Steve Fazzari on Keynesian Economics
Jan 12 2009

Steve Fazzari, of Washington University in St. Louis, talks with EconTalk host Russ Roberts about Keynesian economics. Fazzari talks about the paradox of thrift, makes the case for a government stimulus plan, and weighs the empirical evidence for a Keynesian worldview.

Ricardo Reis on Keynes, Macroeconomics, and Monetary Policy
Ricardo Reis of Columbia University talks with EconTalk host Russ Roberts about Keynesian economics in the classroom and in research. Reis argues that Keynesian models are a useful framework for helping undergraduates understand macroeconomic ideas of general equilibrium. More generally,...
Steve Fazzari on Stimulus and Keynes
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...
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 12 2009 at 8:48am

The guest in his paradox of savings seems to ignore the fractional banking side of savings. His consumer/restaurant case seems to treat money in the bank like money in a mattress.

Let’s look at it this way.

Consumer has $5 a week in his budget for a restaurant meal. He gives it to the restaurant guy. The bank has 0 dollars. Bill Gates can borrow zero dollars.

The consumer now decides to put the $5 in the bank. The restaurant guy gets nothing. The bank now has $5 and now creates an extra $45 to lend. Bill Gates now can borrow $45. Bill uses $35 to build Microsoft and spends $10 at the restaurant to entertain clients.

The restaurant guy now has the $5 he needed to live plus an extra $5 he himself can put in the bank. The bank can now lend out another $45. Another guy wants to start a company that makes cellphone ringtones. He borrows $45. He spends $10 at the restaurant entertaining musicians, he spends $10 on Microsoft software, and he invests $25 in his company.

The restaurant guy now has $20. All because the original salary guy saved $5. Yes, there was a lean time but the restaurant guy is on easy street now.

I don’t see the problem.

Jan 12 2009 at 9:48am

Hi, karl.

This podcast is an effort to open a conversation between Keynesian and classical economists. Fractional reserve banking as you have described it is not an issue on which the distinction between Keynesian versus classical economics turns. Lockstep multipliers or fractional reserves are minor modifications that can be accommodated by either model, not fundamental sources of the critical difference between the underlying models.

You veered awry when you said:

The bank now has $5 and now creates an extra $45 to lend.

The primary question as posed in the podcast was whether the banking sector as a whole has even $5 more than before. If they have $0 more than before, then they can’t create so much as $5 more, much less the $45 more that you are claiming via your lockstep multiplier. It sounds as though you didn’t listen to even 10 minutes of the podcast.

The distinction between Keynesian and classical economics is not about lockstep matters like fractional reserves or multipliers. Multipliers and fractional accounting only distract from the main point and take time to talk through one by one.

The distinction is about deep and fundamental differences between models, distinctions that assume vastly different things about whether or not individuals can form their own reasonable expectations about the future or respond to current or intertemporal price, wage, or interest incentives. The question that underlies it all is whether or not government can help or change anything in times of economic crisis. Offering ad hoc models that assume people cannot think for themselves or that prices don’t adjust according to market forces or that tangential matters play out in some lockstep manner does not help illuminate the underlying distinction between the two primary models.

I’d ask about what’s going on with interest rates. But more on that perhaps in some other comment.

Andy Kneeter
Jan 12 2009 at 10:54am

The real issues are how levered the economy is & how is the best way to direct resources to the most productive areas.

Much of our recent economic growth was the result of increasing personal indebtedness (i.e. – credit card, home, etc…). To perpetuate that growth, underwriting standard of lenders continuously weakened. That’s unsustainable growth & we’ve reached the tipping point. If, according to Dr. Fazzari & Keynes, overall savings & income decrease, because less money is circulating throughout the economy, so be it. It was an unsustainable boom to begin with (fueled in large part by federal government policy, but that’s a different conversation).

How to direct resources to their most productive uses is critical. The historical evidence is clear that free markets are far superior to central governmental planning & spending in achieving this.

The problem I have with Dr. Fazzari & Keynes is there’s no supporting evidence to their arguments.

Economics has made great strides in real-world measurement (thanks in large part to Milton Friedman). Before that, it was primarily theoretical, resulting in all sorts of unfounded, yet reasonable-sounding ideas that end up causing tremendous damage in the hands of misguided politicians.

Jan 12 2009 at 11:13am

Good podcast!

As Russ and Steve point out, there are lots of moving parts in the story where a consumer decides to save more of his income and spend less at a restaurant.

I’m curious about the extreme example where consumers decide to save all their money. I don’t see why all incomes would drop to zero. Indeed, in an economy where every person demands investment goods, there would be lots of employment to be had producing investment goods. Furthermore, producing investment goods would be the only way a former restaurateur could prove to a bank that he could repay a loan to tide him over during his switch to new employment.

But perhaps I’m missing something.

Douglas B. Rasmussen
Jan 12 2009 at 11:43am

I am certainly beyond my depth, but I found this discussion fascinating and so I must ask the following question regarding changes in overall aggregate demand, which, according to SF, is the centerpiece of Keynesian theory.

Regarding the example used in the discussion, if the decision by the family to consume less and save more does not result in a net increase in aggregate savings, because the restaurant owner must lower his savings to keep his current rate of spending, then why is it not also the case that there is no net decrease in aggregate income, because spending by the resutant owner is the same? I thus see no change in overall aggregate demand, and it seems that one has to go to some other consideration to explain overall “booms” and “busts” in the economy.

I may, of course, be confused.

Jan 12 2009 at 11:49am

||You veered awry when you said:

The bank now has $5 and now creates an extra $45 to lend.

The primary question as posed in the podcast was whether the banking sector as a whole has even $5 more than before. If they have $0 more than before, then they can’t create so much as $5 more, much less the $45 more that you are claiming via your lockstep multiplier. It sounds as though you didn’t listen to even 10 minutes of the podcast. ||

Your final comment isn’t entirely fair (or true). Even the host Russ seemed to have a problem trying to understand the guest’s consumer/restaurant example and kept coming back to it.

Let me put it this way.

The consumer gives the restaurant guy $5 for the meal. The restaurant guy pays his expenses and is now left with $2. He puts $1 in the bank. The bank can lend out $10 (lets go with a 10x multiplier).

The consumer puts his $5 in the bank instead of giving it to the restaurant. The bank can now lend out $50.

Or lets work it another way:

You can give your $5 for a service (say a restaurant meal) or you can give your $5 to someone who will make that $5 grow into $20 by creating a more efficient way of doing something.

The bank, as Russ noted, can ultimately be treated like a middle man. You give the money to the bank, the bank invests it.

Jan 12 2009 at 12:36pm

Does anyone know what Keynesians would say happened in the 1970s when we had accelerating inflation and growing unemployment?

Josiah Neeley
Jan 12 2009 at 12:59pm

I thought the podcast was great, and Prof. Fazzari’s illustration of the paradox of thrift was very clarifying. At the same time, I also found myself shouting “No!” periodically while listening to it.

When explaining the paradox of thrift, for example, Prof. Fazzari says explicitly that you have to hold total income constant. Yet his conclusion is that, in the case he provides, total income will fall. This, to put it mildly, is an indication that something has gone wrong somewhere.

Jan 12 2009 at 1:20pm


At the same time the consumer spends $5 less at the restaurant and deposits that $5 in the bank, the restaurant owner deposits $5 less in the bank. The consumer’s bank lends out $50. The restaurant owner’s bank simultaneously reduces its lending by $50. Or whatever you want the fractional reserve multiplier to be.

The extra bank deposit or extra saving of the consumer is exactly canceled out by the reduced bank deposits or reduced savings of the business owner. Fractional reserve banking that would increase one would decrease the other by the same amount.

The offsetting behavior by the business owner is the puzzle presented by Fazzari in the podcast. The question is not about what an individual bank does, but about what businesses, consumers, and the banking sector as a whole do in total and in response.

The key to answering the puzzle is not in what the banks do, but in the response of businesses to reduced demand or other economic phenemena that they might see and foresee simultaneously to the consumer’s choice.

Jan 12 2009 at 1:31pm

Let me think about this $5 a third way.

The consumer gives his $5 to the bank. The bank can lend out $50.

The consumer gives his $5 to the restaurant guy. The restaurant guy pays his expenses ($3), pays for his own personal living expenses and enjoyment ($1), and saves $1. The bank can lend out $10.

Everyone who got paid expenses out of the $3 or got paid out of the $1 for the restaurant guy’s living expenses also do the same. They pay their expenses, they pay their living costs, and they save some of the $4 that’s out there.

The amount of that original $5 is saved is less and less as it gets handed down and handed down but a Keynesians would argue, I gather, that all those tiny tiny fractions will aggregate to the $5 anyway. Then the bank will, eventually, have the $5. As my old boss used to say, if you can’t make it in popcorn, make it in peanuts.

At this point I’m reminded of the second law of thermodynamics that says every time we change energy from one form to another we irrevocably lose some of that useful energy to a form of energy that can do no work. It’s much better to go from oil to gas then oil to gas to using that gas to power a pump that pumps water behind a dam and then using that dam to generate hydro and then using that hydro to charge your GM Volt and driving to work. You lost a great deal of energy in the whole ugly process.

This then makes we wonder is a financial system similar? Is there some loss in the system? The restaurant owner needs to drive his car to the bank to deposit the night receipts. The bank needs to drive all the collected cash to a fed bank. There’s a loss of something there. Electronic payment methods eliminate much of this inefficiency of course so it’s just a small example.

So, I don’t know. It’s instinctual for me to think the consumer saving instead of spending gets the full amount in the bank for lending and if the $5 is disbursed into many hands, the bank won’t actually get $5 in the long run or anywhere near the $5.

If Keynesians have any empirical evidence to support their claim, I would like to see it. Just as I would like to see evidence of my second law of cash-o-dynamics hypothesis.

End of the day, there is some “right” amount that should be saved/invested. The guest said posit a world where no one spent. Well, also posit a world where everyone spent 100% of what they earned on simply consuming.

Jan 12 2009 at 2:01pm

Hi, karl.

What you’ve done in your last comment is clearer than your previous comments. But the deal is this: you’ve described an equilibrium. And describing equilibria is relatively easy. Equilibria are static. Business cycles inherently involve events over time–dynamics.

The question posed by business cycle theory–by Keynesians and classical economists alike–is: what happens out of equilibrium? And also: Are business cycles disequilibria or not? (I.E., Are they maybe just movements toward a whole new equilibrium necessitated by some large underlying change?) If a disturbance causes the economy to move out of equilibrium (say, a consumer suddenly spends $5 less at a restaurant because his tastes change, or more dramatically, if suddenly the whole economy’s balance changes at once because of some large and unanticipated shock), are there natural forces that move the economy back to equilibrium? How quickly? Can we look empirically at the adjustment process and see if we are moving toward equilibrium or away from it?

The Keynesian answer is that at least in some circumstances there are no natural forces that move the economy back to equilibrium. The economy sometimes gets stuck; and only with the aid of government can it return to balance or a path toward balance. The reason could be as described by Fazzari, based on one of Keynes’s original concepts that any attempt to save once equilibrium is breached is instantly offset by the actions of other entities. It could be because of wage or price or interest rate rigidities, or other market failures, or insufficient expectations. But one way or the other, the Keynesian answer is that the natural forces of prices won’t work–in certain circumstances.

The classical answer is that that makes no innate sense no matter how many macro textbooks have repeated it in the last 50 years; and if it’s true, then we need a model that is consistent with the microeconomics with which we do all agree. If the best we can do is say after each new historical economic crisis or resolution that it was a unique event, then proposing solutions is at best a shot in the dark, and may cause even worse long-term or short-term repercussions.

Lee Kelly
Jan 12 2009 at 2:31pm

Suppose that the family stopped spending at the restaurant, but saved the money instead. Fazzari claims that any saving done by the family will be offset with less saving by the restaurant owner. But it depends on why the family is saving in first place.

Suppose that interest rates have increased. With higher returns on saving there is more incentive to. In consequence, the proportion of income saved by both the family and the restaurant owner increases, even if the restaurant owner’s savings decline in absolute terms.

More credit is then available for society as a whole.

Josiah Neeley
Jan 12 2009 at 2:42pm

The more I think about Prof. Fazzari’s example (of the paradox of thrift), the less persuasive it becomes.

Prof. Fazzari is right to note that when a family decides to divert spending on eating at a restaurant to savings, then ceteris paribus the restaurant owner will face a corresponding reduction in income. So the restaurant owner will face the choice of reducing consumption or reducing savings. Prof. Fazzari, however, just assumes that he will respond by reducing savings, and indeed this assumption is pivotal to his conclusion that overall savings won’t increase. So, basically, Prof. Fazzari’s example assumes exactly what it sets out to prove, which is a nice rhetorical trick, but logically speaking not so hot.

Lee Kelly
Jan 12 2009 at 5:15pm

Farrazi talks claptrap. He claims that when people save more, the pool of money available to lend does not change. Apparently, when one person saves some of their income, another loses an equal amount of income to save. Therefore, he claims, the pool of money available to lend does not change because of saving. Farrazi offhandly remarks that this holds under ‘some formal assumptions’. It would have be interesting for him to share those formal assumptions, since they cannot possibly describe reality.

If A saves $10 instead of spending it on B’s services, then why would B save $10 less than before rather than spend $10 less, or save $5 and spend $5? People tend to save a proportion of their income, and it increases when the rewards of saving are greater. If A saved $10 instead of spending it on B’s services because of a rise in interest rates, then wouldn’t B also have an incentive to save a higher proportion of his income? The consequence of either is a net increase in pool of money available to lend.

Right now the U.S. is in a deflationary bust. People have used up too much credit too fast. As the money supply contracts, interest rates rise and people begin to save more (or this would be happening if government did not control the money supply). New savings are then the source of capital for new entreprenurial ventures. By overextending credit, the Federal Reserve has impoverished the U.S.’s future, and now the market is signalling people to stop consuming and start saving, or else there will soon be nothing to left consume.

The Federal Reserve does not actually expand the credit supply, it just fools people into thinking that the credit supply is expanding. The real credit supply is accumulated by real saving, and lays the groundwork to increase future consumption.

Jan 12 2009 at 5:39pm

Call me confused but in the restaurant fable: is the amount of savings fixed but the amount of consumption variable? My reaction is err… narr… how… no.

Can someone point me to an empirical example?

Jan 12 2009 at 6:01pm

I think Prof. Fazzari’s example of the restaurant seems pretty reasonable. Isn’t he basically explaining the identity that includes the velocity of money and GDP, while pointing out that a payment for food (or anything) merely shifts the money from one bank account to another, which will be available for investment regardless of whose account it is in?

However, there are many problems with Keynesian economics and its over-simplification of reality. This podcast led me to realize yet another problematic simplification: that the root cause of economic downturns is aggregate demand. To me, that skips a few steps. It’s not as if people don’t want stuff anymore. The idea that an economy is driven by “aggregate demand” ignores the true role of money in an economy, namely that one’s income is supposed to be reflective of one’s productivity. Money serves as a store of value from production and limits coordination problems in multi-party exchanges of production, more easily allowing for specialization and a more productive economy.

In the recent bubble (and in all bubbles), people were earning money to provide goods and services that were overvalued. Once the bust onset, those people were then not paid to continue to produce those devalued goods and services, e.g. build even more homes, loan money to people with bad credit histories, securitize mortgages, etc. In other words, many people were specializing in production that was no longer valued. This reduced their incomes, which has some spiraling effects into the goods and services that those people purchased, etc. But ignoring why this aggregate demand was reduced – because people found themselves producing goods and services that were no longer valuable – is an important point to understanding that it is not helpful to simply pay idle workers to do unproductive activities in order to reduce unemployment and increase “aggregate demand,” such as in the hole digging thought exercise. Such a program would merely create a government sponsored, unsustainable hole digging bubble at a cost to others – either by taxation (now or later) or inflation, without increasing productivity (in fact, it may even hurt it in the long run by reducing the incentives of stimulus-hired workers to seek other productive opportunities). To the Keynesian, it doesn’t seem that productivity matters, just that people are given money. Keynesian stimulus ideas are partly just wealth redistribution exercises (obviously the programs of a real stimulus program would have more value added than digging holes, but those who are skeptical of politicians think it’s likely that the government will overpay for the value returned by stimulus spending).

Having said all that, I do think there is some merit to investing in public goods during downturns when the increasing supply of idle workers means that the government can employ them on the cheap. But I am skeptical of those who call for nearly $1T in stimulus spending, throwing the caution to the wind with respect to the US government debt, those who are not honest about who really bears the costs of such spending, and those who have too much faith in their models with multiplier effects and don’t discuss diminishing marginal returns of such spending.

Jan 12 2009 at 7:18pm

I think the wrong step Fazzari takes is not realizing that when I spend $100 less on whatever, it’s not a $100 reduction in net income for the business that would have received my money. The business can save, perhaps the whole $100, but a significant portion of it, by not utilizing as much steel as it would have. This is a real saving of resources that can now be put to use elsewhere. Perhaps there is some truth to the Paradox of Thrift, but it is overstated by 1/(profit margin), I think.

Barry Kelly
Jan 12 2009 at 8:56pm

Excellent podcast Russ, best one in some time.

These kinds of talks, where there is some degree of difference between the guests, lead to much more probing and testing of the respective positions, rather than having two folks preaching to one another’s choirs.

I personally found Prof. Fazzari’s specific explication of the Paradox of Thrift solid and enlightening. As a software engineer used to systems thinking, it made a lot of sense.

Lee, Josiah – The reason I expect Prof. Fazzari is suggesting that the restaurant owner is going to reduce saving rather than reduce consumption is because it’s one of the things he’s holding constant.

However, the problem with you two trying to use this to wriggle out of the Paradox is that you then need to explain what happens to the people who the restaurant owner is then consuming less from. Eventually, someone along the chain of transactions is going to be putting less cash into the bank. If you keep “following the money”, you must come to the conclusion that all that reduced consumption eventually gets converted into reduced savings (even if “savings” means “current account”).

Maestro – I’m not sure what you’re disagreeing with. One of the things Prof. Fazzari explained is that saving reduces income in the system as a whole. If everybody saved 100% of their income, there would be no income at all, and all resources would be freed up; but because there would be no demand, the resources can’t be put to use elsewhere. Similarly, if the business has a reduction in income of $100, yes, it will be producing less, and using less resources, but that simply means that either the total amount of resources being used is going to decrease, or alternatively that the cost of the resources will go down – i.e. deflation.

At least that’s the way I understood things from the podcast.

Jan 12 2009 at 9:42pm

One question I would have for Prof. Fazzari (and folks like Barry Kelly) is where did the restaurant come from? In other words, how was it made possible?

Leo from China
Jan 12 2009 at 9:46pm

Isn’t it all about resource allocation? The family didn’t spend another $10 in the restaurant; and yes, the bank didn’t directly get a $10 increase in funding because saving from the restaurant is reduced by $10. But that $10 will be loaned out to some “other”, and hopefully, better productive business, like, a software company? The issue here is where is that $10 spent. Spent on food, drink, or a new tv, it doesn’t really do much good in increasing productivity. But spent on education and new technology, that could lead to actual increased productivity and thus more income.

The spending orgy and over-indebtedness in nonproductive areas lead to the mess today. Market is trying to adjust itself by directing resources away from consumer spending. Americans today spend much less and save a lot more. Yes the banks don’t necessarily get more overall funding, but the money we are not consuming on food and drinks might hopefully be loaned to truly productive businesses. Of course, the banks could decide to lend the money to restaurants or car companies so that they can have more money to “lure” us back into the same consumption pattern via marketing and such.

It’s just all about resource allocation.

Another thing is, which is why I like physical science much more, economists love to create complicated “myth” that simply can’t be verified! It’s a shame, in a way. If some fancy “theory” can’t be explained by common sense, it’s probably not that “fancy” after all. In the case of the Keynesian Economics being employed today to “solve” the crisis, should it not work, they can just say, “hey, the reason it didn’t work is that you didn’t spend enough money!”

Ian Lippert
Jan 12 2009 at 10:05pm

There’s a couple of problems with his restaurant example. First of all, when you spend money at the restaurant not all of that money is going to go to the restaurant owner. The amount of profit the restaurant owner recieves might be something like 5%. The Keynsians respond with a turtles all the way down response. so 5% to the owner, 5% to his workers, 5% to his suppliers, etc. But I am skeptical of this, at some point there has to be costs to the transaction that arent redemable in wages to labour. I cant really articulate it, but I think it has to be true.

If we accept that there are costs not redemable to labour then the act of saving actually increases the amount of wealth in the system because it will be 100% redemable plus interest to the saver instead of 100% minus factor costs. Consumption costs society resources, overall utility will be increased but wages will be less.

The second problem I have with the example is that it doesnt take into account savings that are disbursed into the economy. So yes me saving for my tuition will take money away from the store owner but how much of their current wages are being paid by the students that are dispursing their savings from previous time periods? How does this affect the Keynsian analysis?

Another thing that wasnt touched upon was what the Keynsian explanation was for systemic economic downturns. In this day and age I dont think “animal spirits” is a good enough explanation. It seems like the Keynsians dont analyse how we got here, simply what to do once we get here. But if you cant diagnose the cause how can you come up with any solution?

Josiah Neeley
Jan 12 2009 at 10:41pm

However, the problem with you two trying to use this to wriggle out of the Paradox is that you then need to explain what happens to the people who the restaurant owner is then consuming less from. Eventually, someone along the chain of transactions is going to be putting less cash into the bank.

I thought that too at first, but that can’t be the whole story, as it would violate the assumption that total income remains constant. If total income is to remain constant (Prof. Fazzari’s assumption) and the restaurant owner has lost income, then there must be one or more people in the economy who’s income has increased by the exact amount that the restaurant owner’s has decreased, and they should be able to “make up the gap” in spending.

If you think about it, you’ll realize that there is (at least) one person who’s income has increased is this scenario, namely the person who borrows the family’s saved cash. His income has increased, and by the exact amount that the restaurant owner’s has decreased. He, then, can make up the spending gap either by increasing his own savings or consumption as he so chooses.

Josiah Neeley
Jan 12 2009 at 10:46pm

Josiah – The reason I expect Prof. Fazzari is suggesting that the restaurant owner is going to reduce saving rather than reduce consumption is because it’s one of the things he’s holding constant.

Well, okay. My point is that if you define total income as being total consumption plus total savings, and then assume that neither total income nor total consumption change, then you shouldn’t be surprised when it turns out total savings doesn’t change either.

Pietro Poggi-Corradini
Jan 12 2009 at 10:50pm

I’d like to focus on the other example: the one where the govt pays people to dig each other yards. It seems to me that these thought-experiments can always be cranked up one more level. For instance, the claim was that the newly employed garden-diggers would then go out and buy stuff and this ought to stimulate supply. But the problem I see, especially if the number of new garden-diggers is massive, is that suppliers would then find it hard to hire people to increase their supply because so many good people are busy digging gardens. So whatever extra boost you might get from the new demand, it gets offset by a shift of supply to the left, i.e. in the end the quantity produced is still the same but it is sold at a higher price.

Michael Jones
Jan 12 2009 at 11:06pm

Barry Kelly,

You wrote in your comment to Maestro – “If everybody saved 100% of their income, there would be no income at all, and all resources would be freed up; but because there would be no demand, the resources can’t be put to use elsewhere”

I’m not sure that’s how Keynes would see it. I just finished reading chapter 3 of his General Theory and he wrote that D = D1 + D2. Where D = effective demand, D1 = consumption, and D2 = investment.

Since demand = consumption + investment, if consumption drops due to saving, then investment makes up the difference, and demand is kept constant. If there’s no income, it doesn’t mean that there is no demand, it’s just expressed as investment.

That’s where I didn’t follow Fazzari’s restaurant example. If a customer does not spend his $5 in a restaurant, but puts it in savings, it does not follow the restaurant owner will maintain the same level of spending, and withdraw $5 from savings. Any profit maximizing owner would reduce his expenses as much as he could and perhaps only draw out $2 from savings. So, there is still a net of $3 for the bank to lend out.

I would agree that $5 in savings may not equal $5 in “immediate” investment, since the owner perhaps has fixed capital in the short-run that he can’t immediately sell to offset the $5 shortfall. But in the long-run, that $5 savings will eventually become investment. How can it not?

Jeff Henderson
Jan 12 2009 at 11:23pm

Great podcast. It was a very intelligent, civil conversation that left me with a better understanding of where Keynsians are coming from.

That being said, getting a handle on this alleged paradox of thrift feels like grasping at a slippery bar of soap. Sometimes I think I see it, but then it disappears.

Lee Kelly
Jan 12 2009 at 11:31pm

‘However, the problem with you two trying to use this to wriggle out of the Paradox is that you then need to explain what happens to the people who the restaurant owner is then consuming less from. Eventually, someone along the chain of transactions is going to be putting less cash into the bank. If you keep “following the money”, you must come to the conclusion that all that reduced consumption eventually gets converted into reduced savings (even if “savings” means “current account”).’ – Barry Kelly

My primary message was that although the restaurant owner and everyone else he patronises may have less income, if everyone is saving a higher proportion of their income, then overall savings will increase. How a currency is distributed among them is irrelevent: the restaurant owner may have less, or he may have more. If the reward for saving increases and everyone begins saving a higher proportion of income, then savings will increase overall. More of the money supply will be directed toward investments. Concentrating on the exchange of money between people is to miss the wood for the trees.

More money will be directed toward saving and investing when the rewards for doing so are higher, no different than more money being directed towards bananas if they suddenly tasted twice as good. Farazzi might as well argue that if the family spends money on bananas instead of patronising the restaurant, then the restaurant owner will have less income to spend on bananas, and so consumption of bananas never changes.

Money being saved and invested does not just disappear but is loaned out. It buys capital goods that will eventually increase future output. Given enough time, the restaurant owner can enjoy the same quality of life even if the family does not eat at his establishment any more, and because of the increase in productivity their investing helped achieve.

Jan 12 2009 at 11:52pm

Steve said, “…the source of why the US economy has done pretty well over the last 20 to 25 years – I might think of it as an autonomous drive to more demand.”

I’ve always thought that US citizens are wealthy because of ever expanding efficiency and specialization, not because of a steadily increasing willingness to spend rather than save. That is, US citizens’ ability to have a lot of stuff is based on a real ability to produce a lot of stuff, not on a house of cards of always increasing spending that could collapse at any moment.

A serious question: if income would fall to zero if everyone decided to save all their money, what would happen in the Keynesian model if everyone decided to spend all their money immediately? Would that push income to fantastic new heights, resulting in everyone being wealthier than before? Could we do it again and again?

Jan 13 2009 at 12:00am

Farazzi might as well argue that if the family spends money on bananas instead of patronising the restaurant, then the restaurant owner will have less income to spend on bananas, and so consumption of bananas never changes.

I really like the way you’ve explained that Lee!

Lee Kelly
Jan 13 2009 at 12:43am

‘Would that push income to fantastic new heights, resulting in everyone being wealthier than before? Could we do it again and again?’ – Gary

Investment must at least compensate for depreciation and malinvestment, otherwise future consumption will decline.

It seems to me that this is the U.S.’s problem. Policy has focused on increasing consumption and creating bubbles (i.e. malinvestments). If the dollar loses value and imports begin to dry up, then major reindustrialisation will be necessary. Unfortunately, the U.S. Government has actively encouraged this and continues to do so.

Lee Kelly
Jan 13 2009 at 12:45am

Unfortunately, the U.S. Government has actively encouraged this and continues to do so.

The above should read:

Unfortunately, the U.S. Government is actively discouraging this and will continue to do so.

Lee Kelly
Jan 13 2009 at 1:26am

Many people are worried about deflation. But ask yourself: are these prices sending false signals?

I do not think so. People are beginning to default on or pay off their debts in greater numbers, real-estate prices have been exposed as false, financial companies can no longer evaluate risk reliably. These falling prices are sending very real signals: fewer people should be employed in retail and finance, builders should stop making more homes, governments should rein in their spending, etc.

A Keynesian stimulus would be like enacting price controls on a nationwide level. Injecting money into the economy will not stimulate demand, but rather fool supply for a short time. Before prices adjust it will appear like an economic recovery is underway. By corrupting the price signal, suppliers may be tricked into selling goods for devalued dollars and perhaps ordering more sotck, but this cannot go on forever. Perhaps if the U.S. could afford to go into more debt the supply might keep coming (since any devaluation of the dollar now would be offset by a revaluation in the future), but few people are going to be taking such a risk now. And for that matter, neither should anyone want them to.

Jan 13 2009 at 3:47am

I’m just as confused about Keynesian economics after listening to the podcast as I was after first-year macro.

If I decide to consume less, I’ll invest my extra savings. That savings will be used by other people to consume, or by companies to buy capital goods. The restaurant will have to cut production, and the capital goods producer will increase production. There’s no paradox here.

This is how an economy presumably works in normal times. Fazzari’s claim seems to be that my savings never gets invested (at least during a recession), and so total spending will fall.

Why on earth would my savings never get invested? I wish Russ had asked this more persistently, as this question seems to be at the base of much of the confusion about Keynesian economics. I was waiting for Fazzari to say something like – “in normal times, your extra savings would be lent out to someone else, but during a recession…”

Frustrating podcast…

Gary Rogers
Jan 13 2009 at 4:57am

Another great podcast!

I will be spending some time over the next few days or maybe even weeks thinking through the paradox of thrift but at this point cannot add anything significant. I do agree that a loss of confidence can cause a pro-cyclical decrease in economic activity where consumer saving decreases demand which leads to loss of jobs that leads to more loss in confidence. I see this as a self-correcting anomoly, though, where Keynesians seem to see it as the norm. What surprises me the most is that Fazzari seems to be arguing that there is no benefit to savings. Yes, you can do some thought experiments that seem to indicate this, but experience tells us that savings and investment are critical for a healthy economy.

Economic stimulus can be a useful tool for a temporary loss of confidence in markets, but by definition stimuli promote consumption over savings. If savings are not recognized as valuable and the government continues to add stimulus on stimulus, you end up depleting savings and require continually larger stimuli to make a difference. Sound familiar?

If our economy was healthy right now, I would agree with the Keynesians and recommend a stimulus to get things moving again. However, since I believe our problem is tapped out consumers and an over leveraged financial system from 50 years of successive stimuli, I think adding another trillion dollars to an already dangerous debt is stupid.

Jan 13 2009 at 9:55am

I found this talk to be one of the most interesting yet. I think the guest had many interesting things to say. Much food for thought.

I couldn’t help but notice his apparent sympathy for ‘coordinated’ economies. I find that terminology to be ironic as history as I understand it has shown ‘uncoordinated’ economies to have far more coordination than the ‘coordinated’ ones.

Still, much food for thought. Thanks.

Jan 13 2009 at 10:52am

What drove me crazy about Prof. Fazzari’s analysis was the way he conflated savings and income, as if they were the same thing.

If I save $5 instead of spending it in the restaurant, the restaurant owner *might* invest less (because it is true that businesses sometimes invest out of income), but it can go either way:

(1) The restaurant is well-run and if I eat there the owner invests my $5 to make his business better or puts it in the bank, as I could have

Alternatively, …

(2) The business is struggling to get by because the owner is a bad administrator or he spends all his income on Jack Daniels and expensive motorcycles.

TWO OPTIONS HERE. Whereas, if I put the money in the bank, it is much more likely to go to someone who is going to invest it in ways that will increase the total amount of capital in the economy and maybe even push forward the frontier of technology. In other words, isn’t savings more likely to be invested than income?

Could some economist in this thread explain to me why it is not obvious that the two cases are different?

Also, is it generally the case that economist lump income and savings together as if they were one entity?

Felipe form Brazil
Jan 13 2009 at 11:08am

Lemme try to ask a question about the Paradox of Thrift
At the same time the consumer spends $5 less at the restaurant and deposits that $5 in the bank, the restaurant owner deposits $5 less in the bank. The net results are the same, but then let’s add another dimension to this: Information.
When the consumer spends $5 bucks in the restaurant, and the restaurant deposits $5 in the bank. The information is: The services provided by the restaurant owner are OK. The “flag” attached to the entire process is: All Is Fine in This Arrangement or The actual state of the Economy is working fine.
However, when the contrary occurs, when for the Consumer, he prefers to save the $5 instead of getting that gratification of eating out in the restaurant (that would cost 5$). Although the net result IS the same in terms of money, we now have a NEW FLAG saying: Hey Economy: Time to Evolve! People prefer to save 5$ instead of eating in the restaurant, this sends all sorts of signals and incentives to “evolve” the economy to adapt to the changing needs of the people.
Isn’t it a very important aspect of the system?
As a parable of a biological organism, instead of sending energy to an old and increasingly obsolete organ, we are diverting that energy to the evolution of another organ. The amount of energy is the same, the application is not.
This savings sends signals and incentives, in fact, diverting resources to build a new and improved slice of the economy. It seems to me that this systems in caring valuable information, and that the Paradox of Thrift does not account to that.

Richard Sprague
Jan 13 2009 at 11:37am

The customer stops spending $5 at the restaurant, because he wants to save more for his kid’s college education. The $5 goes to the bank, which ordinarily would try to invest it, but can’t because…

The restaurant, now with less income, can no longer save like it used to. The bank has a net value of $0. This is the Paradox of Thrift.

However….what happens next? The restaurant can’t keep running on low profits, so the owner lays off one of the waiters. Now the restaurant can afford to save the $5 again, and the bank can now invest the extra money.

The bank, always hungry for profitable opportunities, funds a college prep company, which hires the laid off worker. The entire world is better off.

Under Kaynes, however, the outcome would be:

The government takes $5 extra from both customer and restaurant, through some combo of taxation and borrowing, which it invests in an electric car company. The laid off worker gets a job making electric cars.

Is the world better off? Might be, if the government is wise enough to predict and/or conjole the cusomter into prefering an electric car rather than college prep. But without incentives to bet correctly, it’s more likely that the government will simply skew the economy toward something nobody wants, and the customer will end up buying a hydrogen car instead from an importer. Then we are all far worse off.

Jan 13 2009 at 11:50am

Barry Kelly, I don’t see the relevance of his thought experiment. We could flip it around and say that everyone spent all of their income and saved nothing. Would the paradox of thrift tells us that savings would remain the same? With either thought experiment, we’re essentially dividing by zero. If consumption falls to zero, guess what, we’ll all be dead quite quickly.
Gigantic changes in behavior like this don’t fit into the models, and tell us pretty much nothing about what happens in realistic situations.

Madison Classical Liberal
Jan 13 2009 at 2:38pm

Fazzari was incorrect in stating that the restaurant owner would decrease his savings to keep production at normal level. In reality, the restaurant owner would not decrease his own savings but rather decrease his production of food (due to less demand now). His income would fall.

The person could then directly lend this money as a loan to the college he/she will attend in the near future so it can build a better college campus in anticipation of having to serve more students in the near future. The college has purchased a loan (money-capital). The college’s income grew by this amount.

Total income remained the same as savings increased.

Jan 13 2009 at 3:46pm

The problem is the difference in thought between wealth and money. Wealth is measured in goods and not money. So the people can become wealthier even as the amount of money they have decreases, whether in income or savings. I think this distinction is where Keynesians go wrong. His defense of spending by the government misses one point, and that is that the government spending does not actually create wealth as measured in the total number of goods people can afford. In fact WW2 may have on paper been a boon to the economy, the wealth of the individual dropped dramatically. There were rations on almost every product.

Jan 13 2009 at 3:58pm

To isolate the difference between spending and saving/investing in the thought experiment, why don’t we consider this set of alternatives (so we don’t have to think about capital re-allocation between businesses):

1) Family spends $100 in restaurant – cash infusion into business in exchange for immediate goods/services.

2) Family loans/invests $100 in restaurant – cash infusion into business in exchange for claim on future earnings.

Why is it obvious that #1 creates more employment than #2? Why should the owner of the restaurant necessarily prefer option #1?

Ray Mangum
Jan 13 2009 at 6:11pm

Another fascinating podcast. I’m glad you put this back to back with the one on the Austrian view.

I have to admit though, the paradox of thrift is still a big puzzler to me. I don’t think Fazzari adequately responded to the question of why, if personal savings creates no net increase in loanable funds from the bank, it should result in any net destruction of such funds either. Furthermore, he assumes the money I spend at the restaurant would be saved by the restaurant owner. But isn’t the problem that everyone (or most people) in the economy are living beyond their means, borrowing and spending rather than saving?

Also, Fazzari responds to the perennial reductio ad absurdum for Keynesian theory, that the government should hire people to dig holes and then fill them back up again, by saying basically “of course we don’t recommend anything that silly”. But that misses the point, which is that since such a policy is entirely consistent with Keynesian theory, so why not?

Another absurd conclusion: if what I take Fazzari to be saying is correct, I would be doing as much good for the economy, and ultimately myself (since my future employment depends on overall economic prosperity), by spending all my money at the nicklecade instead of saving for college.

Jan 13 2009 at 6:29pm

I am becoming increasingly convinced that beyond the influence of the type of Economics people are exposed to, there is a tendency towards Keynesianism in certain people who simply think differently and can’t ‘get’ the big picture, seen & unseen logic of the Austrian school. As Fazzari demonstrates, some of these guys are extremely bright, but seem to have a glitch in their thinking which can’t process this type of info. I am sure they probably have other thinking skills which we don’t have that are very useful, but trying to reason with them on macroeconomics is like trying to turn an extravert into an introvert or a right handed person into a left hander – they don’t get it, and in fact they won’t ever get it.

Mark K
Jan 13 2009 at 7:20pm

I thought Prof. Fazzari did an exceptional job of defending his arguments. The central question of this podcast turned on whether or not the Paradox of Thrift is correct or incorrect given our present economic circumstances. Whether or not the Paradox of Thrift is true is a measurable effect. The degree to which it is true is whether savings is increased (or decreased) compared to the decrease in consumers spending in response the economic downturn. The Paradox is probably not entirely correct, nor is it entirely incorrect. However, some of the best economic minds are focused on this problem, and a significant portion of them are convinced that Keynesian stimulus is the one of the best measures in the short term. There are several useful government projects that can be undertaken in the short term that are long overdue, such as improvements to infrastructure that will increase commerce productivity in the future. In addition, research and development of alternative energy forms is a government project that could pay big dividends in the future. Even though I prefer the price system to send signals about the relative scarcity of resources, oil prices move so rapidly (and the substitutes are so under-developed) that it is wise to plan for alternatives, particularly since much of the oil supply is in the hands of unstable or unfriendly regimes. I would not argue that long-term government deficits are desirable, and I have always favored balanced or near balanced budgets, but the Keynesians have an argument in the near-term horizon given the economic climate.

Jan 13 2009 at 7:38pm

If the family decides to save, demand falls, but the Fed prints more money because parts of the economy have gone idle, discovered via leading price signals. Thus the bank’s available stock of money in fact rises, and the paradox of thrift is cancelled.

A monetary system with fiat currency works like a barter system in that respect.

Jan 13 2009 at 8:22pm

Hi – I really enjoyed this podcast.

Does the paradox of thrift still apply in this

I get a raise. I put the amount of my raise immediately into savings. I am not forgoing any consumption in order to save the additional amount.

I guess one could say that my boss has less to spend, but hasn’t he given me the raise in the first place because my productivity has increased his income and he is returning a fraction of that additional income to me as a raise?

John Dyer
Jan 13 2009 at 9:45pm

So I listened to some of the growth podcasts at:

But, I’m still confused by all of this. I don’t see how injecting money into the system can cause an increase in productivity. Maybe short term growth but isn’t that a bit of a stretch?

Isn’t my work only able to generate so much wealth? Isn’t the only way to generate more wealth either work more or work more efficient?

Shouldn’t we then be looking for better efficiency and increased productivity so that we really are better off in the long run?

How will the government spending make us better? Or are they just trying to get us to continue to reduce saving and spend 100% of what we make? (Making things appear better in the short term.)


Jan 13 2009 at 10:10pm

I got the feeling that the way the savings paradox was described there was no way that aggregate savings could ever go up. What about down?

The paradox claims that if the restaurant consumer saves instead of consumes, the restaurateur will have to reduce his savings. The inverse situation of the example would be that the consumer spends his savings and the restaurateur increases his savings. The theory doesn’t allow for the reality that aggregate savings does change. Something is missing in the theory.

I think the simplification that the restaurant owner must reduce his savings rate is based on the idea that if he maintains his savings rate he will have to cut costs somewhere else, reducing someone else’s income and their savings rate.

Now, having said that, let’s look at what happens when the restaurant owner finds some cheaper way to run his business. Let’s say he managed his produce better and reduced his waste by half. His income stays the same, but his operating costs go down and he is able to save more. However, he is now purchasing less produce and his produce supplier loses income and must save less. Aggregate savings is unchanged.

What if every business in the world were able to operate 10% more efficiently? According to the Keynesian theory this would be terrible! There would be a chain reaction of reduced incomes and the economy would suffer.

The savings paradox seems to treat the economy as a zero sum game. One person’s gain is another’s loss. How does Keynes explain how wealth increases?

Josiah Neeley
Jan 13 2009 at 10:19pm

I don’t see how injecting money into the system can cause an increase in productivity.

Injecting money could result in increased productivity through what’s known as money illusion. When a person is deciding how much to work, they have to weigh the value of leisure time against what is being offered in compensation for their labor. For example, consider a worker who a) prefers working 40 hours a week for $40k a year to working 45 hours a week for $45k a year, and b) would prefer working 45 hours a week for $50k to working 40 hours a week for $40k. If a worker with these preferences is offered a job working 45 hours a week for $45k a year, he won’t take it.

Suppose, however, that the government injects a bunch of new money into the system, so that the real value of $50k is equal to the real value of $45k before the injection. Now the employer who had offered $45k for the 45 hour a week job can offer $50k, and if the worker doesn’t realize what’s happened he may end up working more than he would have if he knew the real value of what he was going to be paid. Multiply this by enough people in a society, and injecting money into the system can raise overall productivity simply by increasing the total number of hours worked.

Miles P
Jan 13 2009 at 11:20pm

Fazzari argues that if a family saves $100 rather than spending it at a restaurant, the economic impact of the family’s extra saving is canceled by an equal decrease in the restaurant owner’s saving. Thus, a $100 decrease in aggregate consumption generates no increase in aggregate savings. This is a “Paradox of Thrift.”

Maestro points out that the restaurant owner only has to draw down her savings by an amount equal to the *profit* from that $100, which will be less than the full $100 as long as profit margins are less than 100%. Fazzari’s implicit assumption must be that the restaurant owner faces incremental profit margins of 100%.

We make simplifying assumptions all the time in economics, but watch what happens here when we assume the owner faces some incremental cost to providing a $100 meal. For simplicity, let’s say the owner’s incremental cost is half the total price of the meal, $100/2 = $50.

Now, if the family saves $100 rather than spending it, then the family’s consumption still falls by $100. Their saving increases by the same $100. The restaurant owner’s saving decreases not by $100 but by the difference between the price of the meal and her cost, $50. In net, $100 vanishes from the consumption facet of GDP while $50 is added to net savings. There is still a $50 decrease in current GDP, which suggests that that there is still a Paradox of Thrift — or at least half a Paradox.

So why would a society ever be willing to give up $100 of current consumption to increase current net saving by just $50? Answer: if returns on incremental investment (saving = investment) were great enough to justify such.

If we assume incremental profit margins of 50%, as I did in my example, then we assume an incremental ROI of 100%! When Fazzari assumes incremental profit margins of 100% he also assumes in incremental ROI of infinity! The only way to get an interesting Paradox of Thrift is to assume an ROI so high that it is actually rational to burn up a lot of GDP in the transfer from consumption to savings. Only at Fazzari’s end of the extreme does the Paradox seem totally unsolvable.

Jan 14 2009 at 12:27am

Russ, thanks. That was a great podcast.

It failed to move me away from the neoclassical and Austrian perspective on savings and time. As Russ mentions, how can aggregate demand decrease when spending is shifted from consumption to investment? Before the shift, that spending became the income and savings of people working in consumption industries. After the shift, that spending becomes the income and savings of people working in investment industries. Where is the reduction in spending and aggregate demand? Why is investment not a part of aggregate demand? Is demand for investments any less real than demand for consumption goods?

I do not think Fazzari really answered this?

Jan 14 2009 at 5:54am

There are lots of great comments here, several of which cut to the main point of the distinction between the Keynesian and classical models–what happens in the rest of the economy? Here’s how the classical model would be explained in an intro graduate general equilibrium theory class. To do that, we need a three-agent (i.e., three-person) economy. The third agent is not the bank, but the rest of the economy. Don’t get scared–it’s easy and illuminating.

We have three agents–the restaurant-goer, the restaurateur, and, let’s say, a college.

An important thing to note is that, to cut through the complications, we are going to bypass the intermediation of the bank. We will add the banking sector back later.

In Russ’s question, there is an exogenous change in intertemporal tastes, causing someone to decide to buy fewer restaurant meals today, and save in order to do something in the future. The key point is that restaurant-goer has a plan for his savings. For example, suppose he decides to save to send his child to college.

In the absence of banks as intermediaries, he sets out in the economy and finds the college to which he wants to send his child. He approaches them with an offer to invest so that they might have enough chairs and classroom space in the future.

Now, when the restaurant-goer stops buying meals at the restaurant, he simultaneously gives his savings to the college.

The restaurateur experiences reduced income.

The college, however, experiences increased income in the identical amount. It immediately starts building more chairs or adding a professor or classroom space, etc.

The restaurateur could reduce his hours working as a cook and pick up a few extra hours at a chair factory. Or, in a world with even more people, he could fire a cook, who then goes to work at the chair factory.

Either way, the offsetting factor is that current total income in the economy is unchanged. Resources are redirected from one industry to another.

Future income, however, increases. With the new investment, there will be growth in the economy. (What happens to wages and the rate of return on capital in a general equilibrium model with yet two more agents/markets–labor and capital markets–is one step more complicated but doesn’t change our basic story. The details depend on the relative rates of return to capital and labor in chair-making versus meal-making. The ex-cook may find that he earns less in the chair factory, while the chair factory owner might receive more; or vice versa, or they may both benefit a little. Regardless, there is an increase in total future income and output.)

In summary, by by-passing the banks as intermediaries, we can see the difference between the classical and Keynesian model (as put forth by Fazzari).

In the classical model, investment increases by the same amount that consumption decreases. The college sector sees that its future enrollments will increase, and expands its investment activities at the same time as the restaurateur contracts his business.

In the Keynesian model, something intervenes in that classical process. Keynes posited that the decision to buy less today “does not necessitate a decision to have dinner or to buy a pair of boots a week hence or a year hence or to consume any specified thing at any specified date.”

In a classical world with financial intermediaries, the banker would do the restaurant-goer’s bidding. When the additional savings are deposited, it would fish around in the economy for likely investors. It might notice, if there were lots of customers doing the same thing, that colleges–who would presumably be getting more inquiries from guidance counsellors about future admissions–would be asking for more money for investments. The bank would intermediate, clearing the market and funneling the savings to the investors. Any reduced deposits by the restaurateur are offset by his increased deposits from the income he earns from the hours he now works at the chair factory. His income–the total income of the economy other than the one restaurant-goer–doesn’t fall.

In the Keynesian model, something breaks down and the economy gets stuck out of equilibrium, completely off the intertemporal production possibilities frontier. The savings never get turned into investment. Income falls instead of staying the same or rising. There has to be another piece to the story to explain that breakdown, and that is the piece that Fazzari didn’t get into. Perhaps expectations by colleges for future enrollment never increase, no matter how much time ellapses. Perhaps banks systematically can’t figure out what to do. The burden is on the Keynesian model, though, to explain why that might last so long or be so dramatic that the market never clears.

If there is a breakdown in markets so that some market doesn’t clear, say because expectations are never revised, then the Keynesian argument for kick-starting the process with government expenditure has yet another problem. Namely, throwing money at the economy may inefficiently hire people into industries that are not the ones that ultimately have to expand. Hiring people to build roads does not get the economy closer to more chairs for colleges and in fact distracts the economy’s resources and may prolong the pain.

Were I trying to resuscitate the Keynesian model at this point, I might look to the possibility that WWII, which may have been a factor in kick-starting the economy after the Great Depression, may have done so not because it was just a massive government expenditure. It may as a side-effect have ended up training people in such a wide range of job skills, and accustoming people to moving by uprooting people at a time in life right before they started families, that when the war ended citizens were enabled to make more efficient choices. That doesn’t mean that war is a good way to end a recession! It might suggest–if you believe that the economy is stuck out of long-run equilibrium with unemployed resources–that tax cuts, which widely distribute the income kick and enable people to make their own choices about the future, offer more potential benefits than the government’s trying to salvage or outguess the market economy by focusing on particular industries. (Ironically, the logic may even argue in favor of earmarks–because individual legislators are more likely than the Federal Government to have their ears to local conditions and businesses that see benefits to investments. But replacing the market mechanisms of banks and other financial markets with legislators as intermediaries brings in rent-seeking in an inefficient way.)

But that’s enough theorizing. Both the classical and Keynesian models have flaws that don’t match the scant empirical evidence we have. People fight most bitterly when there are no known answers.

Jan 14 2009 at 6:24am

This was a great conversation. You two were the same two in the “boxing match” on Planet Money, right? This was much better. It’s been a long time since I thought about all these moving parts. I was super-engaged as I did my weight-lifting last night, listening to you guys.

I do wish there were more research on how to optimize saving for an economy. You were both right, in ways, but there is an excess of saving. And there is also only so much Investment can do. There is a finite supply (and under supply) of talented people out there.

I also have to disagree that government is always non-optimal. There are a lot of things that ONLY government does that is very, very optimal. Returning to where Mr. Roberts began, the notion that Real GDP is utility, Government is just about the only entity that does large scale land preservation, for example, and that is a very much optimal choice, for us. We don’t need use nearly so much land as we do and our overuse leads to enormous externalities that no one thinks about till it’s too late (see Atlanta). Government buys up land and preserves it as wild or as park, and this does an enormous good in almost all cases.

That’s just one example. Schooling is another good one. Universities. Basic Research. Policing. Public Arts. These are few other things that are really good that the private sector really has no space for.

Jan 14 2009 at 8:26am

I guess what I don’t get is that if consumption is so great and the US has over-consumed by using house equity that never really existed why is there a problem?

Well, actually I do get it. Zimbabwe went through the same problem when the recently new farmers consumed all of the grain and didn’t save enough for seeding the fields the next year.

What I was hoping to hear is how the Trillions in stimuli would help produce and balance the trade deficits. Also how will it help prevent America from getting priced out of the markets when other countries stop lending and inflation takes off.

The theories discussed seem rational when looked at from within a bubble but I don’t see how they can possibly work when competing countries are added into the equation. I would have to think that imbalances that were caused by over-consumption in the past have to be corrected eventually. Taking in (printing) more money to cover the past losses and continuously rolling them up into bigger and bigger balls seems like a scam.

Mark K
Jan 14 2009 at 9:30am

You are one of the most coherent and intelligent posters on this message board, but your supposition that government projects ‘throw money at the problem’ is not a fair characterization. There are few who would argue for example, that the interstate Federal highway system was a project that caused inefficiencies in the market system. To the contrary, the initiation of this project by the Federal government created tremendous efficiencies in a market system that would have not have been able to progress as quickly without an efficient commerce system. I live in DC area, and the Federal project to widen and improve the traffic flow through the Wilson bridge corridor is a fantastic project that will eventually save billions in lost revenue due to avoided traffic delays. Although it may have seemed wasteful or extravagant at the time, the decision to land men on the moon created untold opportunities for technological development. Moreover, the inspiration it created for a generation of scientists, engineers, and poets produced incalculable gains to our society. I became a scientist owing to inspiration created from this voyage, but I also managed an economics degree along the way. I am not confident that any of the projects I mentioned would have gotten done without government involvement.

I think the price system is a fantastic engine for correctly allocating resources, and it should be permitted to work without much government intervention with the exception of laws and enforcement to ensure perfect competition to the highest degree possible. However, I am willing to entertain the idea that at times there are negative cascading effects within the economy that are not self-correcting and require government stimulus to get the economy back on track. That does not mean a permanent government injection of demand, but rather a temporary intervention with well defined projects. During this time, there are several potentially useful projects that can be implemented and impact the economy in a positive ways in the short and long run. On the other hand, I agree with your point, that it is likely some of these projects will misallocate resources within the market economy and produce realized opportunity costs that exceed the benefits.

Jan 14 2009 at 10:04am

Lauren, thanks. That explanation was very helpful.

Brady and Mark, I think the problem is that you are supposing the value of different things without considering the alternatives. Obviously government can invest money in all sorts of different ways, some of which generate positive returns. Its much less clear which of those ways generates the best returns (or positive returns at all) without a profit and loss system. Would consumers rather have a new public park, or the next generation of iPhones? Are moon landings worth more than advancements in standard aviation? Those questions aren’t anything any of us can answer.

As Mark says, the gains from the moon landings were “incalculable”. An ideal economy invests in whatever produces the best marginal gains; when that isn’t calculable, it has a problem.

Also, and perhaps more to the point of this podcast, a huge stimulus may easily exhaust the supply of good investment opportunities. Many of the investments that obviously generate positive returns – courts, police, roads – are already funded. The opportunities remaining aren’t as easy for the voter to judge the value of (e.g., corn ethanol subsidies). Krugman actually said something to this effect when talking about the stimulus, remarking on how the first (I think) $200 billion or so could be spent wisely, but he wasn’t sure about the rest.

Jan 14 2009 at 10:14am

Basic Fizzari’s premise is that if consumer saves $5 more, restaurant owner and employees save $5 less that leads to 0 net saving change. Which means that saving rate is a constant. You don’t need 25 years of tenure as economics professor to know that it’s simply not true. I was surprised that Russ didn’t bring it up.

Mark K
Jan 14 2009 at 10:42am

I think you’ll find that the foundation of the iPhone came from semi-conductor technology that resulted from the space program. And I do not mean ‘incalculable’ in the sense you are implying. I mean that the beneficial economic impact of the moon landings is so intertwined with the economy that it is impossible to separate its effects into the marginal utility of the consumption of product A compared to the marginal cost of producing product A. What is the marginal utility of Einstein’s Theory of Special Relativity? Apparently, it was near zero for everyone but Einstein and a few close friends in the first few years of its existence. Do you expect that the return on such an ‘investment’ is to have a value in nominal terms? It does have economic benefits over time, but it can not be calculated by any economist I know. Or would you have preferred he increased his the calculable output of patents at the patent office instead of day-dreaming about such esoteric things? Can we imagine a world without the Federal highway system? What would the costs or benefits have been to our society over the last 50 years without the highway system? That is not calculable, even though it has had significant value. Just because it is not quantifiable in nominal term does not necessarily make it a ‘problem.’

Jan 14 2009 at 11:25am


Benefits have to be quantifiable in some terms, or we end up choosing investments at random (which was, by the way, similar to the the point that Lange finally conceded to Hayek in their calculation debate). You’re of course correct in stating that the moon landings are heavily intertwined with other technologies, so we can’t say what a world without moon landings would look like. But that is an ex post observation. We (or at least I) am looking for the best process that chooses between these investment opportunities ex ante; e.g. the moon landings vs. advancements in general aviation, from the point of view of someone living in 1960. We can pick and choose different good or bad outcomes from heavy government expenditure, but what we really should be discussing is political processes vs. market or other processes.

The moon landings are, I think, a good example of an investment made for economically irrational (political) reasons that ended up having other benefits. Just because it may have had a positive outcome does not mean we should support an irrational process that may only rarely produce such an outcome. Again, I think democracies can allocate resources towards obvious needs, like courts, police and roads, much more effectively than they can towards things with esoteric benefits like trips to moons or planets.

Of course, I think its also pretty easy to argue that the Apollo program wasn’t a good use of scarce resources (though I’m not going to do that; I really have no idea).

As for your question on the federal highway system, I can imagine a world without it. I can’t easily imagine a world without highways, however thats a different thing from having a federal highway system. I admit I’m not familiar with how interstate highways could be built in the absence of the federal government, though I’m really not familiar with the process the feds go through either.

John Turner
Jan 14 2009 at 11:49am

Mark K:

The semiconductor industry did not arise due to the space program. The first transistor was developed in 1947 at AT&T Bell Labs by John Bardeen, William Shockley, and Walter Brattain. The first integrated circuit was developed in 1958 at Texas Instruments by Jack Kilby. The first silicon integrated circuit chip was developed in 1961 at Fairchild Camera by Robert Noyce. None of these achievements came about due to the space program.

The space program utilized and built upon semiconductor technology, but it was certainly not the driver of the industry, so your premise about the foundation of the iPhone is false.

Jan 14 2009 at 1:25pm

Classical economics made sense to me the first time I saw them explained. Keynesian economics never have, but this podcast was a help in understanding what intuitively seems a fatally flawed theory.

Anyway, during the wrapup I was hoping for a slightly different ending. I’m big on the scientific method. I would have liked to have heard both speakers explain the effect of several likely stimulus actions by the government by applying the knowledge they gained while drinking different flavors of Kool-Aid. Then in a few years, when GDP has remained steady or fallen and the national debt has doubled (isn’t this what happened to Japan in the ’90s?), we could at least conclude that policies based on classical economics would have been likely to yield a better outcome.

Mark K
Jan 14 2009 at 2:20pm

I’ll admit that that statement was over-reaching, but here is an excerpt from an article:

“The biggest technological legacy of the space program may be reliable electronics. In the early 1960’s, when President John F. Kennedy declared a space race with the Russians, electronics were notoriously capricious. The needs of the Apollo program, historians of technology say, speeded up the early stages of the semiconductor industry by perhaps 10 years.”

If the final assertion is true and we put the semiconductor industry back 10 years, the iPhone probably does not exist today. There are unintended consequences to government investment in science and technology, and many of these are beneficial. I would not argue that there are not opportunity costs as a result, or that at times these costs exceed the benefits. Thomas Edison saw the need for national laboratories, and his idea lead to the development of the Naval Research Laboratory, which does military and non-military research on scientific problems. Most endeavors are correctly handled by the price system, but the price system works best when mature technologies can be applied quickly to produce product to public or private consumers. Often, industry is unwilling to perform research unless it is government provided or funded, or if industry develops technology it is proprietary and unavailable to competitors for several years. The opportunity costs to society in these situations are potentially huge.

Jan 14 2009 at 4:16pm

Can anyone answer Lauren’s question or point to an answer elsewhere? What is the mechanism that stops new savings from being directed to buying investment goods? Surely Keynesians have addressed this rather fundamental objection to their theory…

So consumption goods producers see demand for their goods decrease, and reduce production. New savings goes to the bank. This increased savings results in an increase in the supply of loanable funds. This shift results in more real investment (research, production of capital goods, etc…) and a lower interest rate. At the same time, the fact that resources are freed from consumption good production means more resources available for producing investment goods, which means demand for loanable funds will increase. This results in even more real investment and pushes the interest rate back up.

Where do Keynesians see the above story going awry?

Jan 14 2009 at 7:13pm

“The consumer gives his $5 to the bank. The bank can lend out $50.

The consumer gives his $5 to the restaurant guy. The restaurant guy pays his expenses ($3), pays for his own personal living expenses and enjoyment ($1), and saves $1. The bank can lend out $10.”

These are directly contradictory statements.

Assuming the reserve rate is 10% the bank must keep 10% and it may lend out the rest. The way this plays out is that the bank gets the $5 deposit; it decides to lend out $4.50, the person who borrowed $4.50 spends it, the person who gets paid the $4.50 redeposits it back into some bank and the whole cycle starts over again with the bank deciding to lend $4.05. What you end up with is $50 in deposits(bank money), $5 in fed money held as reserves and $45 in loans. But the only way you can eventually end up with that is if the bank decides to lend the maximum amount it can and if nobody stuffs that money into their mattress.

If that money gets held up in someone’s matress the cycle is broken. In the first example you assume that this won’t happen and therefor the full amount of $50 bank money is created.

In the second example the restaurant guy gives $4 to other people to cover his expenses and for his personal enjoyment, these people are then assumed to just hold onto that money instead of redepositing it. This is inconsistent.

If you’re paying via credit card or cheque you’re paying with bank-money and that trivially leads to no change in the amount of money banks can lend.

The money in your account is really just a number denoting how much money the bank owes you; it’s bank debt. But it turns out that bank debt is a very secure asset and therefor people are willing to take bank debt as payment as if though it really was dullish green slips of paper with dead presidents on; but they’re not willing to take an IOU from your friend Steve(who owes you $5 for a beer) in lieu of fed money. If you pay with a credit card or some other means of payment which simply transfers money from one bank account to another, you’re really only changing who the banks owe money to(you can change the reserves at individual banks by transfering bank debt from bank A to bank B, forcing bank A to pay out to bank B. But aggregated over banks as a whole there’s no net effect).

Jan 14 2009 at 9:08pm

Jeff Henderson writes:
…That being said, getting a handle on this alleged paradox of thrift feels like grasping at a slippery bar of soap. Sometimes I think I see it, but then it disappears.

Jeff, I’m right there with you. It must have been Dove (with the 1/4 moisturizer). There were so many times during that Podcast where I was thinking “yeah, but…” I cannot even rememeber the way Fazzari did it know but he even contradicted his own thrift paradox somewhere in the 40-50 minute mark, IIRC. Hard to remember what was said with all the verbal smoke-and-mirrors that that guy unleashed.

The topic of “real” income was dodged by the guest. And Russ must have been so mesmorized (or polite) that he didn’t really touch on the subject of utility.

Zack Johnson
Jan 14 2009 at 9:11pm

Mr. Fazzari states that by saving money rather than spending it causes a loss income… that (I believe this was stated in the podcast) would be returned in the future when that savings was eventually spent. He argues that an increase in spending (by consumers or the govt) increases income and thus economic prosperity. What I dont remember him doing is differentiating between an increase in spending resulting from increased productivity vs increase spending based on leverage. I would argue that an increase in spending based on leverage would have the same net effect as savings. Future economic activity is sacrificed for economic activity today vs the reverse.

In addition, this ability to borrow to spur economic activity only exists because someone else decided to save their money.

To turn his example of everyone saving 100% and not spending anything to everyone spending all that we have. This would be equally destructive an economy. Savings on a personal level as well as on an economy level is necessary to act as a buffer against unforeseen future events as well as an accelerator if an opportunity presents itself.

Balance is the key.

We have an economy that has expanded on consumer spending based largely on debt. To suggest we can in a sustainable manner spend our way out of this by borrowing more money (either via the govt or personal loans) only digs the hole deeper that we as a society must eventually dig our way out of.

Am I crazy? Thoughts?

Cesar Mugnatto
Jan 14 2009 at 10:25pm

I remember reading somewhere that during recessions, the savings rate increases, decreasing again in prosperous times. So a (causal?) link exists between saving and recessions.

What surprised me in this podcast is that neither of the speakers mentioned another basic principle of economics, that being supply and demand. When savers outnumber borrowers (or at least when the money on deposit outnumbers that on loan), that money sits idly in the bank and is not used for productive ends.

When such an imbalance exists, the normal response is to lower interest rates such that it shifts the incentive to greater borrowing and less saving. However, there is a limit to how low interest rates can go (and we are practically at that limit now) and still this has not provided the necessary incentive for businesses to borrow. Clearly, other factors are at play, namely perceptions about the ability to repay loans based on today’s business climate. This caution about borrowing stems from the vicious circle created from potential job losses which cause people to spend less, but in turn, aggravate the problem for businesses.

In such a deflationary environment, printing money to induce inflation may actually be a good thing, if a stable infaltion rate of 1%-2% is the desired outcome.

I think that Mr. Fazzari has some valid points, though he may not have made them very clearly. I suppose that in essence, he argues for a “smoothing” of the natural cycles of business, and that parachute drops, or hole digging/filling projects will prevent markets from overshooting the bottom on the way down. But I’m not so sure I heard him argue that there should be a ceiling to prevent overshooting on the way up…

Jan 14 2009 at 11:07pm

Excellent podcast! Maybe my favorite ever. The step by step discussion was done so well that it had me thinking about the basic economy… what productivity really means without considering a monetary system. And I was excited to see that very point brought up later in the podcast. The idea of the real productive economy and the idea of a monetary economy and what the two mean. It’s a fascinating concept.

Certainly I side with the Keyensian explanation but I have to say Russ’s line of questioning left me wondering how productivity increases occur if not via capital investment.

Jan 15 2009 at 8:37am

Zack. I think that you are correct. The balance seems to be tilted to the side of way too much money borrowed. I think that digging around the hole to get rid of the hole is a bad move.

One thing that I would like answered by Steve is. Where should the money created go to help most? I hope the right answer is Asia or Saudi because this is where the consumable products are coming from.

And why is a recession such a bad thing? Does Keynsian Economics, via money creation and spending, actually cure the problems that recessions naturally clean up?

Jan 15 2009 at 10:48am

Loved the podcast and this one really got me thinking.

I had an idea for an experiment. When the discussion went to the implications of injecting a lot of “printed money” into the economy and measuring the results, it seemed that the hard part is the inability to keep anything constant. What if you did an experiment in a prison, created a currency, pegged to say, the cigarette and then did monetary experiments? You could even have the inmates earn this currency from making license plates or whatever it is they do. Then you would have a controlled environment upon which to do experiments.

Jan 15 2009 at 2:00pm

I really enjoyed this podcast – and what better time to chew over the the Keynesian theory. I found it particularly warming to hear at the end Mr Fazzari and Russ being quite honest about our limited understanding of what really seems to be going on.

Lauren – thank you for adding the clarity to this debate!

I would have liked the podcast to delve a little more into the role of prices and wages. In my economics course we are taught that having upward and downward flexibility in prices and wages gives the economy the room to adjust resources until we are producing at full capacity. However, most economies avoid deflationary scenarios wherever possible. The number of automatic macro stabilisers start to fall if you stick aggressively to this policy. It would have been interesting to see where this line of reasoning ends up in the classical and Keynesian views (especially if monetary policy begins to become less effective).

Pedro, your wrote:
Can anyone answer Lauren’s question or point to an answer elsewhere? What is the mechanism that stops new savings from being directed to buying investment goods?

We have a good example right now – a whole series of flaws in the financial system (over reliance on wholesale financing and securitisation, bad lending to those with high credit risks, lax regulation, inadequate rescue policies etc…the list goes on) resulting in a drastic reduction in the amount of credit available to those who need to borrow.

Jim Gatti
Jan 15 2009 at 2:25pm

It seems to me that the term “money” was thrown around much too loosely during the podcast and obscured what I understand to be the Classical adjustment mechanism. The first thing that has to be recognized is that the money supply is fixed at any point in time by the stock of bank reserves, the banks’ desired reserve ratio, and the public’s desired ratio of currency holdings per dollar of bank deposits. Only the Fed can change bank reserves. The banks’ desired reserve ratio is a function their perceived need for liquidity to handle stochastic deposit flows. The public’s demand for currency relative to deposits is a function of institutional arrangements in the payments system. Since none of these three are likely to be affected by changes in the public’s desire to save, holding all three of those variables constant should not affect the credibility of the following analysis.

Neither increasing nor decreasing the rate of spending alters the ability of banks to lend. If an individual increases the proportion of income that he saves, the question becomes one of what he does with it. If he simply attempts to build up his money balances, somebody else must necessarily have a smaller balance, and the stock of bank credit is unaffected. If all units in the economy increased their rate of saving and attempted to hold the increased saving in money balances, they would be unsuccessful in the aggregate. However, the reduction in spending would have a depressing impact on prices, and the price level would fall until the desired increase in real balances was achieved. This could arguably take quite a long time, and the appropriate response of the Fed should be to increase bank reserves to accommodate the relative increase in the demand for money. It is this increase in bank reserves that leads to the increase in real loanable funds that can be used to employ the resources idled by the increase in saving.

If individuals purchase any other financial assets, the stock of bank credit still does not change. Instead the immediate impact is to increase the supply of loanable funds, driving up the price of financial assets, and lowering their required rate of return. This, in turn, lowers the cost of capital for firms, leading to an increase in investment spending to employ the idled resources.

The problem is in the adjustment process. As Peter Boettke pointed out, it takes time for resources to shift from one application to another and there is frequently a dead weight loss associated with a shift in demand if the resources are less than mobile between applications. That loss is made up of lost output due to unemployment and the costs of transforming physical and intellectual capital so that they may be used in other areas.

I do not see that the adjustment process is any less difficult if the federal government tries to spend a trillion or so dollars regardless of how the funds were raised. You don’t deploy the funds by purchasing goods and services from those firms that have suffered a decrease in demand. You purchase those goods and services required to produce whatever government output is felt to most desirable at the margin (or at least I hope we would). These firms may or may not have idle resources. If they do not, we still face the same problem that Boettke raised and that the Keynesians seem to ignore.

Andy Phillips
Jan 15 2009 at 8:26pm

A quite disappointing performance by Roberts, I’m afraid. No matter how many times Fazzari asserted the Paradox of Thrift Roberts didn’t seem to grasp that Fazzari really was asserting that when the consumer in the restaurant example substitutes unpaid labor for paid and as a result leaves more money in his checking account (the main effect — he is also substituting his food purchases for the restaurant’s, his trips to the grocery store for the restaurant’s deliveries, fueling his stove for the restaurant’s gas purchases, etc.) that there was no net increase in savings. And Roberts kept beginning his responses with locutions like “Well, won’t the increase in savings…” Totally missing your opponents points is no way to effectively refute them… And Roberts never does. Which is a shame, because Fazzari is speaking nonsense. God, I was so missing Friedman, with his unerring clarity in connecting economics with the world of molecules.

I noticed someone further up this thread who caught the basic point that Fazzari was assuming his conclusion by holding the restaurant owner’s expenditures constant. And indeed the restaurant owner is unlikely to notce the loss of one customer. But the probability of that loss being the one at the margin that triggers results is not zero. And Keynesianism is macro-, not microeconomics. The real question is what will happen if a million consumers decide to reduce their expenditures on paid labor, making that money available to them for other expenditures or savings.

Suppose a restaurant requires 1000 customers a month in order to stay in business. Suppose a quarter million five-times-a-month restaurant customers cut back to once a month. The restaurant owners expenditures will NOT remain constant. After some transition there will be approximately 10,000 less restaurants than otherwise. The resources consumed by those 10,000 restaurants will be on the market, prices for those resources will adjust, and (mostly) they will be consumed at a lower price than otherwise, which is a net benefit to the economy. Which must be the expected effect since a previously unused resource (the labor to self-prepare food) has been injected. Someone will get a bigger house or factory for the same price because the labor of the fomer busboy was diverted to construction from waiting tables. Etc., etc.

Jan 16 2009 at 10:20am

Sorry to be late to the party, but wouldn’t money velocity be a critical component of the picture? I also was a bit troubled by the restaurant analogy, but the discussion seems to presume money velocity is held constant. The following link from the St. Louis Fed suggests money velocity in fact is slowing dramatically as a result of the present recession. ( The relevance of money velocity would be that in the event of a loss of demand, injection of Federal funds to stimulate demand could in fact increase funds available for private investment if it succeeds in increasing velocity. Conversely, if money velocity declines, an increase in savings would not necessarily result in significantly increased investment. I’m too new to economics to presume this is a definite conclusion, but I would be interested in what others think about money velocity’s role (if any).

Jan 16 2009 at 10:21am

Appreciate your shows, not matter the content. Thank you!

Scott Wood
Jan 16 2009 at 11:43am

I would like to have asked the Prof. how the savings rate in an economy can ever increase. It sounds to me like, in his world, it is a logical impossibility.

Jan 16 2009 at 3:48pm

This was very fun podcast to listen to. For me, the troubling thing about the Keynesian view is that relies so heavily on a global amortized analysis and ignores component parts which often turn out to be systemic! Take the stimulus checks from last Spring. One claim is that it wasn’t as successful as anticipated because too much debt was paid off with the checks in lieu of new spending. The proponents didn’t account for local choices.

To paraphrase the great Ted Stevens, the economy, according to Keynesians, is a series of tubes. You hook one of the tubes up to a fire hydrant (like a stimulus package or currency printer) and water spills where it’s going to spill and eventually floods the tubes and compartments up to some level dictated by the amount of water. The Hayekian approach is to think of things bottom up, node by node, person by person, transaction by transaction. Yes, there are some emergent systemic phenomena that we can observe and put numbers on, and some policies we can suggest to achieve some widespread outcomes. But we realize that an economy is built and maintained bottom up, not spilt and washed top down.

Despite Russ’s pessimism about the field of economics to be useful over the coming years, there is one giant human artifact that offers hope that the field can eventually get it right. The Internet. Model human economic behavior like the Internet, and you’ll have a much better model. Fortunately, the economy itself is modeling more of itself on the Internet: B2C stores, B2B services, online tax reporting and collection, etc.

Jan 16 2009 at 3:49pm

Lauren – Excellent exposition; if I had a trophy for best answer, yours of the 14th would win. Now I won’t lie awake at night wondering about the Paradox of Thrift.

A few observations:

Austrian economists know that a solid theory of economics must be based on an understanding of how individuals make choices. Austrians have enormous respect for the power of individuals. Keynesian economists, on the other hand, focus on macro quantities like Total Demand, and are far more likely to deprecate the knowledge of individuals. In this podcast, how many times did Professor Fazarri say how “subtle” or “hard to understand” some particular point was? An Austrian might say that the Paradox of Thrift, for instance, is hard to understand because intuitively we know it’s wrong to think that there is something inherently destructive about saving more. A Keynesian relishes complexity because it justifies leaving economics to professional elites who want to make the world over to suit their ideas of how things should be.

Who but an elitist would say in one breath that we need a fiscal stimulus of a trillion dollars and in the next say that we really don’t have enough good cases studies to know whether such an effort will even work? I’m with Lauren that we do know there is a big risk that government will direct spending toward less worthwhile activities than private markets would. Some respondents have pointed out that government does do some things that the private sector finds hard or impossible to do. Fair enough. But I’ve lost count of the number of Keynesians I’ve heard say that it doesn’t really matter how the government spends money as long as it gets spending going. Does any sane person think it’s sound to pay one group of people to dig holes and another group to fill them up? Maybe this is another “subtle” point that the good Prof wouldn’t expect me to understand.

I’d also note that, contrary to the idea that we don’t have enough evidence to say whether fiscal stimulus has much of an impact, there is plenty of evidence that it does not. Milton Friedman wrote often on the subject and won the Nobel for his work on the permanent income hypothesis, which helps explain why fiscal stimulus is not very effective. Russ mentioned that the big tax rebate last spring did nothing. Observe, too, that the stock market – a leading indicator – is not exactly celebrating the new trillion dollar stimulus plan.

Why not take a page from the Austrians and examine why banks and individuals are so cautious now? Obvious reasons are uncertainty over the location and ultimate value of troubled financial assets and dismay over the monetary and fiscal policies that no longer seem to follow any rules.

If we really want to solve the current mess, why not adopt the prescritpion of Holman Jenkins, columnist for the Wall Street Journal? He advocates simply bulldozing the stock of excess houses. That would put a floor under home prices, support the value of securities tied to mortgages, increase confidence, and spur spending. Supply and demand is pretty simple, isn’t it? The only drawback is that Congress can’t siphon off as much tax money as they can under all the other plans out there.

Jan 16 2009 at 5:35pm

Great podcast!

Is the “paradox of thrift” a contention that the expected future velocity of 100 extra saved dollars is less than the expected future velocity of 100 extra spent dollars?

By “velocity of 100 dollars” I mean how quickly it changes hands as part of exchanges that GDP measures.

Of course the 100 dollars will get split up and go to many places. And you have to pull some typical accounting tricks to track the path of “this particular 100 dollars.” But I think my question makes sense. So is that what the paradox of thrift is? A statement about maximizing the velocity of money?

Andy Phillips
Jan 17 2009 at 9:23am

mk — No. The alleged “Paradox of Thrift” doesn’t say spending is superior to saving. It says it makes no difference whether you spend or save. If you save more everyone else must* save exactly the same amount less, so saving doesn’t increase total savings. (*”must” if you adopt what Fazzari calls the “simplest” assumption, that — quoting Roberts’ notes — you “Suppose they keep doing everything same as before, same employees, etc. Spending of restaurant is same as before…” Possible at the micro level, insane assumption for the projected result of the corresponding macro phenomenon.)

Baruch — Are you joking? There is no excess of houses, merely some houses not worth the debt attached to them.

Jan 17 2009 at 3:56pm

A Keynesian view of economics fails in theory and practice. I would have like to have seen this better exposed in the otherwise fascinating podcast.

The very fact that one can speak hypothetically of paying one group to dig holes and another to fill them in – and that this results in a positive measure of economic activity under the theory – itself proves that the theory must be fallacious, however “subtly” so.

Regarding practice and the dearth of empirical evidence to support Keynesian interventions: On the contrary; the entire twentieth century reads like an uncontrolled experiment on the failure of economic central-planning and such interventions.

Jayson Virissimo
Jan 18 2009 at 4:33am

This is in my top five EconTalk episodes. I hope in the future you can get Brad Delong or Paul Krugman to come on the show and have an equally respectful discussion.

Jan 18 2009 at 9:27am

Let me make sure I got that right: Hiring people to produce food that nobody wants (what the restaurant owner in the example does) leads to a destruction of income, but hiring people to dig holes that nobody wants has a positive effect.

Jan 18 2009 at 2:15pm

I haven’t studied economics, but I enjoy the EconTalk podcast.

I heard the following on the Fazzari podcast: Keynesian economics says that an increase in (aggregate) savings means a corresponding reduction in (aggregate) income.

OK, fine. But there seems to be an unstated assumption that more income is inherently good, and less income is bad.

So then isn’t all savings bad? Shouldn’t we be spending as much as possible at all times? I find that hard to accept.

When I choose to save now, I’m doing so with the intention of spending later (perhaps much later). That means more income in the future, right? Why is income today better than income tomorrow?

I didn’t find Prof. Fazzari’s argument convincing at all. Something has to be missing here. Can we have another podcast on Keynesian economics and give it another go?

Jan 18 2009 at 8:17pm

I’d love to see this become a written debate so that both sides can carefully construct their arguments and questions.

Jan 18 2009 at 10:59pm


Please stop going on tangents. I love your show, but you fill in a lot of details when quick statements are usually enough. I understand that many of these complex examples require elaboration, but err on the side of Charlie Rose.

Just a comment, LOVE YOUR SHOW!

Andy Phillips
Jan 19 2009 at 1:17am

Julika — No, hiring people to produce food that nobody wants destroys savings, not income. Just like hiring them to dig holes.

The difference between Fazzari and Xeno is that Xeno knew that the racer would pass the tortoise, and that there was therefor something wrong with the argument he was making. Fazzari really seems to think that saving doesn’t increase savings.

Kristoffer Buus
Jan 19 2009 at 5:12pm

I have been eating my way through all the comments now, it’s fantastic 🙂

Jan 19 2009 at 8:15pm

Haven’t Keynes’s moronic, discredited economic theories done enough damage to the global economy?

Savings are essential for economic development in order to increase production in the future to create wealth, while simply spending on consumer goods cannot create wealth.

The manipulation of interest rates done by central banks along with the fractional reserve banking system is the cause of the business cycle.

The Austrian School of Economics has the only plausible explanation of the business cycle.

Why won’t these Keynesian theories die already!!!

Jan 19 2009 at 11:19pm

I believe one point has been overlooked by everyone. (I skimmed some posts so I may have missed it.) Its this: one response of the restaurant owner to less business is to innovate and get more out of his labor and capital. He might end up with a new chain offering wholesome food at unprecedented, low prices, and save more than before!

Professor Fazzari was clear and appealingly modest, but his argument is frustratingly circular: it implies that aggregate saving and consumption has to be static. If I save more, you have to save less, etc. I’d like to know how some countries can have a 35% savings rate and 10% growth, but if we (the U.S.) save more than 0%, the economy collapses.

It’s so easy to veer awry on such things (thanks Lauren, great phrase).

Jan 20 2009 at 5:01am

Great Podcast. Loved the give and take. Really prompted some thinking.

Lauren get’s kudos from me too.

Pedro on the 14th wrote:
Can anyone answer Lauren’s question or point to an answer elsewhere? What is the mechanism that stops new savings from being directed to buying investment goods? Surely Keynesians have addressed this rather fundamental objection to their theory…

I am going to take a stab at this myself.
In Lauren’s example the college buys chairs and simple goods that the restauranteur can easily pick up the skills for. In the real world this transition produces a large sticking point. The college is more likely to forgo the chairs in place of demanding a new computer system or engineering apparatus (something important to educating the restaurant-goer’s child and very hard to produce).

The restauranteur does not have the skill set to create this product, nor does he or anyone else know how to train him to produce these future goods. So the college holds the cash, waiting for a future product. Thus no investment.

It is hard to parse but ‘Felipe form Brazil’ gave a similar answer to this on the 13th. The market is being incentivized to ‘evolve’ by producing new products. And the college is willingly putting cash under the mattress in order to incentivize the market to produce a computer. The ‘mechanism that stops new savings from being directed to buying investment goods’ is the 1. college’s desire to wait for a computer but even more so 2. the restauranteur’s inability to learn the skill sets needed to produce a computer.

In effect government spending lowers the bar on the skill sets demanded by the marketplace. The college is stripped of its ability to incentivize the market (cash) to produce better skilled workers and thus greater products. This is done via taxes and the result is that the government can employ skills the market can provide.

“I lost my keys over there, but we are looking over here because this is where the light is good…”

Jan 20 2009 at 3:55pm

Casey Mulligan points out on his blog ( that (1) indeed, as we’re constantly told, Personal Conumption expenditure (PCE) is 2/3 of GDP; (2) PCE has indeed been falling since June; yet (3) total spending as of the end of November is UP because of net exports. In short, the Fazzari discussion/interview and theoretical analysis is a purely academic exercise unless we take account of the global context, made more complicated by the widely differing, and presumably shifting, country, regional, and global saving, consumer spending, and investment rates.

Jan 21 2009 at 9:11am


I found two holes (from a micro perspective) in Prof. Fazzari’s restaurant example that you failed to bring up. Perhaps I’m missing something, but here’s what I would have countered with:

* When the family decides to decrease their consumption of restaurant food, they have to increase either (a) their consumption of products at their grocery store of choice; or (b) their consumption of products whereby they can grow their own food. They will purchase at least some amount of a substitute…it’s not a 100% “loss” to the economy per Prof. Fazzari.

* And when the family deposits their savings into the bank, it’s not the same as the restaurant depositing their daily receipts. The restaurant has a regular flow of bills to pay, so they will most likely maintain a checking account with little-to-no interest rate…the bank can’t easily lend out that money long term because the restaurant may need it at a moment’s notice. But the family, who is saving for a college education 10 years ahead, will most likely deposit their savings either in a savings account or a long-term CD (as savings accrue over time)…the bank can indeed lend this money out with a longer-time horizon. Not all deposits are the same, and neither are the economic effects of different deposits.

In summary, it seems to me that Keynesians miss the so-called “second round effects”….or to paraphrase Prof. Sowell, to answer the question “now what?”

Interesting podcast, though!

Sanjay John Gandhi
Jan 22 2009 at 12:04am

There is no such thing as a paradox of thrift. I believe Professor Fazarri (and many other economists) dont get this right-something which took me a while to understand after reading Adam Smith’s “The Wealth of Nations”. The key element is productive and unproductive use of capital/savings.

Let’s take the family-restaurant example Professor Fazarri talks about. Our family has a noble profession-they own a bakery. They make bread, and that’s all they do. They make a lot more of it than they can consume-so they will exchange the surplus bread for whatever-clothes, eating out, vacations, etc. This obviates the necessity of invoking money in our discussions-we will deal with the excess bread only.

In case the family eats out at the restaurant-it is an unproductive use of this excess bread. The excess bread is exchanged for a “nice experience” etc. but tomorrow they are left with less bread and that “nice experience” feeling is gone.

Now instead of eating out at the restaurant the family exchanges the excess bread for A TABLE WHICH THEY WILL USE TO SELL MORE BREAD TOMORROW IN THE LOCAL FLEA MARKET. That is a productive use of capital. They spend the excess bread on this wooden table from the carpenter.

Tomorrow the flea market opens-and Alice from the family is there with the table and the excess bread to SELL MORE BREAD. They sell more bread, and therefore they can make more bread. That increases output of the bakery-they have found a new market for their bread in the flea market-and thanks to the table they bought, they can sell more bread for a couple of years to come.

The table making carpenter benefits. He gets this extra bread, and he is also a thrifty, intelligent man–he doesn’t go out and eat in a restaurant with this additional bread. He goes and buys another small saw with this extra bread, because he knows he can make more tables per week if he has this small saw. Now the saw is a productive use of his capital-he is able to make more tables.

The same is extended for the saw maker, etc. The saw maker doesn’t go to the restaurant either-he goes and buys more steel to make another saw.

No one wants to go to the restaurant-an expense which every man with common sense knows will hardly make them more productive. All they want is to produce more bread, tables, saws, steel etc, etc.

The output of the society has gone up because of this wise use of capital by everyone involved.

You can look at restaurants, vacations, etc. as consumptive expenses, which DO NOT increase the output of the society. They decrease it. Others like putting your excess capital to increase your produce is a productive use of capital-it increases output of the society. Consumptive expenses are unproductive uses of capital-and expenses to increase our production are productive uses of capital or savings.

It follows that a very rich society is a society which consumes nothing, which just spends all their excess output in increasing the output even more. The price of everything is so low-everything is available so cheap-a true mark of prosperity of a society.

You can see this also in my blog at a related post:

Other unproductive uses of capital are Government spending-I will cover this in a post in the near future.

Hope that helps in clarify matters and resolving the “thrift paradox”.


Jan 22 2009 at 12:13am

The Paradox is false because it fails to take into account time.

Originally, it went this way:
On monday the family cashed a paycheck in hopes of eating out later.
On tuesday the family ate out and gave the money to the restaurant owner.
On wednesday the restuarant owner paid his staff, etc.

Now, the family decides to save:
On monday the family deposits the paycheck into the bank for saving.
On tuesday, the restaurant owner is destraught to see his restaurant empty.
On wednesday, to pay his staff, the restaurant owner withdraws money from his bank to pay his staff.

Now, it is true that as of wednesday there is exactly the same quantity of cash in the bank: the family has more and the restaurant owner has less. However, on monday and tuesday there was more cash in the bank than there otherwise would have been. The bank should have responded by finding some way to productively loan the cash out. It would be unlikely, but maybe another customer responded to new lower interest rates to take out a personal loan on tuesday and ate at a local restaurant on wednesday, enabling the restaurant owner to avoid taking his own money out of the bank.

Sanjay John Gandhi
Jan 22 2009 at 9:38am

You can say that all leisure expenses are not going to increase prouction/make u more efficient in your job—and any common person on the street knows that well. They know that taking a vacation won’t help them produce more of whatever they produce at work (software, bread). They know that watching a movie and spending $10 there won’t help them produce more either.

There are small expenses which may be called useful/productive expenses for someone, even though for someone else, who is very poor, they may be called leisure expenses.

I am an Engineer and write software for a company. However, my shirt has holes in it now-it is used an worn out. That is all well; noone in the cool place I work cares if I come with a shirt with wholes-but it makes me feel cold at work! To keep warmer, and to increase output–I will go to Walmart and buy another shirt for $5.
That is a productive use of capital-even though clothes in general are leisure expenses. If I has a really tough engineer and never felt cold at work-I would go to work absolutely naked and write the same quantity of software. The money spent in clothing myself I would use to buy a better mouse to increase my productivity (I have a mouse which I bought used on Ebay 10 years ago—it jumps sometimes-and another mouse would increase my software writing speed).

The best test for an expense being productive or unproductive is done by the person themselves. They know that going to an expensive vacation in the Bahamas is purely for their “feel good”—-and that expense can never be justified as a productive expense, even if they are a billionaire. All entertainment expenses fall in this category. By keeping people employed in the entertainment industry-I actually do the world a disservice. It would be better for me to give my excess produce to a shoemaker, who being a parsimonious person like myself, wouldn’t spend it on a vacation; he would just buy more equipment to make more shoes-or fix the leak on the roof of his house because that really is a NECESSARY expense so he can sleep better and get to work tomorrow well relaxed, ready to produce shoes efficiently.


Dmitry M
Jan 22 2009 at 1:05pm

I have to agree with Paul (Posted January 13, 2009 6:29 PM). It’s fascinating to me how bright minded people like professor Fazzari can subscribe to Keynesian General Theory, unless there are some compelling economic advantages to the subscriber in doing so.
And for those “on the fence” about the logic behind The Paradox of Thrift I suggest F. Hayek’s Paradox of Savings from 1929 where he argued that, when the savings were invested, then production, consumption and aggregate profits would all increase in complementary movements that would bring the entire economy to a new, profitable equilibrium at a higher level of supply and of demand.

Jan 22 2009 at 8:01pm

I can follow Farrazi that when the family spends less on eating out the net effect on economy is not as much as people might think BUT

1) The argument does not consider that in many cases consumers spend on imported goods so a reduction in spending will reduce the retailer income but at the same time reduce the trade deficit with china for example.

2) If I am the family and the net income of my saving is 0 on the economy it is definitely not 0 on me !! So why should I spend the 5 dollars now and eliminate my choice in the future of whether to get another meal or send my kid to college.

I thought this was a greatly educational expose on Keynesianism, I would love to hear more of this person explaining further.

In short retort to the problem of thrift, I would say the consumer’s choice to save is paramount because he is what decides is competitive in the economy.

Give us more Keynes debates/info EconTalk!!!

and God bless freedom

Jan 24 2009 at 5:32pm

This whole discussion of saving vs. consumption by private citizens neglects the reality that no one has been saving money!

What needs to happen is for Americans to rebuild their personal balance sheets and for fast growing international markets to start consuming more.

We’re out of money – the government is amassing debt at a ferocious pace…if the only solution is to build bridges and install solar panels on roofs, all we’re going to do is lengthen people’s commutes and aggravate pigeons.

The whole notion of letting the government save us is mostly bankrupt. It’s so sad that we all work the first 5 months of the year for an institution which has the perverse incentive to spend more every year whether they need to or not.

And regarding the paradox of thrift – it’s a vast oversimplification and a tool for the government to rationalize ever greater spending. What we do know is that buying a burger at Friday’s won’t increase long run economic productivity; which we need for growth. Investing in some capital market however, does elevate that possibility of long run productivity gains –> growth. Too bad most people who invest are really just enriching the mutual fund companies.

Good show though – I’d like to see one on the economics of obesity. You can dwell on the french fry paradox and then debate whether to consume 1 burger now, or 2 in an hour!

Jan 26 2009 at 1:46am

Prof. Fazzari’s answer is only partial. OK, so aggregate demand falls and that causes a slump. But can someone explain why agregate demand falls in the first place? Why does aggregate demand fluctuates?

Short of giving a compelling answer to the ‘why’ question you have done nothing more that describing the proximate cause of slumps whitout touching on the root cause.

Can someone explain what the root cause is according to Keynes?

Jan 26 2009 at 8:38pm

So, if I understand the Keynesian perspective correctly, recessions are caused by a lack of demand. In order to correct the lack of demand the Keynesian remedy is to have government replace that demand by spending lots and lots (and lots and lots and lots) of money, thus replacing the lost consumer demand with government demand.

So take the current stimulus as an example. Consumers are spending as much as they used to which has led to a recession. The Obama economic team wants to ‘fix’ this by having the government spend close to a trillion dollars, thus replacing the nascent consumer demand, thereby stimulating the economy.

Now, if the stimulus was implemented in a best case scenario with all of the money spent in 2009 on projects that are ‘necessary’ – what happens in 2010? Doesn’t this new lack of demand put us exactly where we started? How is this a long term fix. It seems to me that Keynesian stimulus can be nothing but short term in effect.

Am I missing something?

Jan 28 2009 at 4:31pm

Chris, to answer your question “Am I missing something”. Yes…

A finer point would revolve around you reading a bit more on Keynesian Theory and further understanding Obama’s “fix”.

Once this is accomplished, I suggest your example of Obamba and the cause for recession, is not a good or well reasoned one to use.


Sheldon Richman
Jan 29 2009 at 2:52pm

I don’t know why Prof. Fazzari sounds so triumphant when he says that saving instead of spending on restaurants doesn’t increase aggregate income because the restaurant folks lose income. Who said otherwise for the very short run? The point is that the saving will shift resources to earlier stages of production (R&D for example), absorbing the laid-off restaurant workers. So after a brief period the immediate loss of income disappears and in the longer run consumption and real income increase because the structure of production has lengthened, providing more, better, and cheaper goods. If Fazzari is faithfully representing Keynesianism, then what I used to think were parodies of that theory were in fact accurate renditions. Thanks, Russ.

Jan 31 2009 at 11:38pm

It’s a simple matter of Fazzari and Keynesians in general treating all savings as dead money.

Savings under the mattress would be dead money, but not savings in a bank.

It really is as simple as that. There are of course many more details both simple and complex, but it boils down to the Keynesian view is correct only if all savings are truly dead money.

Feb 1 2009 at 12:30am

“Can anyone answer Lauren’s question or point to an answer elsewhere? What is the mechanism that stops new savings from being directed to buying investment goods? Surely Keynesians have addressed this rather fundamental objection to their theory…”

I think this is where animal spirits comes in… that and hoarding. Just as people can become irrationally exuberant about the future, their animal spirits can take the opposite course and become irrationally negative.

When the latter happens, savings in the form of money are hoarded rather than being invested. Maybe underneath mattresses. Or maybe banks build up reserves, refusing to lend. The result, anyways, is savings without investment.

That being said, I’m no Keynsian. Perhaps Prof Fazzari can come back for another episode to clarify.

Feb 3 2009 at 4:26pm

Let’s see how this works with heating oil instead of a meal. I choose to skip heating my house for a time, using 50 gal. less, saving $100. It goes into the bank. The oil retailer sees his revenue (not his net income) reduced by $100 (his net income reduction is very much less). For the rest of the chain we can follow the guest’s logic. But, at the end of the day, isn’t there an extra 50 gal. of oil around somewhere, and isn’t this really what was “saved”? (Of course, thinking in terms of larger volumes, reduced wear and tear on the oil truck, tires, brakes, roads, etc. and many other things are also “saved”.) And, these savings are real unlike the “money” and are available for use in the future. In fact, if Keynesians believe as the guest does, how is “capital” ever accumulated in the first place, if saving by one person reduces saving equally by someone else?

Kevin L.
Feb 3 2009 at 11:54pm

I found the discussion in this podcast totally divorced from reality. Fazzari repetitively claimed that net savings (in dollars) cannot increase. The same argument implies it cannot decrease — the recipients of an increase in spending save some and spend the rest, part of which gets spent and saved etc. until all the extra spending was saved. That is the reason that the (in)famous Keynesian multiplier is finite. (If the average saving rate is 10%, the multiplier, which gives the total $$ increase in spending from $1 of extra spending, would be the inverse of that saving rate, i.e. 10. Now, that conclusion, that the net saving rate for an economy cannot change is completely counter to the historical evidence — up until recently, the net US saving rate had dropped, to even below zero. That is the basis for the debt bubble.

I would like to point out some trivial errors with the analysis given the podcast. The economy is not a closed system. The rising debt of the US was an inescapable result of the net trade deficits. What do the Chinese do with the dollars they are accumulating — they buy US bonds (government and private); trillions of $$ worth in the last decade. If the family saves more by reducing their purchases of imported goods, it is the exporting countries and their banks that have the drop to offset the increase in the US bank.

Another, perhaps deeper flaw, is they ignore the rate at which we spend. In the cartoon economy they describe, they implicitly assume that everyone spends or saves their income at a constant rate. If we get paid once a month, the total demand per month would be the total amount of money in circulation. If we got paid every week, the total spending per week would be the money supply, i.e a rate 4 times higher. This is clearly nonsense in a normal economy but this is exactly how hyperinflation can be sustained. People send their money faster and faster which leads to every greater spending on the same goods and this creates hyperinflation. (The secondary effects are to reduce production, making inflation that much worse).

In the model given in the podcast, there is no real difference between a bank and a restaurant, the loans given out by the bank are (in terms of net demand) the same as the spending by the owner and employees of the restaurant. No sane person would claim that I would reduce net demand in the economy by deciding to go to Burger King instead of McDonalds, but as they discussed it, that is no different than the claim of Fazzari that increased savings by an individual reduces net demand but not increases savings. However, if one recognizes that it takes the bank far longer to “recycle” its deposits into loans than it takes the restaurant (which meets payroll every week), the drop in overall rate of spending by savings (with a fixed money supply) is clear.

Another factor is fractional reserve banking, which some of the earlier posters misunderstand. If the bank has a 10% reserve rate, that means that when I put $110 in the bank, they can only lend out $100, and must hold $10 as “cash” to meet the fluctuating and unpredictable rate of withdrawals. The bank cannot loan out ten times my deposit as stated above unless it could somehow insure that every time the loaned funds are spent (and re-spent) it goes completely from one checking account to another in the same bank.

I find it self serving how economist like to continuously claim how subtle their field is. Economics is not Rocket Science (which in turn is far less subtle than String Theory).

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. What is the Keynesian theory of the business cycle? Why in that view do capitalist economies slump and boom? Direct determinant of output, investment, employment, is amount of money being spent, or aggregate demand. The more people spend, the more people are employed and the more output is produced. Direct cause of current recession is reduction in amount of money people are spending on new houses; fewer workers in the housing market; people borrowing against their houses is reduced, so if people not using their houses as piggy banks any more, businesses won't sell as much and won't hire as many workers. Is there a distinction between short run and long run? Around this time of year in past years, people would argue that we need a good holiday to keep the economy going. Two thoughts, non-Keynesian. If we all decided to work half as much, more time with family or with old people, isn't that good? No inherent value to consumer spending. Second thought: if people don't consume, won't they save and invest? Sympathy to idea of no inherent value to consumption, being less materialistic. Real GDP would be lower, satisfaction might be higher. Coordinated policy: hard to reach that in an uncoordinated economy. If some individuals did that, they may be better off but businesses that had been selling to them would have some problems. How do they get out of those problems? What kind of adjustment would the economy go through? Short run vs. long run. In practice that would not be a coordinated decision. Current economic situation was not a decision to take more leisure, but because we were financially overstretched, unhappy situation with people who want to work being unable to.
6:42Second part of fable: If people spent less and saved more, is that harmful for the economy? Distinction between consumption and savings. Savings and investment are the engine of future growth. Why would savings be bad? Subtle issue in Keynesian macroeconomics. Paper link. Think about family eating out in restaurant. Decides to not eat out so much, specifically deciding to save more--want to go to college, buy a car, take a vacation later. Chapter 16 of Keynes's General Theory quote:
An act of individual saving means--so to speak--a decision not to have dinner to-day. But it does not necessitate a decision to have dinner or to buy a pair of boots a week hence or a year hence or to consume any specified thing at any specified date. Thus it depresses the business of preparing to-day's dinner without stimulating the business of making ready for some future act of consumption.
A little abstract. Family decides to save more and not eat out as much. Restaurant is not selling as many meals; their income has fallen. Forces reduction in income. How does restaurant adjust? Suppose they keep doing everything same as before, same employees, etc. Spending of restaurant is same as before, so the restaurant's saving has to go down. The decision not to spend destroyed income. Decision to save by one group in the economy resulted in less saving by another group in the economy. Cancels out. The paradox of thrift. Fairly standard perspective: behind lower aggregate demand could lead to lower output. When individuals decide to save more, the economy in aggregate doesn't save more. But this happens all the time in the economy. Staying closer to the hearth, healthy food, etc. Didn't cause a massive dislocation of the economy, just readjustment. Some restaurants went out of business, grocery stores started selling more healthy food. Happens all the time; prices steer the economy. Gives people more of what they want. Paradox of thrift--is that really what people are thinking about? The reason it's so hard to understand is that when we individually save more, we see our bank accounts rise, but what we do not see is the income we've destroyed in the economy by not spending. Some people might be spending more and borrowing, demand moving in and out of the restaurant sector. Constant flux and adjustment, overall necessary shift. But from time to time, like now, systematic shifts in spending patterns. Reduction in consumption more broadly now because people are financially strapped. Not just reallocating spending from one sector to another. What kind of market adjustments can occur in that context?
15:27Saving more destroying income: if we save more and there is a larger pool of money for future consumption, isn't that a source of risk-taking and investment? Isn't that going to produce more in the future? If the premise were true, yes, but premise is not true. When we save more there really isn't more to lend out, exactly offset. Pool of saving is not getting somewhere. Challenge of macroeconomics: thinking clearly about a complicated subject. A lot going at a point in time, and a lot going on over time. To get a handle on it, we have to hold a bunch of stuff constant and change one thing. Errors come from neglecting something that has to change. So far, exogenous change in savings--is often caused by something else. Story: change tax policy in a way that makes consumption less rewarding, encourages a shift away from consumption today toward consumption tomorrow. Wouldn't that yield a higher savings rate today, perhaps a lower output rate today? Yes, would encourage savings, say new IRA initiative, consumption tax rather than income tax with deducting saving. Would cause a short-term Keynesian problem, same as family at restaurant. How do you get from the short-term result to the long run result, with more saving? Very subtle issues, core of debate in macroeconomics about wage and price adjustments. Higher saving in the short term destroys income; how does economy adjust? Standard story: unemployed resources, particularly labor. If excess supply, wages are brought down. Unemployment; still insufficient jobs. Wages are lower, so some businesses will cut prices, various subtle effects, goes on until resources are employed. Continue until demand is stimulated by the lower prices. That period of time is the movement between the short run and the long run. Conventional view
23:13Tend to think about quantities and real resources--people, factories, etc. If we decide to allocate a larger proportion toward consumption tomorrow rather than today, won't that keep aggregate demand unchanged? True that restaurants will cut back, but won't construction firm expand? Holding constant here is the confidence and expectations people have about the future--today, trust and optimism isn't there. Setting that anxiety aside, more resources available for other industries when the restaurants cut back. New software designs, etc. What can we hold constant? Standard fallacy is to think we can hold income constant. Say, $14 trillion. Some goes to consumption, some to business investment, some government activities. When people save more, though, we cannot keep income constant. Only way to fill that gap is to have some other process that will restore demand. Paradox of thrift: logical fallacy that you can raise saving and hold income constant. When the families decided to eat out less and save more, the bank won't sit on the money; but the restaurant owner has less income and has to draw down savings deposits. No net flow into the banking system. But no decrease in total income then. There is. What would happen if people stopped spending. Income would fall to zero. There would be no income. Implicit in every mainstream textbook. If we all stop consuming, we take all of our income and try to invest it. Not going to save it by putting it in the mattress; we will try to invest it, maybe through bank as intermediary or us directly. If we all decide to work more for tomorrow and less for today, if we plant the seeds instead of eating the fruit, won't there be more tomorrow? Keynesian macroeconomics distinction: monetary economy is very important to this story. Boettke podcast: one side of macroeconomics says all the monetary side is a veil; another side that says the monetary and real interact. No spending, no income is a monetary kind of system. Robinson Crusoe kind of economy: eats bananas versus saving and planting some seeds. In that case it instantaneously leads to investment. Without foregone consumption there is no growth. Different from monetary system. Market system first sees that people are not going out to eat that often but won't see the decision to send kids to college later. Why won't bank lend the money to the college? The bank never got the net money because somebody else's bank has less. If we have a change in public policy that lowers the ratio of consumption to income, such a change is possible and could lead to more growth and consumption in the future? Only if there is some process that restores aggregate demand and reallocate the resources. In the Keynesian perspective, it's not automatic. Neoclassical growth theory assumes the wage and price changes will work their way through. Radical Keynesians do not consider that process to be automatic. Stickiness of wages and prices--in the mainstream thinking or in the old days. Changed over time. Can't really wait around for wages and prices to adjust; plus there are risks like deflation that might be counterproductive. Monetary policy should step in. Taylor rule, inflation targeting, Fed could help the process along. But monetary policy might not be enough.
36:43Given the dynamism of the market economy in good times, most of the time with stuff changing all the time and not static, constant social and cultural changes and resources allocated and reallocated: What's the source of skepticism that things won't work well? U.S. economy has done pretty well, financial system that facilitated it by lending more. Concern is that particular circumstances can lead that process to falter. While falling wages and prices can have some effects that stimulate demand, they can also have effects that hurt demand. People who have debts in nominal terms will find it harder to pay those debts. If you have $100,000 mortgage and your wages are going down, you may have to tighten your belt with regard to other things. Fed is trying to prevent deflation.
39:16Current situation: let's ignore how we got here. What advice to incoming administration, what's the argument for the stimulus plans? Have to pay a little attention to how we got here. Insufficient spending: Consumer spending is declining, residential construction is in a terrible slump; net exports doing okay but may decline depending on rest of the world. We need more demand. Federal reserve has done what it can: lowering interest rates to encourage people to borrow. Ben Bernanke has creative ideas about lowering long term interest rates. We cannot fully employ our resources; but the government can do that. Increase government spending--they can so far borrow on good credit. Issues about what the government do with that; but basic Keynesian point is that we need more in total. Where is that money going to come from? Let's not do the printing. It get's borrowed. Have to reverse the paradox of thrift. Increase in spending will increase income. Won't increase in spending be offset by less spending by someone else? Government prints more paper and employs more construction workers. Perceived as income. Puzzle: Distinction between Keynesian and monetary stimulus. Case 1: Government issues a tax rebate: get check in the mail. What we did approximately last spring. Make-work project, dig holes and fill them in. What is the evidence that that has any impact on economic activity? Is there a difference between government printing the money versus hiring workers with printed money to dig holes? Latter is considered fiscal policy. Former is called fiscal but is monetary policy. Keynesians believe either can have impact on aggregate demand. No Keynesians actually would say it's a good idea to dig holes and fill them in, but let's consider it as a thought exercise. Hire one group to dig holes and pay them in green pieces of paper; another group to fill in holes. Those workers will feel and will be better off. What Milton Friedman argued was that that attempt would be temporary; there would be an inflation and people would realize they are no better off. Even with Friedman, there was a temporary effect. Issue: question of expectations. If people were aware of the helicopter drop. If Fed prints money and no one knows the Fed has done that, find $100 bill on doorstep; go out and buy a TV or out for dinner; but if they try and do that every week, people would realize there is no increase in productivity in economy. There is a chance that people won't spend it; they will just save it. If you hire workers directly, you create income in the way we measure it. If you simply give somebody money via a helicopter or tax rebate, they might not spend it. Some estimates of tax rebate indicate maybe 20-50% was spent. Problem is too big now, need solution that doesn't rely on people spending what they see as a windfall. It will be there if it works. Simple version of monetarism: automatic link between supply of money and inflation. Link works through demand. Could lead to higher demand, which in some circumstances will lead to higher prices and in other circumstances lead to more production, hiring more workers. Long run would be inflationary.
52:06Differing perspectives on ability of the government to stimulate the economy. A lot of people are arguing for massive government stimulus. Won't be spent on digging holes and filling them back in. But less productive that private sector activity. May not stimulate at all. Investment climate: people are fearful of the future, partly because they are not sure what the rules of the game are going to be. Risk-takers sitting on the sidelines. Pervasive attitude, not just banks but most of us individually. May not have job in a year, not going to buy new car, etc.; cutting back so they can weather the storm if things get worse. Rent-seeking could even make it worse. What empirical evidence that massive increases in government spending and increase in government deficit will have an effect? Animal spirits, Keynes's term. What evidence that government spending can turn economy around. WWII--some debate whether it pulled economy out of the Great Depression. Economy grew at remarkable rates, huge government deficit. Economy was stimulated. Korea, VietNam, periods of low unemployment. Inflation in late 1960s, probably fully or overemployed. Reagan tax cuts of early 1980s, highest unemployment rate since the Depression. Permanent tax cuts. Beginning of consumption boom. Idea was to stimulate savings, but a lot went into consumption. Problem is these things are all anecdotal. As economists we don't have much of a data base. Only a few historical periods in which these kinds of things were attempted. Podcast with Robert Higgs: military side stimulated, other side of the issue. Hard to say let's look at the evidence and decide which view is right.
59:00Hard question: Very difficult to measure these effects, complex system, can't re-run the experiment with alternative or lack of stimulus. Tempting to construct ex-post narratives. One observation. Every observation has unique circumstances. How should a person confront this challenge? Intellectual mothers, different kinds of milk. Bias, how easy it is to confirm our priors. Difficulty of validating one's own theory, don't have luxury of doing controlled experiments, but we still should try. Suggests openness to more historical style of economic analysis, e.g., Higgs. Have to be humble. Have to be aggressive and work hard to get whatever evidence we can. Need to be open-minded. Different kinds of milk, maybe Kool-Aid. Hank Paulson, Ben Bernanke, Larry Summers, Austan Goolsbee, hard job. Humility in last 16-18 months, as a profession struggling to understand this. $1 trillion stimulus will be an experiment from which we will learn something.