Paradox of Thrift

So it wasn't all that long that Keynesianism was pretty much dead and buried. A 1996 article in the Cato Journal, entitled appropriately enough "The Paradox of Thrift: RIP," gloated that Paul Samuelson's seminal Keynesian-oriented textbook had dropped its references to the paradox of thrift. Yes, Keynes was discredited and the triumph of the Chicago school was at hand.

And yet, it appears to be back again. Witness, for example, Paul Krugman's warning in the NYT yesterday that "we're in serious paradox of thrift territory here." And it's not just left-leaning Nobel laureates who see a danger. Solidly-libertarian Megan McArdle likewise argues that the paradox looms.

Megan's argument seems pretty solid. In Keynes' day, the paradox may have been a threat, since a lot of people hoarded actual cash. Taking money out of circulation really did remove other people's income from the mix, which if done widely enough, would ultimately make everyone worse off, as firms cut back on either production or wages.

In the modern world, not many people stuff cash into their mattress. They put in into banks, which then lend the money back out either to individuals who want to buy houses or send Sally to college or to business that want to retool or expand or just plain ol' start-up. The former stimulates demand which in turn drives business expansion and the latter drives expansion directly, which in turn drives wages. All is good.

Only, in the current climate, banks aren't really lending out money. There are lots of reasons for this. Americans, in general, have too much debt and so are not looking to acquire more in a shaky economy. Banks have invested poorly, are over-leveraged and are now hoarding cash to ride out future declines in their balance books. And businesses are not, at the moment, really looking to do any massive expansions. Indeed, many of them are also over-extended, too. The result? People save, and the money kind of ends up just sitting around not being used. Hence, the paradox of thrift.

So my question (or really questions) to DR readers: Is this a stopped clock sort of thing, where a Keynes is actually right about the paradox, but only because of a particularly unique set of circumstances? Or is there some reason to think that the paradox isn't really looming? And if it does loom, what then? The logic of the paradox suggests that it might count as a genuine public goods problem. So is a stimulus (one that actually stimulates, as opposed to the monstrosities that are working their way through congress now) actually warranted? Discuss.

UPDATE (Feb. 5): The arbitrage possibilities of tax cuts hadn't occurred to me. But as Megan points out today, even if people are simply saving whatever they get, it's worth remembering that the government borrows money at awfully low interest rates. Even tax cuts didn't jump-start the economy, they might still end up increasing wealth. Whether that would produce more wealth than infusing cash into the economy is something I'm not even remotely qualified to address.

Share this

Try this

Here's a highly readable, lucid, and, I think, insightful discussion of the paradox of thrift. The underlying problem is not thrift, it's an inability to make good investments. That is what needs fixing, and it can be fixed, but it is not going to be fixed by rewarding the people who demonstrated their inability to make good investments (which is what the government has been doing).

"Thrift" is really a misnomer. A "thrifty" person might be wisely investing his savings. That's often the point of thrift. The specific activity in question is not thrift but what the writer calls precautionary storage. Collecting inert, durable goods as a store of wealth (as opposed to investing in productive capacity). And precautionary storage is a response to, and therefore a symptom of, the aforementioned inability of the financial system to make good investments.

Thanks for the pointer

First, I agree that the paradox is badly named. But that seems to be what we're stuck with, so I used the term anyway.

I'd actually seen Waldman's post, and it strikes me that he's on target. It's one of the better arguments against the whole TARP fiasco (which, as I take it, was always more about restoring confidence than about actually solving underlying structural problems, which, in turn, is why if it had to happen it probably also should have been preconditioned on cleaning house in existing management). At any rate, I should have been more specific in my questions.

Clearly the current stimulus bill is mostly useless. For starters, it has a lot of junk in it that isn't stimulus at all. That should all go right away. But if Waldman is right, then isn't it also useless insofar as it includes things like tax cuts for individuals? After all, if the problem really is that individuals, banks and businesses are so unsure what things are good investments and are thus (in effect) hoarding money, then someone needs to step up and say "Here's a good investment." If Waldman is right, then giving taxpayers (or, at this point, even businesses) funds directly is just going to result in more wealth getting stored. So given that the private sector is (for perfectly rational reasons) failing to invest rather than store, then isn't using tax funds to jumpstart rational investment a reasonable use of money?*

* I realize that many DR readers think that taxation is theft and that there are no reasonable uses of tax funds. But to the extent that you think it's permissible for the state to provide genuine public goods, then, if we really do face a paradox of thrift, a stimulus (in the form of immediate investments) seems to me like it counts.

Tax refunds

But if Waldman is right, then isn't it also useless insofar as it includes things like tax cuts for individuals?

Possibly useless for jump starting the economy - I don't know the answer - but I would like to distinguish two ideas:

Idea 1) The tax refunds are unlikely to do much to help jump-start the economy.

Idea 2) The tax refunds are useless generally speaking.

To clarify this distinction, I'll make a related distinction:

Idea A) Individuals are likely to spend the refund so wisely that they will manage to jump start the economy, and also cure cancer and raise Mother Teresa from the dead.

Idea B) Individuals are likely to spend the refund more wisely than the government would have spent it.

I oppose Idea 2 and endorse Idea B and therefore am in favor of the tax refunds generally speaking. I am, additionally, in favor of tax refunds for moral reasons, but even setting aside moral reasons, I believe there is a strong case to be made for tax refunds aside from whether or not it will cure cancer or solve the Rubik's cube in eleven seconds or jump start the economy. (Of course, insofar as the tax refunds represent higher future taxes then they may be on balance harmful - how good or bad they are depends on what the alternative is - no refund with lower future taxes and same government spending, or no refund with same future taxes and higher government spending.)

Paradox not valid justification for taxes

C.J.Trillian,

I don't believe that taxation is theft, in general. I do believe it counts as theft when done for certain purposes however. For example, paying welfare to people who are not in emergency situations, and will never pay the money back.

This stimulus package counts a theft.

"After all, if the problem really is that individuals, banks and businesses are so unsure what things are good investments and are thus (in effect) hoarding money, then someone needs to step up and say "Here's a good investment."

Calling the savings by the derogatory term hoarding leads to intellectual error. It is injecting an incorrect theory into the interpretation of an observation.

Investments is not the only purpose of money. In fact it isn't even a purpose of money. Money serves as a medium of exchange and a store of value. In effect what the anti-hoarders are advocating is the destruction of money.

What has happened is that the fractional reserve banking system has tricked people into believing they had lots more cash holdings than they actually did. Having cash holdings DOES serve a purpose and is part of the mechanism by which prices (the ratio of money to goods) are set.

Now that the fraud of the fractional system has been exposed it is now apparent to people that their cash holdings are too low. They are reacting properly by increasing their savings. It's the government that it absolutely making the wrong decision here.

As people raise their cash holding the pyramid scheme that is the fractional reserve system will unwind, prices will drop, and peoples cash holdings will go up in value. The bubbles will unwind as they should, etc. Is it really such a tradgedy that gas prices have dropped back down?

As for the people who bought houses at the peak of the bubble, and the banks who enabled them? Well bad investment. Take your losses like a man and move on. The poor people will walk away from their loans who can, and will be in an environment of cheap rentals. Not such a bad deal. Some banks will go belly up and some won't. But it will be the bankers who take the largest hit.

Economic Suggestions by Brad Delong

When you see that Brad Delong is for something run quickly in the other direction. He's part of the reason we are in this mess since he was an adviser to Clinton, and the monetary inflation started with his administration.

You can see from the article that this all comes down to a fear of hoarding. That's why I posted the other comment with a link on the fallacy of the evils of gold hoarding .

"They buy gold, or oil, or art, or whatever, and they keep it, generating scarcity rents for those who can offer perceived value stores, but very little in the way of general income and employment. Precautionary storage, not thrift itself, is the villain of the tale." - McCulley

It's totally confused.

Are people hoarding oil like it says in the article? Absolutely, NOT. Oil prices are falling through the floor. What they are or will be "hoarding" is gold (and short term cash). That is why I posted that other comment that debunks gold hoarding as "the villain of the tale". The real issue being hoarding of what counts as money and a store of value during a bank run.

Part of the fallacy of the paradox of thrift is believing that prices are "upward sticky". Obviously they are not given what's happening to oil, gas, etc.

There are many other contributing mistakes that feed into it. Like believing that in a price deflationary scenario it can be a runaway process. That's nonsense because people can't save that much, they have to eat, heat their homes, etc. Consumption cannot drop to 0, or anywhere near it.

The cause of the price deflation, fractional reserve monetary inflation is totally off the radar of those who believe in the paradox of thrift. The fact that they are prescribing exactly the medicine that caused the problem is beyond their comprehension.

They see a heroin addict in "need" of a fix, in a real bad way, and believe not only that the next fix is the solution, but that if the addict comes back for another fix that the last one "wasn't a big enough stimulus". This is the best way to end up with a dead addict.

The shivers, the cold sweats, the pain of going cold turkey is part of the correction process. You cure an addict by taking away the drugs nto feeding him more faster.

Begs the question

Idea B) Individuals are likely to spend the refund more wisely than the government would have spent it.

As a general rule, I agree with this claim too, which is why I am part of the community at DR rather than at Daily Kos. :-)

But in this particular instance, doesn't Idea B just beg the question? I mean, at least for the purposes of this discussion, we're assuming that McArdle/Waldman/Krugman are right and that we are in danger of something like the paradox of thrift. More specifically, we are assuming that if Waldman is right, then ex hypothesi, people aren't currently using their money wisely. (Or, technically, they're using their money in a way that is wise from their individual point of view but that makes all of us worse off collectively.)

IOW, what Waldman describes is an instance of a genuine public goods problem. Saying that individuals are likely to spend their tax refund more wisely than the government would have spent it just is to deny that we're actually in paradox of thrift territory.

I suppose one could argue that, no matter what, the state will always spend money less wisely than individuals. That is, one can admit the existence of genuine public goods problems and still think that states will muck things up even worse. Unfortunately, there's not much data (on either side) to show that that's true. There haven't been all that many major depressions, for example, and, at least as I read things, really smart economists are pretty much of the opinion that they don't exactly know what kinds of actions are likely to get us out of our current economic woes. As an empirical matter, it's an open question as to whether government spending or individual spending is more likely to fix things.

Not the same

More specifically, we are assuming that if Waldman is right, then ex hypothesi, people aren't currently using their money wisely.

And if government would spend it even less wisely, then my statement is still true, i.e.:

Individuals are likely to spend the refund more wisely than the government would have spent it.

..

IOW, what Waldman describes is an instance of a genuine public goods problem.

A public goods problem, but not necessarily one that is realistically solvable by government.

Saying that individuals are likely to spend their tax refund more wisely than the government would have spent it just is to deny that we're actually in paradox of thrift territory.

No, because your claim that these are equivalent presupposes that government action would be effective.

I suppose one could argue that, no matter what, the state will always spend money less wisely than individuals.

That overstates my position. I didn't say "always". I said:

Individuals are likely to spend the refund more wisely than the government would have spent it.

You took a defensible position (that X is more likely than Y), and turned it into a position that is much more difficult to defend (that X will always be the case and Y will never be the case).

If there are cases where government would spend the money more wisely than we would, then we should let the government spend money in those cases, right? Whoa, whoa, wait a second! That assumes omniscience! It's not enough that there be cases where the government would spend the money more wisely. We need to know with some sufficiently high degree of certainty when that's going to happen.

If government action is superior to private action by $100 1/3 of the time, but is inferior to private action by $100 2/3 of the time, and if we cannot predict in any given situation when that will be, then we should always rule in favor of private action.

Unfortunately, there's not much data (on either side) to show that that's true.

There's not much data to show that the sun will come up ever again, because the future hasn't happened yet. Most of the predictions we make about the future are tied to our past experiences by theory, formal or informal. We should be ready to use theories that have in some way been empirically tested regardless of whether they were tested in precisely the current sort of situation.

There haven't been all that many major depressions

There haven't been all that many babies born with thirteen fingers, so there's not much data for or against the claim that thirteen-fingered babies are possessed by the devil. But still we have very good reason to believe that thirteen-fingered babies are not possessed by the devil. Similarly, it's not necessary that we have experienced many, or any, depressions, to have good reason to believe certain things about them.

really smart economists are pretty much of the opinion that they don't exactly know what kinds of actions are likely to get us out of our current economic woes.

Really smart economists also don't know exactly where Osama Bin Laden is hiding. It does not follow that we should bomb a random location. We should base our actions on what we do know. Bombing a random location has a certain probability that it will kill innocents, and the vanishingly small probability that it will kill Osama Bin Laden is almost certainly much too low to support the decision to bomb, when weighed against the much higher probability that innocents will be killed.

Similarly: government has a pronounced tendency to waste wealth. Since we don't know whether a given government expenditure will get us out of our woes, but we do know that on average, government expenditures are enormously wasteful, then we should act on the basis of what we do know, and not let government spend that wealth.

As an empirical matter, it's an open question as to whether government spending or individual spending is more likely to fix things.

As an empirical matter, it's an open question whether Barack Obama's grandchild will be the second coming of Jesus Christ, but it in no way follows from this that we cannot reasonably say with high confidence that the grandchild will not be the second coming. In fact, it would be unreasonable to say otherwise.

Toward the end you're making some familiar arguments which only seem to recommend what you seem to think they recommend, as I've tried to show by plugging in other content into the same form of argument.

Paradox of Thrift Related Economic Issues

"Do you feel like presenting the case against the paradox of thrift in this entry which was created within the past day (as opposed to being a year old, as the current entry is)?"

I covered the issue of fractional reserve deflation and saving (hoarding)in depth over a Reason.com a few months back. I'll post a link.

It's the related (really the same issue) of the evils of hoarding gold, which is supposedly worse than saving in a bank according to the paradox of thrift.

This is the opening salvo:

"Currency is hoarded instead of invested, and capital decreases"

Errr... currency isn't capital, it's a medium of exchange and a claim against capital. Hoarding gold actually increases capital. You have to produce more goods than you consume in order to obtain the gold. Saving gold frees up that capital for use by other people.

Just search down for Macker from there. One guy liked it enough he suggested I convert my response into an article and post it on a Lew Rockwell.

I think I have another comment out there about the liquidity trap on Marginal Revolution.

Is the paradox of thrift

Is the paradox of thrift something that actual Keynesian economists allege to hold under general conditions, or is it something that they acknowledge holds only under specific conditions, such as those you described above?

My impression is that real Keynesianism is quite different from the folk Keynesianism that laymen get third-hand from journalists and pundits. If you've ever heard the ideas of a libertarian thinker summarized by a non-libertarian, or even by a less-sophisticated libertarian, this should be plausible.

I think that the basic idea behind Keynesianism is that sometimes there are bubbles in cash. Which seems plausible—if there can be bubbles in equities and commodities, why not cash?—but I haven't figured out all the implications. I know that folk Keynesianism is nonsense, but I'm not convinced that there isn't at least a kernel of truth in real Keynesianism.

Real Keynes

The paradox of thrift is actual Keynes (from The General Theory, I believe).

My understanding (and I'm not an economist here, let alone an economic historian) is that Keynes and subsequent Keynesians actually believed that the paradox is something that held under general conditions. And, in Keynes' day, that might not have been a crazy view to hold. At the time, most transactions were in cash, and given the uncertainty of the banking industry, a whole lot of people saved cash by just hoarding it (or storing it in their mattress, if we prefer the less loaded term). If enough people did that sort of thing, it could lead to the paradox as a general rule.

But modern financial markets have mostly done away with passive storage of money. Hence, as I understand it, most economists had given up on the paradox of thrift as a condition that generally holds. What I've seen more recently is people arguing that it might hold in this fairly limited circumstance.

Wrong

"Megan's argument seems pretty solid. In Keynes' day, the paradox may have been a threat, since a lot of people hoarded actual cash. Taking money out of circulation really did remove other people's income from the mix, which if done widely enough, would ultimately make everyone worse off, as firms cut back on either production or wages."

Wrong, wrong, WRONG, Wrong, wrong, wrong, wrong.

The most basic principle of economics is that people trade because of differences in valuations. Party A trades cash for the goods of party B because, A values the goods more than the cash, and B values the cash more than the goods.

Free markets increase everyones welfare because individuals are the best judges of what they value, their own circumstances, their own talents, local conditions, etc. No central committee can know better because they cannot gather the information, and couldn't process it if they could.

If "a lot of people hoarded actual cash" then they did so because they valued the cash MORE than their other trade options. Sure they could have traded their cash for bank deposit slips but THAT WASN'T VERY DAMN SMART now was it. Why wasn't in smart. BECAUSE THE BANKS WERE CHEATING THEM!!!!!!

Deciding not to trade their money for banknotes doesn't "ultimately make everyone worse off". That's garbage that the fractional reserve banks want you to believe. Why? Because they lent the money that they were suppose to be holding IN CASH out to others as long term debt.

The reason people hold cash is that it is valuable for certain purposes. It's needed for emergencies, for when you lose your job, for getting you to the next paycheck, as a medium of exchange, and as a store of value when you want to delay consumption to the future or for big ticket items.

Fractional reserve banks trick people into believing they are holding cash for such purposes when in fact they are not.

People "hoard" cash during bank runs for the same reason that people try to yank their money out of Ponzi and Pyramid schemes once they catch on.

If you and another person both join a ponzi scheme and they manage to pull out their original deposit (or perhaps never invested in the first place) that in no way harms you. In aggregate it can't "make everyone worse off". Forcing the person to participate in the scheme is not a way to solve the problem.

One of the things Keynes got so absolutely wrong is that he confused price deflation caused by productivity increases, with price deflation caused by the collapse of fractional reserve monetary inflation. That's like confusing the price drop in personal computers due to advances in technology with the price drop in a Madoff security.

Keynes did not distinguish between the two kinds of price declines. He essentially argued that lowering computer prices would cause people to hoard money waiting for prices to drop in a runaway process. Which is ridiculous. The very lowering of prices makes the ordering of valuations between money and the goods that can be traded for money drift towards favoring the purchasing of goods. People always have a tendency to want to consume more and the lowering of prices makes that kick in.

Are you waiting for gas prices to drop further so you can do all your driving in the future? Of course, not. Most people cannot postpone most consumption. Only a small amount. The current deflation and bank collapse is NOT about people saving more cash. It's about a pyramid scheme collapsing, fractional reserve banking.

Half-baked arguments

You've by the way highlighted an example of a half-baked argument, specifically:

Taking money out of circulation really did remove other people's income from the mix, which if done widely enough, would ultimately make everyone worse off, as firms cut back on either production or wages.

That's half-baked because it only considers a portion of what needs to be considered. Taking money out of the system will tend to depress wages but by the same token will tend also to depress prices. If my wage is lower but the prices of the goods I buy are lower by the same amount, then I have not been made worse off. Much more needs to be said in order to deal with and maybe rule out this possibility. And it was not said.

Anyway, that's an example of a half-baked argument. It has a serious gap, a gap which threatens to undermine it entirely.

Now, one thing that is repeatedly brought up in discussions about the evils of deflation is sticky prices. Sticky prices flesh out the idea that a deflation, and therefore taking money out of circulation, might have a certain adverse impact. But sticky prices are nowhere here mentioned. The argument goes from hoarding cash to supposed bad consequences without ever talking about sticky prices. That is half-baked. Or, alternatively, it is a mere sketch of an argument worked out in much greater detail elsewhere.

More Half Baked Thinking

"Taking money out of the system will tend to depress wages but by the same token will tend also to depress prices."

It's half baked in another way. It ignores the fact that the person who "took" the money out had to "give" something to take it out. So it almost sounds like the person is cheating the system. The saver had to actually either increase personal production or decrease consumption and then sell the excess goods.

To restate what we are saying another way:
One of most important purposes of the prices is plan coordination. Without such plan coordination modern complex economies can not operate properly. Prices are determined by an iterative process in the economy.

People are specialized in an economy. They tend to produce one good, or class of goods, not a uniform range of goods. One way to save cash is they must produce more of a particular good. This tends to drive down the price of that one good faster than other goods. Thus it becomes harder and harder for them to save. They do, after all have to find buyers in order to load up on money, and must do so in an environment where their own means of income is dropping fastest.

So savings by over-production is a hard thing to do, and is not a run away process. What about savings by lowering consumption?

Look at the example that Megan McArdle gave.

"If we (hypothetically) decide to eliminate takeout from our menu and eat tuna sandwiches instead, we are saving money. But the restaurant loses it. By foregoing spending, we are pulling money out of the economy."


[Actually, it's funny but I always thought of take-out as the cheap stuff and sit down as the expensive. As in McDonald's, vs. Olive Garden. But let's go with the flow.]

She makes it sound like the largest impact of my savings is falling on somebody else, willy nilly, in a concentrated fashion. That's not true. My production is specialized, but my consumption is general. Thus to save substantially my reductions will will be spread over many people.

But who are the many? Well, most people reduce frivolous spending first, when they save. So I'll cut out things like entertainment, then I'll cut back on luxury items, then I will go for more basic models of other items, etc.

In other words, the lower prices will hit producers in industries that cater to things less important to survival, in the present. That, I think, is an important signal to send in an economy in trouble.

So why fret about the fact that there is a signal to produce less take-out and to produce more tuna sandwiches? Certainly, producers who specialized in take-out will go out of business and have to take u p an occupation more in line with what the consumers need. But that is probably only a few. Many producers will be able to easily shift quantities to meet demand. Just stop producing take-out.

In both cases they are coordinating their plans with other individuals in the economy. Hard times so people are planning on spending less on luxury, and conveniences. Prices change and producers adjust accordingly.