A Primer for Investors

Malcolm Mitchell - September 2003

(Click here for the full version [56 KB] in Microsoft Word)

John Griswold of Commonfund supported the research for this article, for which I am deeply grateful. The finished piece is far better than it would have been without Stephen A. Meyer's patient and careful reading, his friendly debate, and the many useful suggestions he made as the text was taking shape. I am greatly in his debt. I hasten to add that all opinions and arguments presented here are my responsibility alone.

Four Points of Clarification
The Long View
Is There Good Deflation?
The Dangers of Deflation
Would the Proposed Remedies Work?
Bibliography and Notes


Thinking about deflation takes us into familiar territory: economic history, the psychology of consumers and producers, and the tangle of interest rates and money supply. We will find, however, that the usual landmarks are no longer reliable guides. Deflation presents fundamentally different concerns from the problems that investment professionals have dealt with over the past half-century, and the greatest danger lies in not recognizing the differences.

The 20th century experience leads us to believe that inflation is inevitable in growing economies. A longer view, however, shows that periods of price stability, which included interim bouts of deflation as well as inflation, have accompanied long periods of expansion.

Economic growth, we believe, is driven by entrepreneurial enthusiasm, which often becomes excessive. We rely on the Federal Reserve to control the inflationary excesses, without extinguishing the enthusiasm. Deflation, however, is likely to produce excessive caution, which, as Alan Greenspan puts it, is "not something with which modern central bankers have had much experience."(1)

The fact is that the sophisticated tools developed by the Federal Reserve to control inflation have never been used in a deflationary environment. The Fed is considered successful when it reduces the rate of inflation. What makes deflation so dangerous, I will argue, is that a general decline in prices, both actual and perceived, disrupts the economy regardless of the rate of decline. In fighting deflation, the only valid goal is a return to inflation.

Deflationary problems, in short, are not inflationary problems in reverse. Having said that, since investment professionals have experienced inflation and understand the efforts to contain it, the best way to describe the potential dangers of deflation is to differentiate them from the inflationary experience. It is one thing to keep bounds around the animal spirits necessary for economic growth, without crushing them. It is another thing entirely to try to wake up those spirits once they have gone into hibernation.

Four Points of Clarification

1. Economists disagree about how the general price level should be measured. Nonetheless, most studies of overall price changes still rely, as I do, on the familiar and arguably sound Consumer Price Index. For readers who want to delve into the debate, the current issue of The Journal of Economic Perspectives carries three helpful articles.(2) They explain, among other things, the distinctions between a "cost-of-living index" and a "cost-of-goods index," as well as the limitations of both.

2. My purpose is to identify the dangers that deflation poses and to weigh the proposed remedies, but not to analyze the causes. Readers can find various explanations for deflation, including Alan Greenspan's: "It now appears that we have learned that deflation, as well as inflation, are in the long run monetary phenomena, to extend Milton Friedman's famous dictum."(3) A dissent from the monetarist theory of price changes, as well as a theory of how demographics and other factors impact prices, appear in David Hackett Fischer's 1996 book, The Great Wave.(4) The point I take from Professor Fischer is simply that history warns us to expect periods of deflation, especially after long periods of inflation.

3. I don't claim to know whether deflation is imminent, and I doubt that anyone else knows. Despite recent declines in some commodity prices and in some leading indicators, the CPI has risen in nearly every month over the past three years, albeit more moderately than in previous decades. We can't assume that because price increases are moderating, they will soon turn into price declines. At a time of violent upheaval in the Middle East, with unknown consequences for the price of oil, it is folly for anyone to try to guess the short-term direction of prices in general.

4. Finally, readers will note that I make no reference in this paper to price declines that China, Hong Kong, and Japan have struggled with in recent decades. It's possible to tease out similarities and parallels between those economies and the U.S. The differences, however, both structural and circumstantial, are in my view far more significant than any similarities, and they make those economies only marginally relevant for determining what will or won't happen here.

The Long View

As children of the 20th century, we have reason to believe in the relentlessness of inflation. The Bureau of Labor Statistics' Consumer Price Index rose from 9.8 in January 1913 (its earliest measurement) to 181.7 in January 2003. It rose more than 30% from 1913 to 1933, despite the collapse of agricultural prices in the 1920s and the depression of the early 1930s. Since 1933, the Index has risen in 67 of 70 years.(5)

Given this experience, we tend to see all of economic history against the backdrop of inflation. It may come as a shock, therefore, to learn that when Adam Smith published The Wealth of Nations, in 1776, neither "inflation" nor "deflation" were yet included in the vocabulary of economic discussion. The terms had long had a physical meaning -- for example, a body swollen with air -- and a psychological meaning -- as in the phrase, His pride was deflated. But inflation would not be used to describe an economic condition until 1838, and even then only as inflation of a currency, or inflation of credit. Deflation, remarkably, would not be used in a modern economic sense until 1920, when a newspaper in Glasgow (where Smith had studied 180 years earlier) referred to "The transition from an inflationary to a deflationary period in prices."(6)

The fact that prices fluctuate hardly escaped Adam Smith's notice. He postulated a "natural price" of commodities, and he argued that "market prices" -- while sometimes "a good deal above [the natural price]" and sometimes "even somewhat below it" -- are "constantly tending toward it."(7) What Smith was describing was in fact the history of prices in England during the century preceding his great work. In the mid-1700s, prices were pretty much what they had been in the 1660s, despite interim spans of what we would now call inflation and deflation.

Alan Greenspan noted a similar but later historical reality in his remarks to the Economic Club of New York in December 2002:
    Although the gold standard could hardly be portrayed as having produced a period of price tranquility, it was the case that the price level in 1929 was not much different, on net, from what it had been in 1800.(8)
David Hackett Fischer, the first historian to systematically mine 800 years of global price data, refers to the period preceding The Wealth of Nations as The Equilibrium of the Enlightenment. Wheat prices throughout Europe "fell sharply after 1661, then fluctuated on a fixed level [until] 1730."(9) As American products entered global trade, their prices followed a similar pattern. The price of Maryland tobacco dropped sharply in the 1660s and then fluctuated around a central range until 1738.(10)

As for Greenspan's example, Fischer calls that period The Victorian Equilibrium, although his dates -- from 1820 to 1896 -- differ from Greenspan's. In fact, both sets of dates are valid. Prices in 1896 were actually lower than in 1820, so the rise from 1896 to 1929 brought prices back to where they had been, as Greenspan says, over a century earlier.

The important truth here is that prices in the U.S. began and ended the 19th century at roughly the same level. An upward spike occurred during the civil war -- prices doubled between 1861 and 1865 -- but it was erased by falling prices over the following 15 years. Why is this important? Because it shows that while the U.S. was a vastly different place in 1929 from what it had been in 1800 -- vastly greater in terms of total wealth, population, land settlement, industry, technology, transportation, gross national product, etc. -- long-term inflation neither played a role in, nor was a necessary byproduct of, that long-term growth and development.

Looking at the 800 years of human history for which he studied contemporary price data, Fischer concluded that "the price revolution of the twentieth century was the fourth great wave of rising prices in world history."(11) The four periods of inflation varied in length (from 80 to 135 years) and in the extent of price increases. They were separated by periods of equilibrium, which also varied in length. Within the periods of equilibrium there were interim price fluctuations, but each period ended with prices at roughly the same level as at the beginning. In total, Fischer's inflationary periods lasted 415 years. His periods of equilibrium, during which prices fell overall as much as they rose, totaled 385 years.

Is There Good Deflation?

My focus in this paper is on the kind of deflation that most Fed officials contemplate when they describe how the Fed would fight it -- that is, a general condition of falling prices for most goods and services throughout the economy, accompanied by economic players' expectations that the condition will persist. I said above that given this focus I plan to ignore questions of how and why deflation might arise. Regardless of the reasons why prices are falling, the Fed has only one set of tools to fight deflation. It is the same set the Fed uses to fight inflation, regardless of what is causing prices to rise.

Nonetheless, many economists insist that the causes of deflation are an important consideration, since it is possible to describe scenarios in which prices are falling yet corporate profits and jobs continue to grow. In such scenarios, falling prices reflect lower costs of production, usually resulting from technological innovations and other sources of increased productivity. This scenario is said to represent "good deflation." On one view, indeed, "The history of capitalism is the history of falling real costs," where real cost is measured as the hours of labor needed to buy the product or service.(12)

All of this may be true, but the potential deflation scenario that concerns the Fed would be signaled not only by a falling CPI -- a condition that, as I pointed out, we are not now experiencing -- but also by falling profits overall and declining employment. Such a situation would be the reverse of the moderate inflation I'll describe below, which may generate growing corporate profits and expanding employment, as well as an increasing CPI.

In inflationary times, some prices usually fall, for both "good" and bad reasons, but they are considered aberrations. An inflationary environment is one in which the sense of the nation and the expectations of economic players are that prices are rising. Wages are expected to "keep up" in every region of the country and in every industry.

The deflation that is feared, and for which I explain below the logic behind the fear, occurs when the sense of the nation and the expectations of economic players are reversed. Most prices -- of products, of services, of labor, of assets -- are falling and are expected to continue to fall. If future data suggest falling prices in general, it will not be good deflation in any sense, for the reasons I lay out in the following section.

The Dangers of Deflation

I propose to examine what would happen under conditions of deflation, a circumstance of which we have no experience, by examining inflation, of which we have had a great deal of experience in the 20th century. Most economists have concluded from that experience that a little bit of inflation isn't bad and even benefits the economy.(13) I plan to look at the assumptions about the psychology of economic players that support that conclusion. This will give us the context in which to consider the psychology of deflation.

Defenders of moderate inflation argue that producers, consumers, lenders, and borrowers can all make allowance, in the course of their normal economic activities, for moderate and reasonably consistent price increases. Let's consider how economic decisions are made under such conditions. How do the competitive, entrepreneurial spirits that propel economies in the first place respond to the expectation of low and stable inflation?

The answer for producers, who are trying to maximize profits, is that they will push the upper limits of expected inflation. If the immediate rate of inflation suggests that I should increase the price of my product by 1%, I will want to see if I can get away with increasing it by 1.01%. If I fail and my sales falter, I have two good responses: I can offer temporary price reductions in order to increase sales; or I can hold my inventory for a longer period until prices rise to mine. If my extra price increase succeeds, however, my profits will increase (by more than my competitors', if they didn't raise their prices by the same amount), and I will be able to expand production.

For both lenders and borrowers, steady, moderate inflation can readily be incorporated into the setting of interest rates. Borrowers accept the rates when they believe they will pay back a fixed loan out of increased revenues (whether salaries or sales). Borrowers actually like a bit of inflation because they repay loans in "cheaper" money, that is, in money that buys fewer goods at the time of repayment than when it was borrowed. The appreciation of homes that results from inflation encourages homeowners to borrow, since the money they pay back will buy less house than the money they borrowed.

Lenders fear inflation for the same reasons, and so they look for higher interest rates, including higher mortgage rates, to compensate them for accepting repayment in cheaper money. However, lenders also find comfort in the fact that repayment is made more certain by the expected rise in the salaries or revenues of borrowers, as well as in the value of assets purchased with the loans.

Finally, consumers are inclined to buy today, because they expect that inflation will raise the prices of goods tomorrow. At the same time, they can expect their future purchasing power to be maintained because salaries will increase. Buying on credit makes sense in inflationary periods, for the same reason that borrowers like inflation. Temporary price reductions work with consumers, because they are seen to be really temporary.

Let me repeat that the above scenarios depend upon economic players believing 1) that the general price level is rising, 2) that the Federal Reserve will act if and when the rate of price increases threatens to get out of hand, and 3) that the Federal Reserve does have the tools to adjust the rate of price increases to acceptable levels. If all this is true, controlled inflation can drive an expanding economy. (We'll discuss the tools available to the Federal Reserve in the next section.)

Let's now turn the inflation scenario on its head. Let's ask what happens to the economy's animal spirits if instead of expecting controlled inflation, economic players face deflation. Can those spirits flourish in a period of general price declines? Is controlled deflation a possibility -- or an oxymoron?

For consumers, the response to expected persistent deflation is to delay purchases, not speed them up, since they will be able to buy the goods cheaper if they wait. What's more, they have to expect their own salaries to drop, or at best to remain flat, and they are thus inclined to protect their assets, rather than spend them.

Borrowing to make purchases does not make sense in a deflationary period, since repayment will be in money that will buy more than it did at the time of borrowing. A home-buyer who borrows to buy a three-bedroom home will repay the loan over a period when the amount borrowed might buy four or five bedrooms. Those people with large amounts of existing debt, relative to their wealth and income, are particularly vulnerable to deflation.

As for the relation between lenders and borrowers, deflation eliminates the one activity they undertake happily together under inflationary conditions -- setting mutually acceptable interest rates. At any reasonable level of expected inflation, or even if the level increases, an interest rate can be found that will satisfy the objectives of both borrower and lender and enable the loan to be made. Deflationary expectations, in contrast, drive interest rates down, and there is a lower limit -- zero. At zero interest, lenders have no incentive to take on any risk at all and will not lend. At any interest rate above zero, lenders conclude that they are guaranteed to make money, and borrowers conclude that they are guaranteed to lose money.

I'd like to elaborate on this last point. Under inflationary conditions, borrowers and lenders find a rate of interest that both believe will compensate them for the expected change in the value of the money between the time it is loaned and the time it is paid back. One or the other may turn out to be wrong, depending upon the actual change in the value of money over the period. Borrowers may lose by paying too high a rate of interest and wind up paying more in total for an asset than the asset becomes worth. But borrowers under inflationary conditions may also win. They hope to, and may in fact, pay less in total for an asset than the asset becomes worth at the time the loan is paid off.

Deflation ensures that borrowers must always lose. At any level of deflation, an asset becomes worth less in the future than it was worth at the time of purchase. If the buyer borrows to make the purchase, the interest paid increases the buyer's loss, dollar for dollar. Under deflationary expectations, borrowers always repay, and lenders always receive, a total amount of money that will buy more goods than it did at the time of the loan.

I am not suggesting that all borrowing would cease with deflation, but that given any other possibility for raising necessary funding, economic players would prefer not to borrow. Under inflationary conditions a real benefit accrues to a borrower if actual inflation turns out to be greater than expected inflation. Under conditions of deflation any potential benefit to the borrower disappears absolutely.

Returning to producers, their deflationary problem is that there is no analogy to pushing the limits of inflation. A lower sales price, lower even than the price dictated by expected deflation, may capture market share, if consumers are buying at all. However, if the price reduction still leaves unsold inventory, the producer can only look forward to letting it go at still lower prices, with correspondingly lower profits.

It seems clear that corporate profits can only be squeezed by deflation, and corresponding stock prices can only fall. Reducing costs to bolster profits involves either paying less for supplies, which exacerbates deflation for suppliers, or reducing wages, which lowers aggregate consumer demand and pushes prices further down.

The important exception to all this is when new products, new technologies, and new services are introduced. These give producers a better chance to flourish under any price conditions -- provided again that consumers are willing to buy at the prices asked, and that they have the cash to do so.

This comparison of the psychologies of inflationary and deflationary expectations leads to the conclusion that under conditions of actual and expected deflation, the economy will not function in any manner familiar to us. The degree of actual deflation -- and whether it's getting "better" or "worse" -- is irrelevant. Prices that are falling across the board and are expected to keep falling, however modestly, alter every thought process and every relationship in the economy. The critical problem of deflation is that once it takes hold, the only solution economic players will look for from Federal Reserve activity is to reverse it and bring back inflation. Falling prices must have a stultifying effect on the spirits of the economy, which cannot be erased until prices in fact begin to rise again.

Would the Proposed Remedies Work?

Since I've drawn some conclusions about the modern world from the history of inflation and price stability in past centuries, I ought to point out two crucial ways in which our modern economy differs from those distant times. That will lead us to a discussion of the tools proposed to fight deflation.

First, the price of "rents" throughout human history is of major importance for Fischer's historical analysis, as it was for Adam Smith's analysis of how a nation's wealth grows. Until the 20th century, much of humanity paid an annual rent to the owner of the ground they lived on, and paid it with the production of that land. Thus important indications of price changes always included the price of renting and the price of the produce, most importantly wheat. Today, the prices of wheat and other consumables are still direct components of general price estimates. But the housing component's impact is derived largely from the asset value of housing, reflecting widespread ownership of homes, and from its critical component, the cost of money.

The second change follows from the first. We tend to believe today, far more than economic players ever believed, that economies are driven by financial institutions. Most historical analyses of our economy place the Federal Reserve system in the dominant role. This is true despite the fact that the system was not established until 1913, its ability to use open market operations to control bank reserves wasn't realized until 1922, and those operations weren't consolidated into a single unit, the Federal Open Market Committee, until 1933.(14)

Everything we think we know about our central bank-dominated, fiat-money economy is based on less than 75 years of experience. They are, furthermore, the same years in which inflation was the principal bugbear. The Federal Reserve is believed to have a major impact on the economy, not only as the agent that controls inflation, but as a potential source of inflation as well. "Dozens of theoretical papers," Paul Krugman writes, "have argued that the temptation to engage in excessive money creation causes an inherent inflationary bias in fiat-money economies."(15) Ben Bernanke, a Governor of the Federal Reserve, bases the Fed's ability to control deflation on its ability to generate inflation: "Under a paper-money system, a determined government can always generate higher spending and hence positive inflation."(16)

Such confidence, however, as applied to conditions of falling prices, has no historical basis. The most that anyone can claim is that central banks seem to have learned how to control the rate of inflation by adjusting interest rates and the money supply. We therefore need to look at how the Fed's tools are used in an inflationary environment and the reasons why they are supposed to work. That will allow us to gauge how well they might work under the fundamentally different conditions of deflation described above.

The Fed's tools are usually divided into four categories, although we'll see that there is considerable overlap among the four:

  • the Fed can expand or contract what is called "the money supply";
  • it can raise or reduce interest rates;
  • it can affect the attitudes, that is the expectations, of economic players;
  • it can affect the exchange rate of the U.S. dollar against other currencies.

    I'm going to consider each of the four in turn. First, however, I need to explain my dissent from the monetarist view of how "the money supply" affects consumer spending. Specifically, I question the idea that when the Fed "monetizes" some government debt, its action somehow puts more money into consumers' hands.

    A widely-accepted monetarist concept, stated in simplest terms, is this: if the money that people have increases, they will increase their spending by some percentage of the increase of money. Furthermore, if spending increases but the output of goods and services does not, prices will rise. So far, these seem to me unassailable propositions.

    The confusion that I see arises over the meaning of people having "more money." For Keynes, who explained The Quantity Theory of Money in 1923, it was the value of the actual coins jingling in consumers' pockets, relative to the cost of some fixed basket of purchases, that determined prices. As more "money" was put into circulation, by getting consumers to carry more coins in their pockets, prices would tend to rise.

    As we'll see later in this paper, recent comments by Fed officials sound as though they are based on Keynes's description of the process. In my view, though, the idea that the cash in our collective pockets controls our spending seems merely quaint. Periodic variations in the amount of bills and coins normally in circulation bear no relation whatsoever to broad changes in consumer spending. The advent of credit cards, the development of brokers' "checking" accounts and "sweep" accounts, and many other financial innovations have fundamentally transformed the "money" that consumers spend. And the transformation will only accelerate, if internet business achieves even a portion of its claimed promise.

    Consumers today base their spending decisions not on the cash in their pockets, nor even on their normal bank accounts, but on a total sense of their spendable wealth -- that is, on that portion of their total wealth that they have not put aside in retirement or similar plans. Spendable wealth may include consumers' income, their bank accounts, their brokerage accounts, their short-term investments, and, on the scale where it pertains, their holdings of government debt. These are all forms in which, in today's financial systems, spendable wealth may be held.

    Note that all these forms of spendable wealth (except income) also exist in retirement funds or similar long-term accounts. In those funds, however, any change in the proportion of wealth held in various forms reflects an investment decision and does not impact consumer spending.

    Finally, the ease with which credit card debt is created or paid down by consumers themselves makes spendable wealth not a fixed but a flexible concept. It responds to consumers' changing ideas of their present and their future, and of course to the desireability of goods and services at the prices offered. It seems to me that this ability of consumers to define their own spendable wealth makes a hash of Fed attempts to influence consumer activity by adjusting "the money supply."

    I'll go back now to consider how the Fed's tools are supposed to affect the economy in an inflationary environment. First, the Fed's ability to expand or contract "the money supply" is based on its buying or selling Treasury bonds from or to private owners. When it does so, the Fed directly changes the aggregate balance in the private economy between bank deposits and Treasury bonds as forms of money. It should be clear that I believe this Fed action in itself does not change anyone's propensity to spend money and thus has little impact on inflationary pressures.

    In buying or selling Treasury bonds, however, the Fed also affects interest rates and the supply of credit that the banking system can offer potential borrowers. This ability to influence interest rates and the availability of credit is the Fed's second and most important tool for controlling inflation. As we saw, economic players can accommodate moderate expected inflation by finding an interest rate that both borrower and lender agree is satisfactory for their purposes. If, for whatever reason, they have trouble agreeing on a rate, and the economy begins to slow down, the Fed can step in and by trading in Treasury bonds -- or changing the Federal Funds rate, or by other means -- move interest rates to a level at which agreement returns and the economy picks up.

    Furthermore, it is possible that merely the expectation of these kinds of Fed intervention is sufficient to return interest rates and the inflation rate to a level at which economic players again take action in the economy. This is the third Fed tool, that is, to use the first two in such a way as to convince economic players that it will do whatever is necessary to achieve the desired result. In my example above, entrepreneurs will be less likely to push their price increases to the limits of inflation if they are convinced that the Fed will continue to act vigorously until it succeeds in reducing the rate of inflation. Thus activity in the private economy may bring about the desired result even before the Fed takes its next action.

    The last Fed tool is its ability to buy and sell non-dollar government bonds. Just as the Fed affects U.S. interest rates -- that is, the dollar price of money -- by buying or selling U.S. debt, so it can affect the price of dollars in other currencies by buying or selling foreign debt. By selling dollars to buy foreign debt, the Fed lowers the value of the dollar and thus makes exports cheaper. This is expected to increase U.S. production for export and in turn generate borrowing and spending by U.S. companies. A lower dollar also makes imports more expensive, which promotes price increases.

    This last tool has been more talked about as a potential counter to deflation than it has been used to fight inflation. And so we are led back to reconsider all of the Fed's tools in the context of potential deflation instead of inflation. How would economic players react to Fed activity in the various forms I described, if we postulate an economic environment in which prices have been falling and are expected to continue to fall?

    I'm going to take the Fed tools in the reverse order from the discussion above, for reasons that will become obvious. First, as to creating rising prices by devaluing the U.S. dollar, that is, buying foreign debt with dollars, it's fair to say that in today's world no one knows what results would follow. Would China retaliate with competitive devaluations to maintain export levels? Would the oil price rise? How much would U.S. import businesses suffer? How would the euro countries respond? The world will not treat dollar devaluations more kindly because we insist that they are only intended to counter U.S. deflation. Every country looks to its own economy and would see the U.S. action as simply a competitive devaluation, and a desperate one at that. It's hard to imagine a scenario in which the Fed could be confident of a successful result from intentionally devaluing the dollar.

    As for the Fed's ability to change deflationary expectations, I've argued that only the expectation of rising prices would have any impact. A smaller expected price decline is still a price decline, and it can't in any way be seen as a benefit for economic players. We saw above that, for example, producers have no fallback if they reduce prices less, thinking that the level of deflation will decrease, and turn out to be wrong. Since no historical evidence supports the Fed's claim that it can turn falling prices into rising prices, economic players would most likely wait to see the turn actually occur before responding to any Fed moves.

    What about the interest rate tool? We've seen that under conditions of actual and expected deflation, no positive interest rate will be seen as anything but a pure loss to borrowers and a pure gain to lenders. Even ongoing and successful businesses would prefer to finance their normal activities through internal cash flow and avoid borrowing. For this reason, giving banks "free" money, or even offering Fed loans directly to the private economy (both of which suggestions have been make by Fed experts), can't create the desired transaction if there are no willing borrowers.

    We need to turn, then, to the Fed's influence on "the money supply." Not surprisingly, we find that economists base their most emphatic statements about the Fed's deflation-fighting ability on this tool. Here is the statement by Governor Bernanke from which I quoted one sentence above:
      Like gold, U.S. dollars have value only to the extent that they are strictly limited in supply. But the U.S. government has a technology, called a printing press (or, today, its electronic equivalent), that allows it to produce as many U.S. dollars as it wishes at essentially no cost. By increasing the number of U.S. dollars in circulation, or even by credibly threatening to do so, the U.S. government can also reduce the value of a dollar in terms of goods and services, which is equivalent to raising the prices in dollars of those goods and services. We conclude that, under a paper-money system, a determined government can always generate higher spending and hence positive inflation.(17)
    This statement is puzzling. No matter how closely we read it, we can't be quite sure whether Bernanke refers literally to physical bills -- that is, cash -- or to some broader notion of dollars. We sense a fudge factor, which in similar arguments is often stated as "liquidity." Another Fed Governor recently wrote, "With prices actually falling, the central bank may have little choice but to expand liquidity quite rapidly in the hope that the liquidity will arrest deflation, even if nominal short rates remain stuck in the neighborhood of zero for a range of monetary growth rates."(18)

    So did Bernanke mean cash as liquidity, or some broader form of liquidity? What he says immediately following the above statement seems to rule out cash itself: "Of course, the U.S. government is not going to print money and distribute it willy-nilly (although as we will see later, there are practical policies that approximate this behavior)."(19)

    We discover a few pages later what he means by the "practical policy" that approximates printing and distributing cash:
      In practice, the effectiveness of anti-deflation policy could be significantly enhanced by cooperation between the monetary and fiscal authorities. A broad-based tax cut, for example, accommodated by a program of open-market purchases to alleviate any tendency for interest rates to increase, would almost certainly be an effective stimulant to consumption and hence to prices. . . . A money-financed tax cut is essentially equivalent to Milton Friedman's famous "helicopter drop" of money. [Italics mine.](20)
    Clearly Bernanke is not suggesting that the government print bills and somehow place them directly in citizens' pockets. (The argument has been made, however, following Keynes, that I will spend more money if I'm somehow convinced to keep extra cash in my pocket rather than leaving it in a bank account. While perhaps true at some level, it seems a fanciful scheme for fighting deflation, especially in an economy dominated by credit cards.)

    Instead, Bernanke suggests a three-step process that is "essentially equivalent" to dropping cash in people's hands. First, the government cuts taxes; second, the government borrows to cover the shortfall in revenue; and third, the Fed buys Treasury bonds, changing the aggregate balance in the economy between bank accounts and Treasury accounts, and also preventing the new borrowing from driving up interest rates.

    If we think about this scheme, it will quickly become clear that the only potential deflation-fighting step is the government borrowing. In the first place, the tax cut itself is no "stimulant to consumption," because the increase in what taxpayers keep is offset by the decrease in what government has left to spend. In fact, since consumers will hold back on spending under deflationary conditions, while government spends all the money it collects, the tax cut alone would probably reduce overall consumption.

    The follow-up to the tax cut -- borrowing by the government -- is the step that could create more spending. The tax cut leaves citizens with more to spend, if they will spend it, and the government by borrowing winds up spending the same amount.

    The third step, the Fed's purchase of Treasury bonds from the private economy, may affect interest rates but does nothing in itself to stimulate spending. A consumer's propensity to spend, as I argued above, is not affected in the modern economy by the form -- whether bank accounts or Treasury accounts -- in which private spendable wealth is held.

    Is it the Fed's buying Treasury bonds that keeps the lid on interest rates? Recall that we are considering a deflationary environment, in which interest rates are already approaching zero. Borrowers are reluctant to borrow, since they can only expect to pay back more in purchasing power than they borrowed. Lenders are certain to receive more in purchasing power than they lent, no matter how small the interest rate they accept. The result is that while the government's deficit spending can to some extent make up for the absence of borrowing by anyone else, it would seem unlikely in such a scenario for interest rates to tend to rise.

    What does this imply for the battle against deflation? It implies that the Fed in reality has no deflation-fighting tools. It is government spending over and above government receipts that forces some private wealth to be spent, yet leaves total private wealth unchanged and available for additional spending. This borrowing and spending, according to the classic formulation, is what generates increased consumption and ultimately inflation.

    It is also precisely what is supposed to result from corporate spending over and above corporate receipts, financed through bond sales or bank loans. Either form of borrowing adds debt to a corporate balance sheet but gives the corporation (and the total economy) more money to spend. The difference between corporate and government borrowing is that corporations have to increase profits in order to pay back loans and stay in business, whereas the government, with its vast revenue-generating capability, can't go out of business.

    This approach to fighting deflation ignores the distributional aspects of government borrowing, spending, and ultimately paying back lenders with tax money that is worth more in purchasing power than the money borrowed. Nonetheless, whatever the justice of the approach, government borrowing can be claimed to be a worthwhile tradeoff if the economy is mired in deflation, and no one else is borrowing.


    While I've argued that the Fed's tools, which many claim will work against deflation, won't, I should add to be fair that every commentator also says the best way to fight deflation is not to let it get started in the first place. That means working to maintain moderate inflation, which suggests deficit spending by the government when corporate borrowing and spending dries up. It may be that the larger debate -- whether the government should aim primarily to fight inflation or to fight unemployment -- is the relevant debate now that the possibility of deflation seems real.

    If it turns out that deflation arises from causes that even the government's deficit spending can do nothing about, we are left with one piece of good news and one piece of bad news. The bad news is that moderate deflation, unlike moderate inflation, is something the economy cannot adjust to. Everything in the economy -- employment, production, corporate profits, and investment returns -- will simply sag. The only hope will be that substantial changes in the larger social structure, or in the global economy, will by themselves finally bring about rising prices.

    The good news, though, is that deflation can't drive prices down to zero. Unlike runaway inflation, which destroys all financial structures and leaves only a barter economy, deflation has some limits. The structures that ultimately lead back to rising prices remain in place, and the government can maintain its role as the guarantor of social stability and material well-being. If history teaches us to expect some periods of falling prices, it also teaches that those periods are limited and that rising prices will return.

    On the other hand, a bout of even moderate deflation in the modern world would probably force economists to reassess their basic principles. Whether that's good news or bad news depends on your point of view.


    Ben S. Bernanke, "Deflation: Making Sure 'It' Doesn't Happen Here," Remarks before the National Economists Club, Washington, D.C., November 21, 2002. Available at

    The Consumer Price Index (CPI) is maintained by the Bureau of Labor Statistics, a unit of the U.S. Department of Labor. Monthly index levels dating to 1913 are available at

    Barry Eichengreen, "Decision-Making in Europe's Central Bank: A Cautionary Tale from the Early Years of the Federal Reserve," Investment Policy magazine, Jan/Feb 1999. Available at

    David Hackett Fischer, The Great Wave: Price Revolutions and the Rhythm of History, Oxford University Press, New York, 1996.

    Edward M. Gramlich, "Conducting Monetary Policy," Remarks at a joint meeting of the North American Economic and Finance Association and the Allied Social Science Association, Washington, D.C., January 4, 2003. Available at

    Alan Greenspan, "Issues for Monetary Policy," Remarks before the Economic Club of New York, New York City, December 19, 2002. Available at

    John Maynard Keynes, Monetary Reform, Harcourt, Brace and Company, New York, 1924.

    Paul Krugman, "Can Deflation Be Prevented?" September 2002. Available at

    Laurence Siegel, "Inflation and Investing: An Overview," Investment Policy magazine, July/August 1977. Available at

    Adam Smith, The Wealth of Nations, Everyman's Library, Alfred A. Knopf, New York, 1991.

    "The Spectre of Inflation," various articles in Investment Policy magazine, July/August 1997. Available at


    1. Greenspan, p. 6
    2. The Journal of Economic Perspectives, Vol. 17, No. 1, Winter 2003,
    pp. 3-58.
    3. Greenspan, p. 1.
    4. For my review of The Great Wave, see Investment Policy, Vol. 1, No. 1.
    5. See the CPI web site.
    6. OED, 2nd edition, Vol. VII, p. 937.
    7. Smith, The Wealth of Nations, p. 51.
    8. Greenspan, "Remarks," p. 1.
    9. Fischer, p. 102.
    10. Ibid., p. 103.
    11. Ibid., p. 179.
    12. Siegel, pp. 15-30.
    13. For example, Bernanke, p. 4: "The Fed should try to preserve a buffer zone for the inflation rate, that is, during normal times it should not try to push inflation down all the way to zero. Most central banks seem to understand the need for a buffer zone. For example, central banks with explicit inflation targets almost invariably set their target for inflation above zero, generally between 1 and 3 percent per year. Maintaining an inflation buffer zone reduces the risk that a large, unanticipated drop in aggregate demand will drive the economy far enough into deflationary territory to lower the nominal interest rate to zero. Of course, this benefit of having a buffer zone for inflation must be weighed against the costs associated with allowing a higher inflation rate in normal times."
    14. Eichengreen, pp. 37-44.
    15. Krugman, p. 1.
    16. Bernanke, p. 6.
    17. Ibid.
    18. Gramlich, p. 4.
    19. Bernanke, p. 6.
    20. Ibid., p. 9.

  •    For comments and questions,
       please e-mail the editor at