Steve Fazzari on Stimulus and Keynes
Jan 24 2011

Steve Fazzari of Washington University in St. Louis talks with EconTalk host Russ Roberts about the economics of Keynesian stimulus. They discuss the stimulus package passed in February 2009 and whether it improved the economy and created jobs. How should claims about its impact be evaluated? What can we know as economists about causal relationships in a complex world? The conversation includes a discussion of the underlying logic of Keynesian stimulus and the effect of the financial crisis on economic research and teaching.

Ramey on Stimulus and Multipliers
Valerie Ramey of the University of California, San Diego talks with EconTalk host Russ Roberts about the effect of government spending on output and employment. Ramey's own work exploits the exogenous nature of wartime spending. She finds a multiplier between...
Nicholas Wapshott on Keynes and Hayek
Nicholas Wapshott, author of Keynes Hayek: The Clash That Defined Modern Economics, talks with EconTalk host Russ Roberts about John Maynard Keynes and Friedrich A. Hayek--their ideas, their disagreements, their friendship and how the two men influenced economists and public...
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.


Jan 24 2011 at 10:34am

The guest may be amicable, but he’s a terribly disoredered thinker.

He concedes various things as “hard” but when the “medicine” isn’t as effective as “the doctor” indicated, then it isn’t that the “medicine” is counterproductive, ineffective or incorrect, its that the disease was worse.

As I listened to this, I couldn’t help but remember the old Steve Martin “medieval barber” skit. In that skit, a parade of “patients” came in with a variety of maladies, only to be told to apply a poultice of sheep’s urine. At the end, he starts discussing the use of the scientific method and then disregards it with a dismissive, “nahhh”.

Even if we had a clear and convincing case for efficacy, keynesians never seem to discuss the potential for political shenanigans (using borrowed money to reward friends) and the distortive effects of these sort of interventions.

By the way, for all the pretenses of scientism, economics on the left is religion.

Jan 24 2011 at 12:03pm

“The other observation is that all those mainstream models are Keynesian in the sense that they anticipated some natural stable relationship between government spending and output–not just output but employment–and those didn’t happen.”

It’s interesting that neither speaker identifies (and justifies) the boundaries of the economies Keynesians purport to model. If you give me a dollar and I spend it across town, have you stimulated the economy on my street or across town? And if “across town” is replaced by “across the Pacific”?

Of course, the stimulus wasn’t “big enough” – how could it be when the economy it would stimulate is global? Look at how nicely our imports from China have risen since the stimulus. How can anyone say it isn’t “working” in a Keynesian sense? It’s just not working HERE. But “here” is not a necessary feature of Keynesian analysis, or, if it is, “here” now means “on this planet,” not “in this country.”

So, Dr. Romer was wrong. Things IN THE US were not “worse” than they thought, they were just not the sort of things that stimulus created in the US would fix.

Jan 24 2011 at 2:39pm

Wow. I hope there are better Keynesian apologists out there than that guy.

I went into the discussion with an open mind, but Fazzari’s logic was surprisingly shallow–where does the money come from to do the Keynesian stimulus? If there really is a free lunch in this form, then why doesn’t the government spend even more money and make the economy hum along even better? Where is the boundary of this logic?

The basic problems with Keynesian stimulus, in my view, are that it promotes a misallocation of resources, runs up government debt and places bureaucrats in positions that should be occupied by entrepreneurs. Fazzari didn’t help me feel better about any of these concerns.

I think that Keynesian stimulus is probably a harmless way of keeping civil servants and politicians busy during normal recessions, but during a so-called “balance sheet recession” Keynesian stimulus looks to me a lot more like early physicians performing bloodlettings, confidently proclaiming success in every case that the treatment doesn’t kill the patient.

Fazzari seems like a nice fellow, but I had really hoped for a defense of Keynesian stimulus with a bit more torque and a broader view of how economics interacts with the other moving parts of the political/financial structure.

I often feel that there must be more to Keynesian economics, and that I just haven’t found the right person to explain it to me in a meaningful way. In a balance sheet recession where everyone has taken on more debt than can be repaid, I just don’t see how injecting more credit into the system makes any sense.

Notwithstanding the arguments for more stimulus made by the Keynesians right now, it seems to me that what we are observing is simply the crash that follows a boom facilitated by easy credit, exactly as von Mises so clearly described it. I can’t believe anything in this complicated world could be THAT simple, but maybe it is, and the Keynesians are the modern equivalent of the clerics who wouldn’t look through Galileo’s telescope for fear of what they might see.

The clerics in the example above were probably not bad people, and probably didn’t think of themselves as closed-minded. Their problem was that they were working within a paradigm that was beginning to see the assumptions on which it was premised being invalidated. I think we may be in the process of watching the same phenomenon unfold with respect to Keynesian economics.

If there is a Keynesian economist out there who can dismantle von Mises’s arguments more methodically, I would love to hear him/her interviewed.

Jan 24 2011 at 3:18pm

One point about Christina Romer’s projections. I think it is undeniable that Romer painted a rosier picture than would have happened with or without the stimulus. In fact before the stimulus had even been voted on her projections had failed to anticipate the scale. If you consider the fact that it was enacted part way through Feb. and even by August only 19% of the money had been spent I am doubt full much at all had gotten out the door in the month or so following the its passage and by then the unemployment rate was already over 9%.

I think it is basically impossible to argue that her unemployment projections with (obviously) and without stimulus were way too low. The only way around this would be to argue that the business community was so freaked out by the mere POSSIBILITY of a stimulus package that it drove up unemployment. Also it shows that aside from any “regime uncertainty” effect, unemployment had already neared its peak before much stimulus at all had gone out the door.

Jan 24 2011 at 3:58pm


I am a little confused. You seem to fault Fazzari for not given a Keynesian response to questions and issues he was never really asked about. If anything that would be a fault of Russ for not prodding him on the topic (no offense Russ, I know you do what you can with the time allotted, maybe another interview cold cover more).

That is like declaring he has no good explanation for why the Giants won the 2010 World Series. Sure he MIGHT not have a good response, but he also was never asked about it, hardly a fault on his part.

“If there really is a free lunch in this form, then why doesn’t the government spend even more money and make the economy hum along even better?”

Keynesian DO argue the stimulus should have been larger. The reason they don’t recommend constant government intervention of this sort in non-recessionary periods is because demand does not need boosting and it would more likely lead to more inflation rather than any employment increases (I believe the NAIRU deals with this).

“Where is the boundary of this logic?”

Good question, one that he was never asked so maybe e-mail him rather than assuming he lacks any response.

As far as the debt worries, I do not know what he would answer, but what I would point out that it is much easier to payback debts in general when an economy is growing rather than shrinking or stagnant.

It is ALSO harder to pay off debts when there is deflation. Also I think it may be hard to argue that there is an “over abundance” or expansion of credit right now.

Many people point out that the difference between debt projections of various people and organizations has to do with how well they expect the economy to perform. Even if we were not to rack up much debt, if we experience a stagnant and/or shrinking economy and/or deflation what debt we DO have becomes ever harder to service.

“Notwithstanding the arguments for more stimulus made by the Keynesians right now, it seems to me that what we are observing is simply the crash that follows a boom facilitated by easy credit, exactly as von Mises so clearly described it. I can’t believe anything in this complicated world could be THAT simple, but maybe it is, and the Keynesians are the modern equivalent of the clerics who wouldn’t look through Galileo’s telescope for fear of what they might see.”

One question that I have that I don’t see explained by the Austrian BCT is why does unemployment in recessions such as this effect all sectors of the economy? Why if resources are misallocated to one industry doesn’t that industry only experience lay-offs as they transition into new more well funded industries. If it is merely “frictional” unemployment, why is there not the same “frictional” unemployment during the build up of the bubble as people leave other industries to tradition into the artificial inflated capital goods market.

It seems to me aggregate demand explains much better why the economy and employment as a whole slump and persistent unwanted unemployment occurs. Why in the Austrian plan does the market not just pick up the slack from the misallocation in the unrelated sectors of the market?

andrew fischer lees
Jan 24 2011 at 9:34pm

I just checked – can anyone find stimulus funding by industry? I don’t think it’s publicly recorded anywhere – I just downloaded the 71Mb CSV file, and among the hundreds of columns there’s no industry code variable!

The BLS seems to understand that industry-by-industry accounting is useful for unemployment. Why wouldn’t stimulus funding, which has employment as its aim, use the same accounting methodology?


p.s. as an economics research assistant, I was partially supported by stimulus funds in 2009…WHILE IN KENYA.

Jan 24 2011 at 10:32pm

I’m no authority on Austrian economics, but it seems to me that if you distort both demand and the structure of production by driving the interest rate below the “natural rate” – not all sectors of the economy will be distorted equally.

The distortions will be everywhere, but will be most dramatic in those sectors that are extremely credit intensive, like commercial and residential real estate, and the least dramatic in the sectors in which consumers finance their consumption with earnings, rather than credit. Hence the grocery business will suffer less than the SUV, condo, and McMansion markets.

When the credit intensive projects start to fail and the supply of credit available to finance them contracts, so does the effective demand for them and employment in the credit-intensive sectors declines dramatically, and this has important secondary effects on less credit sensitive sectors.

This seems to fit with the empirical picture of what we’ve seen in both the boom and the bust. Do you disagree? If so – what makes the Keynesian demand-as-uniform-aggregate story more compelling to you?

David B. Collum
Jan 24 2011 at 11:01pm

1. John Hussman published an excellent article that sheds light on the relationship of monetary expansion, interest rates, and monetary velocity…

2. Not all are talking about recovery. It has a semblance of a mirage. Rhetorically, can you really claim the economy recovered from the 2002 recession given the profound damage that surfaced in the subsequent severe beating? Are you sure that a cyclical recovery going forward has meaning if we crash and burn again? I personally am not a believer in cyclical recoveries in secular downturns. They are headfakes indicating that the corrective measures–purging the rot–was incomplete. The monetary policies are said by some to be damaging. (As an aside, I am not sure why recessions are said to stop as the expansion phase begins; seems to me that the second half of the lousy economy has commenced.)

3. We the savers carried the first portion of this stimulus on our backs in the form of curve steepening measures forcing rates to unacceptably low levels (well below what a free market would mandate.) How unstimulative was my 0.0% return on my capital? The second payment will come in the form of inflation, which is being hidden for now by flawed statistics (CPI malarky) and the uneveness with which inflation rears its ugly head. The bill will come due, however.

4. People wrote about the disastrous mortgage bubble voluminously before it broke. To say nobody saw this mess coming is patently, demonstrably false (and these were not just “stopped clocks”.)

5. There is a logic to Keynesian economics at the zeroth approximation. Government should not be spending when the private sector is bidding for the resources. It is simply more cost effective to carry out government projects–ideally necessary ones–when the labor and materials are cheap as found during a recession. There is, however, a total asymmetry to Keynesian economics in practice. No politicians (nor central bankers) have shown the will to pull in the government spending when times are good. Just look at the municipal and state mess we are in. Those huge malinvestments occurred during the affluent times, only to be revealed to be flawed when reality struck. Keynesianism is an abstraction that is brilliantl on paper but not so hot in practice.

6. The bottom of Keynesian thinking–the absolute mindlessness of the stimulus–has got to be “Cash for Clunkers”. It took people who desperately needed to solidify their balance sheets, coaxed them into getting rid of debt-free vehicles, and hurled them back into over consumption and debt. Meanwhile, it destroyed capital by destroying perfectly useable cars. If I said what I really thought about the program, my post would be vetted. Let’s just say this was cock-eyed Keynesian thinking from the depression era reappearing. Back then, they destroyed food while people were hungry to try to force prices up. There was a much simpler solution to the GM problem: GM should have been sent through bankruptcy, contracts terminated, and the pieces auctioned off to any capitalists convinced they can make a credible car at a profit without subsidies. If that problem cannot be handled, it’s impossible to imagine a more complex problem can be solved.

7. When our monotonically rising debt normalized to GDP turns around, we are going to fall very, very hard. That day is guaranteed to come; it’s arithmetic. I blame Keynesian logic for this unsustainable multi-decade trend.

8. Finally, I sure hope we are not going to be force-fed claims that nobody saw the problems with a half-quadrillion dollar, wholly opaque derivatives market with a doubling time of 2.5 years when it finally hits the ultimate wall.

Jan 25 2011 at 11:30am

The question that I want answered clearly by a Keynesian is why would deficit spending by Government increase total spending in the economy if it is funded through borrowing. I could see that if the central bank had bought up all the existing Gov. debt and could not find anything else to buy but we are far from that.

Is the answer that fewer available Gov. bonds would cause more people to hold cash or Gov. guaranteed demand deposit and the banks not lending the increased amount of money? IMO that would point to a problem that Gov. money and bonds are out competing everything else in a bad economy. That is bad and would lead me to suggest the private creation of money.

Jan 25 2011 at 2:35pm

The main problem I’ve always had with the whole Keynesian thing is that it seems, to me, that it’s attempting to address the symptoms instead of the cause. Doesn’t artificially inflating aggregate demand simply put us back in the same economic position that caused the whole downturn?

Rondy K. Smith
Jan 25 2011 at 3:23pm

Quoting Russ Roberts: (with ** my remarks)

“One of the themes listeners are aware of here is I’m increasingly skeptical of the scientific nature of economics. This would be an example of where I don’t think there are any data that would let us resolve this dispute.”

**It will never be resolved. How could it possibly be known what the severity of a recession might be when they can’t even determine whether a recession has happened until after the fact?… Yes, this recession was worse than anticipated. To my mind the real unemployment rate is probably closer to 20% than to 10%. If we didn’t have Baby Boomers retiring in droves, it would be 20% or more. And, yes, this whole debate over Keynesianism versus Monetarism has to be philosophical because neither side can prove their case empirically. The true nature of money and how market participants us it in crisis conditions has long been overlooked by theoreticians on both sides.

Jan 25 2011 at 4:41pm

I was surprised that there was no discussion of the “multiplier”.

I saw a piece in the WSJ that the stimulus ,in fact,had a fractional multiplier.Thus a $1 spent through stimulus had an $0.80 impact on the economy (GDP?).

I am not sure why this happened,but perhaps recipients merely hoarded the money or paid off debts.In some cases municipal layoffs were deferred perhaps resulting in a negative productivity.

Jan 25 2011 at 8:12pm


“The distortions will be everywhere, but will be most dramatic in those sectors that are extremely credit intensive, like commercial and residential real estate, and the least dramatic in the sectors in which consumers finance their consumption with earnings, rather than credit.”

Since the main issue as I see it is misallocation, isn’t the distortion the fact that certain goods and investments (such as capital goods) appeared to be more attractive as compared to other goods which seemed less attractive. If this is the case than demand and employment from the inflated sector would go down and demand and employment in the suppressed sector should go up, correct?

“so does the effective demand for them and employment in the credit-intensive sectors declines dramatically, and this has important secondary effects on less credit sensitive sectors.”

What are the “secondary effects” that cause decreased employment in less credit sensitive sectors?

“If so – what makes the Keynesian demand-as-uniform-aggregate story more compelling to you?”

I don’t want to put words in Fazzari’s mouth but I wouldn’t be surprised if he agreed with your description somewhat as well.

What it leaves out, however, is the effects of the paradox of thrift on savings, investment, and employment; The nature of the liquidity trap; as well as the nature of the “secondary effect” to which you refer that hurts the economy as a whole.

As people realized housing was overpriced, demand dropped precipitously in this sector (which in turn resulted in a drop in prices) as well as the sectors and investors involved in the securitization of these loans. This puts people in this sector out of work (although the trillion dollars given to banks seem to have insulated them from much loss). Because this was such a large part of the economy it lowered demand throughout the economy (in aggregate). The decrease in demand put downward pressure on most other sectors of the economy, causing them to lay-off workers, which intern decreased demand even more.

Now there is plenty of money at the banks, but they are skittish to invest because no-one wants to invest in, let’s say, a coffee shop if no-one’s buying coffee (could be any other business however). This results in tons of money essentially being sat on while people on involuntarily pushed out of work for extended periods of time.

What is wrong with this picture?

Jan 26 2011 at 2:31am

Not sure I understand exactly what you mean here:

“Since the main issue as I see it is misallocation, isn’t the distortion the fact that certain goods and investments (such as capital goods) appeared to be more attractive as compared to other goods which seemed less attractive. If this is the case than demand and employment from the inflated sector would go down and demand and employment in the suppressed sector should go up, correct?”

To my way of thinking driving the interest rate below the natural rate and diverting capital into credit-sensitive sectors would increase demand for the the output from those sectors (more McMansions) and increase employment from the said sectors. Have I misunderstood what you are saying in the above?

The secondary effects that I’m thinking of are reduced demand in sectors that generate the inputs required to produce the Mcmansions (loggers, copper workers, freight haulers, title companies, etc, etc, etc) which translates into reductions in employment and wages in those fields that’s proportional to the magnitude of their dependence on demand generated by the malinvestment. I’d also expect to see significant tertiary effects in sectors that rely on discretionary spending, such as travel, leisure, dining.

As far as the paradox of thrift is concerned – was Keynes really equating transferring real savings into bank deposits and other financial assets with hoarding physical currency? Money that investors convert into T-bills, etc are being used to finance both investment and consumption, no? If Keynes was talking about anything other than hoarding physical currency then I have a great deal of difficulty with the “paradox of thrift.” How do you understand the concept.

Anyhow – interesting points. Thanks for contributing them.

Gary Rogers
Jan 26 2011 at 5:37am

You had a good discussion of how stimulus spending can be different depending on how it is deployed, but it is just as important to look at where the money comes from. I wish the discussion covered more of to the other half of the transaction, which is what happens when you take the money from the private sector of an economy to fund a government stimulus. This, too, can take different forms through taxation, borrowing or printing money and a discussion of the merits of each of these would be in order. Keynesians always ignore the funding of their stimulus and talk about spending multipliers. I would like to have heard a discussion of the unseen side of their policies.

Jan 26 2011 at 8:39am

It was an interesting discussion, I have always found it fascinating that so many economists fight over models when they do not have enough convincing evidence (and yes, that holds for Austrian economists also).

I do have a question about evidence. Right now, across the Atlantic, UK is going through massive and substantial spending cuts. Will the performance of economy not settle the Keynesian vs Austrian question?

Jan 26 2011 at 8:47am

Floccina –

“The question that I want answered clearly by a Keynesian is why would deficit spending by Government increase total spending in the economy if it is funded through borrowing.”

If the Gov’t announces a plan to build a road to my town, I borrow some money and buy a bulldozer. If the gov’t doesn’t announce that plan, I don’t borrow any money to buy a bulldozer.

The money I borrow is newly created fractional reserve, bank-made money. It arises through lending that my bank is willing to do because the Gov’t is going to build a road that will create a demand for my dozer services.

Then my workmen buy new boots, or a car to get to work. (To what extent this multiplier benefits Americans as opposed to other global providers is not Keynes’s issue.)

In short, borrowed money can be new money. Indeed, all money is “borrowed” in some sense, as it all represents an entitlement to future output, i.e., it is a form of debt.

Lee Kelly
Jan 26 2011 at 9:38am


Nominal spending is equal to the supply of money multiplied by the velocity of money. If nominal spending rises while the supply of money is constant, then the velocity of money must have increased. Since fiscal policy does not alter the supply of money, it follows that it must alter the velocity of money. Since the velocity of money is just the flip-side of the demand for money, fiscal policy can only increase nominal spending if it reduces the demand for money.

How could fiscal policy reduce the demand for money? In times of great uncertainty, risks rise and interest rates fall, and the relative desirability of holding money increases. Since aggregate cash balances cannot equal more than the supply of money, attempts to increase cash balances with a constant money supply will inevitably be frustrated and nominal spending must fall. Government bonds are normally very safe and liquid assets — from the perspective of particular households and firms, they are like money that earns a little interest. By increasing the supply of government bonds, households and firms can be enticed to exchange their money holdings for government bonds. Supposing the government does not just hold the money it borrows in place of its bond holders, the demand for money will fall and nominal spending will increase.

However, I think there are many problems with fiscal stimuli — do not mistake me for an advocate. Primary among these problems is the central bank monetary policy, because, in principle, the central bank has almost complete discretion over the level of nominal spending. If the central bank is not increasing (or decreasing) nominal spending, then it is because it does not want to. It follows that any change in the demand for money instigated by a fiscal stimuli will be offset by tighter monetary policy, and the overall effect of fiscal policy will be to transfer resources into the public sector without increasing nominal spending.

Emerson White
Jan 26 2011 at 5:28pm

“Why a carpenter right now is struggling to find work when a computer programmer in 1999 doesn’t is a little hard–we don’t fully understand the process.”

I think the answer to this question is fairly obvious. In the tech bubble we overvalued technology and technology startups, but we didn’t over produce computer programmers. We weren’t done integrating computers into our lives and businesses and we didn’t have an excess of computer programmers, we just hadn’t found the most profitable ways to use them.

the housing bubble however did produce an excess of houses and an excess of carpenters (which is exacerbated by the excess of houses). There was no innovation in housing that fueled the need for new housing, we were building the same houses that they built 50 years ago and adding new kitchens and bathrooms. The Housing bubble was driven by an exogenous bundle of financial “innovations” (scare quotes because it was really new ways to make poor decisions) while the tech bubble was driven by real endogenous technological innovation.

Rondy K. Smith
Jan 26 2011 at 10:21pm

Quoting Mr. Fazzari

“For example, if you talk to a classically or neoclassically oriented economist about whether markets are all in equilibrium; we measure a lot of unemployment but it’s really voluntary in some sense, and ask what happens when government deficits go from 2-3% of GDP up to 10% of GDP, they say: My gosh, interest rates are going to skyrocket. Well, they didn’t. We don’t want to get into the arcane details of different models, but that is a fairly direct prediction of a Keynesian prediction when you are operating below full employment. Inflation slowed during this period of time even though we had all this expansionary monetary policy.” (end quoting)

The monetary expansion was being substituted for by an increase in a phenomenonal form of money I have deemed as “Perceived Liquidity.”

This component acts to buffer the monetary expansion and thus it has no linear modulation of economic output. It also subverts the ordinarily expected demand for money, for nominal money. The demand thus subverted, the cost of money, interest rates, do not rise. The same thing happened during the Great Depression when rates got as low then as 4/10s of 1% on deposit accounts, much like today. Probably the most significant difference between then and now has been the advent of money market mutual funds from about the 1970s. Had Paulson not guaranteed them, they would’ve caused an even worse financial collapse than the banking system caused in the 1930s.

David B. Collum
Jan 27 2011 at 11:32pm


When the government stimulus plan ends, the guy with the bulldozer has now has a bulldozer, a debt, and no customer. What happens next? Follow that logic down the line.


Jan 28 2011 at 9:19am

Nemoknada and Lee Kelly, Thank you for you answers.

Jan 28 2011 at 4:44pm

For a school of thought that (justifiably, IMO) prides itself on its better-than-mainstream understanding of time and change, I do wish that more Austrian-minded participants in this conversation would exhibit a little more consistency in their historical-mindedness. For example, Russ’ parable of chemotherapy actually undermines rather than supports the point that he wishes to make. He rightly observes that the modern oncologist doesn’t know whether or not his radiation treatment worked until he evaluates its effects post facto. And yet, despite the fact of that very real uncertainty, the contemporary oncologist is *not* using the kind of techniques that, for example, Superheater mocks in the first comment above. Obviously, the pre-modern “oncologist” (or barber) didn’t possess any degree of a priori certainty about the efficacy of his leeches and crude homeopathic remedies either — although he probably had some confidence that (his) ministrations were likely to result in more symptomatic relief and/or a higher survival rate than the available alternatives.

Is there widespread consensus today, especially among medical experts, that chemotherapy is more “effective” than bleeding and/or sympathetic magic? If so, how did we get from there to here? Is that consensus predicated on the “mainstream” experts’ beliefs or assertions that they now know everything about human biology, or at least enough of everything to have total certainty about the results of their actions as well as their inactions? Not to my knowledge. Is that mainstream consensus so overwhelming that the very possibility of sustaining belief in homeopathic remedies, or faith healing — or the idea that intervening to help the sick is itself unnatural, and contrary to the proper Darwinian order of things? Not entirely it seems. Does that mainstream consensus turn out to be incomplete or just plain wrong in some way from time to time? Fairly often in fact — which is why it’s always better if a range of views that are outside of the mainstream coexist…

Does anyone today actually wish that the medieval barber had never stepped up — that no one had ever attempted to intervene in human health outcomes until they had first achieved a priori perfect knowledge of how every anatomical mechanism and emergent biological process works under all possible conditions? Does anyone actually think that we should give, e.g., individual peripheral nerves veto power over what is and is not permissible for doctors to do to save the rest of the body, in whole or in parts, from prolonged injury and/or death?

Superheater may be right that “economics on the left is religion.” But people who live in their own esoteric houses of worship should probably not cast broad anti-religious aspersions so lightly.

Yeaster Doozer
Jan 28 2011 at 7:04pm

Nah, this keynesian vs. neo-classical debate is getting a bit tiresome. Time to give ‘Fragglenomics’ a chance (

Jan 29 2011 at 9:19pm


Sorry it took so long to respond. I’ve been very busy.

“Have I misunderstood what you are saying in the above?”

Yes and no. My point was that not only does this funnel money into over investment in certain sectors, but it pulls investment away from other sectors. Why when the bubble bursts don’t we have a symmetrical growth in demand for non-capital intensive goods, after all they have been underproduced. Why does the economy as a whole suffer? Also, what good does large scale underutilization of labor provide for the economy?

“The secondary effects that I’m thinking of are reduced demand in sectors that generate the inputs required to produce the Mcmansions (loggers, copper workers, freight haulers, title companies, etc, etc, etc) which translates into reductions in employment and wages in those fields that’s proportional to the magnitude of their dependence on demand generated by the malinvestment. I’d also expect to see significant tertiary effects in sectors that rely on discretionary spending, such as travel, leisure, dining.”

A couple of things. Your description seems to leave out the economy wide effects. The overnight lending and commercial paper markets and growing demand for cash reserves didn’t just effect McMansions, it effected every business and person that requires any form of credit or loan. This is all as a result of a large increase in demand for money. It is true that Keynes didn’t venture much into the causes of depression (although followers such as Minsky made compelling arguments and described a business cycle theory that in some ways eerily relates to the facts of the current recession) but he focused attention on this as the cause of the general slump. The vast majority of the economy contracted after all, some areas more than others.

“As far as the paradox of thrift is concerned – was Keynes really equating transferring real savings into bank deposits and other financial assets with hoarding physical currency? Money that investors convert into T-bills, etc are being used to finance both investment and consumption, no? If Keynes was talking about anything other than hoarding physical currency then I have a great deal of difficulty with the “paradox of thrift.” How do you understand the concept.”

People seem to have a lot of trouble with this concept. If I pay you $10 dollars for such and such a service, that is $10 that does not go into my bank’s savings account. Now let’s say you put it in your bank account. Has total savings increased? No. The $10 that I would have saved is now the $10 you saved. Obviously you too could spend it on down the line, either way it is likely to be saved at some point.

Now in normal times this is fine. The money you put into the bank gets loaned out to someone who needs it to fund a business, house purchase, etc. Currently, however, banks and individuals want to hold onto that money as reserves incase of a sudden loss of income or the need to pay off debt, etc. So in one way or another the money just sits there, or in the case of banks largely goes to buy Treasuries at near 0% interest, which is not far from putting it in a mattress.

Jan 30 2011 at 2:27pm

I completely agree that the best way to examine economic models and their efficacy is through historical analysis. My argument with Keynesians would be the typical critique using the 70’s stagflation but I would also add this….

Since the great depression, the majority of our policies have been aimed at driving demand. Keynesian ideology is obviously the major defense of these policies.

To me, I see the outsourcing of production as the natural result of these policies. I would also point out that the policies of Reagan and Volcker had the opposite reaction to what orthodox Keynesians would have predicted.

Sorry for my brevity but I’m on an IPad. Someone else extrapolate for me 😉

Jan 30 2011 at 2:33pm

I think we would need to have a written debate to really get at some of the issues, otherwise it’s too easy to be imprecise about the arguments. How about seeing if Cato would do a Cato Unbound issue on the topic?

Jan 31 2011 at 7:25am

Re: AHBritton @ January 29, 2011 9:19 PM

“The money you put into the bank gets loaned out to someone who needs it to fund a business, house purchase, etc. Currently, however, banks and individuals want to hold onto that money as reserves incase of a sudden loss of income or the need to pay off debt, etc. So in one way or another the money just sits there, or in the case of banks largely goes to buy Treasuries at near 0% interest, which is not far from putting it in a mattress.”

In the case of banks that now seem to be disinclined to lend at levels that would be more consistent with their observable reserve ratios, the aforementioned “paradox” is resolved by acknowledging how deeply and thoroughly the system has been crippled by the self-delusion of perfect risk management through financial engineering. For a resumption of “normal” lending activity to be (merely) possible, not only do banks have to lay on much more capital to cover existing risky-to-demonstrably-worthless commitments that were previously deemed to be risk-free; they also have to replace large quantities of the same kinds of faux capital assets that are currently counted on the other side(s) of their ledger(s).

Of course, knowing what most of them have now, belatedly and unhappily learned from first-hand experience, banks also have to further recalibrate their perception of individual lending transaction risks in light of the much greater level of systemic uncertainty imposed by other known and unknown/invisible lending institutions. Thus the relevant Paradox is not Thrift but rather Trust: how does one operate when the prudential calculations that would have previously favored action (lend!) now recommend inaction (hoard!).

Perhaps Japan provides some clues as to possible answers, other than the obvious (accept relative stagnation as the new normal)…

Jan 31 2011 at 8:42pm


Just to clarify, the actual “paradox” is the fact that saving $10 increases your savings, but not savings as a whole.

In addition it removes income from those whom it theoretically would have been given to, therefore decreasing either their savings, or their ability to spend that money and create income for somebody else. So if everyone saves, savings as a whole will likely decrease because income as a whole must decrease.

I agree with most of what you said. What I would add (possibly disagree) is that banks aren’t lending because they have bad assets and the “climate” for investment is very poor. Part of the reason the investment climate is poor is because there is so little spending as to reduce the likelihood of receiving an adequate return on investments. The reason there is so little spending is because people are nervous and saving (creating more capital in the bank which isn’t invested) and ALSO credit is so tight that the benefits normally derived from credit cannot be had, and exacerbating peoples reluctance to spend.

This ends up being circular as you can hopefully see.

Finally regarding Japan Milton Friedman’s solution was just to print more money, paraphrasing him he said that after all the central bank could print as much money as they wanted at almost no cost.

My problem with this is that the printed money I believe ends up unfairly distributed solely to banks. In addition it does not guarantee this money is lent or spent. Finally, no additional value is actually being produced with this money.

It seems to me (and my mind could be changed) that despite its problems it makes more sense to just hire people, either to help fix roads and other government activities, or just to perform some other job, fund musicians, and artists like the WPA did or what have you. As the economy recovers slowly reduce this program.

I know you would probably argue about the uncertainty created by government, that government can’t allocate resources, etc. My response is that this seems like its possibly an argument for a more effective governmental institution than against my proposal necessarily. And given our imperfect government it is likely that at least something of value would be created, as opposed to nothing when the money is just printed.

Sri Hari
Jan 31 2011 at 9:50pm

Keynesian economics can work if government stimulus uses capital which is obtained from the market and not by any quantitative easing, does not distort the market forces for creative destruction and also not destroy the ability of market to understand risk. The biggest challenge of quantitative easing by Feds and stimulus package by the government is the complete masking of underlying risk and sustainability – the fatal flaw of Keynesian pump priming. And it invariably fails in the long-term. But works wonderfully well for short-term political cycle.

Austrian economics requires a sustained downturn to repair the economic system to weed out big businesses which depend upon the support of policy makers to keep their un sustainable business models alive. In the end, political rhetoric combined with short-term economic success drives policy-so we are stuck with the discredited Keynesian economic model for ever, until we find a leader who passionately believes in a sustainable economic & business model-read Austrian, and has the capacity to bring the voting public to accept the medium term pain. Ain’t happening with Obama!!

Feb 1 2011 at 5:53pm

I do not understand them discussing the stimulus without mentioning the source of the money.

Feb 1 2011 at 8:33pm

RE: AHBritton’s remarks @ January 31, 2011 8:42 PM

You mistake me sir. I agree with you on almost all counts. I’m actually a pro-coordinationist (of the distributed pan-prudential variety, to be specific). If there was any point of (transient) disagreement, it stemmed from my interpretation of your prev. suggestion that what banks are holding much more of now is individual private deposits (from the hypothetical population segment that still has lots of surplus cash that could be used to consume, but is banked instead). However, your subsequent comment about “printed money” and the unfairness of its current distribution mechanism/outcome erased that concern. Austrians may be right about the non-neutrality of (discount window/QE) money, but the notion of liquidating the entire edifice of modern banking, government, et al. to dislodge the current mix of incompetents, crooks, and competent non-crooks, just so a new group of transcendently indifferent uberfolk can impose their own sincerely and transparently self-serving arrangements does not strike me as a very appealing political project.

John Bolton
Feb 2 2011 at 11:15pm

The principle of giving people money to increase demand seems straightforward, but it raises a big question for me. The question is, “where does the money come from?” If the money is reallocated or borrowed, then doesn’t that result in a decrease in demand in the areas of the economy where it otherwise would have been spent? If the money is “printed”, then isn’t the result simply inflation?

Feb 3 2011 at 2:33am

John Bolton,

Two things. If it is simply printed it is inflationary, however that effect is delayed.

Second, it is only inflationary when the demand for money is low, which is not our current situation as made obvious by the current CPI and inflation numbers.

In addition, despite the negative sides of inflation, and obviously one doesn’t want inflation in the double digits for prolong periods, inflation actually helps in paying off debts as debts our usually denominated in US currency. Not to mention a devalued currency makes buying imports from the US more attractive to other countries…. but like I said it’s a trade off.

I hope that helps.

human mathematics
Feb 3 2011 at 10:06am

I can’t comment on the Selgin podcast so I’m leaving my response to that here. This stuff where you interview conservatives / libertarians / neoclassicical economists / Cato people is always sort-of interesting because of the topics you explore.

But, the adversarial discussions are always more enlightening and interesting. The people you agree with tend to get too many softballs.

Feb 3 2011 at 1:34pm

John Bolton,

One more thing, to respond to the question ” If the money is reallocated or borrowed, then doesn’t that result in a decrease in demand in the areas of the economy where it otherwise would have been spent?”

This all depends on where the money comes from. If borrowed from overseas it is theoretically the foreign country forgoing consumption now in the hopes of more consumption in the future through investment.

If it is borrowed domestically it could theoretically crowd out other investments. In this case it is important to look at the economic climate. In the current economic climate with banks hoarding money and then loaning it to the government for almost nothing, it could be argued that they wouldn’t be investing it in businesses and such privately anyways, even if they weren’t giving it to the government.

Because of the poor economy and people defaulting on loans banks would rather just hold onto the money or invest it in a rocksolid government treasury (whether the US or elsewhere) for 0% interest rather than loan it out and potentially suffer great losses.

The question is whether it is crowding out investment, or putting money to use that would otherwise be sitting around in bank vaults (not necessarily literally), or loaned to a foreign government.

Feb 4 2011 at 5:10am

One of the things I think would have been interesting to discuss is the difficulty in developing a federal stimulus plan when multiple levels of government are involved. For example, if the federal government initiates an expansion of government spending but the state and municipal governments largely contract then would this not negate many of the beneficial effects of the federal stimulus package?

Richard W. Fulmer
Feb 7 2011 at 1:58pm

The discussion reminded me of a lecture I attended on petroleum reservoir modeling. One of the tasks of a reservoir engineer is to create a three-dimensional picture of an oil reservoir in order to identify the most promising places to drill production wells. These models are extrapolations of a relatively few data points taken from exploratory well core samples, and depend on the judgment of the engineer and his knowledge of the region’s geology.

The speaker displayed a series of slides depicting what the reservoir might look like if the exploratory wells had been drilled in West Texas, the Louisiana Gulf Coast, or Oklahoma. In each case, the same data points were consistent with very different reservoirs. The models differed because of the geological differences between the three locations.

Historical economists try to reconstruct what happened and why based on a finite set of data points. Their answers depend upon their models of how the world works, and the same data can support a number of very different scenarios.

The analogy breaks down in (at least) two ways: (1) the reservoir engineer used exactly the same data in order to come up with his three different models, while different economists will choose different data because their models dictate which data they believe is relevant, and (2) the engineer bases his judgment calls on objective truth – we really do know the geology of the three regions in question. Economists have no such knowledge to guide them.

Feb 16 2011 at 7:30am

I would have liked Russ to further explore Hayek’s point that a stimulus might actually retard recovery. It is a very difficult process for capital and labor to reallocate, new machinery be built, firms formed, labor to retrain into new skills and find new uses, etc.

Kaynesian models seem to grossly underestimate this difficulty.

It’s easy to see why, If capital is just a curve on a graph labeled K. Next to K they see this other line—“aggregate demand”. Gee that looks awfully low—wouldn’t it be neat if we could just move it up?

Even if one isn’t an Austrian, everyone agrees some areas of the economy (housing) have been wildly over invested and expanded.

Is it possible that Kaynseian stimulus actually retards the process of reallocating capital and labor?

Thomas Woods uses the example of a circus. Say a circus comes to town and this causes a sudden boom to the local restaurant. The owner becomes very excited and expands rapidly. Sure, a few odd duck economists point out to the owner that many of the customers are wearing clown shoes, but demand is booming so he rapidly expands. A lot labor and capital is now being used in the restaurant instead of other sectors of the town’s economy, like the hat factory and the lumber mill.

The circus then leaves. Demand collapses in the town but they still have this gigantic restaurant. The mayor decides to start handing out money to all the town’s citizens . Although the restaurant isn’t directly being subsidized due to an uptick in demand it is barely kept going, they don’t have to lay off any waiters or cooks, although the workers are making much less.

Sadly, recovery is only slowed by this process—because no one on the restaurant staff goes back to their previous jobs at the hat factory or the lumber mill, nor do they retrain or start new businesses. Capital remains over invested in the restaurant, productivity doesn’t rise—new wealth isn’t created they just sort of slump along, eventually the stimulus runs out and the restaurant gets torn down and everyone finds new jobs but it actually took longer than if the mayor did nothing.

Aren’t some carpenters not retraining, when the stimulus is letting some construction firms that really should no longer exist limp along until the stimulus stops and then they finally die off or get caught up in another boom?

Haven’t we seen this reflected in banks? Why would a bank quickly sell all it’s REO houses at a incredibly low value—because it needs cash—when it can wait out the crisis a bit and hope to unravel its portfolio more slowly and not take such bad losses.

Government stimulus also seems to favor construction since it is focused on infrastructure projects. Wouldn’t construction be the last place you would want to stimulate since too much capital is allocated to construction to begin with, supply vastly out numbers demand?

Wouldn’t it be better to stimulate dentists or fast food restaurants or anything else before construction?

I wish these guys had explored the recession of 1921! Huge disaster over 12% unemployment in the first year, but a very quick recovery without stimulus. Also I always thought the 70’s in the US and the 90’s in Japan supported the anti-Kaynes view. Wish this was explored more.

What about all the horrible recessions in the 19th century that ended relatively quickly without stimulus?

I think libertarians are a bit too hard on the paradox of thrift. Here is why:

Not all saving represents the same level of investment. If I am scared and invest in treasury bonds, put it under my mattress, or invest in currency or commodity reserves surely that represents more “thrift” and less investment than if I buy stock in a small cap firm. In a non-recession no problem, but the idea is government can force some of those savings into more “investment”. i.e. the police bust down your door in the middle of the night, rip open your mattress, take your money and hand it to your stock broker. The idea is at least logical.

I’m not talking about a social safety net issues here just fiscal stimulus. I don’t want to ignore how much it sucks to loose your job.

Feb 21 2011 at 8:52pm

My problem when folks talk about Keynes is they never get much past the idea of “if you spend government money, does it create more jobs.”

Keynes went into interest in all kinds of interesting ways which I think is really important. His arguments are much more subtle than he’s ever given credit for.

Don’t ask me to explain them. I only fully follow them when I’m actually reading them.

I just think you need to go deeper, because Keynes went a lot deeper in the General Theory. It would be great to hear more about those parts. Because there is so much math they are hard for me to follow, but I could tell it was really subtle and insightful stuff.

Comments are closed.


About this week's guest:

About ideas and people mentioned in this podcast:Articles:

Podcasts and Blogs:



Podcast Episode Highlights
0:36Intro. [Recording date: January 20, 2011.] Steve is a self-avowed Keynesian, not one who finds it politically expedient. Philosophical, methodological, economic differences between us. We actually have a pleasant relationship, don't yell at each other, one we really appreciate. Very much so. Somewhat rare for people who disagree about their economic theories. Want to revisit some of the things we talked about in the past and see where we are in the economy. Thought we'd start with that. We are right about at the second anniversary of the Stimulus Package that was passed in February of 2009. At the time it was $700 billion and something; I think now it's accumulated to be a little over $800 billion, maybe $862 billion. Give me your assessment of what its impact has been. Has it worked? If not, why not; and how would we know? First off, let me say it's a hard question. In economics we all have this problem: What's the counterfactual? What would things have looked like had we not done this? That's the gold standard of being able to assess a major policy like this, and we don't have that. My own sense is it's done more or less what they said it would do in terms of the marginal effect. I don't remember the exact number of jobs they were talking about. Of course there is no exact in this kind of thing. In my assessment we would have had a much deeper recession without the stimulus package. Probably not talking about recovery at this point. If you put the two things together--the fiscal stimulus and the monetary actions--without those things I think we would be looking at a much, much deeper and more persistent problem. Facing some serious problems. In that sense, I think the stimulus has done what it was supposed to do. Well, let me back up on that. I think the stimulus has done what I would have predicted it to do, but it hasn't been adequate to turn the economy around in a significant way, and I think one legitimate criticism of the Obama administration's way of discussing this at the time, and I think this was the way they thought about it when they passed it, was that this was going to be a substantial help to the economy and that it would turn things around. I think they thought it was the right thing to do at the time. But I also think they believed that in a year or so of that passage the economy would be starting to recover fairly robustly. I bet they were pretty confident that by the 2010 mid-terms the economy would be looking fairly bright. I think they were wrong at the time. The evidence turned out to be wrong. I don't think it's because the stimulus failed. I think it's because a mainstream Keynesian perspective underestimated the problems that we face.
4:20Let me challenge you on that. Picky point, but interesting methodologically. That has been the standard defense--the one you just gave: there were predictions made at the time that if we didn't pass the stimulus unemployment could reach as high as 8.5%, whereas it went over 10%. And when confronted with that, the people who made those predictions, one of whom is a first-rate academic economist, Christina Romer--working for the President, political role--ex post, she said: The situation was worse than we thought. Of course, that's possible. How would you know? What I find fascinating about these discussions is it's like saying the chemotherapy wasn't strong enough. Sometimes you can go in and look and see how bad it really was. Sometimes you can find out that you really misestimated, underestimated the severity of the problem. Here, what could possibly help us understand that other than the fact that the stimulus doesn't seem to have been as successful as it was forecast at the time? Good question, hard question. Really comes down to this issue of what is the counterfactual. In order to jump to the end of my thinking on that good question, on some level I think this is a matter of judgment, much more than we'd like it to be. You have to look at the basic logic of the underlying theory; you have to look at evidence, both current and in the past, which is going to be sketchy; you put all this together and there's going to be different ways of looking at the world, make a judgment call. In this particular case, I think part of what happened--the Obama administration said they would reach some number. I think it was eight and a half. Without the stimulus and then a lower number with the stimulus. I think 7%. It goes up to the low tens. I think if you look at data, that number understates the severity of the problem, what's happened in the work force. It probably effectively went higher than the low tens. I suppose that's prima facie evidence that things are worse than even what we thought was the worst-case scenario when they were making this forecast. I also think the persistence of this downturn, that things turned around so slowly. I was going back, looking at what people were saying; I found something, I think the spring of 2008, a little less than a year before the stimulus package passed; a statement by Ben Bernanke along the lines of: the economy might be in a recession--and as we know now we were in a recession starting late 2007--but by the second half of this year--2008--we will start to see growth resume. Well, the second half of 2008 was when we had the most spectacular failure since the Great Depression, and the economy literally fell off a cliff. Bernanke appointed by the Bush administration, and he's making these kinds of claims--I think there was a widespread tendency across the ideological spectrum in economics to underestimate the severity of what actually happened. I agree with that. Just a footnote on the 2008 story--of course, in the spring of 2008, Bush threw a so-called stimulus program of tax rebates. Not cuts in rates, but checks that had a disappointing impact on consumer spending but at the time was thought to help ameliorate the problem; and then few people foresaw the financial catastrophes. The other footnote I would add is that I don't think the economy went off a cliff in September or October 2008. Financial markets certainly struggled and it certainly made it harder for the economy to recover. I do think there is a tendency to overstate the seriousness of the situation we are in. I don't know how we would measure it. Let me back off of that a little bit. Maybe we tend to use the "off a cliff" metaphor a little too easily. I'll take that as a friendly criticism. That said, if I want to try to defend the position that if we look at economic activity--not just financial markets--GDP numbers and maybe more importantly the labor market in the fourth quarter of 2008, first quarter of 2009, it was awful. We were tracking Great Depression numbers at that point. On the employment data it's remarkable how closely we tracked them. Not for such a long time. By the spring or summer of 2009 we were starting to see things were stabilizing and we were pulling off that awful path of the Great Depression. But there was a period of approximately six months when things looked very dark.
10:10Let me come back to this question of how would you know if things would have been worse. The alternative hypothesis of course is that the stimulus didn't work or made things worth. We non-interventionists have our own areas where we wave our hands. We talk about expectations and regime uncertainty, that government regulation intervention discourages risk-taking. And while I think that's true, we don't have a theory of the magnitudes of those things. It's always easy to invoke those; as some commentators have observed on my blog and maybe in podcast comments: there's always uncertainty. There's always changes in regulatory activity. So, just finding them convenient to invoke when things are going badly is a little bit of confusing causation with correlation, potentially. You can get some evidence on regulatory uncertainty. You can survey people--not the best method. It's one method Bob Higgs has used, WWII, Great Depression, interventions of the Roosevelt era and how people were alarmed by them. You don't know whether those responses were real or whether they were just strategic in how they answered. But I think you have the same problem as a Keynesian. When you say things were worse than we thought, there's no--what I'm trying to get at here is this metaphor of the health of the economy. Unlike the health of a body where pre-x-ray, pre-catscan, you'd have to do an autopsy to find out if things were worse than we thought--and in an autopsy you could actually look and you could see that the cancer had spread more widely or the tissue around the arteries were more hardened. I don't think we have anything remotely like that that could add content other than the empirical outcome. Which you invoke with: look how unemployment went. Let me go back to the point I was making; spring of 2008 and spring of 2009 or winter of 2009 in terms of the information people had. But what I'm invoking here as evidence that things are worse than many expected is that people were making public statements about what they expected the economy to do. I take them more or less at their word, that their forecasts were honest and not strategic. The forecasts for some time turned out to be grossly positive. So, there was a sense of disappointed expectations. I think we'd come out of this period of the Great Moderation period of macroeconomics: recessions are troublesome, we don't want to underestimate their social cost, but they tend to be relatively mild and relatively short; and in particular monetary policy is our mechanism that can mitigate the worst sides of the business cycle. I think this was the view coming in, the view Bernanke was reflecting in the spring of 2008. Might have stretched credibility a bit further by the spring of 2009 and late 2008 or when the original Obama policies were being settled right after the 2008 election. Christina Romer was reflecting, too: this economy will come around on its own but it could use some help. But they were too optimistic. I think the data show they were too optimistic. I think the evidence that things were worse than they thought is that the forecast came out differently. I suppose you would say that's a bit of ex post reasoning. This argument is subject to that criticism. Steve, you've got to let the host do some of the work here! You made a sensible, self-aware, humble comment so I thought I would follow your lead. People were making their forecasts and they turned out to be wrong. Pretty badly wrong; we're not talking about a little. Let's invoke the private sector. I think the point I could be making applies to the mainstream macroeconomic forecasting community. Absolutely. So, everybody's too strong. Maybe a few fringe Keynesians which I will sometimes associate myself with saying: We really were saying ahead of this whole thing, look out, it's going to be different this time. If you cut across the mainstream of the forecasting community, I think there was a reasonably tight sense that more or less corresponded to what Christina Romer was saying in January of 2009. Two criticisms. The left's criticism: We knew that the stimulus wasn't big enough but they couldn't get it through Congress so they settled for what they got and put a good face on it, but they knew all along it wasn't going to turn out very well. Don't know if that's true. Don't find it compelling. The other observation is that all those mainstream models are Keynesian in the sense that they anticipated some natural stable relationship between government spending and output--not just output but employment--and those didn't happen. Two interpretations. One is the models didn't apply to this particular time period, or the underlying situation was worse than we thought. Stuck. One of the themes listeners are aware of here is I'm increasingly skeptical of the scientific nature of economics. This would be an example of where I don't think there are any data that would let us resolve this dispute. I think it's philosophical. I don't know how to characterize it precisely, but I don't think if we had all the time in the world, you and I could come to an agreement on this. I suspect you are right. I have some sympathy. Question if there is a scientific economics; I join you in that questioning without having come to a personal conclusion on the answer. But to some extent my point about judgment applies. I think we are more in the world of a historian than the physicist. Our approach is going to be to pull the world together such as it is. There will some qualitative, some quantitative evidence; what people said, survey kinds of things. At the end of the day this is not going to be definitive and people are going to have to make a judgment. Much the way you might make a judgment about what are the forces that caused WWII or what was the motivation of the Civil War. Well said. Evidence: part of it, too, is the definitive one- or two-sentence story of this is what happened, we're not going to have here. But there are other aspects of the Keynesian story that fit. For example, if you talk to a classically or neoclassically oriented economist about whether markets are all in equilibrium; we measure a lot of unemployment but it's really voluntary in some sense, and ask what happens when government deficits go from 2-3% of GDP up to 10% of GDP, they say: My gosh, interest rates are going to skyrocket. Well, they didn't. We don't want to get into the arcane details of different models, but that is a fairly direct prediction of a Keynesian prediction when you are operating below full employment. Inflation slowed during this period of time even though we had all this expansionary monetary policy. There are these other pieces that don't directly say here we can see the evidence that the stimulus worked, but they do fall in line with the Keynesian perspective on what's happened. I'm not going to defend that particular version of the classical and neoclassical market that all markets are in equilibrium all the time. Of course, a monetarist--and this is an example of the unscientific nature of economics--will have a story, one of which is the banks are holding all their cash; people are anxious about the future; that's why we haven't seen the inflation, we're going to have it. You can always tell a story that fits the facts better than another story. I think that's what historians do. They tell stories. They marshal the evidence, and some stories are more convincing than others. Our understanding of history and economics moves forward in fits and starts and sometimes doesn't move forward at all. Just different story-telling. You and I come down on different sides of that judgment call, but I think the way we reach the judgment, we agree entirely.
20:17Let's go back to the fundamentals of stimulus in Keynesian economics. Steve is a superb Principles teacher. I want to think of this as a textbook story for the moment. When we teach our students our models of macroeconomics, the Keynesian model talks about an insufficiency of aggregate demand and that government spending can fix that under certain circumstances certainly through borrowing money and spending it. I've posted on Cafe Hayek a lot of my puzzlement over these kinds of arguments. I don't mean this doesn't make sense; rather I'm skeptical about it so I want to try to get at where my skepticism has some traction or where I'm misinformed or misunderstanding. Tell me what the argument is for why the stimulus should have worked. Forget whether it actually did; talk about the logic of it. In some respects I'm going to go to some of the topics we just mentioned, the issue of demand or aggregate demand. It doesn't really matter whether the demand comes from the public sector or the private sector. I often have a discussion with people who are somewhat conservative ideologically, businessmen; and they say you sell more when people are coming through the door. Everybody says yes, of course that's true. Any sensible economic model is going to say you only produce what you can sell. You start from that intuition, though: the number of people who walk through the door and put their money on the counter is an important factor in how many workers you'll hire. More controversial point is there are times if you just look at the private sector itself, or you hold government constant, and demand--willingness to lay money down on the counter--falters from the private sector, that there is just an insufficiency of demand spread broadly throughout the economy. Without some outside force intervening either to gin up private sector demand or to create demand in the form of something like government spending, that business will simply not be able to sell enough output to fully employ the labor force. At the most basic Principles level, that's the simple intuition of Keynesian economics. Let's talk about some different forms of stimulus, taking that structure as correct. I always think of various ways we could increase spending. I think it's important to mention--I don't know if this bothers you or not, but it bothers me--when people confuse spending in nominal terms and income in nominal terms with well-being. Not just because money isn't all we care about--which is true--but also because we care about what our spending and access to goods and services actually is, not the dollar value attached to it. It's easy to increase total spending in the economy by printing money. Here are the experiments I have in mind. Case 1: we print money and give it to people. They find themselves holding more money than they expected, and they go out and spend it. Case 2: the government creates a project that has real economic value. It builds a road that serves an area that is underserved and so is a productive use of resources. The third would be it pays people to dig ditches and holes and fill them back in--a somewhat satirical example that Keynes used. Those three all increase aggregate demand. They strike me as very different, though. Are they different? Yes. Let me be clear. Hiring one group of people to dig a hole and another group to fill them back up again is one of the great red herrings of these debates in the past. I am not in favor of such fiscal policy. I heard Joe Stiglitz say live that even though that is not the best use of government money, it would still improve the economy. I think I'm inclined to say that at least from a debating point of view, that's not a good issue to even engage on. Smart. Let's talk about number 2: increase in government spending that produces a project of value. Let me argue that there are two things that come from that. One, we don't have to talk about Keynes or even macroeconomics. We can talk about public economics, which in the profession is microeconomic. Cost-benefit analysis. There are some things it's better for the government to do than others. We could have substantial differences about the details, whether this or that project meets the criteria; we probably would not disagree very much about the basic framework for making the evaluation. It could be, and many on the left argue a microeconomic argument, that we are underinvested in infrastructure, education, green technology. They may not make the argument this way, but if I were trying to clarify the underlying argument, I would say recession or not, it would be useful for the government to engage in this. Or not--and that's a debate one could have. But the argument often presented is: It's a twofer. You get a beneficial outcome like a green technology or a road you wouldn't produce privately but which has net benefits, and you stimulate the economy. Yes; it's that second stimulate-the-economy part. Here, I go back to my simple Keynesian intuition that if more people are coming through the door, that business will produce more and hire more people. There is no reason the demand has to come from the private sector. The government building a road which hopefully is of social value independent of the state of the economy--if it turns out that private sector demand is insufficient and we have involuntarily unemployed resources then the government's money is as good as anybody's money. Construction company will be motivated to employ those unemployed resources just as well. But that wouldn't be any different in Case number 1. In Case 1, the case where we just print money and give it to people, there are going to be people who show up at those stores asking for more goods. Shouldn't Zimbabwe have a booming economy? Nice rhetorical flourish, hold off on Zimbabwe. Let me broadly agree with you. There is an allocational difference. Basically saying if the government stimulates demand through an infrastructure project, that's helpful in bringing resources into use. If the government stimulates demand by putting more money in people's pockets and they will take that money to the stores and spend it, that also stimulates. The Keynesian part of that, to the first order, is similar. Allocational difference. And important issue, much understated--allocation matters. I am strongly of the view that the economy needed in the past years and still needs more demand stimulus. I don't have a strong opinion on whether that demand stimulus should be money in people's pockets that lets them on their personal, decentralized individual preferences decide how that money will be spent, so which kinds of businesses will be stimulated, versus a government allocation policy where it's a bureaucratic process that decides that. When I say bureaucratic, I suppose for your listeners, that's sounds like a nasty word. I don't mean to imply that. Political process that will decide how the allocation goes. In the broad sense, whether you print money and put it in people's pockets or whether you have government infrastructure programs--as a first order effect qualitatively the same. Could argue one more effective than the other; different dynamics. The kind of twofer argument people on the left would make is that we need more government activity, so this is a particularly good time to create these public goods which have been in deficiency for years. All sorts of nuance.
30:48So here's the question, then. Let's now talk about the practical aspects of the $800 billion or so--by the way, it hasn't all been spent. People have said that the stimulus was poorly designed even from a Keynesian perspective, and what it was spent on was not particularly useful. So let me talk in terms of stimulating--forget the political side. About a quarter of the lost jobs from December 2007 have been in construction and manufacturing. A lot of jobs; a lot of carpenters out of work. A lot of tool and die people out of work. As a result of that--I may have this wrong, may be half. But hundreds of thousands of jobs have been, maybe more than a million, lost in those two sectors alone. Employment is way down. Part of it: when you build way too many houses, as part of the thing that kicked the problem off, you are going to find that the demand for building new houses--just had the recent data on housing starts, horrifically bad--if you are a carpenter it's going to be a long time between jobs unless you are one of the ones who is still working. Manufacturing: we've got a combination of effects, part of a long-term trend that started in 1950 away from manufacturing, partly I think because of tariffs we put on steel, partly because of increases in productivity, partly computerization playing a major role in reducing labor costs. In both cases you have people sitting around thinking: Should I go back to work with the skill that I have--say, carpentry, tool-and-die operator--or should I take a lower paying job and suffer with what I've got. The stimulus itself didn't go to create jobs for carpenters or tool-and-die operators, to take two examples. A lot of it went to the states to continue paying unemployment benefits, which is a nice thing if you are unemployed. It went to keep teachers and other state employees from being fired or taking pay cuts even, which obviously is politically attractive. I know that Washington University in St. Louis got some nice funds to increase medical research. But those aren't going to get to the folks who are unemployed. The way Arnold Kling talks about this on EconLog: There isn't a factory that produces GDP. We have very diverse economy. Sectors that are hurt are not going to be helped much by spending in other sectors. The other example I would give is geographic unemployment. Chart on metropolitan unemployment--very high in urban areas that had the housing boom and bust. California, Florida, Nevada, Arizona, Michigan--all have above average unemployment; others have below average unemployment. Yet the spending went out in different ways all over the country. Isn't it wrong to just think about insufficiency of aggregate demand? Largely I agree. Additional flexibilities in the economy by which the stimulus can be more effective than your last few minutes of comments might suggest. Would it be better for Keynesian designed stimulus, basically trying to raise demand, to be targeted to those occupations and regions that are most hard hit? I would say, Yes. I think that would be better. You can imagine political problems. If you just said: California, Florida, Arizona, Nevada that we really need to help. Politically tough sell. Moral hazard issue: this is where people were the craziest and we are actually rewarding them after the fact. Rewarding might be a little strange--not a happy outcome--but we know that's a bit of an issue. Practical problems, but if you think of the economy as just producing homogeneous goods--which I call Stuff--GDP, not realistic. I accept the point that there is an allocational aspect. Let's go further. Our last podcast may have touched on this, too. I'm going to be the man on the left and be the one arguing: There's flexibility in the economy! We don't have Leontief technologies where there are a fixed number of people in every place. There's substitution that goes on. If a government project is building new highways that otherwise wouldn't be built, my guess is that many construction workers could find jobs there. May not be able to make as much money, but certainly the possibility and likelihood that people will move in those directions. You mentioned research funding coming to Washington University--public goods thing we could get into--but there aren't too many Ph.D. scientists who are losing their jobs because of the problem in construction, so what's the point? The Ph.D. scientist needs staff, help to take care of the infrastructure of the building; may need to build new lab space. Spillover. You do get a little bit of the best being the enemy of the good in this kind of debate. Interesting empirical question. We can actually look and say: This program funded by the stimulus, and see how many jobs are created. Whereas the key question is how many net jobs are created. Were those people already working, from a different industry but already employed? We can also see to some extent if carpenters are working in those projects. My guess is no, but that's my bias; I'm willing to be agnostic. It's harder than that. Suppose we build a new research lab and hire an administrative person who was previously employed in legal office. Now there's an opening in that legal office. I think this does get at the heart of what we talked about earlier--how hard it is to know what's going on. How do you know, if you are running a restaurant, that your customers coming in have a job because of the government stimulus program? You don't know and you don't care. Irrelevant unless you expand your restaurant and the stimulus ends--unless it's the first example of printing money--you may find that that was not a real change, a temporary change and you've over-invested. Nominal changes can have short run effect but unlikely to be sustained. But certainly a business person has trouble figuring out where the demand is coming from. To some extent, shouldn't worry about it too much unless it turns out to be really nominal and hence unsustainable.
40:37Only thing I want to add to this discussion about the GDP factor is one issue that doesn't get talked about is if in fact you are stimulating sectors where there is full employment, where resources are fully engaged, whether it's lumber for building or steel for the new lab you are building or Ph.D. researchers, pushing up the demand for those things means the effect isn't on quantities but on prices. Implicit in the GDP factory story is there is this slack that the extra demand soaks up and therefore pushes people into the labor force. If you are pushing people into areas where there isn't slack, you are going to have price effects that are going to be seen rather than quantity effects. I agree with that. The Keynesian stimulus argument requires that there be underemployed resources in some general sense. This is why Keynes titled his book the General Theory, because he argued that should you have enough demand to get the economy to full employment, the novel aspects of his book no longer apply. Unemployed resources are critical. If you break this down to the sectoral level, I agree that you could certainly have price effects; though they could not be simple rise in price that's fully inelastic. Could imply some expansion which then could draw unemployed resources from other parts of the economy. All pretty complicated. Let's turn to the labor market. One of the things that mystifies me: you said earlier that there's this classical or neoclassical view that all unemployment is voluntary, all markets clear. Certainly the Austrian part of me would say that's not true; the neoclassical part of me would say there is imperfect information, people may not know the opportunities. Carpenter is unemployed--for all he knows, the demand for new housing may come back in 3-6 months, but it might be ten years, in which case he should probably look elsewhere. You can't foresee the future perfectly. An expert can't. Difficult decision. I don't want to debate about whether that's voluntary or involuntary. Obviously it's not ideal. Change of all kinds puts people in situations where they are unclear. That's part of the reason they are unemployed. Another reason is they don't know about every single opportunity out there every single minute. They don't know there's a job two states away or maybe across town. All kinds of imperfections. Let's put that aside. It's certainly true that--example from the tech world: In 1999 if you were an internet person, you were a skilled computer person, and your firm shut down--a startup that didn't make it--you were a little upset, disappointed, but you'd find a new job pretty quickly. In 2001 when the tech bubble collapsed, that wasn't true. But it didn't last very long, people found jobs. Today, we're in that world for a whole bunch of people. Hard to understand why it's working for some folks. For example, if you have a college education, the unemployment is about 5%, but if not it's about 15%. Thinking out loud here. We don't really understand those differences very well. Yes. I agree, something we need to pay more attention to. I think you are going to find that unemployment is up substantially even for college-educated people. It is. It's increased across the board. I think the gap has widened but not so much. It just isn't that much hardship for a college education. I guess it depends what you majored in and what your skill set is, even within that college educated group; I think there are a lot of people who if they lost their job they'd find another one pretty quickly. Right now, in the middle of a horrible recession. It's not a bad time to lose a job for some types of people, but for others it's a very, very bad time. Bit of an aside. As a professor I talk to my undergraduate students; they go through here, looking for jobs--can be a little bit of a downer conversation. Should be somewhat encouraging. If you look at the dynamics of the labor market, there are a lot of jobs. Even in the worst of the recession, a lot of jobs created every month. Monthly report on the first Friday of the month is the difference between the jobs created and the jobs lost. In a recession, jobs lost dominate, but if you look at the absolute number, there are a lot of jobs created. Even in good times, can do the symmetric thing--there are a lot of jobs lost every month. That's one of the things I find fascinating about this: We tend to think of the economic as this organic creature like a person's health. The JOLTS data, nice project where the government actually collects not net jobs but the creation. Jobs are created when existing firms expand or new firms come into being; destroyed when firms contract or disappear. In any one quarter, good times or bad, millions of jobs are created and millions destroyed. The difference is in healthy times, the number created is bigger than the number destroyed by a few hundred thousand in a given month. Or a little more or less. In a recession those numbers are different; harder to find a job. In particular last month net job creation was 103,000, which is not very cheerful when unemployment is as large as it is. We don't really understand that very well. Microeconomic dynamic. Very complex forces. There is interesting work along those lines: Olivier Blanchard, Peter Diamond who won the Nobel Prize this last year--but the paper I'm thinking of is the Blanchard-Diamond Brookings paper some years ago. This stuff has stayed fairly arcane. Another natural experiment that is puzzling: I think about the return of American soldiers after WWII--massive inflow of labor; similarly in the 1960s and 1970s huge increase in the number of women who decided they wanted to work rather than stay at home. They pretty much all find jobs. The unemployment rate doesn't go to 20% or 15% ever in those times. The economy creates jobs in a strange, organic, not-quite understood fashion; and all of a sudden it's not doing it. For me, one of the puzzles that Keynesian economists have to face is: I can go through with my students a detailed discussion of why demand may not be adequate to get us to full employment and the problems associated with that. Demand is not automatic. We can't rely on it to be sufficient really at any point in time. But, you look over the long sweep of history in the United States and you see employment growing, production growing. Influx of women into the labor force--our baby-boom generation coming into the labor force. There is some discussion of why these factors might have led to a somewhat higher than average long term unemployment rate in the 1970s and 1980s, but in retrospect those fluctuations look pretty small. One has to address that. Where does that demand come from? Where is the growth of the demand? There has to be an economic process which is generating over longer sweeps of time demand growth that seems to be following up this resource. Fascinating question. I'm doing some research on it, but not quite ready for the public yet.
51:15Fundamental question: In good times, people find through the division of labor, competition, emergent complex phenomenon called an economy, ways to use their skills that people value; people pay for those, swap goods and services; it all works pretty well, and then once in a while it doesn't. Why a carpenter right now is struggling to find work when a computer programmer in 1999 doesn't is a little hard--we don't fully understand the process. Let's turn to the question of economics as a profession. Do you think this crisis is going to change either macroeconomics or how we teach it at the textbook level? Good question, hard one. Macroeconomics has changed a fair amount in the past 30 years in various respects. Have to say yes, it will change it. How is harder to say. Much more emphasis on financial instability of various kinds. More emphasis on moral hazard issues in financial markets, possibly a kind of bounded rationality. A lot of models based on rational expectations kind of assumption, which is that people basically know everything there is to be known about the distribution of outcomes, so anything is the result of a so-called shock that was fully unanticipated. Just read The Big Short, interesting the kind of group think in financial markets. Book is more about the pariahs who went against the group. We might see a broader perspective about how expectations are formed, how people deal with uncertainty work their way into macroeconomic models. From my Keynesian perspective, pretty good years for Keynesians. Some pushback against a resurgence of Keynesian ideas; but that's been going on for 30 years. The fact that the debate has gotten more intense reflects that at least on the surface a lot of the Keynesian stuff looks more plausible. Wondering, somewhat expecting that the younger people at top universities are going to think about moving more back in that direction. Might see a rise in the copies of The General Theory sold. Our colleague Hyman Minsky, largely Keynesian framework--his book Stabilizing Unstable Economies is back in print after 20 years out of print. You asked about the economics profession. I think there will be more interest in the issues. My comment on the group think: bailouts, costs, a lot of people made a great deal of money. Something wrong with financial markets either inherently or my claim is it's bad policy. I agree with you that there's going to be an attempt to mingle finance and macroeconomics more closely. Most of the macroeconomists I know don't know much finance, and most of the financial people I know don't know much macro. Partly specialization, how we get rewarded for being specialists. There are people who do a little of both. I think macroeconomists were blindsided--to use another Michael Lewis phrase--by the role of the financial sector. We haven't had a financial crisis in our lifetime. Yes, they understood about the role of banks in the Great Depression, but the shadow banking system and role of leverage and how that worked its way through the housing market caught most mainstream economists by surprise. Not where people were looking. I think that will change. How do you really know things in economics: Current conditions as indicative of how the world was and always will be. Not just this year. There was a whole generation of macroeconomists who grew up, mid-1980s through 2007, Great Moderation period--includes us--in which monetary policy seemed to be very effective at stabilizing the economy. Start to think this is the way it always is and don't pay attention to other factors. Multi-decade-long period. Blindsided by the changes. Minsky the exception in this particular case. Going back to my groupthink explanation. A lot of people forgot how complicated the world is. Sad that in the midst of such a hard time for so many people, it's an incredibly interesting time to be an economist. Have not been more intellectually stimulating in my years at Washington U. If you study tornadoes, you don't really want to see anyone's farm get busted up, but if it's going to happen you kind of want to watch.