Robert J. Barro on aggregate demand

There has been a recent kerfluffle over whether Robert Barro rejects the notion of aggregate demand, which he had written with quotation marks as “aggregate demand.”  Scott Sumner surveys the back and forth.

I say use The Google to find out what Barro really thinks and indeed he has written a whole piece on the topic (jstor), namely “The Aggregate-Supply/Aggregate Demand Model,” from the mid 1990s, and here is the abstract:

In recent years, many macroeconomic textbooks at the principles and intermediate levels have adopted the aggregate-supply/aggregate-demand (AS-AD) frame- work [Baumol and Blinder, 1988, Ch. 11; Gordon, 1987, Ch. 6; Lipsey, Steiner, and Purvis, 1984, Ch. 30; Mankiw, 1992, Ch. 11]. The objective was to allow for supply shocks in a Keynesian framework and to generate more satisfactory predictions about the behavior of the price level. The main point of this paper is that the AS-AD model is unsatisfactory and should be abandoned as a teaching tool.

In one version of the aggregate-supply curve, the components of the AS-AD model as usually used are contradictory. An interpretation of the model to eliminate the logical inconsistencies makes it a special case of rational-expectations macro models. In this mode, the model has no Keynesian characteristics and delivers the policy prescriptions that are familiar from the rational-expectations literature.

An alternative version of the aggregate-supply curve leads to what used to be called the complete Keynesian model: the goods market clears but the labor market has chronic excess supply. This model was rejected long ago for good reasons and should not be resurrected now.

If you read the paper, you will see three things.  First, Barro is fully aware of “AD-like” phenomena and does not reject that notion.  Second, Barro seems to prefer the IS-LM model to AS-AD, albeit with some caveats about possible false predictions of IS-LM and also noting in footnote two that he prefers his own presentation in his 1993 text.  Third, Barro’s criticism is (whether you agree or not) that AD-AS collapses too readily into standard rational expectations models and doesn’t really provide an independent foundation for sticky price macroeconomics.  In a nutshell “The AS-AD model is logically flawed as usually presented because its assumption that the price level clears the goods market is inconsistent with the Keynesian underpinnings for the aggregate-demand curve.”

Krugman had written this:

If you read Barro’s piece, what you see is a blithe dismissal of the whole notion that economies can ever suffer from am inadequate level of “aggregate demand” — the scare quotes are his, not mine, meant to suggest that this is a silly, bizarre notion, in conflict with “regular economics.”

I believe that is not a good characterization of Barro’s views and it is also an object lesson in the importance of the Ideological Turing Test.  I would cite not only this piece, but also forty years of journal articles, many of which study the importance of nominal shocks and demand, albeit without (in general) using textbook AD-AS terminology.  Indeed, Barro working with Herschel Grossman is one of the founding fathers of quantity-constrained Keynesian sticky-price macro and he is still citing this work favorably in his mid-1990s piece; see for instance Barro and Grossman (1971, 1974) and also their book from 1976: “This is a textbook on macroeconomic theory that attempts to rework the theory of macroeconomic relations through a re-examination of their microeconomic foundations. In the tradition of Keynes’s General Theory of Employment, Interest and Money…”

On the UI issue, I would note that the multiplier from transfers is likely unimpressive relative to the multiplier from government consumption.


This is a debate for specialists, but I get the first word. I was under the impression that the nutter Austrians reject AS / AD curves, instead insisting each industry and/or person has their own supply or demand curves and they don't really aggregate (I know it sounds absurd to me). Thus Austrians deny that a layoff in say the auto industry will affect the rubber tire industry, since each industry is more or less different. At least that was my possibly flawed understanding after debating these kooks in Mises forum (for which I got kicked off, despite not using profanity or saying anything vulgar).

As for this debate, it seems a variant of the above: the Rational Expectations school is trying to bifurcate the AS/AD curves into various components, possibly to achieve the same thing (deny Keynesianism of its force).

Nothing nutty or kooky about macro-economics. That's why there's so much unanimity even with the field itself.

perhaps the rubber tire industry has other customers. Atari going bankrupt didn't doom the video game industry. Instead, they made games for other consoles.

Atari going banckrupt doomed video game manufacturers who specialized in Atari games.

Suppose that during a boom investors invest in roughly three classes of capital: a commercial complex selling ice cream at the North Pole, a commercial complex selling sand in the Sahara and they also invest in capital improvements for growing corn in Iowa.

During the boom the economy grows fast: people are hired at the North Pole facility and they use their wages to buy many goods, including corn. People also get paid at the Sahara complex and they also use their wages for buying corn among other things.

When the inevitable happens and the North Pole and the Sahara complexes go broke, most of the people hired there get fired; the newly unemployed stop buying not just sand and North Pole icecream, but they also buy less corn, from the healthy industry in Iowa.

"See," says the Keynesian,"the real problem is aggregate demand. If we look at total capacity utilization, we see completely unused icecream shops, completely unused sand selling complexes and only partially used corn production facilities -- clearly the economy is producing far bellow potential. If we plot the potential GDP under the assumption of full use of icecream shops at the North Pole and full use of sand selling facilities in Sahara, we get something far larger than current GDP. We need to increase aggregate demand until we get full capacity utilization."

The Keynesian would have a point, if all the industries would be the same as the corn industry, which suffers during the bust even though the investment there was basically sound. But what is the use of concepts like "aggregate demand," when the real problem is that investments at the North Pole and in Sahara need to be either used for different purposes or scraped altogether?

In theory, stimulus could be used wisely: the government could invest in selling icecream in Sahara and either reconfigure the North Pole investment into an entertainment facility or ignore the previous investment altogether. But it would do little good to extend the investment in sand selling in Sahara or to invest more in icecream selling at the North Pole.

Yet, Keynesians have the same justification for both categories of investment: they are good because they increase "aggregate demand."

But maybe, just maybe, we need much more refined models of the economy as a whole; models that allow us to ask questions like "what sectors can absorb investment in a sustainable/profitable way?" or "What is the extra knowledge that the government has which it can use to make profitable investments, knowledge that investors that sit on hoarded liquidity do not have?"

If we take into account that the plague of overly simplified macroeconomic models is not restricted to Keyensian macroeconomics, but it infects other things, like growth theory, the case for refining the models is irrefutable.

@Alex K: that was awesome, you kooky Austrian! :-) I agree with your analogy but you have to also figure that there's some sort of linkage between the employees at the North Pole, the Sahara, and the Iowan corn farmer. One affects the other. I think Keynesianism is essentially an argument for preservation of the status quo rather than a painful "restructuring" of the economy: i.e., an argument to kick the can down the road. Usually that works. It's like driving a car using your rear view mirror: it actually is better than driving blindly, since most of the time you can interpolate how the road will curve based on how it's curved in the past (until you get to a fork in the road--and that's called a structural problem; are we at that fork now? Or is this simply a lack of AD? That's the 64T $?!)

Interesting point, Ray. You've made me think that maybe this is what is at the core of the Libertarian/Keynesian disagreement. Keynesians are trying to manage the economy, whereas Libertarians are trying to create a revolution that will bring about a paradise. Kind of like Communists, Jehovahs Witnesses, or those Hale-Bopp guys.

The real problem with your argument, Michael, is that it lures libertarians into the complacent thought that all opponents of libertarianism make arguments as stupid as yours.

However, generally this is not the case: most criticism of libertarianism is less stupid than yours.

When you come to a fork in the road, take it.
Yogi Berra

There are good arguments for various palliative measures. Presumably, we don't want the unemployed to die in the streets, so there is a humanitarian case for unemployment insurance. I didn't even mention money and nominal price rigidities , which may be an important part of the story (although I don't think it's the only important part.)

What we don't have, is a strong argument that the multiplier is high, regardless of what the government invests in. The full extent of this problem can not be elucidated in highly aggregated models.

As for Austrians, they are/were wrong in many details, but I find that behind the sometimes incorrect arguments, there are intuitions that are basically spot on. If you don't want to be influenced by such intuitions, you are missing out.

" Presumably, we don’t want the unemployed to die in the streets, so there is a humanitarian case for unemployment insurance."

UI hasn't been around forever, before it was adopted was there a lot of dying on the street? If there was it must have been before photography or even art because there's no record of it. And is that, as fearless leader BHO might put it, " a false choice"? It's either UI or death on the streets? Those are the only two options?

People all the time arbitrarily designate quotes as "scare quotes" when instead they are often meant to emphasize that the particular phrase in quotes is the point in contention.

Pointing out that the term is in contention is precisely the point of scare-quotes.

My point was that Krugman's statement: "the scare quotes are his, not mine, meant to suggest that this is a silly, bizarre notion" seems an extreme interpretation of the authors intent.

I believe this is a confusion over Krugman's true objection; Krugman's (broad) complaint is that, as he sees it, people like Barro will deny the obvious (such as AD being a thing) when it suits their agenda, and in particular that in recent years many conservatives have taken such positions in the context of policy debates, for this and only this purpose, in defiance of what he sees as common sense and Economics 101 - and that they would of course switch back if it suited their new policy agenda. Barro having previously made other statements, or even actually believing differently than what he wrote, in no way contradicts this model, regardless of its truthfulness. It is also fair to assume that most of Barro's current readers will not choose to learn about such details of Barro's in-depth views when reading his current article.

In other words, Krugman is projecting his own tactics on to his opponents and then attacking them for their (projected) shortcomings.

Normally that would be painfully childish, but its different when you are in a liquidity trap. It just has to be.

The IS-LM model is definitely an improvement over AS-AD. Expect to hear more about it once Stanley Fischer is confirmed as Vice Chair of the Federal Reserve. BTW, the "rational expectations" school will lose credibility once more economists realize that most humans don't think rationally.

Clever use of RE to discredit RE!

Daron Acemoglu also doesn't like the AD/AS model.

I'm not clear why it is obvious that AD is a thing.

I just flipped open my old copy of Money, Employment and Inflation. Barro and Grossman have a whole chapter, 28 pages, on the Dynamics of Aggregate Demand. In short, unlike Krugman, Barro actually did the hard theoretical work of trying to make it work, and he concluded that it cannot be fixed.

And to members of a certain political party, AD will never be high enough until everyone either has a self-actualizating job or a manufacturing job (oddly enough the growth in service sector doesn't count). Tax cuts and monetarism is the most effective way to boost AD in terms of bang for the buck. Non-defense, non-tax cut govt. spending offer a much lower ROI.

I'm sorry, the whole "Idealogical Turing Test" seems to be nothing more then another way for libertarians to 'explanabrag' that no one understands their views because no one has the same "deep" understanding of the topic they (the libertarians) do.

Yeah, basically. Very disappointing to see Tyler invoke this--it's pretty close to an ad hominem, and discredits his other worthwhile posts.

You guys are joking, right? You seriously dont see the value in being able to understand someone else's point of view?

Funny how you're criticizing us for being close-minded when Tyler is the one essentially saying "they don't understand anything because they're blinded by their ideology!" I think libertarian is an ideologically bankrupt and wholly inconsistent religion, nowhere near worth taking seriously, but I still come to this blog every day. So, yes, I see the value in understanding another person's point of view. Tyler is demonstrating that he does not.

Uh Michael, a situation where you are both close-minded and blinded by your ideology is a totally consistent one. And I don't think Tyler accused *you*, since he wasn't addressing *you* in the first place. Maybe you have some other disorder, like confusing yourself with Krugman? Or maybe you reflexively feel the need to defend Krugman because you want to identify yourself as a member of his tribe? Seems like acute mood affiliation to me. Peace.

"Funny how you’re criticizing us for being close-minded when Tyler is the one essentially saying “they don’t understand anything because they’re blinded by their ideology!”"

You are daft. TC is clearly saying that he doesnt think that Krugman is stating, or understands Barro's argument.

Understanding the other side of arguments is clearly important. I just don't think the ideological Turing test can be applied to libertarians, particularly with Bryan Kaplan's huffy "We [libertarians] learn his[Krugman's] worldview as part of the curriculum. He learns ours in his spare time - if he chooses to spare it." attitude.

I don't actually understand your complaint. You seem to agree that it's important to be able to understand the views of people you don't agree with. But, um... with some kind of specific carve-out for libertarians? Or what? Are you just complaining that, in fact, Krugman does understand the views of those he disagrees with? Do you agree that it's important that Tyler be able to understand Krugman, but that Tyler shouldn't mention it? I genuinely can't tell what you're saying.

Is this really your reaction when an inaccuracy is being pointed out? Nothing above implies what Krugman is missing is "so deep"--it's pretty clear there's a discrepancy between the way Krugman portrays Barro's views and Barro's actual views. And an extra tip: Barro is not a libertarian.

Whether or not Barro is a libertarian is irreverent.

Re-read my comment.

I'm saying that invoking Kaplan's concept of an "ideological Turing test" is silly, because it's just a libertarian huffing that no one really understands them, or their views, because... something. They're never particularly clear, but it usually ends up being something about no one else understanding the topic as well as the libertarian does.

As for the daily Krugman-gotcha-game, I really don't care why people are cheering on 'their' op-ed writer over potentially inappropriate use of quotation marks. The charitable read is that Barro meant to use the marks in a way to (implicitly) invoke an explanation of a concept to lay readers. Uncharitable: he thinks the concept is meaningless and is showing his contempt.

Personally, I think it's the former. For all I know, it could be the latter, Barro doesn't clear his op-eds with me before he publishes them. Either way, this is the sort of thing that really only matters to the kind of people who get prickly when they feel people are being improperly labeled as 'libertarian', which sort of implicitly reinforces my original point about this silly Ideological Turing Test.

"Whether or not Barro is a libertarian is irreverent."

Assuming you mean irrelevant, I agree. That would make your introduction of libertarianism into the discussion unnecessary--it simply makes it look like you have an axe to grind.

Nor does the "Ideological Turing Test" necessarily have anything to do with libertarianism. It simply refers to one's ability to accurately state the views of one's intellectual opponent, regardless of who's debating. In this case, it is invoked to highlight that Krugman did a poor job of stating Barro's views. If you feel that Cowen is just playing a "gotcha-game", I would say you're being defensive--both Cowen's and Krugman's blogs are meant to be places to debate and discuss economics arguments, including the specific claims that other blogs are making, and it's not a "gotcha" to do just that.

Yes, I meant irrelevant.

Libertarianism in invoked in Kaplan's original "no one is capable of understanding me" justification of the ideological Turing Test (implicitly cited in the original post). Because I don't consider it anything more than a way for libertarians to explanabrag how they're tragically misunderstood because no one else is as smart, clever, insightful, whatever, I don't consider it a valid argument. Therefore the entire post boils down to "Paul Krugman said something I don't like". And, speaking for myself, if that's what you want to see in blogs, great. I think it's a silly slap fight between adults, and really the sort of thing that really only appeals to people who about the definition of a 'True Scotsman'.

Brian M says: Invoking Kaplan’s concept of an 'ideological Turing test' is silly." I agree, noting that you are the only commenter here to invoke Kaplan's[sic] name. You grind an irrelevant axe here.

The problem with Barro/Krugman is that they are looking at two very similar circumstances (1928 and 2008) but with entirely different responses. In 1928 inequality reached an historic high, the financial system soon collapsed and, along with it, so did the economy. The government did not intervene, inequality plummeted and stayed low for decades, the economy fell into depression, economic misery evolved into tyranny, and, well, everybody knows the rest of the story. In 2008 inequality once again reached an historic high and the financial system soon collapsed, but unlike before, the government intervened, inequality dropped somewhat but soon recovered and is once again approaching an historic high, and the economy did not collapse. Choose your poison: financial and economic collapse without government intervention, a precipitous drop in inequality, and eventual economic recovery and a long period of shared prosperity; or financial but not economic collapse and continued high levels of inequality coupled with continued financial and economic instability. I suppose I'm in the Barro camp when I comment that excessive levels of inequality are self-correcting, but I'm in the Krugman camp when I comment that the correction is too painful to permit. I call mine the Gospel of John school of economics: I see, see, see.

Sorry. Aggregate Demand is simply, to use the terminology of the General Theory, that part of monetary transactions with a non-zero elasticity of production, i.e. those transactions which lead to the production of new goods and services instead of reshuffling second or third hand stuff like shares, bonds, existing houses and the like. To use more modern terminology: final demand, household consumption+exports+investments+government consumption. You can actually estimate this and use input-output models to make rather depandable short run preductions about the influence of additional demand on the use of labour, copper or the production of carbon dioxide. You can't of course explain it by assuming that the whole household sector behaves as if it's the neoclassical consumer, long ago ditched by marketeers who actually tried to sell something to real consumers. Let's please, please get rid of that utter nonsense about ill defined (if at all!) utility of representative consumers with eternal lives. Economics can be a science, too.

These neverending arguments could be summarized as "I didn't say what you think I said."

Is it a reading comprehension problem, a conveyance of ideas problem, or is the only thing they know how to do is duel straw men.

Utterly out of context (or maybe not):
As a non-Economist who is fascinated by socio-economic discussion, but seriously concerned about the Economy in a multi-generational way:
Why do so many Economists (or at least ones that we hear about) spend so much time analysing data with the intention of trying to predict/ understand/ generalize the data of a near-dysfunctional (and barely improving, except for occasional dips and peaks) system rather than trying (at least occasionally) to model a perfect (however you define it) system (or at least provide insights on its construction) that is very highly flexible, scalable, and gradually implementable? With climate change and nuclear-blast predicting super computers recently numerous and accessible (to many PhDs and division heads at least), why not model such a system based on the perceived behaviours, wants, and expectations of the public (ultimately in quest, of course, of the 'absolute good': a post-scarcity society)? Perhaps a model that allows a functional system that doesn't matter whether there are 100 architects/ bakers/ bankers or 10,000, because it dynamically re-balances the relationship (and adjusts accordingly the job description somewhat and nature of the service relationships) with the underlying knowledge of the skill sets of each individual and other resources available. Certainly there is enough flexibility in each person's skill set, the relationships within a system, knowledge of all units, and the time to make resources available that we could have a bubble-free system where everyone improves, is utilised to a high level, and responds to people's desires to retire, change, plateau, or other -- similar to a very large construction site project management program. Perhaps a Star Trek like utopia - but why no rough models or deep analyses (supported by simulations such as the highly data and computationally intensive carbon lifecycle analyses) beside popular publications? (Krugman may have insights)(where are the forward-looking Econometricists?)(Are we so bound to the perceived freedoms of an unplanned society?)

Sumner continues to bring it:

Comments for this post are closed