "his popularity with the electorate was deep and personal... rarely did his popularity dip below 50 percent; it often exceeded 70 percent, an extraordinarily high mark."

But a look at Gallup polling data brings a different perspective. Through most of his presidency, Reagan did not rate much higher than other post-World War II presidents. And during his first two years, Reagan's approval ratings were quite low. His 52 percent average approval rating for his presidency places him sixth out of the past ten presidents, behind Kennedy (70 percent), Eisenhower (66 percent), George H.W. Bush (61 percent), Clinton (55 percent), and Johnson (55 percent). His popularity frequently dipped below 50 percent during his first term, plummeted to 46 percent during the Iran-Contra scandal, and never exceeded 68 percent. (By contrast, Clinton's maximum approval rating hit 71 percent.)" - http://www.fair.org/index.php?page=1832

BTW, speaking for myself, as a lefty over 30 I have fair amount of positive things to say about Eisenhower, Nixon, & Ford, but far, far, far fewer for Reagan, George The First & George The Second. Which is another reason why I think this post is beyond stupid....I have some historical background that does not match the author's study

Great post. I wrote up a response on my blog here: http://dlpfc.wordpress.com/2009/01/07/politics-and-perceptual-delusion/ that touches on the issue of whether liberals are as bad about this as conservative.

My take: there's a case to be made that liberals are less prone to this particular bias, but that it doesn't make any sense to credit them as "better" than conservatives because of that.

Another point to contradict Ezra's point the left is really the same as the right.

Now Ezra's a little young to remember this, but the left was highly critical of Clinton while he was in office.

So again, this post should be filed under circular.

The point Ezra makes here was made well before Bartels and Achens: it goes at least back to John Zaller's classic 1992 book The Nature and Origins of Mass Opinion. It covers a lot of ground, but the most pertinent point here, which Zaller illustrates with many examples, is that high-information voters (of all political persuasions) are much more polarized than low-information voters. (High-information is defined as people with a good general political knowledge, rather than those with specific information about the topic under discussion.)

So, taking the example of the Vietnam War, at the start of the war low-information voters (liberals and conservatives alike) broadly supported the war. As time went on, those low-information voters (liberal and conservative alike) turned against the war. High-information liberals turned even more strongly against the war, although they too had tended to support it at the start. But high-information conservatives supported the war even more strongly at the end than they had at the start.

The reason is precisely the one that Ezra gives. People who know a lot about politics are more concerned that their views should be consistent with each other, and to explain away things that might contradict their overall world-view. People who know less in the first place are less likely to be concerned about inconsistency, and so to screen out underlying realities.

The bottom line is a rather depressing one: the more informed you are, the more likely it is that partisanship will trump reality. Another depressing conclusion is that we may all decry partisanship, but partisanship is a problem precisely because of people like us, the sort of people who read political blogs like this one.

Speaking as someone who works in this field, is a liberal, and a skeptic:

What makes us think that Democrats are perceiving the economy any more accurately than Republicans in this chart?

That is, their estimation of economic performance is more correct, but it may be so for the wrong reasons: specifically, just as Republicans view the economy as doing badly during the reign of a guy they hate, so Democrats might view the economy positively for similar reasons -- reasons that have nothing to do with their perceptions of reality.

So not only do I hypothesize that the chart might be reversed during the Bush II era, but that still wouldn't prove that the Democrats were perceiving things accurately. Presumably they would rate the economy as having done worst under Bush than would Republicans, but that too might be due to partisan bias. What you really need to look at is judgments under, say, the second term of Reagan, when the economy improved. My bet is that Democrats would judge it as having done worse than Republicans. But what would that prove? What is the objective judgment of how well the economy really did -- medium, ok, or great? GDP doesn't cut it: stocks were up, but the middle and lower classes, not so much. Macroeconomic indicators are much less objective than Bartels and Achen suggest, even under Clinton.

In short, partisan bias is a really hard issue to pin down. The Bartels & Achen paper shows that a) low information voters are essentially random in their judgments, and b) high information Republicans have an tendency to view the Clinton economy as worse. But does that prove that that judgment is biased? And does it prove that the Democratic judgment is unbiased? Nope. In fact, until we see something like the Reagan data, it is isn't even partial evidence that Democrats are less biased (assuming a default hypothesis that both sides are biased again the opposite president).

I read a study where researchers took truthful statements that were negative about Bush and Kerry. Both sides were pretty much equal in being wrong and believing whatever was in their respective candidate's favor.

So if one side or the other claims some type of superiority, think again

The more pertinent question would be: Who was more apt to bail out before the dotcom boom went bust, Republicans or Dems? Who was more apt to bail out before this last bubble? Only an fanatical idiot would have let partisan blinders draw them into obvious, ruinous debacles like those.

Yes, there is the question of deciding what the baseline objective facts are, which is not so simple as the post suggests. There are also expectations. During both the Reagan and Clinton booms, I felt that the deregulatory fervor was creating short-term prosperity at the expense of long-term stability, and therefore viewed the economic direction of the country less favorably than most. Since I felt this way under both Reagan and Clinton, I would not call it a partisan bias, but perhaps an ideological one, and one that I believe recent events have largely vindicated. But if my pessimistic assessment were judged during those two presidencies by the facts at that time, one would have to say that my position was poorly supported by fact. I do think it is more rationale to take expectations into account in making evaluations despite their speculative nature, as no economic policy can really be judged by its immediate impact alone. But the speculative nature of expectations is such that one's intellectual models will loom much larger in them than in one's evaluation of the present.

Ezra,

I can believe this was true in 1996, but given that its over a decade later and infromation consumption patterns have changed dramatically, How can this survey say anything meaningful at all about what is currently happening?

Did they control for the local economy in the place subjects spent the Clinton years? If not, EPIC FAIL.

http://www.gamegoldme.com/