Russ Roberts: Why Keynesians Always Get it Wrong (and Most Economists Too)

"Economics as practiced by most of the profession...is full of hubris and should be much more full of humility," says former George Mason University economist Russ Roberts, NPR commentator, host of the popular podcast Econ Talk, blogger at Cafe Hayek, and a research fellow at the Hoover Institution.

Throughout the fiscal crisis, Roberts has attacked economic stimulus plans promulgated by left-of-center Keynesians and right-of-center monetarists. Such plans, he says, almost always fail to deliver the desired results because they're predicated on a fundamentally flawed understanding of "the economy" as a single entity that can be directed one way or another. (This is one of the themes of the "Fear the Boom and Bust" rap videos Roberts co-created at Econstories.tv.)

Which isn't to say that there aren't better or worse policies to follow. We may not be able to control the economy says Roberts, but we can make it easier for it to function by following what he calls "general principles that are pretty timeless." These include a government that lives within its means and investors who are held accountable when their investments go belly up.

Reason TV’s Nick Gillespie sat down with Roberts to discuss the failure of macroeconomics, his move from the free-market bastion George Mason University to Hoover Institution, and his new online project, “The Numbers Game.”

About 6 minutes.

Produced by Joshua Swain, with help from Amanda Winkler.

Scroll below for downloadable versions and subscribe to Reason TV's YouTube channel to receive automatic notification when new material goes live.

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  • Mike Laursen||

    Russ rules!

  • sarcasmic||

    aye

  • The Late P Brooks||

    Because they are collectivist who believe in "Society"?

  • 0x90||

    People who deal in absolutes always get it wrong, sooner or later.

  • T o n y||

    Or in the case of Austrians, always.

  • 0x90||

    * whoosh *

  • ||

    Still pretending economics is a science, Tony?

  • Calvin Coolidge||

    Propose hypotheses, gather evidence to test them, confirm or reject. Reformulate as needed.

    If that isn't science, then what is?

  • Redland Jack||

    I think the problem might be that economics is really three different things: Micro, macro, and econometrics.
    Since metrics is basically statistics, it is relatively solid.
    Micro principles are agreed upon by almost all economists. It can reasonably be called a (social) science.
    Macro, on the other hand, is still in an immature phase. There is no consensus in the field. Results are driven by assumptions that are sketchy. Most models can't be broken down into micro principles. At least in the departments I was in, macro is held in much lower 'esteem' than micro (or metrics).

  • BarryD||

    "Sketchy assumptions" or "ulterior motives"?

    Paul Krugman has convinced me of the latter.

  • Redland Jack||

    I think they go (somewhat) hand-in-hand. In macro, it is pretty easy to get the results you want by changing your assumptions. If you have a hardcore political bias, you can probably drum up a model that supports it (consciously or subconsciously).

  • Calvin Coolidge||

    See my post below on Keynsian "macro" models, which are what most people think of when they think of economics.

    The Lucas Critique in the late-60s first pointed out that moving curves around in a macro model often violates fundamental microeconomic principles.

    Still, it is difficult to conduct a controlled experiment in economics. Or, at least, tens of millions tend to starve or be butchered whenever a Savior of the People comes along to try it.

  • R C Dean||

    Micro principles are agreed upon by almost all economists.

    Consensus?

    It can reasonably be called a (social) science.

    So, not really a science.

  • Redland Jack||

    I guess so. I mean, we talk about the 'law of supply and demand', which I guess doesn't have the same absolute nature as the 'law of gravity', insofar as you could, conceivably, raise children in an environment where raising the price caused them to desire more of the object. Or, more plainly, it only holds as long as humans behave like humans. In some far future, maybe people will be entirely different...
    And, yeah, it could (pejoratively) be called a consensus. For instance, I think almost all microeconomists would agree that utility increases with wealth (and probably at a decreasing rate). I think the only economist who don't agree with most of the micro principles are the Marxists. So, it wouldn't be impossible for the consensus to be overthrown by new research (though the same would hold true for physics, biology, etc.)

  • OldMexican||

    Re: Redland jack,

    I mean, we talk about the 'law of supply and demand', which I guess doesn't have the same absolute nature as the 'law of gravity'

    Oh, yes it does.

    insofar as you could, conceivably, raise children in an environment where raising the price caused them to desire more of the object.

    Depends on what you mean by "price." It also depends on their opportunity cost.

  • Redland Jack||

    You can, theoretically, have Giffen goods. I'm not sure what empirical support is available for them, though...

  • OldMexican||

    Re: Redland Jack,

    I think the problem might be that economics is really three different things: Micro, macro, and econometrics.

    It is not, it is not and it is not. There's only economics.

    Since metrics is basically statistics, it is relatively solid.

    It is also pointless, since we're talking about people here, who are (and will always be) fickle.

    Macro, on the other hand, is still in an immature phase. There is no consensus in the field.

    Oh, there is a consensus, all right. Keynesians and Neoclassicals think that lumping together different demand schedules and calling it "aggregate demand" is a way to do science. Austrians all say they're full of shit.

    Austrians have been proven right.

    Most models can't be broken down into micro principles.

    That's because most models are bunk.

  • The Derider||

    LOL, "Austrians have been proven right"

    Austrians reject empirical analysis, so they certainly don't think any data will prove them right or wrong.

  • OldMexican||

    Re: The Derider,

    Austrians reject empirical analysis,


    That is not true. Austrians simply do not rely on empiricism to formulate theorems.

    so they certainly don't think any data will prove them right or wrong.


    That's a very clumsy strawman, Joe.

  • The Derider||

    Not relying on empiricism to formulate theorems = rejecting empirical analysis.

    They don't care if their theories match reality. At least they're up front about it.

  • OldMexican||

    Re: The Derider,

    Not relying on empiricism to formulate theorems = rejecting empirical analysis.


    Historians rely on empirical analysis all the time. Yet a mistake by some of them is to derive a theory of history through events that already happened, from choices made by different people already dead. Case in point: Marx and his totally debunked theory of history based on class struggle.

    Also, what you propose is a false dilemma: You can formulate theorems based on a prioristic axioms and not empiricism. Or, how do you think most of mathematics works, Joe?

    Unfortunately, it is clear you're nothing more than a dumb socialist with even less knowledge and insight than Tony, which is saying much.

    They don't care if their theories match reality.

    They do match reality, precisely because they explain human action - i.e. reality, Joe.

    For instance, people will choose from different preferred choices. Now, tell me that that is not true. C'mon, Joe, tell me that ain't so.

  • ||

    Austrians reject empirical analysis, so they certainly don't think any data will prove them right or wrong.

    Joe we have piled empirical studies on you left and right that prove stimulus does not work and that our debt hurts the recovery and that more taxes will not work and that spending cuts are necessary.

    Yet you have ignored it all.

    You really should not be making spurious claims about who and who does not reject empirical evidence. You are the worst offender in that regard.

    Hell at least Shrike actually makes an argument.

  • The Derider||

    No, you certainly haven't piled empirical studies that prove stimulus does not work.

  • OldMexican||

    Re: The Derider,

    No, you certainly haven't piled empirical studies that prove stimulus does not work.

    Oh, so you're living in another planet. I see.

  • Redland Jack||

    I'm not the world's biggest fan of econometrics, but it can often be a useful tool, albeit more of a 'practical' tool than an 'academic' one. (That is, I wouldn't want to base any theories on the results of econometrics, but if I had to decide whether my business should make 2 phone calls of 30 minutes to clients or 3 phone calls of 20 minutes, I'd probably use econometrics).

    I hadn't thought that the Neoclassicals spent much time with aggregate demand, but I never really cared much for macro, and it has been 8 years or so since I was in an economics department.

    I'd agree that if a model can't be broken down into micro principles it's probably not good. But just because most models are bad doesn't mean they have to be...

  • OldMexican||

    Re: Redland Jack,

    That is, I wouldn't want to base any theories on the results of econometrics


    That's the point. The difference between the empirical sciences and economics is that a) things don't have will, whereas humans do and b) because things don't have will, they follow rules that can only be observed and derived through empirical experience and mathematical models, whereas human behavior can be anticipated - since humans are rational - and thus does not require either empirical observation or mathematical models.

  • AZ||

    Well, the "gather evidence to test them" part can't really be done for macro, due to only having one world, with near infinite complexity. There's never a control against which to test ideas, and there are third (and fourth, fifth, etc.) variables and directionality issues all over pretty much any statement of causation.

  • The Derider||

    Also true in Geology and Astronomy, and yet these are sciences which make highly accurate predictions.

  • OldMexican||

    Re: The Derider,

    Also true in Geology and Astronomy, and yet these are sciences which make highly accurate predictions.


    Geology is not a predictive science.

  • The Derider||

    Yes, it is.

  • OldMexican||

    Re: The Derider,

    Yes, it is.


    No, it is not, you nitwit. A lot of geology happens because of LIFE, and LIFE cannot be predicted.

  • Redland Jack||

    It's also the case that most things in nature are far less complicated than people.

    E.g., when you build a model that includes people, and they learn about the model, they may change their behavior, thus invalidating the model.

  • T o n y||

    I dunno if I ever pretended that. Obviously you're suggesting that in the absence of a science, faith is an appropriate placeholder.

  • Restoras||

    You have continually insisted that there is plenty of empirical proof of your economic theories, whatever they are, while dismissing those economic theories that lack empirical proof. Presumable, when you use the term empirical, you are referring to data that can be perfectly explained by obervable, and testable, phenomena.

  • Restoras||

    I know you won't read this, but I'll suggest it anyway. The author (an Ivy Leaguer, you should really love him since he is part of the credentialed class) explains how the underpinnings of much economic theory from the 20th century, at least the theories that purport to be science through the use of equations that can only be used in the complete suspension of believable circumstances, is tremendously flawed.

    Origin of Wealth

  • OldMexican||

    Re: Tony,

    Or in the case of Austrians, always.


    You only say that because you want thievery security.

  • Calvin Coolidge||

    Absolutely!

  • Calvin Coolidge||

    It is the long run, and Keynes is dead.

    His models are still taught in Econ101, because they are simplistic and teach the concept of economic modeling without actually requiring any mathematical background.

    Unfortunately, his models are also fundamentally flawed and have been discredited even among New Keynsians since the 1960s.

    Equally unfortunately, people who only took one economics course ever, such as Mr. Robert Reich, still swear by the models, as their conclusion that government should spend, spend, spend whenever the economy is in a downturn is politically convenient to them.

  • BarryD||

    "politically convenient "

    That sums it up.

  • Tak Kak||

    You really won't learn much actual Keynes in 101, he won't come into the picture until Macro. And even there it's more New Keynsian IS-LM curves and whatnot.

  • Redland Jack||

    It probably depends on whether you're under a quarter system or a semester system. Under the semester system, Econ 101 is typically a combined micro/macro.
    That being said, I agree with you that it is typically Keynesian cross stuff (if I'm recalling the proper term).

  • Restoras||

    As has been pointed out elsewhere today, Keynes also advocated for significantly restrained government intervention in the boom times, and the repayment of everything that was borrowed. I love how all these "Keynesian" assholes conveniently forget that.

  • The Derider||

    Umm, we don't.

    Remember Al Gore and his social security lock box? Same principle.

  • Virginian||

    Oh, yeah the lock box full of IOUs.

  • BoscoH||

    Russ is awesome. Probably the most important person in figuring out the financial crisis. I wonder if Russ and Nick are working out together. Seems that The Black Izod has replaced The Jacket of late.

  • triclops||

    Russ is awesome, I look forward to Econtalk every Monday.
    I wish I could have gone back in time and told myself to go to George Mason University and major in econ. Some pretty amazing peeps there.

  • Redland Jack||

    Yeah. In hindsight, I probably should have gone to graduate school at George Mason instead of Berkeley.
    It might shock you, but Berkeley doesn't have a lot of variety on the macroeconomics front (micro, on the other hand, was fine).

  • The Derider||

    The fact of the matter is that Austrians predicted massive inflation as a result of the stimulus and quantitative easing, and it simply has failed to materialize.

    How convenient that they refuse to use empiricism to formulate their theories.

  • Jordan||

    All that liquidity won't stay locked up forever. It's just not possible. Meanwhile, Keynesians predicted economic recovery as a result of the stimulus and quantitative easing. And Austrians predicted the housing bubble in the first place, while Keynesians encouraged it.

  • OldMexican||

    Re: The Derider,

    The fact of the matter is that Austrians predicted massive inflation

    World, may I present the idiot that still believes inflation = higher prices.

    We already have the inflation, Joe - it's called continuous creation of money.

    We haven't seen the money distributed rapidly because the banks are keeping it in reserves while it receives interests from the Federal Reserve, but the moment the Fed stops paying interest on that money, the banks will lend and you will have your [even] higher prices.

    Besides that, we DO have massive inflation. You just want to believe the fudged numbers from the government but the housewife knows there's high inflation, and so does John Williams of Shadowstats.com

    The problem is not with the Austrians but with your total lack of knowledge of how the banking system works. The inflation is THERE and is growing.

  • The Derider||

    OK, so you think all the economic indicators showing low inflation rates are part of a government conspiracy.

    Good luck with that.

  • OldMexican||

    Re: Th Derider,

    OK, so you think all the economic indicators showing low inflation rates are part of a government conspiracy.


    You really don't know how government calculates CPI, do you, Joe?

  • Rufus J. Firefly||

    Have you seen the price of food lately?

  • The Derider||

    If inflation =/= higher prices, then how would a housewife be able to detect high inflation?

    You're a moron.

  • OldMexican||

    Re: The Derider,

    If inflation =/= higher prices, then how would a housewife be able to detect high inflation?


    The same way one detects electricity by turning a light switch.

    Electricity is NOT light, yet we know the effects when we turn on the switch.

    You're outmatched, Joe. You lost the spark a long time ago.

  • robc||

    If inflation =/= higher prices, then how would a housewife be able to detect high inflation?

    By looking at the change in M2, the same way anyone else would detect it.

  • Mr Whipple||

    Price inflation lags monetary inflation by 3-5 years.

    See: An Inflation Primer, Palyi (1960)

  • Mr Whipple||

    And yes, Palyi uses empiricism. Not that that really matters since Economics is not a "hard" science.

  • ||

    Don't quote studies for Joe. His reactions to them are predictably groan worthy

    I can even write his response for him:

    "Oh yes this stuff interests me a lot. Now i will complete ignore it and say "It is only one study and therefor does not count""

  • The Derider||

    How about something from a peer reviewed economic journal from the past 2 decades?

  • ||

    And goal posts go ZOOOOOOOOM!

    Then again, consider that Derider believes HuffPo and Democratic Underground to be esteemed, peer-reviewed journals of economics.

  • robc||

    There is no such thing as price inflation. Inflation is monetary.

    "Price inflation" is just changes in prices.

  • robc||

  • Danno||

    The economy is non-linear which means it cannot be predicted like house takings at a casino. Ask the dummies to predict the weather and when the Sun will blow up.

Click here to follow Reason on Instagram

GET REASON MAGAZINE

Get Reason's print or digital edition before it’s posted online

  • Progressive Puritans: From e-cigs to sex classifieds, the once transgressive left wants to criminalize fun.
  • Port Authoritarians: Chris Christie’s Bridgegate scandal
  • The Menace of Secret Government: Obama’s proposed intelligence reforms don’t safeguard civil liberties

SUBSCRIBE