Arnold Kling  

What the Klein Study Proves

The Cognitive Capture Dilemma... Kid Takes on Ben Barber...

[update: Daniel Klein emails me to credit Zeljka Buturovic for the study.]
Dave Leigh writes,

People who frequently shop at Wal-Mart scored much better than those who never do (2.24 to 4.24 questions wrong, respectively). This isn't much of a surprise to me... the snoots who won't shop there aren't likely to know a thing about saving a buck. It's pretty clear that the people who shop there aren't stupid about economics... but their detractors may be.

He refers to the study by Dan Klein. I think what makes the study obnoxious to many people is Klein's use of the terms "right" and "wrong" to describe what many people would prefer to think of as matters of opinion about propositions in economics. I think that the study has merit, regardless.

Suppose, as a worst case, that Klein's characterizations are not defensible. As Leigh points out, the study nonetheless shows interesting correlations between economic beliefs and other characteristics. I think it really challenges Bryan Caplan's view that the unwashed masses have the most dangerous economic views from a free-market standpoint.

In the past two years, I have given talks at two churches. At both, I tried to explain some basic realities of health care, namely (a) that as individuals we would like unlimited access to medical services without having to pay for them, but collectively this is not sustainable; and (b) health care does not break down neatly into necessary vs. unnecessary care, but rather that even among beneficial treatments there is a continuum of benefit-cost ratios, starting with some services that have very high benefits relative to cost and ending with services that have low benefits relative to cost.

At the University of Vermont, an outpost of the established Church of Unlimited Government, I received nothing but pushback from the audience. The typical question from the audience asserted that people have a right to health care, and that settles the issue.

In northern Indiana, inside a Methodist Church where I spoke to many people of Tea Party sympathies, people got what I was saying right away. They asked more constructive questions (I got pushback, but it was of a different nature and on different issues).

I think that libertarians ought to be very agnostic about the relationship between educational attainment and grasp of economics. That relationship may not be as strongly negative as Klein claims to have found (I will disclose here that I tend to believe that Klein is right.) However, it is certainly not strongly positive, as it would have to be in order to justify confidence in giving power to the educated elite.

Comments and Sharing

COMMENTS (22 to date)
Brian writes:

The reason why the adherents to the Church of Unlimited Government believe what they believe contrary to the evidence and common sense is because they understand correctly that the current political system allows our government, especially the Federal government to enact laws that can justify anything. Therefore, what they wish for can become law and can place whatever mandates it wants upon individuals and businesses.

There is almost no pushback from the other branches of government if it is held by the same party and judges can always be replaced.

The only area of law that has withstood the overreach of government has been the right of free speech and assembly. Although with the enactment of hate crimes laws, it too has been diminished.

Alex J. writes:

Brian: Campaign finance. Also, chilling effects on researchers due to pervasive government funding in academia.

Steve Miller writes:

I'm not sure why this study would trump Bryan's findings based on two large data sets, the SAEE and the GSS. I found pretty much the same thing as Dan Klein in my GSS study when it comes to left/right patterns, with the notable exception of trade/immigration issues. However, the Caplan finding re: education held. Education is definitely correlated with higher economic literacy and more "free-market" views in the GSS. So how much weight should we put on the eight question Zogby poll versus the batteries of questions in the SAEE and GSS?

Arnold, you've been making it sound like this is a contest between Bryan's idea versus Dan Klein's study, but it's clearly a contest between different data sources, and what reason is there to think one is better than the others?

Rebecca Burlingame writes:

Two things come to mind with your post and I will mention the serious one first. I have met dozens of sincere and intelligent people over the years who had a sincere desire to heal people but get pushed aside because their views are considered too alternative. Here is a huge part of our country's untapped potential. Such people would have had plenty of willing consumers as well.

Now for the lighter side. Why is it that in Wal-Mart, one can find coffee of all kinds but - at least where I live, almost no whole beans? Invariably I have to go to Starbucks afterwards. Do only the snooty types out there like to use whole beans and grind them at home?

Richard A. writes:

Here would be an interesting question --
Labor shortages are caused by employers underpaying their employees.

1. Strongly Agree
2. Somewhat Agree
3. Somewhat Disagree
4. Strongly Disagree
5. Not sure

My observation is that too many Republican officeholders get a question like this wrong.

Boonton writes:

Ironic isn't it that the left has enacted a major health care reform that seems to embody the economic truths you talk about. Expansion of benefits has to be paid for with cutting out other benefits that are high cost, low benefit. On the other hand, Tea Parties argued against it with exactly the economic ignorance that you say is so damming....that health care is a 'right' (if you're 65 or over) and even the most modest attempts to find savings is throwing poor grandma in front of a death panel to be 'rationed' into a coffin.

This isn't, though, anything new. Go back a few years and you have what would now be Tea Partiers clapping for a President whose contribution to health care reform was more entitlement without even a pretense of paying for it.

This reminds us of the difference between rhetoric and reality. Do Tea Party types 'get' the economic truths you are talking about or do they like projecting the idea that they are 'grownups' who 'get it' while those on the left like to project the idea that they 'care' but reality doesn't always indicate that this mask is quite true.

Coupon_Clipper writes:

Boonton- Great point.

Arnold, I don't think you can ignore how heavily flawed this study was though. As much as it's tempting to believe these results, there's a huge bias in the methodology, as was pointed out the first time this study was mentioned on econlog (or was it MR? I can't find the post).

Basically, here's the short version of the critique: If you knew nothing about economics but you were a Limbaugh ditto-head, you'd get most of these question right. But if you knew nothing about econ but were a lefty, you'd get these wrong.

You could imagine coming up with legitimate econh questions that have the opposite bias (Does monopoly represent a market failure? Which side of the Laffer curve are we on?) and you very well may get the opposite results.

Matt C writes:

I think your points a) and b) are a great pair of ideas to start off with when talking about U.S. healthcare. If your audience can't "get" those two facts--and I'm not surprised that many can't--there's not a lot of hope for further useful discussion. Just excuse most of them from the room and carry on talking with whoever is left. :)

Nikonman writes:

"The typical question from the audience asserted that people have a right to health care, and that settles the issue."

I've always found this position interesting, because there are not many other "rights" that force other individuals to give something up. In other words, for people to have a right to health care, doctors and hospitals have to forgo at least part of a payment, and taxpayers have to forfeit more of their income in taxes. If someone has the right to, say a liver transplant at a total cost from beginning to end of around $1 million, then a lot of people have to give up something to provide that right.

On the other hand, my right to free speech doesn't cost anyone else anything, and my right to freedom of association doesn't cost anyone else anything, and my right to freedom of religion doesn't cost anyone else anything. My rights to life, liberty, and the pursuit of happiness don't cost anyone else anything.

I guess one could argue that my right to rule of law requires taxation to support law enforcement, but that is a right that everyone enjoys, including free riders. That is more of a general welfare argument, whereas health care doesn't really fall under the general welfare umbrella, because it is very specific to the individual who receives the benefit.

What are some other examples of specific individual rights, that would not be included under general welfare, that cost someone else something?

Ted writes:

Can we stop using this obvious fallacious research? I already stated the more subtle problems in a previous post (, but now I'm just looking at the plain methodology and it's even more awful. For one, their sample isn't representative of the population. There aren't enough Hispanics of women, for example. Also, the fact they used an e-mail survey kills the results entirely. E-mail surveys are known among anyone who does polling to be the absolute worst research method ever. It's well known that internet-based polling of any kind produces absolutely awful results. Zogby, whose poll was used here, almost always performs worse than anybody's by large margins (often more than 2 magnitudes of error compared to more traditional methods). Particularly because Zogby's audience is pre-selected because members sign up for access to the surveys. That's not even scientific. That's simply a joke. Furthermore, it's well known that internet polls like Zogby's tend to draw salient voters who have strong opinions going either way. Strong opinions are often ideological and so given the questions would give a bias reading of the liberal population. There is a reason professional pollsters don't take Zogby seriously and nor should any study. It's amazing to me this was even published without anyone pointing out that Zogby's methodology is probably the worst out there and cannot be used under any circumstances.

Stop citing this study as though it means anything. It's actually garbage. The research design is so flawed I have to assume the authors either deliberately biased this research or they are just tremendously ignorant of how problematic their research design is. Someone should do a real study with a legitimate methodology because the question in and of itself is actually an interesting one.

Rebecca Burlingame writes:

There are times in life when debate can seem like how many angels are dancing on the head of a pen and it is okay. But this is not one of those times. In the not so far off future, most of us are not going to have adequate access to health care if this is not settled. And ultimately it is fairly simple. If most of us do not have the right to heal in some capacity, we certainly do not have the right to health care. The crucial surgeries people rely upon could happen styled in settings borrowing from factory models and most people with any intelligence could be trained for specific functions just as factory workers were once trained in our country.

Chris Koresko writes:

Coupon_Clipper: If you knew nothing about economics but you were a Limbaugh ditto-head, you'd get most of these question right. But if you knew nothing about econ but were a lefty, you'd get these wrong.

Of course the other reasonable interpretation of the above fact (assuming that it's true) is that Limbaugh spends a fair amount of time talking about economics, and he's usually correct (or at least arguably so) when he does. So the Limbaugh ditto-head actually knows something about it.

That said, I wonder how hard you would have to work to make a set of survey questions that left-leaning non-economists would tend to do better than right-leaning ones. Has anyone tried to do this?

Coupon_Clipper writes:

Chris, that's a very good point. But I would point out that there wasn't (to the best of my recollection) a single question that an economically illerate lefty would've gotten right and an economically righty would've gotten wrong. And I was able to provide two such questions off the top of my head!

That said, I wonder how hard you would have to work to make a set of survey questions that left-leaning non-economists would tend to do better than right-leaning ones. Has anyone tried to do this?

That's a good idea. I hope someone tries this.

Thanks for the reply.

John V writes:


Yes, you could make such a set of survey questions.

The problem with such a set is that the questions would revolve greatly around special case scenarios of market failure or break down and would probably be confusing in how they are worded because you'd have to account for a lot to set up the question AND word it to make sure it played toward lefty sensibilities so they had a better chance of "feeling" the right answer. It would almost have to sound like a riddle in order to not sound like a question with an obvious answer.

Of course, the other problem is that none of this necessarily means that economic righties would get them wrong. Just the heaviness and subtly of the question would probably put up red flags and make them think about it....which is the whole point.

Also keep in mind that this would no longer be about a grasp of basic economics. It would be about complicated and special-case scenarios where markets break down in a way that plays toward lefty sensibilities.

AND, as a libertarian myself, the problem I have with this last point is that the solutions to such problems that advanced econ may point to are not necessarily workable or reliable because: 1.of the narrowly focused models that produce them and, more importantly 2. because of the tool (govt.) that would have to legislate/implement them.

John V writes:

Here, coupon-clipper:

Can someone turn this:

...into such a question?

Miguel Madeira writes:


- Immigration creates unemployment?

- Tax cuts usually make deficits bigger?

- A balanced budget is better?

- If you restrict the supply of a product with a very rigid demand (like an addictive substance) the total spending buying this product will increase or decrease?

Miguel Madeira writes:

More suggestions:

- It is better for economic efficiency that we subsidize corporations to employ people instead of subsidizing unemployment?

[I imagine that the the correct answer is "Not sure", but probably many conservatives will mix the concepts of "efficiency" with "production" and will answer "yes"]

- A positive trade balance is a sign of a robust economy?

[Again, I imagine that the correct answer is "Not sure"; in Portugal, most right-winger will answer "yes" - the portuguese right-wing is obsessed with exports; I don't know how is in the US]

- A tax on pollution that forces many companies to close, or, at least, to reduce their jobs is good for the economy?

[Again, the correct answer is "not sure" (if the tax realy reflect or is lower than the cost of the negative externalities of these pollution, the economy - properly measured - is better now; if the tax overestimate the cost of pollution, the economy is worse; however, I imagine that conservatives will answer "Strongly Disagree"]

Paul writes:

Workers, tradesmen, small businessmen have no one, no party looks after them at all. I don't care what the parties say. The Democrat party is the party of union, or soon to be, government workers, or as I know them, the labor benefactors of the welfare plantation. They are the overseers that tend to the day to day needs of the party and follow the orders of the elite. The 'clients' are the crop of the plantation. The reason and source and purpose of the Democrat workers existence. So they see that the plantation is always breading more dependency to keep the plantation flourishing.

The GOP finds workers annoying. Tradesmen as above their station in life, and small business and their mewing petty concerns as positively threatening. The GOP is a Imperial/big finance/big gov party. Self supporting workers, tradesmen, businessmen are to the GOP what blacks are to the Democrat party. To be patronized, lied too, bought off against their own interest.

So, Wal-Mart shoppers might not of gone to elite prep schools, but they know they are getting it from both parties, and they know, slowly, that there is no free lunch, although the bills are dumped on them. They can not afford to be economically deluded. The elites don't want them, and the welfare swamp will destroy them as it has so many others.

I suspect most of the counter-intuitive results of the Klein study (e.g., the lack of correlation between education and enlightenment) were due to the question about licensed professionals. That put licensed professionals on their guard and may have even caused some of them to give deliberately wrong answers to help discredit the study.

Boonton writes:

I think your points a) and b) are a great pair of ideas to start off with when talking about U.S. healthcare. If your audience can't "get" those two facts--and I'm not

I notice on the right a lot of rhetoric is little more than burning straw men. In the so-called Church of Unlimited Government that thinks health care is a will most likely find that they too 'get' those 'facts'. If, for example, you asserted in your speech that since resources are limited gov't medical programs like Medicare should not cover certain things like plastic surgery except in some exceptional circumstances (say serious disfigurement or reconstructive surgery for women who've had mastectomies). You'd almost certainly find few who will 'boo' you even though if you really honestly thought unlimited access to medicare care was a right there should be no limit on grannies getting face lifts and breast implants paid for by Medicare!

The fact is the left understands these 'points' and 'facts' more often than you think. There's usually a lot more context, though, that you just opt to ignore. Likewise for all the 'understanding' the other side has, why is it that the Tea Party almost universally announced that any cuts or cost containment in Medicare was 'rationing' and 'death panels'? Even better, how many Tea Party endorsed candidates in the farm belt are running against subsidies and import quotas? Hmmmmmmmmm. You get an A on the multiple choice Econ 101 test. You get an F on application of principles!

Miguel Madeira writes:

Even more suggestions:

- "If, maintaining all other things equal, the state raise more money in taxes, there will be less money avaliable for private investment" (the "wrong" answer is "Strongly Agree": if the state receives more taxes and all the rest - including government spending - remains equal, there are less deficit, then less crowding)

- "Politics to reduce the difference between richs and poors make everybody poorer" (in these case, both "strongly agree" and "strongly disagree" are "wrong" answers - usually these politics reduce total wealth, but only in extreme cases reduces the wealth of everybody)

- "Societies with less differences between richs and poors usually have less economic growth" ("wrong" answer - "strongly agree")

And the cherry in the pie:

- "Taxes used for paying the Welfare State are a violation of natural rights" (like the question about exploited workers, the wrong answer is "strongly agree", because concepts like "exploitation" or "natural rights" does not make economic sense).

I wonder why, with so many question where the wrong answer is the intuitive right-wing answer, the authors only made questions where the wrong answers where the intuitive left-wing answer? Intellectual dishonesty or intellectual limitation (I suspect that is the second case)?

Chris Koresko writes:

@Miguel Madeira:

I find I have problems with most of your suggested questions. Most of them seem to have no clear answer, in some cases because the questions are not well posed.

For example: "Politics to reduce the difference between richs and poors make everybody poorer" Here the answer could be, "Obviously no such policy is going to make absolutely every single individual poorer, so the answer is no." Or it could be, "Strong redistribution reduces economic growth, and over the long run it will make the great majority of the people worse off in an absolute sense than they would have been with a more laissez-faire policy, so the answer is yes."

It is better for economic efficiency that we subsidize corporations to employ people instead of subsidizing unemployment?

[I imagine that the the correct answer is "Not sure", but probably many conservatives will mix the concepts of "efficiency" with "production" and will answer "yes"]

It's not obvious to me what the real answer to this question is, but I think most conservatives would focus on the word "subsidize" and argue that economic efficiency would be better if neither of these were subsidized.

I think if one were to make a serious attempt to measure the economic literacy of right- or left-leaning non-economists, it would make more sense to construct a list of textbook scenarios for which the correct answers are known, and preferably are derivable by straightforward arguments (which probably means we'd have to stick to micro for the most part).

Examples which come to my mind (without takin my own suggestion to consult a textbook) would be:

  • A tax on gasoline falls most heavily on which group: (a) people who own stock in oil companies; (b) people who work for oil companies; (c) people who buy gasoline
(As I understand it the correct answer is (c), since the elasticity of demand is low)
  • A small increase in the Federal income tax rate will likely cause the tax revenue to (a) increase; (b) decrease; (c) increase or decrease, depending on what the tax rate was before
(I think the answer is (c) here; this is essentially a question about what side of the Laffer curve we're on)
  • An increase in the minimum wage will likely cause (a) an increase in the number of employed workers and an increase in their average income; (b) a decrease in the number of employed workers and a a decrease in their average income; (c) a decrease in the number of employed workers and a an increase in their average income
(I think the answer is (c). Does anyone here disagree?)
  • The effect of laws requiring people who ride in cars to wear seatbelts is (a) a reduction in the number of traffic deaths; (b) an increase in the number of traffic deaths; (c) a decrease in the rate of pedestrians killed by automobiles
(This somewhat violates the requirement of being derivable by a simple argument and almost falls into the category of trivia. But I understand the correct answer to be (a).)
Comments for this entry have been closed
Return to top