Red Reality, Blue Reality
Andrew Gelman and John Sides write about partisan bias in assessing economic conditions:
A good example comes from the research of Larry Bartels. He analyzed a 1988 survey that asked “Would you say that compared to 1980, inflation has gotten better, stayed about the same, or gotten worse?” Amazingly, over half of the self-identified strong Democrats in the survey said that inflation had gotten worse and only 8% thought it had gotten much better, even though the actual inflation rate dropped from 13% to 4% during Reagan’s eight years in office. Republicans were similarly biased about the Clinton-era economy: in 1996, a majority of Republicans thought that the budget deficit had increased. This partisan filter was also evident after the Democrats’ retaking of Congress in 2006. Research by Alan Gerber and Greg Huber shows that Democrats became much more optimistic, and Republicans more pessimistic, about the national economy.
As Ezra Klein pointed out some time ago, Bartels’ research actually suggests that the problem has odd interactions with levels of political information:
For one thing, the gap grows as levels of political information grow. But more interesting is that along most of the hierarchy, Republicans show the pattern you would assume—the more they knew, the more accurate their perceptions about the state of the budget deficit. But among very high-information Republicans, perceptions start getting less accurate. This is presumably because very high information Republicans were able to familiarize themselves with sophisticated arguments about why the apparent improvement wasn’t real improvement.
I assume the way this works is akin to how I was slower than most Americans to recognize that violence in Iraq was dipping not despite being better-informed about the situation in Iraq than most people but precisely because I was well-informed. Well-informed and suspicious! So I was keenly aware of all kinds of problems in the data and its presentation that undermined the dominant—and, it turns out, correct—narrative about an improving security situation.
16 Responses to “Red Reality, Blue Reality”
Register
RSSLogin
Topic Cloud
ArchivesBlog RollAbout Matt YglesiasContact Me
Donate
April 17th, 2009 at 8:53 am
Maybe, but with Republicans might a large part of that disconnect be more because the highly “informed” are getting most of their “information” from sources such as Fox News and Limbaugh?
April 17th, 2009 at 9:05 am
Exhibit B: charles/Mixner.
April 17th, 2009 at 9:15 am
Amazingly, over half of the self-identified strong Democrats in the survey said that inflation had gotten worse and only 8% thought it had gotten much better, even though the actual inflation rate dropped from 13% to 4% during Reagan’s eight years in office.
Some (but not all of this) can be attributed to the fact that way we measured the CPI changed during that time (and continues to evolve). You cannot compare early 80s inflation with late 80s/early 90s inflation, and neither of this is comparable with modern inflation.
The numbers all mean totally different things and no one has come up with a good conversion (there are methodology problems in Jim Rodger’s Shadow Statistics). It is like saying 30 feet is longer than 2 miles because 30 is greater than 2.
April 17th, 2009 at 9:17 am
I presume these numbers hold things like class and income constant? Because otherwise, if you assume that on average political information increases with income, those would be fairly accurate readings of the Clinton years’ impact from each group’s perspective, wouldn’t they?
April 17th, 2009 at 9:23 am
What’s interesting now is that Republicans, led by Kudlow, are insisting that everything is really fine–such is their contempt for even quasi-statist solutions to the economic crisis. This is, in a sense, admirable: it is the triumph of ideology over a straight partisanship that (when practiced from the opposition) usually paints everything in as dire terms as possible.
April 17th, 2009 at 9:23 am
The problem isn’t being correct or incorrect but normative. Had inflation gotten “better” during the Reagan presidency? The rate of inflation certainly was lower at the end but was inflation “better.” Not if judged by the cost (high unemployment) to bring down the rate. Similarly, was security in Iraq “better” in 2008? The number of deaths certainly was lower as compared to the prior two-three years but was security “better.” Not if judged by the cost (ethnic cleansing) to bring down the violence. Why do researchers use normative terms (such as “better”)? Maybe because intuition tells us more than statistics. After all, Matt’s belief that security had not really gotten better in Iraq may prove to be more accurate than the gruesome statistics.
April 17th, 2009 at 9:25 am
Matt’s admission in the last paragraphs is a bracing reminder of the kind of self-examination needed to stay honest, no matter how good your intentions.
I think it’s in “Notes on Nationalism” that Orwell observes how little intellectuals, left and right, expected the British Army to win the series of battles at El Alamein, in the summer under Auchinleck and in the fall under Montgomery. And the reason, he realized, that he had been able to see victory was as likely as not was that he wanted it to happen. By-and-large, the hatred of the “blimps” was so great among Britain’s intellectual classes that they didn’t.
I know I was concerned lest a tactical success, however many Iraqi and American lives it might save in the short run, would only embolden a wrongheaded policy, both locally in the conduct of the Iraq war and more generally going forward. And to go along with that, I convinced myself it was unlikely to succeed anyway. That was partly because the tactical change reported, “the surge,” looked to be more of the same. The Ricks-documented tactical changes, vastly more informed command, and above all, the buying off and truce with the Sunnis, were not on my radar. I needn’t have worried, it turns out, because we can see now that the current policy began with the replacement of Rumsfeld by Gates after the ‘06 election, and the recent election endorsed the approach, which Obama had already identified as his own. But in the Spring of ‘07 it was easy to expect, because I in a sense wanted, it to fail.
April 17th, 2009 at 9:26 am
Great post — and the analogy to Iraq could be expanded to cover a whole range of other topics.
This is why it’s so hard to shake confirmation bias: the harder you try to be objective, the more you know, and the more you know, the harder it is to be objective.
April 17th, 2009 at 9:28 am
I like that Yglesias can identify when his biases cloud his thinking. You don’t see enough of that on the internet.
April 17th, 2009 at 9:32 am
It isn’t just that highly informed people are more able to seek out information and viewpoints that confirm their ideological preconceptions. It’s also that informed people are more likely to have an emotional investment in those preconceptions, and thus have a greater incentive to selectively seek out information and justifications to support them.
In the case of Matt Y. and security in Iraq, the relevant problem was that he had a lot invested in the (pretty superficial) “incompetence dodge” idea. Thus, he had a strong incentive to resist a conclusion that abandoning really stupid policies in favor of less stupid policies could have a significant positive impact on post-invasion Iraq (even in the face of emerging evidence to the contrary).
April 17th, 2009 at 9:39 am
I’ve found that libertarians are often some of the worst offenders in this regard. Many of them seem to believe that adopting a simplistic set of dogmatic principles (”government bad”) has given them a unique and universal insight into every political and economic issue; admitting limitations to those basic principles would cause the entire pretension to come crumbling down.
Furthermore, many self-proclaimed libertarians also seem to believe that adhering to those dogmas makes them manly, rugged individualists. Hence their otherwise puzzling lack of concern about runaway military spending, officially sanctioned torture, or poorly planned attempts to transform other countries through invasion and the application of massive violence. Identifying with state violence provides an alternative way of making oneself feel manly and tough, so libertarian ideals can be safely disregarded in those areas.
Once one develops that kind of emotional attachment to a political worldview, acknowledging contradictory realities is simply not an option.
April 17th, 2009 at 10:36 am
N,
What makes it worse is the Democrats and even Republicans are aware of the possibility of confirmation bias and partisan blinkers. Even if they aren’t sufficiently cautious about those problems, they understand in theory that it is possible for someone to succumb to those pitfalls.
In my experience – more than I’d care to admit to – with libertarians, I’ve found that they consider their own narrative about partisanship (their non-partisan status, viewing both sides as similar) to be an inoculation against the possibility of confirmation bias. So, when they see a statement or argument that reinforces that view, they have no buffer at all, no recognition that they might be more inclined to credit it than it really deserves because it’s what they want to hear, because they believe that what they want to hear is, by itself, a defense against buying into a partisan narrative.
April 17th, 2009 at 10:45 am
Uh, there’s another explanation for this phenomenon of high information respondents being seemingly divorced from reality: they are intentionally misleading the poll interviewer because they know what they would like the poll results to be. That is, if I’m a high info Dem (as I am) and someone calls and asks me to take a survey, I do. If it’s a political survey, I always answer them in ways that make the conservative candidate or policy look worst/most unpopular. I don’t really care whether the message statement the interviewer reads me are true or not. This is my small contribution to screwing over Republicans … which unfortunately also results in “innacurate” survey data.
April 17th, 2009 at 1:37 pm
Shorter Matt: I was wrong because I had too much information!
April 17th, 2009 at 3:38 pm
C.S Lewis in That Hideous Strength (which I know many people have major issues with, but put that to the side for the minute) wrote that educated people are easier to manipulate through propoganda than the uneducated because the educated believe what’s in “their” papers. In modern parlance, they cocoon. In contrast, according to Lewis, the common folk mistrust everybody and thus resist propoganda better.
Not sure that I buy that, particularly given recent experience. But Lewis had the disadvantage of writing 30+ years before Fox News.
April 17th, 2009 at 4:51 pm
“Well-informed and suspicious! So I was keenly aware of all kinds of problems in the data and its presentation that undermined the dominant—and, it turns out, correct—narrative about an improving security situation.”
Yglesias’ almost unnervingly wonkish grasp of major policy issues is the reason I began following his blog, but it was really the moral integrity of his writing that has propelled him to the top of my must-read short-list. Left or right, there are few other talking heads that are as consistently honest and self-critical.
But what I’ve come to discover through his writings is that– beyond the valuation of capital-J Justice as an end in and of itself–intellectually honest writing offers a secondary appeal which has less to do with ethics per se than with rational intellectualism itself. Simply put, Yglesias is not always doubling-back and contradicting himself in order to cover his ass, professionally speaking. And so his cumulative writings become nuanced instead of self-contradictory. Thus on a purely amoral, dispassionate, and dare I say cerebral level, Yglesias’ writing is just inherently more interesting than anything than the writings of a hack like, for instance, Bill Kristol could .