Saving Supply and the Natural Interest Rate

Examining the business cycle based upon the assumption that the supply of saving, the demand for investment, and the natural interest rate are unchanging is like putting on blinders. It is certainly possible that an excess supply of money will be matched by an increase in the real supply of credit and push the market interest rate below the natural interest rate. On the other hand, what happens if there is a decrease in the supply of saving?

Suppose the market interest rate was initially equal to the natural interest rate, and then the supply of saving decreases. At the initial market interest rate, the quantity of saving supplied is less than the quantity of investment demanded. That part of income not spent on consumer goods and services is greater than spending on capital goods.

The natural interest rate, where saving and investment are equal, is now higher. Assuming the market interest rate rises with the natural interest rate, the quantity of saving supplied increases and the quantity of investment demanded decreases. In the end, the amount saved and the amount invested are both smaller. The composition of demand has changed, with more demand for consumer goods and less demand for capital goods.

Of course, these inclusive aggregates--consumer goods and capital goods--mask important changes in the composition of demand within the aggregates. If consumer goods are homogeneous, then the increase the quantity of saving supplied is simply an offset or dampening of the initial decrease in the supply of saving. Equivalently, the decrease in the demand for consumer goods due to the higher market interest rate is just a partial offset or dampening of the increase in the demand for consumer goods due to the initial decrease in saving supply. However, since consumer goods and services are an aggregate, it is likely that there are shifts in the composition of the demand for consumer goods and services towards those with smaller interest elasticity of demand. For example, the demand for durable consumer goods like automobiles may decrease, while the demand for restaurant meals increases.

Similarly, if capital goods are all homogeneous, then the increase in the market interest rate simply decreases the demand for them. But capital goods are not homogeneous. And the particular purposes to which they are put are not independent of the interest rate. Those capital goods with higher interest elasticities of demand will see the greatest reductions in demand. In particular, capital goods that will generate consumption in the more distant future, perhaps because they are more durable, will especially lose value due to the higher interest rate.

Perhaps I have missed some of the subtleties of Austrian capital theory, but I hope this gives the proper flavor. The change in the composition of demand, both between aggregate like consumer and capital goods, as well as amount different consumer and capital goods, are important. Presumably, the demands for consumer goods and services with less interest elastic demands rise, as do the demands for capital goods specific for their production. The demands for consumer goods with demands that are highly interest-elastic fall, as do the demands for a variety of different capital goods depending on their interest elasticities of demand.

The relative prices of the goods and services with lower demands fall and the relative prices of the goods and services with higher demands rise. The change in relative prices and profitabilities results in changes in production and the allocation of existing resources, including labor. Employment expands in "consumer goods industries" and shrinks in capital goods industries. Or, more precisely, it rises in consumer goods industries where demand is less interest elastic and in closely related capital goods industries, while employment shrinks in capital goods industries where demand is more interest elastic. An entire integrated structure of production is developed based on this new pattern of demands through time.

It would seem that this reallocation of resources would be associated with both structural unemployment and capital losses for many capital goods that cannot be shifted to the direct production of consumer goods that do not have interest elastic demands. For example, a need for another stove in the restaurant may be an added demand for capital, but the robot welders on the auto assembly line will provide little help. Perhaps they can be shifted to the stove factory, but perhaps not.

High unemployment, firms failing, losses on capital goods, perhaps even capital goods abandoned, all appear to be consequences of a decrease in the supply of saving and the necessary increase in the natural interest rate. Can the economy recover? Yes, but it can only recover gradually. Labor needs to be reabsorbed in those sectors of the economy with growing demand--consumer goods with low interest elasticities of demand and related capital goods. New, more appropriate capital goods must be constructed. Eventually, unemployment will fall and output will recover--at last partially.

The reason for this discussion isn't to claim that the current recession, or any recession, has been the result of a decease in the supply of saving. My point was rather to suggest that there is little or no difference between this process and what advocates of the Austrian theory of the business cycle describe as a recession.

Perhaps more interesting, suppose rather than starting with a decrease in the supply of saving, what if this was a return to the composition of demand and allocation of resources from the recent past. Suppose the supply of saving rises, stays relatively high for a few years, and then returns back to its previous level. How different, really, would be the pattern of demands, production, and employment than the traditional "Austrian" cycle theory?

For example, suppose that the collapse of the dot com bubble left households poorer or fearful and they responded by saving more. The Fed, miraculously, adjusts the market interest rate to the new natural interest rate. A new pattern of demands and a new allocation of resources, including labor, develops. Capital goods specific to this pattern of demands are constructed. Then, after some time, households have rebuilt their wealth a bit or just no longer fear that this stock market crash will have the consequences of the one in 1929. The supply of saving falls. The pattern of demands and the allocation of resources must change again.

Focusing particularly on the "malinvestments" ("mal," only in the sense that they are inappropriate to the pattern of demands associated with low saving,) what are we to make of the entrepreneurs who purchased these capital goods and suffered the losses when saving fell again? Isn't this just an entrepreneurial error? And what was the nature of that error? Wasn't it the assumption that the temporarily low level of "the" market interest rate (and natural interest rate,) would be persistent and justified the production of capital goods that will only remain profitable if enhanced levels of saving persist?

Austrian cycle theory shows that an effort to use monetary policy to impact real interest rates, the composition of demand, the allocation of resources, and the functional distribution of income between labor and capital can at best only appear to work in the short run. In the long run, it is likely to have disastrous consequences. This insight has value. However, trying to explain actual business cycles using this theory, in a world where hardly anyone is using monetary policy for that purpose, is fraught with danger.

In particular, in a world where the demand to hold money, the supply of saving, and the demand for investment can all change, a pattern of thought that identifies increases in the quantity of money or decreases in market interest rates with an excess supply of money or a market interest rate pushed below the natural interest rate can be highly misleading.

Final Sales of Domestic Product: First Quarter 2010

The GDP report came out, and as usual, the headline is the real GDP figure. Real GDP increased at a 3.2 percent annual rate in the first quarter of 2010.

However, I always scroll down to see what has happened to total cash expenditures. My preferred measure is Final Sales of Domestic Product. The value of the first quarter 2010 was $14,568 billion. The annual rate of increase was 2.51 percent.

The good news is that cash expenditures have nearly returned to their peak value before the recession of $14,583 billion in the 3rd quarter of 2008. The shortfall is less than .1 percent.

On the other hand, if cash expenditures had continued on its 5 percent trend growth path of the Great Moderation, it would now be $16,311 billion. The current level is 11.3 percent below trend.

Even with a modified growth path of 3 percent, changing at the peak in the third quarter of 2008, the target level would be $15,723 billion, and so the current value is 7.63 percent too low.

Worse, a 2.5 percent growth rate will never return to a higher 3 percent growth path, much less a 5 percent one. Surely, it is better for cash expenditures to rise rather than fall, but the current growth rate remains too low.

I believe that the Fed should commit to returning Final Sales of Domestic Product to the 3 percent modified growth path next year. It is now the second quarter of 2010, and so, one year from now, in the second quarter of 2011, cash expenditures should be $16,324 billion. That would be an increase of 12 percent from last quarter, or averaging a 9.6 percent annual growth rate for the current and next 3 quarters.

Russ Roberts on the Crisis

Russ Roberts of Cafe Hayek has an excellent paper on the Crisis--Gambling with Other People's Money.

Roberts blames the crisis on a pattern of creditor bailouts by government. The government has developed a policy of making sure that people who lend money don't suffer losses. On its face, that is absurd. Banks take losses from bad loans all the time. Robert's point, however, is that the households and businesses that lend to financial firms are shielded from these losses. He emphasizes that government policies are aimed at limiting the losses of any financial firm to stockholders of that particular firm.

For commercial banking, his account rings very true. Deposit insurance explicitly protects insured depositors from loss. Insured depositors have no reason to worry about the risks taken by their banks. Losses are borne first by the bank's stockholders and then by uninsured depositors and other creditors. However, FDIC has made it a habit to protect all of the uninsured depositors and other creditors from loss as well by using the "purchase and assumption" method of resolution for insolvent banks. FDIC guarantees enough of the bad assets of a failed institution, so that some larger, hopefully sound bank, will then take the remaining assets of the failed bank as well as take responsibility for all of its liabilities.

Roberts emphasizes that the GSEs, Freddie Mac and Fanny Mae, borrowed by selling bonds that had an implicit federal government guarantee. As expected, the lenders were protected from loss when Freddie Mac and Fanny Mae became insolvent. Harder to explain was lending to stand-alone investment banks, like Goldman Sachs, Morgan Stanley, Bear Sterns, and Lehman Brothers. None of these borrowed using insured deposits. There was no implicit government guarantee... oh wait. The government bailed out all of the creditors of these institutions except for Lehman Brothers.

Roberts builds a convincing argument that bailing out creditors leads to perverse incentives. However, I disagree with his claim:

The punishment of equity holders is usually thought to reduce the moral hazard created by the rescue of creditors. But it does not. It merely masks the role of creditor rescues in creating perverse incentives for risk taking.

I think that making equity holders take losses does reduce the moral hazard created by bailouts. For example, the initial TARP proposal was for the federal government to buy "toxic assets" from commercial and investment banks. This would bailout the banks for making bad loans, and so, the equity holders as well.

Fortunately, that plan was dropped. Instead, the TARP money was used to purchase stock in the banks. This resulted in some dilution of ownership, but it still involved bailing out stockholders.

If the various commercial and investment banks had been given the more standard FDIC treatment, all creditors would have been made good, but the stockholders would have been left with nothing. Bush and Paulson made moral hazard worse.

Robert's point, of course, is that even if the stockholders had lost everything and only the creditors were covered, there is still serious moral hazard. Before deposit insurance, banks needed to maintain higher capital ratios to attract depositors. The way to make sure that depositors or those holding the deposit-like short term commercial paper of investment banks insist on high capital ratios is for them to lose some money in banks that fail.

I think Roberts fails to emphasize the "bubble" element in the housing market. While he discussed a variety of government policies that were aimed at raising housing demand and so the equilibrium prices of homes, the "bubble" is the mistaken projection of past price increases into the future. In my view, this is what leads to the overshooting and an inevitable collapse.

If we think about "greater fool" investors, who understand that past price trends should not be projected into the future, but who believe that there are some "fools" who do and that profits can be made at their expense, the resulting "game" is very much like poker. This is especially true when the goal is to fleece not only the fools, but also others seeking to sell out to greater fools. And it is in that poker game, that Robert's point about gambling with other people's money rings true.

I strongly disagree with part of Robert's conclusion:

Be aware that the Fed is certainly part of the problem and may not be part of the solution. The Fed created the artificially low interest rates that helped inflate the housing bubble. The Fed then raised interest rates too quickly with disastrous effects for the adjustable-rate mortgages encouraged by their low-interest- rate policy. Monetary policy should not be left to any self-proclaimed or publicly anointed maestro. Following an automatic money growth rule or the Taylor rule would have avoided much of the pain. Somebody needs to hold the Fed accountable for funding exuberance.

I am not sure that the Fed's excessively low interest rates were part of the problem. If the natural interest rate falls and the Fed properly allows the market interest rate to fall with it, then housing prices will rise.

Worse, following "the Taylor rule" or anything like it is a mistake. Interest rates should depend on supply and demand, not manipulations of the quantity of money by a central bank based upon observations of past inflation and estimates of an output gap. Further, a money growth rule is equally bad. Money is an economic good, and its quantity should adjust to the demand to hold it.

Naturally, I favor a target for a slow steady growth path of aggregate cash expenditures. To the degree possible, market forces should be allowed to freely adjust both interest rates and the quantity of money consistent with maintaining that stable macroeconomic background.

In conclusion, I must ask, what "crisis" do we mean. I believe that the Fed's policy of interest rate targeting was responsible for the real crisis that has left total cash expenditures about 10 percent below their long run growth path. I won't deny that a long standing policy of creditor bailouts was responsible for the solvency problems of leading Wall Street investment banks, and many commercial banks as well. The degree of malinvestment in housing, as well as the inevitable disruption to production when insolvent financial institutions are reorganized might also be laid at the door of this policy of creditor bailouts. But these things pale against the failure of the Federal Reserve to do its job--keeping aggregate cash expenditures on a stable growth path.

Speculative Bubbles and the Natural Interest Rate

One problem with "Austrian" Macroeconomic thinking is excessive focus on scenarios where the natural interest rate, and so, the supply of saving and demand for investment, is given. The "classic" scenario is where the demand to hold money is assumed given as well, and then an increase in the quantity of money pushes the market interest rate below the natural interest rate.

Suppose instead that investment demand increases because of a speculative bubble in housing. The "cause" is the projection of past price increases into the future. People invest in order to profit from those expected price increases. If there are people who will be willing to pay high prices for the houses in the future because they want to live in them, then there is nothing irrational about this investment plan. It is only a speculative bubble if the reason people are paying higher prices for the homes is that they plan to sell them to someone else at a still higher price, but that person will only pay that price because they plan to sell to still someone else at a yet higher price. It is a speculative bubble if there is no one at the end who is willing to pay the high price in order to live in the house.

Suppose that instead of this kind of bubble behavior, people are investing in houses now because radio messages are received from outer space. Alien settlers are on their way with ships full of great consumer goods. They will exchange those for houses when they arrive. Assuming the messages are true, then purchasing a home is a great investment project. It will generate large amounts of future output. (And, when the messages turn out to be a fraud, the result will be somewhat like the bursting of the speculative bubble.)

So, what are the effects of an increase in investment demand? Investment demand is made up of investment demands for all sorts of different types of capital goods by a variety of different firms. (This analysis includes residential investment demand by households along with the rest.) If the demand for residential investment increases, the sum of the demands increases as well.

At the existing level of the natural interest rate, the demand for investment is greater than the supply of saving. The natural interest rate is higher. Assuming that the market interest rate was equal to the natural interest rate before, and then rises with the natural interest rate, then the quantity of investment demanded will fall and the quantity of saving supplied will rise. At the new equilibrium, the amount saved and invested are both higher. This increase in the amount saved is equivalent to a decrease in the amount consumed. Less is spent on consumer goods and services. The increase in the amount invested is an increase in the purchases of capital goods. There is a shift in the allocation of resources.

The increase in the market interest rate to some degree chokes off the increase in the demand for residential housing. However, it also reduces the demand for all other types of capital goods. Naturally, there will be an increase in the derived demand for sawmills, forests, brickyards, and the like. But like the homes themselves, these added derived demands will be dampened by the higher market interest rates. Still, there are presumably some capital goods less directly related to housing production that will also see their demands fall due to the higher interest rate.

The increased market and natural interest rates imply an increase income from capital. Other things being equal, that would be an increase in aggregate income. How is that possible? If the basic identify of macroeconomics is that aggregate income is always equal to output, then how can those earning greater capital incomes do so when output is limited by the existing productive capacity? It is because the new homes are worth more than the other capital goods (or currently produced consumer goods and services) being sacrificed.

In the scenario of a speculative bubble, this greater value is an illusion by assumption. With the alien immigrant scenario, the value of the bonanza of products they will bring is being pulled into the present.

What about labor? Those constructing new homes to take advantage of their high prices demand more labor. Wages rise so that those producing other capital goods and consumer goods will use less labor, freeing up the labor needed to construct new homes. Assuming that the quantity of labor supplied rises in response to higher real wages, the quantity of labor used in production rises. This increase in labor input provides a second, more tangible source of added current production.

The result of the great profits that can be generated from housing production (or more exactly, the expectations of those profits) results in higher real wages and higher real capital incomes. Because the homes and capital goods used to produce houses are so valuable, real output rises just from the shift in the composition of output, and then it also rises because higher real wages draw out more labor input.

This real boom is consistent with scarcity. Leisure, other capital goods, and so far, current consumer goods are being sacrificed. Housing and capital goods used to produce housing are being produced. The illusion of created by the speculative bubble (or the space alien settlers) is motivating these sacrifices and creating the real boom.

Is it conceivable that current consumption could rise in this boom? To the degree that higher real wages bring forth an increased quantity of labor, it is obvious that current consumption can rise. As the workers earn higher incomes, it is likely that they will increase their current consumption.

Even ignoring this effect, it would be possible for those receiving increased capital incomes to expand their consumption. Resources, including labor and existing capital goods, can be stripped from various capital projects, for example, dams, and used to produce consumer goods and new homes. Income, output, and consumption can all expand because the houses being produced are so much more valuable than the dam project.

So leisure and alternative capital projects are sacrificed, and real wages, employment, real labor income, real capital income, the real capital stock (including houses,) aggregate real income, aggregate real output all expand. If the cause was the alien immigration, then when they arrive, the past sacrifices are shown to be justified. The sacrificed capital projects (the dam) don't produce output, but the trade goods brought by the aliens compensate for the sacrificed products.

But suppose the aliens don't arrive. Suppose it turns out that the radio messages were a fraud. Everything must go into reverse. Real wages must be lower. Employment will be lower. The real capital stock will fall. All of those houses aren't so valuable after all. Real capital incomes must fall. The production of consumer goods must fall.

Of course, if the cause of the great profits in housing was a speculative bubble, then it was always a fraud. That is, it must eventually end.

In my view, the best macroeconomic environment to for the necessary readjustments is to keep total cash expenditures on a stable growth path. While real wages and real capital incomes may need to fall (and certainly grow more slowly for a time,) stable growth in nominal incomes imply that this will require price inflation to a higher price level. If the failure of nominal wages to keep up with inflation results in people quitting work or dropping out of the labor force, this is the proper response. The decrease in labor input will reduce the productive capacity of the economy, but at the same time that the loss in real income reduces the demand for consumer goods. And, of course, the lower real interest rates may well cause reduced saving out of that reduced income, and at the same time increase the profitability of those investment projects that had been crowded out by the housing investment.

I hope that my analysis of the impact of an increase in the demand for investment has been suggestive. Could something like this have occurred during the housing boom? Suppose central banks are targeting the market interest rate so that increased investment demand results in an excess supply of money? Certainly more analysis is necessary. But recognizing that the natural interest rate can and does change is a good start.

Recalculation--Two Versions

Arnold Kling has claimed that the current recession is due to a need to "recalculate." There are really two versions of the argument. As Frank Knight said of Keynes, I will argue that what is correct isn't new. And that what is new isn't correct.

Looking back at the housing boom in the past decade, there is good reason to believe that there has been substantial malinvestment. Too many homes have been built. Various capital goods have been produced that are specific to the production of housing. There is a similar problem with human capital--too many skilled carpenters.

With hindsight, fewer houses should have been produced. Fewer capital goods specific to housing construction should have been produced. Less carpenter skills should have been developed. Instead, other goods and services should have been produced, including both physical and human capital specific to other industries.

While those are all sunk costs at this point, it remains necessary to expand the production of other, more valuable goods. This involves shifting labor and other resources away from housing construction to the production of other goods and services.

During this process, the unemployment rate for labor will be relatively high. There is a longstanding term for this type of unemployment--structural unemployment. If the unemployment rate is relatively high, the employment rate is relatively low. The productive capacity of the economy is depressed.

Added to this problem is the fact that human and physical capital specific to housing construction are lost. New capital goods must be constructed and new skills developed. Labor productivity should be depressed even after labor is absorbed. The productive capacity of the economy will only gradually recover.

Personally, I think the term "recalculation" is inappropriate to describe this adjustment process. Perhaps "redeployment" is the better term. A shift from goods that were being overproduced to the goods that are currently being underproduced is a redeployment of resources.

That such a redeployment is likely occurring is important to keep in mind. For many years, monetary and fiscal policy was aimed at stabilizing unemployment or real output. If structural unemployment is currently high and productive capacity low, seeking to use expansionary monetary and fiscal policies to return unemployment to where it was before this redeployment began would be a mistake. Similarly, seeking to return employment and output to their pre-redeployment growth paths would be a mistake.

However, when Kling began using the term "recalculation," he sometimes described a quite different scenario which is not a redeployment of resources from what was overproduced to what is underproduced. Rather, he suggests that we must pause, and leave resources unemployed while we determine what to do with them. Of course, there is no "we" making conscious decisions, but Kling argued that "the market" must similarly pause to decide what to do, and so there is a "recalculation."

I believe that this conception of recalculation is inconsistent with the most important principle of economics--scarcity. There are not enough resources--land, labor, and capital--to produce all the goods and services that people can use to achieve their goals. It is certainly possible for a particular good to be produced in such large quantities that isn't scarce. Or even that it be produced to the point that some of it becomes garbage and someone must be paid to dispose of the excess.

It is conceivable that all tangible goods and services could be produced in quantities where none of them are scarce, but that isn't the situation that faces the world today. Billions of people in the world could use more of any number of existing goods.

Under conditions of scarcity, overproduction of any one good means that some of the good is worth less than its opportunity cost, which is the other goods or services that the resources could have been used to produce instead. To say that homes were overproduced is not to say that they are not scarce much less that some are garbage. It is rather to say that the other goods and services that could have been produced are more valuable. And to say that fewer homes should be produced today is not to say that no one could use the additional houses, much less that they will need to be demolished. It is to say that the resources that would be needed to continue to produce new homes at the rates typical during the boom have more valuable alternative uses. The other goods or services they could be used to produced are more valuable than those additional houses. And so, there is a need to redeploy resources.

Suppose, on the other hand, that the economy doesn't face scarcity. I don't want to tease out all the implications of this assumption--zero prices and wages--for example. No, assume that people decide that they don't want to purchase new homes, and there is nothing else that they want to buy. And so, until it is possible to figure out some new good that they want to buy instead of houses, there is nothing for people to do. They must wait around.

In a world of scarcity, this really isn't an issue. There are many things that people want to buy instead of houses. They bought the houses because they valued them more than the other things they could have instead used their income to buy. If they decide houses are not so desirable to buy, they buy other things instead.

Of course, the analysis above ignores leisure. It is possible that scarcity exists because leisure is valuable. People work in order to earn income to purchase new homes. They decide they no longer want to do that, and so they work less and earn less and purchase fewer new homes. Fewer homes are produced, so total output is lower. Fewer people work because they prefer more leisure, and so, there is less employment.

When someone comes up with a new good people want to buy, then people will go to work to earn the income to buy it. And firms will hire workers to produce it. Then why is output and employment low now? Because no one has determined what good they want instead of new homes. We know we don't want a bunch of new homes. But we must decide what we do want. We must "recalculate." Meanwhile, we enjoy leisure.

Unfortunately, this appears to be a very poor explanation of involuntary unemployment. It is rather an explanation of long vacations taken by people now have everything they need. It doesn't fit in well with people cutting back their purchases because they have lost their jobs. I personally know people who are new entrants into the labor force. They are currently unemployed and would like to work. The reason they would like to work is to purchase any number of existing goods and services.

Of course, if Robinson Crusoe finished up his house, and didn't want a second home, and had nothing else he wanted, then he would simply take it easy. He would take a vacation and have all the goods and services he could use.

But perhaps the problem is one of distribution. Some people, call them "the rich," have all the goods and services they can use. They don't want new homes. And so they buy less. And they work less. Meanwhile, other people, call them "the poor," have many goods and services they they want to buy. And so they work, earn income, and buy them.

Total output and employment fall because the rich don't want new homes and there is nothing else they want. So they work less and earn less income. Meanwhile, the people who face scarcity continue to work, earn income, and purchase goods and services. Once "the rich" discover some new good they want to buy, they will go back to work, and earn income to buy it. And firms will hire "the rich," so that they can produce this new good desired by "the rich," as well as continuing to produce the goods the poor continue to buy.

Unfortunately, this still appears inconsistent with people who face scarcity being unable to find work and earn income to buy the things they want. Of course, those "poor" people who produced houses no longer have customers. But those rich people who were producing goods for the poor no longer want to work. So the firms producing those goods have less labor. There is a need to redeploy resources. "The poor," who used to provide labor to produce new homes for "the rich," must shift to producing goods for "the poor." "The rich" reduce how much they work and how much income they earn, because they don't have anything they want to buy.

But, of course, I am imagining that people who have nothing they want to buy have no interest in earning income, and so they work less. Suppose that instead they save. They continue to work and earn income, saving that money so that they will be able to purchase these new goods and services that they might want to buy if and when they appear.

Note that this version of recalculation is very similar to naive Keynesian economics. Add the notion that the reason people save more is because they can't think of anything to buy, then we have an increase in saving, less spending on consumer goods and services, and so less derived demand for labor. The supply of labor isn't reduced. And so, a surplus of labor.

The approach is also very similar to Marxism. There are unemployed people who need goods and services. But they have no income. The income is being earned by "the rich" who have nothing they want to buy. They simply save the money. While coming up with some kind of new luxury good for the rich to consume might help, an alternative is to redistribute income to those who need it. They will spend the money they receive, increase the demand for goods, and create more employment.

But increased saving should allow for additional investment--spending on new capital goods. These capital goods will allow for the production of additional consumer goods in the future. However, in the recalculation story, no one knows what consumer goods will be demanded in the future, and so what capital goods should be purchased. In fact, "the rich" were purchasing new homes as a speculative investment. Once they no longer want to buy homes for that purpose, and there is nothing else for them to buy, they will continue to save, without there being any investment.

If people want to save more than they want to invest, interest rates should fall. And, in fact, during the current crisis "the" interest rate has fallen nearly to zero. But apparently, this lower interest rate has not motivated "the rich" to consume more. They have nothing more to buy. And the lower interest rate hasn't motivated them to purchase capital goods. Because they don't know what capital goods are appropriate.

But how can people save without there being matching investment? They must accumulate money balances. And so, "the rich" want to earn income, and save it, in order to be able to purchase new consumer goods in the future that might be developed. They don't want to invest in capital goods until they know what consumer goods that people like them will want to buy and so what particular capital goods are appropriate.

And so, they work and earn income, and save by accumulating money. They aren't accumulating money because there is a certain amount of real balances they want, and once they have accumulated those, they can go back to purchasing one of the many goods and services they have been sacrificing in order to build up those money balances. They aren't sacrificing goods and services to add to money balances. They are accumulating them in case some new good arrives that they will want to buy.

I don't want to claim that no one could possibly behave in this fashion. I called them "the rich," for a reason. To me, it is only the very rich that might be in this situation where they have everything they want to buy but rather than enjoy leisure continue to earn in order to be able to buy something that they might want in the future. Because incomes are so skewed, even if there are a small number of people like this, this behavior could conceivably have a significant effect on the economy.

As explained above, the argument is Keynesian in that it suggests there is too much saving. It is Marxist in that their appears to be a problem of distribution. (I am dismissing the possibility that anyone would be so blind to the reality that the vast majority of people have an entire list of goods and services that they would buy if they had additional income.)

I think it is obvious that having people who face scarcity--want to work to earn income to purchase goods and services--remain unemployed because other people want to work to accumulate money because they might want to buy something in the future is unacceptable. Making those facing scarcity, "the poor," pause until someone comes up with something "the rich" want to buy is a coordination failure.

It seems to me that the solution is that "the rich" need to work less. There is no point to saving if it isn't funding investment. If people don't want to invest, then they shouldn't save. If people are only working in order to save, then if they save less, they have no reason to work. And, so, if the market system is going to provide proper coordination, there must be a signal for these sorts of people to work less in this situation.

In my view, the market signal is obvious. The real interest rate from holding money must turn sufficiently negative to clear markets. Those who want to work now and save because they might want to buy something in the future, and they don't want to invest in capital goods because no one knows what particular capital goods are appropriate, need to pay in order to save. And if they don't want to pay, then they should work less.

In conclusion, redeployment of resources, including labor, is something that happens all the time. The notion that it always happens at a constant rate is implausible. It seems likely that the aftermath of the housing bubble requires more redeployment of resources than usual. This might raise unemployment and depress output for some time. However, redeployment means redeployment to somewhere else. There are plenty of scarce goods and services to produce, and so, there should be sectors of the economy with rising prices and profits, and rising production and employment.

The notion that it is somehow necessary to "pause" and have some industries shrink, nothing expand, and just wait until someone figures out something to do, is inconsistent with scarcity. With certain Marxist assumptions (the rich don't face scarcity) and Keynesian assumptions (people want to save but no one wants to invest because of uncertainty,) it is possible to explain such a "pause." But in the end, it always comes down to monetary disequilibrium.

Cowen on Say's Law

Tyler Cowen argues that even in aggregate demand is too low, supply still matters.

In a world of unemployed resources, imagine a company called Apple invents a device called -- improbably -- an iPad. That's a positive supply shock. People will be lead to spend more money. The inventors and their employees will have more money to spend. There will be a positive multiplier throughout the economy, as analyzed by W.H. Hutt. First aggregate supply went up and then aggregate demand went up. It's all one step on the way to economic recovery.

Why don't the people buying the ipads spend more on ipads and less on other goods? Naturally, one would think they would spend less on substitutes for ipads.

From a flow perspective, Cowen is assuming that the ipad consumers are choosing to reduce saving. This raises the natural interest rate. If the market interest rate is unchanged, say because of central bank policy, nominal and real expenditure expands. This expansion ends if and when higher current income raises saving again and so the natural interest rate equals the market interest rate.

From a stock perspective Cowen is assuming that the ipad consumers choose to hold less money. Given the quantity of money, nominal and real expenditure expands. This process ends when real income rises enough so that the demand to hold money again rises to the existing quantity of money.

The flow and stock perspectives are consistent, though perhaps the flow approach is more appropriate if the central bank adjusts the quantity of money to target market interest rates. In both, nominal expenditure and aggregate demand depend on the quantity of money and the demand to hold it.

Perhaps it is sensible to assume that the introduction of a new good will cause a decrease in the demand for money. And so, it will raise aggregate demand.

On the other hand, surely most of the effect will be a shift in the composition of consumption and unchanged desired money holdings.

Ignoring this immediate effect, consider the multiplier Cowen describes. What I always call the "Yeager" effect becomes important. The increase in ipad production raises output. Assuming money is a normal good, the increase in income raises money demand. If the quantity of money is unchanged, the resulting shortage of money results in people reducing nominal expenditures. If this Hutt/Keynes multiplier results in those with greater income spending more, and additional production in response, the demand for money is rising at each step. In other words, this process is impossible.

The true "multiplier" effect is just the fundamental proposition of monetary theory. Individuals with excess money balances spend them and people accept them in payment not because they want to hold more money, but in order to spend them in turn. The process ends when demand for money adjusts to the quantity.

If the demand for money changes, and the quantity of money is unchanged, the fundamental proposition of monetary theory suggests that one person adjusts spending and then another, until the demand for money returns to its previous level.

If there is insufficient aggregate demand, a increase in supplies of goods and services that causes lower prices will increase the real supply of money, which raises aggregate demand. And, I don't disagree that the introduction of a new good would plausibly cause a substitution away from many other things, and that one of them is money, and so it would help a bit.

But monetary disequilibrium-- the quantity of money and the demand to hold money--is central to understanding inadequate aggregate demand.

Malinvestment and the Liquidity Effect

Austrian theorists typically downplay or dismiss the "liquidity effect" of monetary disequilibrium on market interest rates. Instead, they champion a loanable funds explanation of interest rates, with an excess supply of money impacting interest rates by creating a matching increase in the supply of credit. This is sometimes called the "injection effect" of money on credit and interest rates.

New Keynesian economists, on the other hand, make the liquidity effect of money on interest rates central to their very understanding of money. It seems that the dominant perspective somehow identifies money with the short term interest rates set by central banks.

This divergence in perspective hit home recently due to two books I am reading. The difference is obvious in the introductions in Steve Horwitz's Microfoundations and Macroeconomics: An Austrian Perspective, and Michael Woodford's Interest and Prices.

Horwitz criticizes IS-LM analysis in a way that I found too dismissive of the liquidity effect of monetary disequilibrium. Similarly, I recently read a chapter by Roger Garrison explaining the "loanable funds" approach to interest. While he mentions that saved funds might not be lent out, his approach is to give reasons why the hoarding of money should be minimal.

Woodford, on the other hand, defends his modeling strategy of ignoring the quantity of money and just treating Federal Reserve policy as setting short term interest rates. He mentions that a money demand function can be appended to the model, and that would determine the quantity of money. Presumably, the model "determines" the quantity of money that creates whatever liquidity effect necessary to keep the short term interest rate where the central bank wants it.

Woodford also discusses cashless payments systems, and argues that if such a system actually develops, then the quantity of money will become meaningless and the only way a central bank could implement policy is through targeting the interest rate. He explains that a "channel" policy, where the central bank sets a lending rate and pays interest on reserve balances will keep short term rates in the channel between those rates, even if there is no money.

Cashless payments systems have been a special interest of mine, and I believe that they do have money. Woodford, however, seems to identify money as some asset or assets that will be subject to a liquidity effect. To him, "cashlessness" seems to mean the absence of an asset that people will hold even though its yield is less than the yield on other sorts of securities.

In my discussion of the disequilibrium version of the Austrian cycle theory, I described what I understand to be the injection effect. Given appropriate institutional assumptions, an excess supply of money will be matched by an increase in the real supply of credit. This increase in the real supply of credit will result in a lower market interest rate. I have asserted that this effect is ephemeral, because it only exists as long as there is an excess supply of money. And doesn't an excess supply of money mean that people are holding more money than they want to hold? How long can that last? People holding more money than they want to hold spend it.

Suppose that this injection effect is understood in the context of the liquidity effect. The liquidity effect also depends on institutional assumptions. The most important one is that the nominal interest rate that can be earned on money is "sticky." With conventional, tangible, hand-to-hand currency the nominal interest rate is stuck at zero. To the degree that issuing deposit money requires banks to hold vault cash made up of zero interest currency, or clearing balances with interest stuck at zero or some other rate (or even sticky) then some of the characteristics of the interest (or lack thereof) on currency will be partially transferred to the interest rate paid on deposits. If government regulations prohibit the payment of interest on various classes of deposits, and the "rents" provided by this anti-competitive regulation can only be dissipated through quality competition, an even greater stickiness is assured. And, of course, it is possible that competing private banks would offer deposits that pay interest that is only periodically adjusted. The stickiness of the interest rate paid on money could be a market phenomenon.

If there is some source of stickiness in the interest rate paid on money, then changes in market interest rates impact the opportunity cost of money and should be negatively related to the demand to hold money. If, on the other hand, the interest rates paid on monetary assets moved with other interest rates, then this effect would not exist. If the interest rates paid on money are sticky, then the interest rate elasticity of the demand for money could be significant in the short run, but then dissipate over time.

Even if the demand for money was perfectly inelastic with respect to interest rates, perhaps because the interest rates paid on money adjusted instantaneously, there would still be an injection effect of an excess supply of money. However, suppose that the interest rates on money are sticky, so that as the increase in the real supply of credit reduces market interest rates, it also reduces the opportunity cost of holding money. The amount of money demanded rises to match the expanded supply. People are satisfied with the existing quantity of money. The only problem is that the market interest rate is below the natural interest rate. That is, saving is less than investment.

In an earlier post I discussed the liquidity effect and explained my view that an excess supply of money exists when the quantity of money is greater than what the demand for money would be given the existing interest rate paid on money and a market interest rate is equal to the natural interest rate. That market interest rates may be distorted so that they do not properly coordinate saving and investment is one of symptoms and negative consequences of monetary disequilibrium.

If the institutional conditions hold for both an injection effect and a liquidity effect, it seems plausible that a persistent effect on market interest rates is possible. Only as lower market interest rates begin to impact savings (and so consumption) and investment, does the monetary disequilibrium begin to impact aggregate demand for final goods and services. And, of course, this plausibly impacts the composition of demand, in the end, based upon the interest elasticities of demand for various goods and services.

Perhaps this is just a situation where I have unusual intuitions, but the idea that people are simply holding more money than they want seems unlikely to last long. People will spend the money. But that people may take time to respond to changes in interest rates is much more plausible. And at least one requirement for malinvestment to be a problem is for monetary disequilibrium to have a persistent impact on market interest rates.