Arnold Kling  

How I Think About Keynesian Economics

With All Due Respect, 2... Central Planning for Health Ca...

By the way, my panel appearance on jobs is now up on C-span (I start about 45 minutes in. Also, I saved some of my thoughts for the Q&A, which starts about 64 minutes in.)

In this post, I want to try to explain Keynesian economics to people who are inclined not to believe it. That is, to people who think "If the government spends money, it has to come from somewhere else, so Keynesian economics is just an example of fallacious reason that comes from observing the seen and ignoring the unseen."

It helps to think of economic activity as outsourcing. That is, if you iron your own shirt, that is not economic activity. If you outsource ironing by paying someone else to do it, that is economic activity. If you eliminate ironing by buying permanent press shirts, that is also economic activity (Nick and I call the cost-saving innovation Economics 2.0).

A recession is a decline in economic activity. People are not outsourcing as much as they did when the economy was booming. Economists have noticed during this recession that global trade figures have plummeted. Indeed. That illustrates a decline in outsourcing, a decline in economic activity. In Adam Smith's terms, the extent of the market has declined. Not because of explicit trade barriers, but because....well, I would say it is because the market needs to recalculate, but that is not a very Keynesian notion. Keynes would have said it is because consumers are attempting to hoard, and businesses do not have the "animal spirits" needed to convert this hoarding into investment.

The thinking behind Keynesian economics is that when you are in a recession, if the government runs a bigger deficit, recipients of the money will want to increase their economic activity. This will draw people out of non-market activity (unemployment) and into the market. The newly-employed will want to outsource more of the satisfaction of their needs, and this will create a virtuous cycle (the multiplier).

Imagine that all of us were chefs, each with a different specialty. In good times, I patronize others' restaurants and other people patronize mine. That is economic activity. In a recession, for some reason we stop going out to eat. I don't enjoy eating my own cooking every meal, but I don't think I can afford to go out. Since I am not patronizing your restaurant, you think you have to cut back on eating out, also. Economic activity declines.

Thinking about the economy in these terms, the idea of using government deficits to boost economic activity makes perfect sense to me. The reason that I am doubtful about Keynesian policies today is that I think that recalculation is important. Nobody, least of all the government, knows the best way to align the work force to satisfy needs. The recalculation problem has to be solved gradually, by trial and error. It is as if we had a bubble in Mexican restaurants, and now it is not clear whether they need to be replaced by more Indian food, more French-Asian fusion cooking, or whatever. I do not think that central planners in Washington know the answer any better than the market. So I am not confident that a Keynesian approach is going to get us very far this time.

Comments and Sharing

CATEGORIES: Macroeconomics

COMMENTS (20 to date)
N. writes:

I understand, and agree with, the recalculation story.

What I would like to see is a well thought out rebuttal.

John V writes:

Yes. Having first come into my layman's understanding of cycles, recessions and whatnot through an Autrian-style approach, I instantly visualized a computer defrag screen or something to that effect. It was the idea of seeing little pieces moved around one by one from where they were to somewhere else. It isn't random but very purposeful and intricate...and it takes time. IOW, recalculation as you say.

To the extent that I understand "Capital Structure", what I take from it is that it's not some generic, homogeneous lump like clay to be molded or "k" if the world deals with one single good. Sheesh. It's a bizarre puzzle of so many different actions and activities and production and sale processes that come to be in their form based on millions and billions of choices and decisions based on exponentially more factors of cost and comparative cost, time and place and zillion other tiny reasons which form layers upon layers on decisions and preferences and so on...all tied to previous layers. Mind boggling.

SO with this in mind, the layman's understanding of the Keynesian Concept and the multiplier also struck me as laughably simplistic and missing so the point of being a if capital structure has now become the liquid-metal terminator in Terminator 2: one lump of homogeneous mass that isn't dependent on intricate placement of pieces. When that terminator broke into pieces, it just flowed back together into one body. Sorry, the "catallaxy" (sp?) doesn't work that way. If it did, Keynesian would be a no-brainer easy-button to any plunge in spending. But it isn't. What gets me is how Keynesian economists have convinced themselves that this "composition" issue really doesn't matter. Unreal.


Your explanation of how Keynesian economic works fails to address the very criticism you point out. That is, "where does the money come from?"

To be a true Keynesian, you would support both deficit spending in a recession, and running government surpluses in good times. Thus, you can tap the surplus funds for deficit spending, to "recalculate" the slack.

However, when the government runs perpetual deficits, they only way to increase (deficit) spending is to borrow more. Naturally, this is money borrowed from the private sector (or other governments) - money which would have been used, whether spent or invested, in other ways. Thus, we are merely substituting government spending/investment for private spending/investment.

Brad Fults writes:

This is an interesting explanation, but doesn't it still suffer from “observing the seen and ignoring the unseen”?

When people decrease their fancy restaurant budget, they must still eat, so their fast food or canned food budgets likely increase, no?

The same applies to the shirt pressing: if they stop taking their shirts to the cleaners, they will have to use their own iron & ironing board, which they may not yet have and must buy, or which will wear out from use and thus require replacement.

It seems to me that you can argue the only time that economics activity *ceases* in a recession is when the people go back to living off of the land—using water from their well, beef from their own cattle who feed on their own grass which is planted from their own seeds, etc. The amount of people even in a position to pursue such a course is tiny now compared to the 1930s, so the same remedies shouldn't be expected to work (assuming anyone thinks they worked then).

I agree with your idea of recalculation, but I think you might be able to make a stronger argument. Namely, that the decline in economic activity *is* recalculation, not just a prelude to it. In other words, it's only a decline as pouring water from one glass to another causes a decline in the water level of the first glass. No water is lost, it's just shifted; the loss is entirely dependent upon your perspective.

But the water glasses represent instantaneous transfer and “perfect information” which seem obviously unrealistic, so the lag we see with unemployment is due to “sticky” employment and wages after all. The drop in production, however, is a conscious recalculation undertaken by the producers in reaction to the market's demand shift.

Maybe I'm not saying anything new, but I figured I'd ramble a bit.

Carl The EconGuy writes:

Keynesian economics, as opposed to the economics of Keynes (to which Keynesian economics bears little resemblance), works only if government bonds are treated as net wealth by the private sector. See Barro on that issue. Also, the empirical tests don't bear out the basic hypothesis -- we're much more nearly Ricardian than Keynesian, as a matter of empirical fact (just ask yourself if you've seen any interest effects of public debt lately). See Barro's article in the WSJ just the other day -- he finds no discernible multipliers. Why anyone is peddling Keynesianism these days is truly a mystery, as is the fact that anyone publicly professes belief in it.

here's the latest Barro article in WSJ

and you can find a lot of his earlier writings, both popular and technical, on his homepage:

Lord writes:

A Keynesian wouldn't care what Washington does with it, only that they do do something with it. Once it transfers from government hands it is the market deciding where it goes. In this, it is just another form of monetary policy, increasing the amount of money in circulation.

I wouldn't discard recalculation though. We have just had a large energy shock, one that makes trade less favorable, one that will make efficiency more worthwhile. Everything is being repriced according to the amount of energy, and particularly the form, it uses. It is just it is easier, it is even possible, to undertake these changes only if the economy is growing, at least in nominal terms.

John V writes:


When you say:

"Thus, you can tap the surplus funds for deficit spending, to "recalculate" the slack." fail to understand the significance of recalculation. Arnold didn't explain what that really meant.

It's not question of how much you have in storage. That quantity of money is irrelevant to recalculation. Re-calculation is a decentralized process and by its nature is impossible for the government to figure out. That's the point...waving tons of money at a recession, whether the money is borrowed or from a slush fund, CANNOT account for proper recalculation.

Greg Ransom writes:

Note well.

There are not production goods in your story.


No capital. No capital goods.


The very CORE of the coordination problem of a capitalist economy across time simply doesn't not exist in Keynes' "world".

Smith's "extent of the market" is crucially shaped by rival production processes, taking alternative lengths of time, using different technologies, and requiring different repair and replacement schedules.

When the "extent of the market" over extends itself systematically across production time (e.g. very long-term production good of residential housing) inevitably the "extent of the market" retracts, and all of this production across time structure my be "re-calculated".

None of this exists in Keynes -- and none of it exists in the brain of 99 out of 100 tenured macroeconomists, who block to the door to tenure to any grad student daring to think otherwise.

JPIrving writes:

@John V.

Great analogy regarding recalculation. Thoughtful, well written comment too, I would just add that it is *possible* that New Keynesian models using aggregates like "K" ect could still be useful in thinking about how an economy works, and how to get it out of a recession.

The statistician George Box said: "...all models are wrong; the practical question is how wrong do they have to be to not be useful." So while the "Austrian" story is certainly an accurate depiction of how our quasi capitalist economy works, it is at least conceivable that models need not replicate such a level of detail to still be useful. Certainly not for making useful forecasts, but perhaps in meditating upon the effect of policy proposals.

Too bad we can't go back to 2007 and put Sumner in charge of the Fed...that's the only way I see of making any progress on this Recalculation vs All Other Theories debate.

Snorri Godhi writes:

The restaurant example helps, but I am surprised to see no mention of taxation, either in the post or in the comments. The first difference between eating in your own restaurant and some other restaurant is that, in your restaurant, you pay neither sales tax (except on the raw materials of course) nor income tax. Therefore, if I wanted to encourage people to patronize each other's restaurants, the first thing I'd think of is cutting taxes. (And I'd be astonished to hear that nobody has thought of this before me.)

Nathan Benefield also makes a good point, one that I have seen mentioned before.

Lee Waaks writes:

The problem I have with Keynesians is that they seem to explain a recession/depression by assuming a recession: the very thing they are trying to explain. "Animal spirits" don't get us very far as an explanation as to why investment/spending suddenly collapses. If we view the economy solely through the lens of consumer spending, Keynesian solutions look logical and government spending might be the sensible solution. But if we emphasize the importance of saving and the capital structure (as Greg Ransom does above) -- and the potential for government to foster malinvestgment, then Keynesianism looks awful.

Travis writes:

Russ Roberts had a great podcast in early 2009 with a Keynesian expert (I believe his name was Fazzari).

Fazzari basically traced the multiplier impact to this conclusion: Keynesianism works when a previously unemployed person is put back into the work force (through the policy) and spends money that otherwise would not have been spent.

Made sense to me while I was listening. But I began to wonder how unemployment benefits might cancel out this spending in the modern era. It seems logical that if I was temporarily (at least perceived) hired to dig a ditch by the government - that would induce me to spend on food, shelter, clothes. But aren't those the same things I'm buying with unemployment?

[The Fazzari podcast is at --Econlib Ed.]

Lukey writes:

All I can say Mr. Kling is that you frame your comments as if they will address the Keyenesianism skeptics' suggestion that these analyses fail to consider the "cost" side of the equation and what that does to depress future economic activity (possibly by so much more than current activity is enhanced that it makes Keynesian stimulus undesirable when one considers the long run) and then you go on to completely ignore the cost side in your analysis. I remain a skeptic.


[Comment removed pending confirmation of email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog.--Econlib Ed.]

Duke writes:

The invisible part of the equation is that the economic activity is tied to the income and obligations of the participants, and their wealth. We have cheaper goods, and no jobs in many cases. For a wile lenders will lend before they see that the borrowers can't repay. Then for a while people can sell assets to purchase necessary goods. After that a solution must emerge. People with out an address have a hard time taking a job even if it comes available. There is a level of destruction beyond which return is very challenging.

I mention individuals but this is extending to municipalities, states, and sovereign countries. China will likely keep their dollar peg until Japan and Europe are wiped out. Then let if float so we can slowly pay our debt. Only then will jobs return.

Our approach has turned out to be, borrow money from taxpayers who can't pay their own debts, then lend it to banks that can't find qualified borrowers. They in turn buy government bonds paid by the people the they wouldn't lend to. Why not just pay the banks the interest as long as we can afford it and cut out the inescapable obligation to the tax payers who are clearly unqualified to assume it. The other two options would have been save the banks and break them up or let them fail [probably the scariest option of all - which may turn out to have been the best in hind-site].

Brad P. writes:

"Imagine that all of us were chefs, each with a different specialty. In good times, I patronize others' restaurants and other people patronize mine. That is economic activity. In a recession, for some reason we stop going out to eat. I don't enjoy eating my own cooking every meal, but I don't think I can afford to go out. Since I am not patronizing your restaurant, you think you have to cut back on eating out, also. Economic activity declines."

It still doesn't make sense to me.

It ignores uncertainty which is probably the reason we stopped going out to eat in the first sense.

Mostly, though, the automatic response to this situation is for the second chef to lower his prices, which, assuming you are not the absolute best chef in the world, would combine with your own growing displeasure to skew the opportunity costs back to where you now prefer to return the second chef.

It seems the scenario you present would require a devolution of the division of labor. Is that the case? Is that the cause?

Also, it seems that general tax cuts would be a solution to the recalculation problem.

Joe Calhoun writes:

By saying that Keyenisan solutions won't work this time because of the large recalculation required, you seem to imply that Keynesian solutions have worked for past recession. It seems more likely to me that the current large recalculation is required because previous Keynesian responses didn't allow small recalculations to occur. Don't recessions always have micro foundations? If you ignore those micro considerations how can you ever truly solve a recession?

genghis writes:

there is almost a perfect inverse correlation between government spending and income.
So, government only makes things worse by spending more.

Farcaster writes:

I think an important consideration is how the Keynesian stimulus is delivered and the availability of productive savings opportunities.

If our economy is primarily services (e.g., using the dry cleaning example above) using the money for welfare and tax cuts just has a temporary benefit and eventually the money runs out and we're back to square one.

However, let's say the $787 billion ARRA was devoted entirely to building 150 nuclear reactors at $5 billion each. These are massive construction projects that could pay back the government via a dividend, essentially paying for themselves much like TARP has with many banks.

Further, these would generate a significant number of jobs and reduce utility costs for individuals in the future, freeing up more money to spend on other things besides essentials.

In economic parlance, they would have a huge multiplier.

Only in the U.S. could we borrow $1 trillion from China (or print the money) and have very little to show for it.

The $1 trillion hole in the economy must be filled with productive investment that makes us more competitive.

James Christopher writes:

I also started off coming into a more Austrian view, primarily through getting involved back in 2007 in the Ron Paul campaign, and his encouragement to "study Austrian economics".

A quick note on that -- one thing I have found is that many people who get into that, find themselves reading more general articles about government, recessions, spending, etc, and find themselves agreeing with the conclusions -- so often that they end up seeing themselves as subscribers of Austrian thought, even though they may not know that much about the underlying theories. They then find themselves ridiculed -- and they don't know why -- the conclusions make so much sense -- obviously government is wrong, and you cant spend your way to prosperity, etc etc. But that's a different story.

But anyways, I agree with the point made about recalculation, most "experts" you see seem to think that government must arbitrarily prop up this industry or that industry -- "We must not lose jooobbbss!!"

I disagree about the government spending aspect, for one simple reason. Regardless of what theory you look at, or what models you use, it still always comes down to real capital (which is obviously finite... something which government doesnt seem to understand), how it is flowing, and most importantly.......... what it is being used for. And I think the Austrians (they arent the only ones of course) are correct in focusing on this as their foundation -- and more importantly, not losing touch of it.

Anyways, as pretty much everyone reading this website would probably take as self-evident (but government and many "experts" seem to be disconnected from to the nth degree), economies only grow in one way -- as technology advanced, and capital is used more effectively, i.e. more efficient business models, more effective agricultural methods, etc.

So as someone before said, "where does the money come from" -- but more importantly -- Where is the money... or more correctly, the real capital, being allocated to, as a result of spending?

Following the government policy, when think that spend, spend, spend is the key, regardless of whether it involves spending $5000 on a toilet seat (3 laborers, 5 managers, and 3 boss-overseers to install it... and of course, gotta have labor union consultants, and a few other bureaucrats sitting around to make sure it goes correctly)

or whether it is simply encouraging people to spend in general.

Someone throwing a big party, and consuming lots of resources to a non-productive end helps keep the stores their purchase afloat for awhile, and so forth and so forth as it's multiplied.

The paper numbers increase, and economists obviously go to great lengths to prove this. But is this fundamentally misguided? Regardless of the numbers, we still have to deal with physical reality -- the numbers look great, but what has actually happened to the capital?

Does a party with 500 people, who consume a massive amount of food, etc, do any good for the economy? No, it is a complete waste.

Not only have resources been used that the market could have allocated to something that is actually productive, but in the case government trying to loosen credit to encourage spending... it is, as the Austrians would say, propping up this unsustainable level of consumption.


Anyways, I'm not a professional economist or anything, but I tend to agree with what Richard Fenyman said about Physics, which is that if it cant be explained to a class of freshman, then we dont really understand it. I tend to think that this applies in most areas -- the fundamentals are always very simple.

And I think that many economists, specifically the ones that government listens to -- mostly keynesian/neo-keneysian, if Im correct -- are very far detached from the simple physical reality.

Abstract theories can't override basic physical reality.

Comments for this entry have been closed
Return to top