Arnold Kling  

In Defense of Macroeconomists

PRINT
How Persuasive Is a Reputation... Notes from the Monterey Tea Pa...

Menzie Chinn makes the case.


How can one incorporate the idea of the importance of the banking system (separate from its importance to the money creation process)? One very simple static model is provided by a twenty year old paper by Bernanke and Blinder.

I am not persuaded that macroeconomists had a good model for this crisis. I think that Peter Schiff comes across as much better than the average macroeconomist.

I'm supposed to do a bloggingheads next week witih Mark Thoma on macro, and I am going to be making the case that macro is really off base. I think that one of my issues with macro is that it tries to look like physics, with laws and constants.

But picture Isaac Newton scribbling down the laws of physics every day, and every night God comes in and erases the scribblings, changes the laws, and fiddles with the constants. That is what macro is like.


Comments and Sharing


CATEGORIES: Macroeconomics



TRACKBACKS (1 to date)
TrackBack URL: http://econlog.econlib.org/mt/mt-tb.cgi/1738
The author at Belligerati in a related article titled Quote of the day writes:
    But picture Isaac Newton scribbling down the laws of physics every day, and every night God comes in and erases the scribblings, changes the laws, and fiddles with the constants. That is what macro is like.In Defense of Macroeconomists I... [Tracked on April 16, 2009 1:48 PM]
COMMENTS (17 to date)
Greg Ransom writes:

Hayek invented a word for this. It's called "scientism".

Arnold wrote:

"I think that one of my issues with macro is that it tries to look like physics, with laws and constants."

The fundamental rule of economics is that you MUST be scientistic in your work and you must never question that approach. Never.

Never reflect on it, never second guess it. Never think hard about it. Keep the blinkers on, and never mention the naked emperor prancing about proclaiming what a "real" scientist he is.

Greg Ransom writes:

Imagine if Darwinian biology was dominated by "scientists" who thought that Darwinian biology had to look like Newtonian physics with laws and constants.

In fact, philosopher of biology Michael Ruse once attempted to put Darwinian biology in the classic scientistic "covering law" model of science -- it was a gigantic failure.

When modern macro was created, the most influential economists thought this positivist / covering law picture of science WAS science -- read Bruce Caldwell's work on this to get some idea. You can find Paul Samuelson talking about Mach and "operationalism" and econometrics and all sorts of philosophy of science whooee -- the picture of "real" science that Samuelson had in his head.

It turns out the Mach / positivist / Friedman / operationalist / covering law picture of science does not even apply to the simplest models of Newtonian physics, it's most plausible application.

It's a real joke when applied to complex phenomena or sciences of complex phenomena such as Darwinian biology or neuroscience.

Greg Ransom writes:

Arnold writes:

"I think that one of my issues with macro is that it tries to look like physics, with laws and constants."

Hayek discusses this issue in his essays "Degrees of Explanation", "The Theory of Complex Phenomena" and in his Nobel Lecture "The Pretense of Knowledge".

pushmedia1 writes:

How does one study a system if one denies there exists a system? Are you saying there is no macroeconomics? there are no general equilibrium effects?

Keith writes:

I'm no economist so maybe I'm missing something, but simple logical models that predict general trends and general results make more sense to me than what I perceive to be the current methods. Based on my limited experience, there sees to be more emphasis put on empirical statistics and mathematical modeling instead of basic theory.

I'm a layman but something like Friedman's Free to Choose, Read's I, Pencil or Hazlitt's Economics in One Lesson makes much more sense than something like Keynes General Theory. I've tried (with limited success) to understand things like Game Theory or statistical models. All of that makes me wonder why macro exists at all when micro seems to make more sense in simple logical terms. I can't mark it down as my lack of intelligence, as very few who subscribe to current methods foresaw what has happened.

I'm inclined to accept the "Austrian" explanation (which Schiff agrees with) as it's the only one I've heard that seems to accurately describe in simple terms what has happened so far.

HH writes:

Arnold, you are basically mirroring an argument I made earlier this year:

http://akingsmind.wordpress.com/2009/01/26/like-physicists-around-a-god/

Lee Kelly writes:

Although macroeconomists are generally very competent at what they do, most do not seem to know what they are actually doing.

Modern economists have inherited a set of traditions and assumptions about the economy and its method of study produced by those Hayek called "contructivist rationalists", or "scientists" in the sense that Greg Ransom mentions. These traditions and assumptions are now deeply embedded in the way that modern economists think, regardless of whether they would actually agree with those ideas if explicitly discussed. The result is that macroeconomists literally don't know what they're doing. (They might as well offer economic policy advice based on a game of Monopoly, because it could hardly create a more absurd situation than the one we are currently witnessing.)

Lee Kelly writes:

Although macroeconomists are generally very competent at what they do, most do not seem to know what they are actually doing.

Modern economists have inherited a set of traditions and assumptions about the economy and its method of study produced by those Hayek called "contructivist rationalists", or "scientists" in the sense that Greg Ransom mentions. These traditions and assumptions are now deeply embedded in the way that modern economists think, regardless of whether they would actually agree with those ideas if explicitly discussed. The result is that macroeconomists literally don't know what they're doing. (They might as well offer economic policy advice based on a game of Monopoly, because it could hardly create a more absurd situation than the one we are currently witnessing.)

Lee Kelly writes:

There are more questions than problems, and more answers than solutions. For rational inquiry to begin there must be some standards by which problems and solutions can be recognised. Which standards are chosen will define the methods and focus of the inquiry. Macroeconomists inherited their standards from positivists. Of course, nobody is a positivist these days, because almost everyone understands the shortcomings of positivism, right? Unfortunately, the traditions laid down by positivists of the early 20th Century have remained implicit in the profession.

What is counted a real problem or acceptable solution is dictated by these unexamined traditions. One of these standards manifests itself in a preoccupation with what is flatteringly called "empiricial work". That so many of the measures used have so little--if anything--to do with what they purport to measure seems to not phase many. Macroeconomists just keep playing the same positivist game, and they must, for the profession demands it. Retaining the trappings of science is more important than finding the truth.

Troy Camplin writes:

To paraphrase Hayek: a system cannot understand another system as or more complex than it is. Stuart Kauffman also argues for a certain kind of lawlessness that comes about with emergence into new systems. In other words, the reductionist program cannot work to explain anything more complex than the least complex elements of reality. The laws of physics can teach us about physics. For chemistry, we need the laws of chemistry, which turn out not to be reducible to the laws of physics. Same with the laws of biology. Etc. Worse, if we apply Hayek's insights, we can never understand the economy, because it is more complex than we are, being made up of us as constituents.

I look forward to seeing the two of you debate.
Mark has quite a bit of background in money and banking but at times I find him nothing more than a cheerleader for Paul Krugman and just as much a political hack.

As far as Macroeconomics/Micro, I just wonder if we could not say the same about Microeconomics. There is always someone that can come by and challenge the given paradigms about the science of business and thus change all the models.

Case in point would be someone like Mohammad Yusuf. With {in his presentation} is a blend of non-profit and profit motives together.

Of course you are right that somehow Macroeconomists are looking for the holy grail of models when in fact every economy is probably based on another model. I remember once you said the problem with Macroeconomics is not that there was a Great Depression but that there were not a 1000 ones to compare and contrast with as well as have much more data to use in analysis.

Just my 2 cents...

Kenny writes:

Comets don't change direction when you talk about them and they're not all trying to predict the future either.

Jacob Miller writes:

I get frustrated when macro models wander too far away from the underlying productivity of the individuals, or reality. Sometimes models simply need to return to the "100 people on an island" example.

In the Modigliani Miller financial models the real action is in the violation of the assumptions. It seems that sometimes in macroeconomics we end up believing that "my model" is the end all be all when we really should be focusing on the violations of the assumptions.

Greg Mankiw made a great point that took this approach back in December when he wrote, "Willy-nilly spending is a good way to stimulate the economy only if the outcome is judged by the wrong metric."

Vangel writes:

I agree that Schiff comes across better than most economists because he does his analysis on the basis of what he has learned from the Austrian School. The fact that many mainstream economists are very good at mathematics does not mean that their efforts to make so many simplifying assumptions so that they can use mathematical tools to make forecasts is valid. As far as I am concerned mathematical economist models are just as credible as the mathematical models used by the financial sector.

Alan Furth writes:

Blind faith in the power of mathematical models was not only the cause of the failure of Macroeconomics to predict the crisis, but also one of its fundamental causes. As Felix Salmon argued in a recent article, the triple-A ratings given to CDO's were computed by a "formula that killed Wall Street": David X Li’s Gaussian copula function.

In line with Hayek's legendary argument about the incapacity of math to capture the full complexity of economic phenomena, there's a recent essay in The American Magazine by Jerry Z. Muller that brilliantly addresses the relevance of this issue for understanding the current crisis.

I comment on these ideas and provide links to the articles in a recent blog post:

http://alanfurth.com/a-whole-new-mind-for-finance

Greg Ransom writes:

Here are some "violations of the assumptions" to start with:

1. hetereogeneous production processes exist

2. hetereogeneous & multiple consumers exist

3. hetereogeneous production process take more or less time

4. entrepreneurial learning can't be put into a mathematical construct.

Jacob Miller writes:

"It seems that sometimes in macroeconomics we end up believing that "my model" is the end all be all when we really should be focusing on the violations of the assumptions."

Carl The EconGuy writes:

Modern macroeconomists are like the UN global warming doomsday predictors -- massive simulations based on a preconceived and non-tested framework that yields the desired conclusion that "government works." Economics is indeed like physics, but not like classical physics of constants and clearly testable hypotheses about processes. Modern physics is beyond that. Big physics, cosmology and unified field theories, and little physics, subatomic particles, both address issues that cannot be tested or observed, and so the judgment about what's good theory relies on secondary criteria -- in other words, at a conceputal level, macro is out of its depth. Macroeconomics these days works only inside a narrow range of issues with huge DSGE models with hundreds of equations and parameters, for the medium sized models -- there are bigger ones. But the underlying conceptualization is still Keynesian, meaning that consumption and investment functions are real, except for some randomness, and not based directly and explicitly on individual actors' decisions based on *perceived* wealth -- which may be very different from *real* wealth because of a lack of proper asset pricing models stabilizing asset valuations around equilibrium market prices. Without incorporating that difference into behavioral equations, the modeling is not going to get anywhere outside a narrow corridor of random walks. The problem is that, if that difference were incorporated into the model, there wouldn't be any GEs at all when financial markets crash. That means that, for big disturbances, the macro models are useless. Modern macro is like a physician who only knows how to diagnose and treat burst appendices at a time the plague hits. In that sense, it's not much better than old macro, but it's technically a huge success. But the theory isn't any better than global warming theory, and not nearly as pretty as physics models.

Comments for this entry have been closed
Return to top