David R. Henderson  

Murphy on the Paradox of Thrift

Back to Basics... New Commanding Heights...

Fellow blogger Arnold Kling has weighed in on the paradox of thrift. Arnold states:

Robert P. Murphy and Bryan seem to me to be confusing things.
Murphy starts with an illustration of the paradox of thrift from Keynesian economist Steve Fazzari. I don't like Fazzari's illustration, so perhaps my quarrel is with him.

I'll think through Arnold's exposition. But Robert Murphy didn't like Fazzari's illustration either or, more exactly, didn't agree with it, which is why he wrote his devastating article. I highly recommend it. I'll admit my bias: as editor of the monthly article on Econlib, I commissioned it. But see what you think.

I'm guessing that Bob Murphy can find the flaw in Arnold's exposition of the paradox of thrift too. Stay tuned.

Comments and Sharing

CATEGORIES: Macroeconomics

COMMENTS (10 to date)
simone writes:

Murphy's points are on target. The simple point on marginal propensity to consume is devastating for the Keynesian story. When Murphy adds time the story is complete. Arnold needs to revisit his view.

Maniel writes:

What we are experiencing these days is a debt-based economy. What that means is that the people (the private sector) and their governments (the public sector) - federal, state, and local - have purchased on credit. This is excellent - we can finance homes, automobiles, health care, early retirement and just about anything else we can imagine - until the bills come due. It appears that for many individuals, that day has arrived. As long as those people are in debt, having to devote major portions of their income to service that debt, their participation in "the economy" will be somewhat subdued.

So behold - folks are beginning to save; how very logical and pragmatic. Government has yet to get the message. Some in government appear to believe that their role is to get people to spend money they don't have by artificially stabilizing prices and institutions using money that the governments don't have either.

Like it or not, we are on our way toward an equity-based economy. When we finally get there, life will be good.

Snark writes:

Wow! Paradox of Thrift, Liquidity Trap, spending multipliers... Keynes lives on!. When he dies again, will we find ourselves re-examining the possibilities of Labor Theory of Value and Dialectical Materialism.

Ressurection is apparently not the exclusive preserve of religion.

Bill writes:

The paradox of thrift is no paradox at all.

It is simply proof that Keynesianism and its prescriptions are wrong and should be discarded in any serious discussion of economics. Murphy correctly skewers the static nature of Keynes arguments that quickly fall apart in a dynamic environment. It is a 3 dimensional model which gives bad answers in the reality of the 4 dimensional world.

Bill Woolsey writes:

The Paradox of Thrift is just the fundamental proposition of monetary theory where the implications for the supply and demand for money are pushed into the background. Yeager called it, "The Keynesian diversion."

The fundamental proposition of monetary theory comes from the early Keynes. The individual adjusts his or her money holdings to meet his or her demand for money by changing spending. However, for the entire economy, the demand for money adjusts to a given quantity of money through changes in nominal income.

So, I can increase my money balances by spending less. Given the quantity of money, however, we cannot all increase our money balances by spending less. All of us spending less reduces nominal income, however. This reduces the demand for money until it meets the existing quantity. (This not about how much people want to borrow or lend!)

Keynes was working in the Cambridge tradtion, where money demand was assumed to be a fraction of nominal income (k). The thought experiment above is an increase in k. The more modern way to describe this reality is that for the economy as a whole, given the nominal quantity of money, an increase in the demand for real cash balances results in a lower price level until the real quantity of money rises to meet the demand. If prices (including nominal incomes like wages) are sticky, then lower real output and real income causes the real money demand to fall to meet the real quantity of money.

The fundamental proposition of monetary theory transforms itself into the Paradox of Thrift, if savings occurs through accumulated money balances.

If prices (including nominal incomes like wages) fail to drop smoothly and rapidly to clear markets, then the drop in nominal income involves a reduction in production and employment. If this were the sole impact, then there is no increase in wealth. Individuals have saved, by consuming less and accumulating larger money balances, but the impact is to reduce real income and employment so that they are satisfied with the existing amount of money. There is no aggregate increase in nominal or real wealth. The real and nominal quantity of money are unchanged.

If markets "clear" in the long run, the result of the added money demand is a lower price level and unchanged real output and employment. While the economy as a whole has not added to its nominal wealth, the lower price level implies that it has saved and aggregate real wealth in the form of greater real money holdings have increased.

The most realistic short run scenario, I think, would be a mixture. Lower output and lower prices. And so, while an individual can save by accumulating money balances and add to his or her nominal and real wealth by the amount saved and the money accumulated, in the short run, the economy as a whole only turns part of this added saving into additional real wealth. This is the increase in real money balances due to the lower price level. The lower output and employment is not an increase in wealth and realized saving.

The Paradox of Thrift will not occur if all of the saving is done by accumulating nonmonetary assets. In other words, if the fundamental proposition of monetary theory does not apply.

Similarly, if there is an effort to save by accumlating money balances, and the quantity of money rises, then nominal and real wealth increases.

For debt financed government spending to fix this problem (of saving through the accumulation of money) people have to be willing to hold government debt in place of money. If the scenario is that people are trying to save by accumulating money balances, and the government borrows, then perhaps, people will be willing to save by holding that newly issued government debt. No longer does the fundamental proposition of monetary theory apply. There is no need for nominal income to fall to get money demand down to the existing quantity of money. (This is where ricardian equivalence kicks in: maybe expected future taxes partly causes people to save more, and try to accumulate even more money balances.)

The liquidity trap type arguments that suggest that an increase in the quantity of money won't solve the problem require that people are currently saving by accumulating existing government debt (which causes no problem in and of itself) and by accumulating money balances (which generates the difficulty.) If there is an attempt to solve the problem of saving by accumulating money by creating new money by purchasing government bonds (conventional open market operations) then the reduction in the availabilty government bonds to purchase results in savers choosing to save by accumulating money even more. It is a special case, which may be applicable right now. The "solution" is "unconventional" monetary policy.

_Everyone_ (among the economists) understands that if additional saving by an individual leads to additional investment in capital goods, then the individual and the economy have both saved more. I would hope that everyone understands that if some individuals save and other individuals dissave, then the economy has a whole has not saved. Consumption has been transferred between households now. Also, if we think about what saving "in the aggregate" is supposed to mean--more capital goods produced, there is a sense in which saving through an increase in the real value of money balances is really an illusion.

The "naive" classical notion that as long as the money is put into a bank and currency is not hoarded under the matress, there is no problem, is _wrong_. The fundamental proposition of monetary theory applies to money "in the bank" as well as to currency. We cannot all accumulate more money "in the bank" by spending less. Because of the regulatory structure of the banking system, and perhaps becaues of the way unregulated banks might operate, if additional saving results in someone moving currency into the banking system, or even funds from checking accounts to saving accounts or CD's, then, this would increase the quantity of money. Depending on how these thinges are defined, then it could be that if saving by accumulating money includes a change in the form in which money is held, the banking system will increase the quantity of money.

There is a free banking theory that suggests that by spending less, the variance of net clearings fall, and that if banks are not subject to reserve requirements, they will incease lending and so increase the supply of money. I find this argument (from George Selgin) plausible. But note--it requires that the quantity of money increase.

The "classical" theory about saving and investmnet is that there is a market price that coordinates saving and investment. Saving occurs through accumulating nonmonetary assets. This added demand results in higher prices for them and lower yields. Firms finance investment in capital goods by selling financial assets. The higher asset prices and lower yields signal them to sell more financial assets and purchase more capital goods.

Keynesians try to explain why this process sometimes goes wrong. The interest rate signals fail to corrdinate consumption now and in the future (all those t's by Murphy) with the production of consumer goods now, and the production of capital goods now, and so the production of consumer goods in the future.

Most of these Keynesian arguments amount to the claim that at very low interest rates, people will start to save by accumulating money balances. And so, we get the "paradox of thrift." But it remains true that the problem is really just an aspect of the fundamental proposition of monetary theory. As Yeager explained, it really is a "Keynesian" diversion.

fundamentalist writes:

The confounding of savings and hoarding was one of Keynes’ mistakes that many people picked up on very early. I’m surprised that it survives today. It just shows how little progress has been made in macro over the past 80 years. Keeping savings fixed at $100 when people spend less is the definition of hoarding, not saving.

Savings and investment are closely linked. If people save (consume less), the prices of consumer goods fall and their profits decline as well. With lower profits, wages are relatively higher, so employment in the consumer goods industries falls. Lower profits and relatively higher wages persuade consumer goods makers to buy labor saving equipment to boost their lower profits. The increased demand for labor-saving equipment persuades capital goods producers to borrow more, invest and increase employment in their industries. If savers have put their money in the bank, interest rates fall, too, making it possible for capital goods producers to borrow. Labor will switch to the capital goods industries, but it will take time.

Total income will drop only if people hoard, not if they save. Hoarding took place in the Great D because people were afraid of banks and paper money, so they demanded gold and kept it locked up or buried.

quanticle writes:
Savings and investment are closely linked.

Not always. For example, look at the current situation. Here banks are so scared to lend that any additional deposits that the consumer makes will be kept in the vaults rather than lent out for new investment. Functionally, in the current situation, putting money in the bank is no different than stuffing money underneath your mattress.

Bob Murphy writes:

As I said over on Arnold's thread, I think his (i.e. Keynes') argument boils down to:

(1) Assume investment is determined by animal spirits, not by interest rates (and hence savings).

(2) Assume drops in consumption don't amplify animal spirits.

(3) Note that savings = investment by accounting.

(4) Conclude that drops in consumption don't increase savings.

That is a logically valid deduction, but I don't think it's a good model of the economy. (Arnold doesn't either, I take it.)

I think people are focusing too much on the restaurant example per se. What I really wanted to get across in my article was that even on its own terms the Keynesian story doesn't make much sense. If the paradox of thrift is ultimately due to the fact that investors won't invest more (no matter how low interest rates) if the demand for their product has fallen off, then why would a random spending package fix that?

The other thing going on here is the difference between nominal and real income (and saving etc.). I omitted that from the EconLib piece in the interest of simplicity, but in retrospect maybe the argument needs to go there to plug all the holes.

fundamentalist writes:

quanticle: "Not always....Here banks are so scared to lend that any additional deposits that the consumer makes will be kept in the vaults rather than lent out for new investment."

Savings and investment are still tied more than people realize in the current situation. People lost a lost of savings in the asset price collapse, especially real estate. Reduced savings has resulted in less investment. Cash balances have fallen as people try to make payments on debt with less income. People are trying to rebuild cash and savings. When uncertainty lessens (that is, when the state decides to quit rescuing every failed business) people will want to hold less cash and they will re-allocate some cash to savings in assets. Businesses will begin to borrow again and the recovery will begin.

Mr. Econotarian writes:

If the family spends the money instead of saving it in the bank, the bank falls below its capital requirements and is taken over by the FDIC, the equity holders are wiped out and the bond holders take a haircut.

Comments for this entry have been closed
Return to top