Truth and Economics

by Russ Roberts on January 26, 2009

in Podcast

As far as I know, no prominent market-oriented economist has come out in favor of a trillion dollar increase in government spending as a way to improve the economy. Every market-skeptical economist that I have heard is in favor of it on the grounds that it will improve the economy. Each side claims to have empirical support for its position.

What does this tell you about economics as a science? What should a non-economist conclude?

In this podcast with Robin Hanson, we explore these questions and a few others. It's a little bit off the beaten track, style-wise. It's more of a lengthy confession by me of how my views on these issues have changed, followed by Robin's response. Robin has some very interesting things to say.

Be Sociable, Share!



86 comments    Share Share    Print    Email


John V January 26, 2009 at 9:28 am

What an excellent topic! And very pertinent too.

And who better to talk about it.

I will listen to it later today. But it sounds excellent already.

Thank You.

Jeremy January 26, 2009 at 9:47 am

What does it tell a non-economist? It tells me that economists are paradigm warriors, just like every other discipline in social science, and that the intellectual ego dominates all aspects of the ivory tower. It tells me that you can make statistics spin any story you want it to (so why again does having empirical evidence mean that your story is "true"?).

Actually, my problems with this type of mentality is a major reason I left academia.

Thanks for the podcasts; I listen to them at the gym. I enjoyed your discussion with BdM about Iran (is that man EVER wrong?) and your discussion with Fazzari about Keynsian economics.

Question: "Where is the empirical evidence that a massive stimulus of the economy is likely to work?"

Answer: "We, as economists, do not have a large database…"

Translation: There is no empirical evidence!

Martin Brock January 26, 2009 at 10:04 am

Frankly, I concluded long ago that "economic science" is a contradiction in terms. To be kinder, the science of economics is like the science of chemistry, when alchemy was still considered a legitimate part of it. A more scientific core exists, but the economic alchemists don't know that they're alchemists and neither do most other people, even most professional "economists".

To be fair, many people, even including many professional "physicists", still don't understand Quantum Mechanics, and some mathematicians say ridiculous things about implications of the Incompleteness Theorem.

I agree that the nominally "Keynesian" stimulus being discussed is alchemy; however, many nominally "market oriented" economists are also among the alchemists. They promise to make gold from lead by other means, but their promises are no more persuasive.

I once read an article on Social Security reform at the Heritage Foundation's web site (hopefully not by an economist) declaring the benefits of "private accounts" containing Treasury securities with a yield guaranteed to exceed the rate of inflation. This "market" (or "private property") solution would cost taxpayers more for the old age pensions and wouldn't address the Medicare problem at all.

In reality, the "market oriented" proposals for Social Security reform ignore the real economics of the problem as much as, if not more than, the Social Security system itself, because payroll taxes have never been remotely like "investment for retirement", either by design or in effect or by historical precedent, only in superficial appearance.

Payroll taxes are far more like the support of aging parents by their children, the tradition they actually replaced. Elder support is the purest form of consumption, the furthest thing imaginable from investment.

Replacing a massive amount of the purest consumption with nominal "investment", while pretending that the purest investment (child rearing) is "consumption", is economic nonsense, yet the "market oriented" economists preach this incredible gospel routinely.

Martin Brock January 26, 2009 at 10:17 am

It tells me that economists are paradigm warriors, just like every other discipline in social science, …

Paradigm wars are not unique to the social sciences, but they are more difficult to resolve in the social sciences.

Mathieu Bédard January 26, 2009 at 10:24 am

Very interesting premises. I can't wait to listen to this one.

JP January 26, 2009 at 10:24 am

I think the micro economist can claim what the macro economist can not, that they have sufficient data and therefore understanding. Thus the former can claim something near the domain of science while the latter must, if he is being honest, claim the mantel of an artist.

EvanM January 26, 2009 at 10:26 am

It tells me that, as everybody knows, Economics IS a science, and the market-skepticism view is the consensus view by scientists worldwide.

I mean, there's all kinds of proof, and if you deny it, then you're an anti-science, anti-progress firebrand, and you're probably funded by Big Oil, or the Heritage Foundation, or something.

And if that doesn't work, my Harvard Professor said so.

Martin Brock January 26, 2009 at 10:33 am

Question: "Where is the empirical evidence that a massive stimulus of the economy is likely to work?"

The answer depends on the meaning of "work". Economists suppose that "work" means something like "reorganize resources for real growth in the least time", but politicians mean something else, and so do most people frankly. For them, "work" means busying people while providing them paychecks, regardless of any real economic output as measured by a market.

In reality, these days in the U.S., real productivity can shrink quite a bit and no one much will care, as long as the lowered entitlement to consume is spread widely enough. Real GDP growth will almost certainly fall as the baby boom ages, because labor force growth soon will fall to nothing. GDP itself could even fall. That we hardly discuss this obvious, irrefutable and inescapable fact at this time is the best evidence I can imagine that "economics" is more an extension of politics than it is a science.

I'd like to see the average age of applicants for unemployment compensation and the average age of persons reporting unemployment in surveys over time. I'd also like to see the average of persons in the "long term unemployed" category excluded from official "unemployment" statisitics?

Is this age rising? If it is, are we observing "early retirement" masquerading as "unemployment". If not, is an unprecedentedly large population of persons in their fifties and sixties crowding younger workers out of particular labor markets? As many boomers realize that their retirement expectations are unrealistic, will a crowding out problem worsen? Why aren't we asking these questions at this time? Everyone knows what's happening. The denial is so thick, you could cut it with a knife.

Martin Brock January 26, 2009 at 10:41 am

And if that doesn't work, my Harvard Professor said so.

I suppose your Harvard Professor has a research grant from the Commerce Department and would like a post at the Fed or the World Bank.

cpurick January 26, 2009 at 11:33 am

Man, I've been dying for someone to ask something like this. I've been baking a theory of my own:

When a serious student of economics chooses a political philosophy he will be libertarian (or conservative) because it's consistent with what he already knows.

But if a liberal studies economics he will be a Keynesian, because it's just about the only thing in economics that's not inconsistent with what he wants to believe.

cpurick January 26, 2009 at 11:43 am


Unless Keynesians want to admit that they're prepared to put America through something like World War II, the track record of straight "public works" spending ain't all as hot as they crack it up to be. If they seriously apply Keynes to this recession, it will be depression.

Like pulling off a band-aid as fast as you can, I just hope they've inflicted their damage by next November.

Martin Brock January 26, 2009 at 11:59 am

"Like astronomy and number theory, pretty, fascinating pictures but little utility …" How much "hard science" falls into this category? The science is "hard" but marginally useful.

General Relativity is a good example. As far as I know, the only practical application it has, to date, is in the GPS, or so I'm told; however, I suppose GPS engineers would have constructed useful corrections without GR and would have accounted for these corrections theoretically somehow, so even in this "hard science" domain, it's tough to know how much observation supports a particular theory.

Mark Denovich January 26, 2009 at 12:04 pm

So "economists" who fail to understand markets, also fail to understand the absurdity of this fiscal stimulus package.

What exactly is the mystery here?

Greg Ransom January 26, 2009 at 12:57 pm

Hayek's point is that economists are looking in exactly the _wrong_ place for the empirical facts grounding the explanatory power of economics.

Empirical fact:

individual knowledge is limited

Empirical fact:

the world presents us with a place in which economic efforts are massively coordinated

Empirical fact:

there is a logic of valuation, which human beings use to order their affairs.

Empirical fact:

This logic of valuation extends to production resources

Empirical fact:

Most production resources are non-permanent

Empirical fact:

Relative prices transmit knowledge beyond that known to a single mind


From these empirical facts, you get Hayek's enormously powerful empirical explanation of the trade cycle.

Most of the "data" you find in "macroeconomics" is trace data which needs to be explained.

It isn't stuff governed by simple two-variable "laws", or and law-governed piece of statistical mathematics.

And these social constructed data points are _not_ governed by the pseudo-science of "Keynesian economics".

Jeremy January 26, 2009 at 1:13 pm


Thanks for the link. I have bookmarked it!

I also read your interesting story about Obama here:

I have got to say, it scared me!

Greg Ransom January 26, 2009 at 1:14 pm

The biggest issue here is that most all economist have a deeply false picture and conception of "science" — one inherited from philosophers and the philosophical tradition, not from science (you can actual trace this historic fact).

By the lights of the economists, Darwinian biology would not be considered science. Nor would linguistics.

Most economists do "science" wrong — exactly why Hayek invented the word "scientism".

Current January 26, 2009 at 1:56 pm

Martin Brock, the complaint about Social Security and other programs is how prices are factored into things.

The cost of these programs will be paid for in taxes. Where I live in the UK this is already the case since our baby boom happened earlier.

Whether they are paid out of taxes or from investments which are similar to taxes (like regular dividends from shares) is superficially irrelevant.

The real problem is what is factored into the market prices for investments now. Markets cannot easily factor in the effect of the tax increase that will be spread across the economy.

It is true though that some economist don't explain this.

dg lesvic January 26, 2009 at 2:20 pm

Economics is not an empirical but a logical, or, more precisely, praxeological, science.

It starts out empirically.

"Empirical fact enters into the theory, but only at the level of basic axioms," such as the disutility of labor, the variety of resources, and the goods on the shelves. But the theory itself is "a priori to all other historical facts."

Rothbard, America's Great Depression, P 81 and P 305, note 4.

Since they are all complex, "Every historical experience is open to various interpretations and is in fact interpreted in different ways…History can neither prove nor disprove any general statement."


"Action and reason are congeneric and homogenous…two different aspects of the same thing. That reason has the power to make clear through pure ratiocination the essential features of action is a consequence of the fact that action is an offshoot of reason…Logical thinking and real life are not two separate orbits. Logic is for man the only means to master reality. What is contradictory in theory is no less contradictory in reality."


Mike Farmer January 26, 2009 at 3:24 pm

As a non-economist I conclude that economics is not a viable guide for central planning. Human intereactions are too complex, and if interactions are left to the free choices of free people, without violating the rights of others, then the economy is dynamic. All an honest economist can do is try to be on the edge of the change, interpreting what's happening to help people make the best decisions. But, to me, trying to plan the economy based on economic model is useless and counter-productive. A good economist is probably more valuable predicting the negative consequences of bad economic models than creating a working model to plan and control the economy.

Lance January 26, 2009 at 3:34 pm


How about Martin Feldstein or Kenneth Rogoff? They've advocated stimulus proposals, yet they're pretty much aligned with free markets.

Randy January 26, 2009 at 3:42 pm

Mike Farmer,

"A good economist is probably more valuable predicting the negative consequences of bad economic models…"

Agreed. Because politics is exceptionally prone to irrational exuberance.

dg lesvic January 26, 2009 at 4:38 pm

Mike Farmer,

You're a better economist than you give yourself credit for, for you instinctively understand what economics is and isn't, and who is and isn't an economist.

Since we live in a world of scarcity, we must all economize. But that doesn't make us all economists. Economics implies something beyond Home Economics, the problems of Ma Kettle in her kitchen; it implies Political Economy, the problems of nations and social classes in the market and the political arena, and not its passing data but eternal truths.

The data by itself is a meaningless jumble. It is only through the logic of economics that we can make sense of it. But while the logic is essential to an understanding of the data, the ever changing is irrelevant to the eternal logic. So though the macro-economic plotters of the data and "governors" of the nation's economy use economics, they do not contribute to it. And, however more complex the data of the overall economy than that of the corner grocery store, they are still more like Ma Kettle, adjusting to changes in the data, than Adam Smith, discovering eternal truths, and not economists, like Smith, but economic "managers," like Ma Kettle, though on a larger scale, and, unlike Ma, doomed to failure.

Oil Shock January 26, 2009 at 6:48 pm

They have got "data" to prove it. Empiricism gone wild.

The longer a trader’s ring finger,” suggests Agora Financial analyst Sam Buker in Whiskey & Gunpowder this morning “the more money he’ll make.” Sam is citing a Cambridge study in which the U.K.’s leading minds went around London measuring digits of various investors, which might help explain why the pound is tanking.

“Those with the longest fourth digit made over five times the money of their less-well endowed colleagues. The average salary was $537k. The long-ring group netted a healthy $828k, while the shorties had to make do with $145k.

Dr. T January 26, 2009 at 8:02 pm

What does this tell you about economics as a science?

I like economics, but the claim that it is a true science irks me. The scientific method, the foundation of natural sciences such as physics, chemistry, and biology, almost never is used in economics. At best, economics barely qualifies as a social science. There are observations, hypotheses, more observations, correlations, models, and conclusions, but there are few useful experiments that can challenge the conclusions.

Economics should just define itself as itself. There is no need to claim it is a science; it is misleading to do so.

cpurick January 26, 2009 at 8:08 pm

Hmm. I distinctly remember hearing that the ratio of the ring finger to the index finger is a strong indicator of whether you're attracted to men or women. I don't think it matters if you're a man or a woman, if you get my drift. Apparently it has something to do with estrogen in the womb. Longer ring finger means you'll like women, longer index finger means you'll like men.

Not surprisingly, Oil Shock's study says the high testosterone crowd makes all the money.

This no doubt irritates some women, especially ones with really long ring fingers.

Mike Farmer January 26, 2009 at 8:40 pm

dg lesvic, thanks, that makes a lot sense.

Randy, irrational exuberance, indeed. I'm dizzy with all the plans put forth. As a real estate broker, I have no idea how to predict in the midst of all this manipulation. Buyers and sellers are probably a little confused, too.

Martin Brock January 26, 2009 at 8:51 pm

… the complaint about Social Security and other programs is how prices are factored into things.

My complaint about Social Security is about how costs are not factored into things. The historical precedent for Social Security is not "saving for retirement". The precedent is "supporting elderly parents"; however, because the system is dressed up superficially to resemble "saving for retirement", it is extremely inequitable for parents and extremely disrespectful of children.

In the popular imagination, Social Security says, "Give money to the government, like an investment, and you earn entitlement to benefits when you retire." In reality, the program says, "Support your parents collectively, and you may expect all children collectively to support you." Given the history and the pay-as-you-go nature of the program, it can't be saying anything else.

Social Security is nothing like investment in non-human capital, and no economist with a clue will claim to know that replacing Social Security with such a system is even possible, because no historical precedent for universal retirement by purchasing non-human capital exists. In reality, labor is the most valuable resource in practically all productive endeavors. I'm not asserting a labor theory of value, but the labor theory of value is approximately true insofar as labor is most of the value of everything human beings value. That's not just my opinion. It's the opinion of Ludwig von Mises and Julian Simon and many other, classically liberal economists.

Since we all have parents but we don't all have children, Social Security is obviously a boon to people who support few or no children. Deadbeat dads (and moms) do far better under the system than supportive parents, because they "earn" the support of their children by supporting their parents, not by supporting their children. These non-supportive parents may use earnings not directed toward the support of their children to buy entitlement to other rents imposed on children, like Treasury notes or real estate yielding more conventional "rent".

Thus the people earning the least of this rent imposed on children are the parents who invest hundreds of thousands of dollars in their support and therefore may purchase less of the entitlement. This consequence of the system is undeniable, yet the denial is palpable.

Martin Brock January 26, 2009 at 9:02 pm

Furthermore, the true libertarians who opposed Social Security at the time of its enactment opposed it for the reasons above, not because its "rate of return" was too low. One doesn't earn a "rate of return" on the support of one's parents, because this support is itself a return on investment. Expecting a "return" on this support (or on payroll taxes) is like expecting a return on your mortgage payment.

dg lesvic January 26, 2009 at 11:53 pm

Dr T,

You wrote,

"The scientific method…almost never is used in economics.

No other method is used, by real economists.

The Scientific Method is the method of the controlled laboratory experiment, of a single factor of change in an otherwise unchanging environment.

The economist uses that method no less than other scientists, with only this difference. His "laboratory" is within rather than outside his own mind. For, while he cannot create an actual environment of no change, he can imagine it.

And this method of imaginary constructions, the specific method of economics, is as universally valid, objective, and scientific as any other.

Babinich January 27, 2009 at 5:46 am

Mike Farmer on Jan 26, 2009 @ 3:24:46 PM

"As a non-economist I conclude that economics is not a viable guide for central planning. Human intereactions are too complex, and if interactions are left to the free choices of free people, without violating the rights of others, then the economy is dynamic."

Does 'Central Planning' becomes more practical to social engineers if free choice goes cockeyed because of the implementation of incentive altering legislation?

I'd answer this question with a resounding "Yes"!

Mike Farmer January 27, 2009 at 6:48 am

If you are asking if social engineers use central planning to correct the negative consequences of their incentives, I'd say yes, thus the vicious cycle of intervention. But this is not a condemnation of free choice, but rather condemnation of social engineering, central planning and "incentives".

Mike Farmer January 27, 2009 at 6:51 am

One too many "but"s.

Paul Ralley January 27, 2009 at 7:32 am

I have a backlog of econ talks to listen to, but I shall promote this to the top of my list.

Thanks for this great service, it gives me a massive consumer surplus!

Mark Yannone January 27, 2009 at 10:08 am

Before you argue about truth, look at the authority. If government has no authority to act in a certain way, then there is no need to debate its effectiveness. For instance, we have no need to argue about the effect of violating the Second Amendment. Your economic analysis of a violation of the Bill of Rights is of no interest.

That takes care of 82 percent of the issues, and I've got a study to prove it.

ChrisF January 27, 2009 at 10:57 am

What if trying to figure out what happened to the economy is equivalent to trying to figure out why a single wave in the ocean happened to be bigger than its neighbors?

Macroeconomics is concerned with discerning how broad ranges of people will act under certain conditions. So, for example, if you put a tariff on, say, French wine, there is a view that this will cause French wine producers to complain to their local representative in the French Government, who will talk to his other friends in the French Government and they will, collectively, decide to respond by imposing a tariff on American cars. That's a lot of predicting. What if the wine makers just say "Good Riddance" and don't lobby? Or, what if the French Government is busy dealing with some other crisis?

The point is that macroeconomics is all about looking at the behavior of a chaotic system and trying to figure out what a steady state will look like given a certain set of inputs. Sometimes, this is possible. But, I suggest that sometimes it is completely impossible. The ocean is chaotic enough that it will never be possible to understand why any specific wave happened and, even if you could figure that out for one wave, it won't help you with the next wave.

dg lesvic January 27, 2009 at 12:47 pm


You wrote,

"The ocean is chaotic enough that it will never be possible to understand why any specific wave happened and, even if you could figure that out for one wave, it won't help you with the next wave."

Economics isolates each factor of change from every other, telling you the effect of each by itself.

Isn't that worth knowing?

Greg Ransom January 27, 2009 at 1:41 pm

This is a fools version of "science". It essentially excludes most of astronomy and most of Darwinian biology, two of our most outstanding sciences.

This model of "science" has done a tremendous amount of damage in economics and the social sciences.

Someone wrote:

"The Scientific Method is the method of the controlled laboratory experiment, of a single factor of change in an otherwise unchanging environment."

dg lesvic January 27, 2009 at 3:26 pm


If that was directed to me, I don't know what you're talking about.

ChrisF January 27, 2009 at 4:20 pm

dg lesvic –

My argument would be that it may be impossible to identify the change due to any one factor. A butterfly flaps its wings and there is a hurricane on the other side of the world.

The standard approach for what you're doing is "Holding everything else constant." That works very well if, for example, you have F(x,y) = x + y and you want to hold y constant to see what F does when you change x. But, let's say you have F(x,y) = (x if y is odd, y if y is even). If you hold y constant, you will never discover F. It's even worse if F is a function of 1000 variables or 10,000,000 or 4 billion, each one of which is a person.

dg lesvic January 27, 2009 at 7:08 pm


You wrote,

"…it may be impossible to identify the change due to any one factor."

For instance?

dg lesvic January 27, 2009 at 7:14 pm


Ignore my last post. That was no good.

Here's the solution to your problem.

F equals ABC times XYZ divided by CRAP amd multiplied by BULLSHIT.

dg lesvic January 27, 2009 at 7:28 pm


You said that my approach works very well if there are no constants and doesn't if there are constants.

That's what I've been saying, too, so where is the disagreement?

You may be right about the butterfly. We may never know all of the effects of its actions. We may never all of the effects of a given human action. We may never know all there is to know about the world. And probably won't. Nay, certainly won't.

But does that mean that we can never know anything about it? Does that mean that all scientific inquiry is vain, and should be halted?

All I hear from people like you and Martin is that we shouldn't think.

Then, what should we do, join the Taliban?

dg lesvic January 27, 2009 at 9:18 pm


I wrote;

"You said that my approach works very well if there are no constants and doesn't if there are constants."

I meant,

You said that my approach works very well if there are no variables and doesn't if there are variables.

Sorry about that.

Previous post:

Next post: