This excerpt was originally published in Money For Nothing, pages 71-99, London: Nicholas Brealey Publishing Ltd, 2003.
On the eve of the Great Crash any investors who had been worried by the musings of Roger W. Babson had received considerable reassurance from an altogether more distinguished source, Irving Fisher: mathematician, inventor, eugenicst, health enthusiast, crusader, public figure extraordinaire, and professor of economics at Yale University; indeed, one of the greatest economists of the twentieth century. In 1929 he opined: "Stock prices have reached what looks like a permanently high plateau." Then the stock market plunged -- taking a good part of Fisher's reputation with it. What made it worse was that Fisher stayed optimistic. In 1931 he praised President Herbert Hoover for his "calm reassurances to business."3
But when thinking about economics of the Great Depression Fisher was on the ball. He even came up with a new theory: debt deflation. The key problem was excessive indebtedness and how this interacted with falling prices. In essence, by cutting back their spending to repay their debts people would cause prices to fall, thereby increasing the real value of what they owed. As Fisher put it:
The very effort of individuals to lessen their burden of debts increases it, because of the mass effect of the stampede to liquidate in swelling each dollar owed. Then we have the great paradox which, I submit, is the chief secret of most, if not all, great depressions: the more debtors pay, the more they owe.4
Mind you, this was a bit late for his own finances. Fisher had made a fortune out of inventing a card index system (now called a RolodexTM) and in 1925 his firm, which held the patent, merged with his main competitor to form Remington Rand and then Sperry Rand. His optimism on the eve of the crash cost him dear. He is estimated to have lost $10 million and thereafter he was never out of debt, much of it to his sister-in-law, who bailed him out. So great was his predicament that in order to save him from eviction, Yale University had to buy his house in New Haven and rent it back to him.
Perhaps this is one of the reasons why the theory of debt deflation, which Fisher did so much to pioneer, was pretty much neglected for so long -- until the experience of the 1990s, notably in Japan, breathed new life into it and highlighted the hazards of falling prices.
Why Worry about the D-Word?
In order to be sure about the harm that deflation may do to incomes and living standards, we first need to be careful about what we mean by the term. Falling output, which is often referred to as deflation, obviously harms incomes and living standards, but that is not the same thing as falling prices, even though the two do sometimes go together. What I mean by deflation is falling consumer prices -- and not just some prices. At low rates of inflation there are always some prices that fall. Over recent years we have become accustomed to the prices of electrical goods, telecommunications equipment, and even clothing and footwear falling year after year. However, because other prices have carried on rising, the overall level of prices has continued to increase. This is not deflation.
Deflation occurs when the general level of prices falls. In the extreme form prices of everything fall: goods and services, shares, houses, and labor. This is what happened in many countries in the 1930s. It is what happened in the 1990s in Japan. And it could happen in much of the world right now.
As Figure 3.1 shows, in Japan the recent downward trend of consumer prices has meant that the consumer price level in 2005 is the same as it was 12 years earlier. And other prices that are traditionally more flexible and respond more quickly to shifts in demand -- the prices paid to Japanese producers, as well as the prices of land, houses, and shares -- have fallen much faster. This is deflation.
Should it be feared? In principle, deflation may be perfectly all right. In 1955, the US price index briefly dipped into negative territory, with the annual rate of deflation reaching 0.7 percent, without serious ill effects. And for much of the nineteenth century deflation was endemic but apparently benign. From 1812 to 1896, producer prices fell on average by 2 1/2 percent per annum and consumer prices by 1.8 percent per annum. And there were some spectacular examples of falling prices. For instance, the introduction of new glass-making machines allowed the price of a dozen glass goblets to fall from $3.50 in 1864 to 40 cents in 1888. Yet real GDP increased at an average rate of 4 1/2 percent, 2 percent in per capita terms.5
Interestingly, this period of prosperous deflation saw the same combination of technological advances and increased globalization that is evident today.
[Figure 3.2 UK and US price level, 1790-1939 (1800=100)]
This has led some commentators to draw a distinction between "bad deflation" and "good deflation." "Bad deflation" occurs as a result of asset-price collapses or bad monetary policy. "Good deflation" occurs as a result of increases in productivity growth or reduced costs stemming from an increase in international trade. Accordingly, recognizing these favorable factors at work now, these commentators are able to be optimistic about the future. If Europe and North America experience deflation, they say, it will be the "good" sort.
I find this analysis unconvincing and unhelpful. What happens if depressed asset prices, poor monetary policy, strong productivity growth, and cheaper imports occur simultaneously? What happens, indeed, if "good deflation" sets up forces that lead to "bad deflation"? This distinction will not do. To get the measure of deflation, we have to get to grips with those two human factors dominating economic life: what is in people's heads and what is in their institutions.
Expectations and practices
Rather like inflation, if everything in the system is adjusted to it, deflation need not be bad at all. In fact, according to some distinguished economists of yore, including the celebrated Milton Friedman,6 a moderate rate of deflation may be positively desirable. His idea was that if the price level was stable, and still more so if it was rising, people would hold too little of their assets in notes and coin, because they pay no interest. After all, they would be holding this money for nothing in return. This was "suboptimal" because money is costless to produce but helps to save resources by obviating the mood for frequent trips to the bank. The solution was a regime where prices fell at the real rate of interest, say between 2 and 4 percent a year, thereby giving money holders due encouragement to optimize their money holdings.
OK, I admit it. In relation to the major issues at stake, this argument does rather seem like small change. It is the sort of debate that gives economists a bad name.
Older economists had seen different merits in a moderate rate of deflation. They argued that if prices fell at the rate of increase of productivity, this would impart stability to the wage structure. Rather than having to negotiate wages upwards all the time to benefit from the continual upward march of productivity was sit there and let falling prices do the job for them.7 This regime was referred to as "the productivity norm." Advocates of the productivity norm had a point but, as I shall show in a moment, when thinking about deflation there are surely other, more important considerations.
Modern economists and central bankers have tended to argue that the ills of deflation are rather like the ills of inflation; that is to say, the uncertainty over money values that confuses consumers about relative prices and increases the perceived level of risk to do with all transactions that are spread over time, as well as so-called menu costs; that is, the costs incurred in altering slot machines and printing price lists.
While this is more like it, it is surely not quite right. It is still too comfortable a vision of the evil that deflation can bring. In short, it is a picture of the effects of deflation when all of the financial and economic system is adjusted to it. But typically the system won't be adjusted to deflation; and it certainly isn't now. On the contrary, companies and private individuals have scarcely considered it at all when framing their decisions. Until recently, many of them would hardly have encountered the word deflation, let alone known what it meant.
Moreover, in marked contrast to the second half of the nineteenth century, deflation would come into a world riddled with debt, debt built up by households, companies, and governments with scarcely a moment's thought to the possibility that its real value could be driven up by a fall in the general level of prices.8
The result is that if, as I believe, some of the leading industrial countries do come to experience deflation, the effects will be profound. There will be a significant impact on corporate finances and levels of corporate failure, major effects on all asset classes, and it will cause disruption to the financial system, with serious implications for the solvency of financial institutions and the health of the real economy. In short, deflation poses a potent threat to the stability of the whole economic and financial system. Not good deflation, but deadly.
Deflation the destroyer
On the face of it, falling prices could leave companies pretty much unscathed -- and many economic commentators, particularly those of the market fundamentalist bent, still thick on the ground in the US, take a relatively relaxed attitude to the prospect of deflation. They are wrong. In practice, deflation would be a machine for the destruction of much of corporate America, not to mention the corporate sector of just about every other developed country.
Although the downward flexibility of wage and salary costs has increased in recent years, it is still low in relation to selling prices, particularly as more products seem to have become commoditized. The result is that deflation will hit profits severely and unless it is extremely mild or short-lived, it is bound to bring on a wave of insolvencies. In this way the deflation of prices can easily lead to the depression of real output and employment.
In addition, the balance sheets of most companies are currently structured to lose out from deflation. All companies carry some real assets in the form of fixed capital, land, buildings, or inventory, and the market value of all of these can be expected to fall in nominal terms under deflation. Apart from shareholders' funds, however, almost all of the liabilities will be fixed in nominal terms, either short-term bank borrowings or debt of various kinds. So typically the corporate balance sheet is unbalanced in relation to falling prices. The assets fall but the liabilities don't, with the result that the value of shareholders' equity falls.
Admittedly, if a company's debt finance is on variable interest rate terms there may be some relief. At least the interest rate on the debt can fall, reflecting the weaker trend of prices. Up to a point this means that a company holding real assets financed by fixed monetary liabilities with variable interest is hedged. After all, if the rate of change of the prices of the assets on the balance sheet moves from +2 to -2 percent but short-term interest rates fall by 4 percent, the company is no worse off.
However, there is a little problem about the number zero. Interest rates cannot fall below it, yet there is no such restraint on the rate at which prices can fall -- including asset prices.9 Thus real short rates are liable to rise in a period of deflation, thereby raising the cost of borrowing for individuals and the cost of finance for companies.10 This means that even variable-rate funding offers no protection against anything other than the mildest deflation. After tax the increase will be even greater because nominal interest payments are allowable against profits when computing a company's tax bill, but no such allowance is made for the extra financial burden implied by falling prices under conditions of deflation.
Furthermore, as I pointed out in Chapter 1, the trend of recent years in America has been for companies to increase the gearing on their balance sheets by issuing more debt and even buying in equity. Accordingly, companies have restructured their balance sheets in a way that makes them more vulnerable to deflation.
Sharing in the corporate pain
Despite the acute dangers of the corporate sector, the historical evidence from the interwar period suggests that equities could hold up well in a period of deflation. This evidence is highly misleading, however, since in those days equities were lowly valued before the onset of deflation. And, of course, Japanese equities have done appallingly badly during the deflation that has lasted from the beginning of the 1990s to today. The Nikkei Index peaked at just under 39,000 at the end of December 1989. In mid-2005 it had not quite regained the 12,000 mark.
As they represent the ownership of streams of income that are variable with respect to changes in the general level of prices, you might easily think that equities should respond directly to the rate of inflation/deflation. On the usual simplifying assumption that everything adjusts to the change in the inflation environment, the result falls out that the price behavior of equities should directly mirror movements of the general price level, leaving real equity valuations unchanged. In that case, as long as you think in real terms, everything is tickety-boo.
In practice, though, matters are unlikely to be quite that simple. How equities react to deflation is clearly intimately bound up with how the corporate sector responds to it -- and there, as I argued above, the answer is bound to be negative.
Moreover on past form, at first at least, bond yields are unlikely to fall sufficiently to reflect the full change in price prospects. Accordingly, with the onset of deflation real bond yields may rise. The markets use real bond yields to discount prospective future corporate earnings when setting the current value of equities. So higher real bond yields brought on by deflation would cause equity values to fall. (They would also temper what should otherwise be a favorable response of commercial property values to the onset of deflation.11)
Swings and roundabouts
You might think that since the balance sheet problems of companies represents a redistribution of wealth through a mismatch of assets and liabilities, other parts of the system -- notably the personal sector and the financial sector -- must be gainers. In a narrow sense that is right. Personal and financial-sector holdings of company debt will rise in real value as deflation proceeds.
The constituency of the gainers will extend much more widely throughout society to the holders of all monetary instruments, including deposits in banks and savings institutions. Admittedly, in times of very low interest rates brought on by deflation their interest income would fall to virtually zero, as it has in Japan, but with prices falling year after year who needs interest? The real value of savings increases at the rate of deflation.
In principle, the gains would offset the losses. In some cases, indeed, the gainers would be the same individuals as the losers. So where is the problem? It is simple and stark. Who are companies except proxies for their ultimate owners, individual shareholders? It is impossible for a society as a whole to gain by the corporate sector becoming worse off. But it can lose a very great deal if the extent of the corporate sector's problems is so acute that companies are forced to cut production, fail to service their debts, or go under.
Moreover, there is the little problem of the institutions that stand between the gainers and the losers, taking deposits from the one and making loans to the other. As borrowers fail to service their debts and cannot repay, while the assets against which the loans were secured fall in value, the solvency of these financial institutions is impaired. Eventually many would go bust.
The benefit of a high real rate of return on your savings is of no avail if the institution your savings are with goes bust. You will have saved all the money for nothing. In this way the pain of borrowers is transferred to savers, and the distress of both depresses consumer spending more, while the bankruptcy of some, and the feared bankruptcy of all, leading institutions causes the stock market to plunge still further, thereby destroying wealth and imperiling the solvency of other financial institutions, thus giving the deflationary spiral another twist.
Accordingly, the potential gain of the financial sector from deflation is a mirage. Worse than that, the financial sector ultimately becomes the repository of much of the pain caused by deflation, in the form of missed interest payments and bad debts. As the Japanese experience confirms, in deflationary conditions the financial sector is likely to perform particularly badly.
Furthermore, in anything other than a completely benign form of deflation where the rate was modest and it was fully anticipated, severe dislocation in the financial sector would be likely to restrict the supply of credit and contribute to a collapse of confidence and a deterioration of credit-worthiness throughout the economy. Needless to say, all of this would be decidedly negative for equities.12 Besides these calamitous consequences, the fact that Joe Soap's bank deposit would be worth more in real terms pales into insignificance.
A Cautionary Tale
This account of the ills that deflation can cause is not a fairy tale. Nor is it even a historical tale of the 1930s. On the contrary, it is a fair description of the deflationary process in modern-day Japan. After the collapse of the bubble economy, the 1990s saw a wave of financial bankruptcies caused by the interaction of collapsed stock markets and persistent deflation. 1997 saw the failure of Sanyo Securities, Japan's seventh-largest broker, Hokkaido Takushoku Bank, one of Japan's leading 20 banks, and, most spectacularly, Yamaichi Securities in the largest corporate collapse in history. In 2000 two large insurance companies, Kyoei Life and Chiyoda Mutual, went bust. And this has been during a relatively mild deflation and in an economy not noted for brutal financial adjustment.
Nor is such experience confined to the peculiar conditions of Japan. The perfect example of a disaster brought on purely by a change in the inflation regime is the venerable British financial institution Equitable Life, which for hundreds of years was used by the British professional middle classes for long-term saving. Now it is closed to new business and many of its policy holders are in a state of despair. Their plight has been brought about by the unforeseen transition to a regime of low inflation (well, unforeseen by most people).
Back in the inflationary era, the Equitable gave some of its policyholders the option of taking out a fixed-rate guarantee on its pensions, which effectively shielded the taker against the chance that long-term interest rates might fall dramatically and thereby cut back their pension entitlements.
At the time, the Equitable evidently took the view that a dramatic fall in long-term interest rates was unlikely given that inflation was ingrained in the system. Accordingly, it decided not to hedge its exposure. It continued with this stance even as inflation and long rates plunged, thereby raising enormously the value of the guarantees given to some policyholders. In the end, other policyholders in the so-called with profits fund, who were effectively taking an equity state in Equitable's investment strategy, footed the bill. Complex and extremely expensive legal action ensued, which bit into the value of policyholders' assets even more.
To cut a long story short, the Equitable is now a pale shadow of its once proud self and many a prospective pensioner is having to come to terms with much diminished prospects for their retirement. And this was simply from the transition to low inflation. The Equitable saga could prove to be a children's tea party compared to what might happen with the onset of deflation.
Not so Safe, after All
In the last chapter I emphasized the dangers posed to pensions by the bursting of the stock-market bubble. These problems would be far worse in deflationary conditions. Where pensions are set in relation to a salary, they are specified as either fixed payments that can rise at a certain minimum rate per year. Under current arrangements in most countries they cannot be cut, regardless of what happens to consumer prices, wages, or the value of investments. This is a classic case of mismatch between assets and liabilities.
In order to provide for these pensions, funded schemes rely on their investments to rise. If the value of their investment falls, however, then once they have used up any cushion in the fund they are forced to find money from outside to be able to carry on paying the same pensions. This may mean increased contributions from existing members or a greater contribution from the sponsoring company.
Deflation also worsens the vicious interaction between pension under-provision, the stock market, and the economy, which I highlighted in the last chapter. It weakens investment performance but boosts the guaranteed real value of pensions, thereby transferring wealth to pensioners. As the size of this transfer mounts, it worsens the position of the corporate sector and thereby weakens share prices still more -- and so on and so forth. Nor will there be an escape for the system as a whole if the threatened companies seek to get out of their predicament by slashing costs and employment levels, for that will simply intensify the deflationary pressure in the economy.
Pensions paid by so-called defined-contribution schemes do not experience quite the same problems, but they do experience problems. It is simply that the people who bear the impact of the losses are different. These schemes are a vehicle for stock-market investment where the prospective pensioner directly bears the investment risk. They will only get the pensions they thought they would if the stock market performs as expected. If deflation sets in, it will not. Admittedly, in this case there are no financial institutions in the middle whose existence may be imperiled. Nevertheless, the prospective pensioners will be decidedly worse off in money terms and, if they have considerable liabilities fixed in money terms, for instance through a mortgage (see below), they could be the ones whose solvency is in danger.
When occupational schemes are not funded, which is true in most of continental Europe, deflation is again deadly. Prospective pensioners will expect, and in many cases be entitled to, certain fixed sums, but as deflation proceeds, even if there is no real weakness in the economy, company revenues will be under downward pressure and company profits will probably fall, thereby making the burden of paying out the pensions all the greater. In this case the added burden is again felt directly by employers, who are liable to react by cutting costs, thereby intensifying the deflationary spiral.
Most people would naturally tend to assume that whatever problems deflation might cause for private-sector pensions, at least their state entitlements would be immune. I am not so sure. After all, under deflation tax revenues are likely to fall and so if the government cannot also cut its expenditure, its financial deficit must widen. In acute cases, the result will be a prospective level of debt that is unsustainable. Accordingly, prospective state pensioners would be right to be concerned about whether they will ultimately get what they think they have been promised. This is, after all, exactly what worries many people in Japan today.
There is a way out, but it hardly offers much succor to pensioners: namely reductions in pensions in line with falling prices. This has already happened in Japan, where in February 2003 the monthly pension was cut by 0.9 percent, in line with deflation in the shops. It would be open to other governments to adjust the nominal value of state pensions downwards or, if this were not deemed to be compliant with the existing "uprating" legislation, to pass new legislation allowing pensions to be tied to movements of the consumer price index, both up and down. There would, of course, be enormous resistance to this from pensioners, and perhaps even from the public at large. If this pressure succeeded in ensuring that state pensions increases had a floor at zero while the state's tax take shrank in line with falling prices, deflation would cause the problem of burgeoning state pension liabilities to intensify.
You might think that a regime of fluctuating prices, where the average inflation rate was zero but years of inflation alternated with years of deflation, would escape this difficulty. However, if state pensions could not be cut, it would not. Pensions would remain constant in the years when price levels fell and rise in the years when it rose. In such a regime, the real level of state pension (and its real cost to taxpayers) would ratchet up over time.
Redistribution between pensioners
Still, it's an ill wind and all that. The shift to deflation from a previous history of inflation would bring substantial gains for those who had retired on fixed pensions set earlier. In America and Britain, the outright winners would be members of defined-contribution pension schemes who retired when annuity rates were high, reflecting the markets' expectation of high inflation, only to find that deflation set in.
Equally, there is a category of outright loser. As I stressed above, if the countries of Europe and North America are set to experience deflation, this is probably not going to be a new regime of complete stability -- a steady state with a constant rate of deflation continuing year after year. After a period of deflation, these countries would probably undergo a bout of inflation. In fact, if the policy objective of the authorities were directed to maintaining a certain rate of inflation, or even to achieving and maintaining a certain price level, then after a period of deflation, policy would be set deliberately to deliver price increases. (After all, in inflation-targeting regimes it is explicitly recognized that inflation can be too low.)
The swing from deflation to inflation poses a really serious risk for pensioners. Suppose they retire at a time when the financial markets are worried about deflation, when annuity rates are set to reflect that, and when corporate and state uprating practices are also set to reflect it, but when, just around the corner, there lies a bout of inflation, perhaps stimulated by the authorities and perhaps virulent. Pensioners caught in this position would be in a similar situation to all those people who retired on fixed pensions in the postwar years only to find that those pensions' real value was substantially eroded by inflation. The problem is especially acute now that people typically face longer periods of living off a pension due to both earlier retirement and longer life.
Until recently this danger might have seemed pretty remote for most people working in large companies in the UK, because they would have been members of an occupational pension scheme whose retirement benefits would have been tied to final salary and subject to uprating with inflation, at least in accordance with statutory minimums and often uprated more generously. Now, as I emphasized in the last chapter, companies are rushing to close such schemes. Switching from final-salary to money-purchase arrangements may or may not be a good thing overall, but one aspect is clear. Compared to final-salary schemes, money-purchase schemes transfer risk from the employer to the employee -- including inflation/deflation risk. In the UK, the way things are going, most people working in the private sector are going to be exposed to this risk -- and it can be fatal.
Pensions and society
In the 1970s the plight of pensioners whose real incomes were ravaged by high inflation excited considerable public sympathy, but the onset of deflation would bring the reverse. Imagine a situation where pay was static or falling for most of the population, while those living on pensions saw their incomes rising at 3 percent per annum. There would be considerable resentment. The high and rising real value of pensions currently being paid out to the retired would be paid at the expense of those still working. Some British companies are already expressing surprise and dismay that they are increasing the incomes of their pensioners faster than the incomes of their current employees, some of whom are on pay freezes or enjoying increases of only 1 or 2 percent.
These income-distribution and fairness effects would also be bound to cause considerable political difficulties. We are familiar with the idea of the population aging and gray power emerging as a political force. Yet we should not underestimate the continuing power of those in work, who will continue to be more numerous in the electorate even in the most graying of countries, and will continue to be capable of exerting a degree of economic power.
After the last bout of serious deflation in the interwar period there was a deliberate policy of reducing the real incomes of rentiers -- those who live on the income from their investments -- by cutting interest rates. In today's society, if we were to undergo a significant amount of deflation, pensioners would be the major beneficiaries and there might well be a comparable movement to reduce their entitlements. The scene would be set for a battle between the generations. How the problems of the pension system will pan out in the future is a subject I discuss in Chapter 7.
Because of their long life, pensions are perhaps the most sensitive part of the economic system to a bout of deflation. However, mortgages must come a close second. Someone who takes out a mortgage is engaging in a transaction whose value is determined by the interaction between one fixed money amount, namely the capital sum borrowed, and up to three variable amounts: the rate of interest (unless they take out a fully fixed mortgage), the rate of increase of their income, and the rate of increase in the value of the property.
The majority of people in the Anglo-Saxon countries have substantial mortgage debts. In the UK, the ratio of mortgage debt to GDP is 60 percent. Admittedly in continental Europe such debts are typically lower, but they are still high. The average ratio of mortgage debt to GDP across the eurozone is now 40 percent, and in Denmark and the Netherlands it is above 60 percent.
Deflation would raise the real value of these debts, making them more burdensome to repay. In the first instance mortgage holders might feel little affected, as nominal interest rates fell and their earnings still went up. But this is just the initial stage. In a full-blown deflation earnings would first stop rising and then would actually fall, as would the price of property. However, as I pointed out above, official interest rates cannot fall below zero and that means that mortgage rates could not plausibly be lower than, say, 1 percent and in all likelihood would be much higher than that. The problem is clear: a fixed burden of debt and debt interest payments while the ability to finance and repay debt is falling year after year.
This mismatch is the route to mortgage misery -- and could lead to disaster for the economy as a whole. Millions of people in most of the world's leading countries would be facing financial ruin. Many would surely default, as they did in the 1930s. Even in the UK recession of the early 1990s, thousands of homeowners were unable to keep up their mortgage payments and lost their homes. As the value of the properties fell below the mortgage debts secured on them, there were many cases of people posting the keys of their property through the lenders' letterboxes and simply walking away. Why carry on forking out all that money for nothing?
Admittedly, then the pain was caused because interest rates had risen to insupportable levels. But pay was still rising. Under deflation, interest rates would fall to rock bottom and no more, while pay would continue to fall. This would bring exactly the same squeeze but from the opposite direction.
So in current circumstances deflation is to be feared -- whatever the experience in Britain and America at the end of the nineteenth century, and whatever the textbooks say. But is it a realistic prospect outside Japan? Deflation is not a uniquely Japanese phenomenon. In 2002 China, Hong Kong, Singapore, and Taiwan also experienced it. In China this was the second bout of generalized deflation in five years and it came after only a short intervening period in which prices barely rose at all (see Figure 3.3).
Together, these five countries account for about 20 percent of the global economy and an equivalent portion of world trade. (Their share of world output is probably nearer 30 percent on a purchasing power parity (PPP) basis.)33
Moreover, the extent of the deflation in some of these countries is notable. In China, goods prices declined by 10 percent over five years. Hong Kong's measure of the overall price level (the GDP deflator) declined by 16 percent in five years.
Outside Asia, Switzerland, Sweden, Argentina, and Saudi Arabia have briefly experienced deflation in recent years. In the EU, the rate of price inflation has already touched very low levels in a number of countries, including Britain, France, and Germany. In the United States the rate has not yet fallen to very low levels, but until recently the country enjoyed strong economic expansion. Given the strength of the American boom in the late 1990s, inflation was remarkably low.
Moreover, the signs are that the American economy is now highly prone to deflation. In Q2 2002 the GDP deflator, just about the broadest measure of price pressures, was up only 1 percent over the previous year, representing a drop of 1.3 percent from the inflation rate prevailing at the cyclical peak, some six quarters earlier. That rate of inflation deceleration is about double the norm in previous business cycles.34 In early 2003 the rate of core inflation -- that is, excluding the volatile elements of food and energy -- fell to 1.5 percent, a 37-year low.
The prices of goods leaving British factories have been falling for years -- as have the prices of goods sold in British shops. It is only the continued inflation of service prices that has kept deflation at bay in the UK.
Moreover, Asian deflation has a way of spreading into Europe and North America, as prices are driven lower by the relocation of manufacturing, especially to China, where costs are a fraction of what they are in the developed countries. For many products, once production is shifted to China the ultimate decline in prices might be 70-80 percent. So great is the overhang of surplus labor in China that in the manufacturing sector China's prices are going to become global prices. What is more, these efforts are now starting to spread into the service sector, as European and North American businesses outsource to countries such as India such services as call centers, email help desks, software design, accounting, network management, billing, telemarketing, and transcription and translation services.
The zero margin
The weight of this evidence leads to a clear conclusion. For interest rates it may be, but for inflation zero it is most definitely not a magic number. Complete price stability is an illusion. Policy makers have peddled a fantasy that we could have gently rising prices for ever. But we have never had this, apart from the extraordinary period of the first 25 years after the Second World War.
In Europe and North America this generation may be unused to deflation, but previous generations regarded it as part of normal experience. In the UK deflation has alternated with inflation for centuries, leaving the price level broadly stable over long periods. In 1932 the price level was marginally lower than it had been in 1795. Moreover, there were some periods of very rapid price declines in the twentieth century. In 1921 prices fell by 21 percent in the UK and 10 percent in the US.
Until recently, this long historical experience with deflation was not only unknown by most of the population but disregarded by economists and policy makers. It occurred under a system that tied the value of money to gold (or some other precious metal) and that therefore could not be managed. There were periodic gluts and shortages of the precious metal that precipitated bursts of inflation and deflation, about which the monetary authorities could do nothing.
In the 1930s, however, countries abandoned the Gold Standard and afterwards national central banks managed their own monetary systems based on paper money and credit, the supplies of which were potentially unlimited. Central banks could readily create money for nothing. Accordingly in a modern monetary system, the experts thought, deflation was now impossible. The historical experience with deflation could be left to the history books. Then came the Japanese deflation of the 1990s (which I discuss briefly below). Now the historical experience of deflation no longer seems so irrelevant.
Even so, given that just about the whole world experienced prolonged low inflation in the 1950s and 1960s without any break into deflation, it is legitimate to ask what it is that makes deflation a serious possibility now. Why shouldn't the world simply carry on with a very low rate of inflation?
There are three factors that over the medium term are tending to cause countries to operate below full capacity, and these are closely associated with the main themes of Part II. First, as I argue in the Finale, large parts of the world are wedded to a policy of substantial current account surplus, thereby sucking demand away from the rest. Meanwhile, in the developed countries the intense competition unleashed by globalization is reducing the level of unemployment or unused capacity necessary to stop the economy from boiling over. Yet outside the US and the UK this change is barely acknowledged by the policy establishment.
Moreover, as I argue in Part II, the combination of globalization and a faster rate of knowledge discovery should unleash a higher rate of productivity growth. There are clear signs of this in the US, where productivity growth has remained very high even in the downturn, although outside the US this improvement in economic potential is hardly acknowledged. To be fair, outside the US there is not yet much sign of any improvement in the data. Nevertheless when the effect does appear, on all past form the central banks are likely to be slow to acknowledge it. The upshot may well be that policy will be run too tight, given the enhanced potential, with the result that the economy continually operates below capacity.
These factors give us the likelihood of unused resources, but why should this produce deflation as opposed to simply loss of output and increased employment? In The Death of Inflation I gave the answer. Prices are no more downwardly flexible than they used to be, so that weak aggregate demand will more readily result in falling prices. Globalization, a more competitive and more unstable business environment, and the growth of variable pricing practices have all contributed to a situation in which prices can more readily fall across wide sections of the economy.
With trade unions much weaker, it is more conceivable that even wages could fall. In the Untied States and Britain, but also increasingly elsewhere, a larger proportion of pay is now made up of variable components such as bonus and performance awards and profit-related elements. Even if "the rate for the job" is still sacrosanct, these variable components such as bonus and performance awards and profit-related elements mean that total earnings can fall -- even in nominal terms, if there should be an adverse shock to aggregate demand. And there are an increasing number of wave cuts in particular sectors. In 2003, American Airlines agreed to $1.8 million worth of wage cuts with its unions.15 In the UK, Clifford Change, the world's largest law firm, cut its pay rates for lawyers.
In Japan over recent years average earnings have been falling.16 In 2002, average monthly pay fell by 2.2 percent. There are also many striking examples of wage cuts. Public-sector workers employed in Nagano recently accepted pay cuts of 12 percent over three years and Nissho Iwai and Nichimen, two trading groups that plan to merge, have asked their unions to accept pay cuts of 20 percent.
If my fears about the immediate outlook for the world economy prove justified and some industrial countries show next to no growth at best, one or more of them will experience deflation. In short, the countries of Europe and North America are one recession away from deflation. My only substantive doubt is whether that recession occurs sooner or later. But as I argued in the last chapter, there are powerful forces acting to bring it sooner.
Countries at Risk
Because of the extent of the loss of wealth in the stock market and the high level of flexibility for both pay and prices there, the United States is one of the most likely candidates for deflation. But until recently there has been a good deal of complacency about the prospect, even in high places. Glenn Hubbard, chairman of the US President's Council of Economic Advisers, dismissed the prospect on the basis of four spurious arguments.17 The first is that US productivity growth has been strong, thereby underpinning strong growth in real incomes that in turn would sustain strong growth in consumption. Yet if consumption is strong because productivity growth is strong, this says nothing about the extent of inflationary or deflationary pressures. Indeed, the fact that productivity growth is strong is a priori an argument in favor of the prospect of deflation, because strong productivity growth requires that aggregate demand increase at the fast rate to prevent prices from falling.
The second argument is that the surge in US house prices is fully sustainable and underpins strong consumption growth. The reason for this is that high house-price growth stems from high levels of immigration and land shortage and high transaction costs deter speculative behavior in the housing market. Yet this completely ignores the points I made in the previous chapter about the vulnerability of the US housing market and the way the housing market works.
The third argument is that falling prices are not always bad. This is true, but so what? It is not an argument against the possibility of deflation occurring.
The fourth is that a sustained decline in prices that magnifies the real burden of debt is unlikely. But this assumes away the very thing whose likelihood is under debate.
One can only hope that the President receives better advice on other issues. At least the Fed is well up on deflation risk and the markets are also keenly aware of it. Indeed, you could say that in 2003, at least deflation entered the mainstream of financial America and it has stayed there.
But unfortunately, awareness is not quite so acute elsewhere, even though it should be. Germany is in front of America in the deflation stakes because its starting point is so much lower, with core inflation below 1 percent in the first half of 2005 and dipping as low as 0.3 percent at one point. And the country looks uncompetitive. Moreover, since its partners in the eurozone are inflating only at very low rates, for Germany to regain competitiveness fairly quickly, prices there will have to fall. So far from being an accident, deflation in Germany could be said to be preordained. This situation is uncannily similar to what happened in the UK after it returned to the Gold Standard in 1925 at the old (now too high) parity.
Against this, German price and wage structures are more inflexible than in the US, so a downturn will probably have to be that much more severe to deliver falling prices. On the other hand, the German authorities are not able to set monetary policy to forestall the threat of deflation but have to rely on the ECB, acting in the interests of the eurozone as a whole.
This is a particularly alarming prospect that seems hardly to have been considered by all those advocates of the European single currency -- or by many of those who worry about deflation. The main reason why conventional economists are so confident that deflation can be easily stopped is that the central bank can simply print money. But what happens if you haven't got a central bank with money-issuing powers? Germany still has the Bundesbank, but it can do virtually nothing. It is now no more than a branch of the ECB. Meanwhile, the ECB is not exactly the sort of central bank to undertake the radical steps that may be necessary to stop deflation. It seems to be frightened of its own shadow.
What is of special concern in the European case is that although monetary policy is conducted on a eurozone basis, monetary and financial institutions remain largely national. In Germany, for instance, the banking system has a preponderance of German debt on its balance sheets. The pensions and insurance systems are largely Germany: German providers and German takers. In other words, there is still a heavy concentration of nationally based credit and systemic risk in a monetary system that is being run on transnational lines.
Admittedly, deflation in Germany and France would not be able to go very far before the ECB was obliged by its remit to sit up and take notice, because these economies are so weighty in the eurozone economy and their inflation rates are an important component of the eurozone's inflation rate. But deflation could get going in these countries, and even more so in the smaller countries, while the ECB continued to be worried about inflation for the eurozone as a whole, not least because the fast-growing countries of the eurozone, especially Greece, Spain, Portugal, and Ireland, tend to have a much higher inflation rate for a systematic reason. Essentially, their price levels are catching up with those of central Europe. Arguably, their high inflation rates are therefore a different kettle of fish and should be excluded from the ECB's target, or the target should be raised to accommodate them.
Nevertheless, the point is that as things stand the ECB makes no allowance for this factor. The result is that if it succeeds in getting the eurozone's average inflation rate below the target of 2 percent, it will virtually guarantee that the German inflation rate will be just above zero. Pushing average eurozone inflation much lower, either by accident or design, would tip Germany into deflation. And this is not only a problem for Germany. Belgium, Austria, and Finland also have very low rates of inflation, and are closely tied into the German economy. They too are in acute danger of falling into deflation in the next couple of years. Even the Netherlands, which has tended to have a higher rate of inflation, would not be far behind.
The problem of divergent inflation performance has intensified after 10 further countries joined the EU in 2004, all of which tend to have high inflation rates for similar reasons. As and when some of these countries join the eurozone, the zone's inflation rate will rise, arguing for tighter monetary policy -- which means even lower inflation in Germany and the rest of core Europe.
At first sight it is hard to see the UK being high on the list of possible deflators, but in fact there is a significant deflation risk there too. For a start, the UK's inflation rate has been extremely low -- just about the lowest in the EU. On the common European harmonized measure, HICP, UK inflation has frequently been less than 1 percent. Furthermore, the UK is particularly exposed to a slump in residential property prices. As I stressed in the last chapter, UK house prices have risen a long way, look overvalued according to the usual yardsticks, and are particularly important as a determinant of consumer spending and therefore aggregate demand.
The UK's deflationary risk would be greatly increased if the country decided to join the euro at somewhere near the exchange rate ruling during the summer of 2005. The exchange rate of the pound against the euro was then much too high. If entry proceeded at that rate it could condemn British exporters to years of decline, and the overall economy to years of stagnation and falling prices. But as things stand, British entry looks a most unlikely prospect.
Forseeing and Forestalling
So deflation is potentially deadly, it has occurred in Asia, and it could easily appear in the West. But could it easily be stopped by the policy makers? This is not the place to review the detailed policy options to defeat deflation, but some points of principle need to be established here.
The behavior of central banks is critical to the possibility of deflation in the modern world because, in contrast to the Gold Standard, monetary policy is now managed. Indeed, at least until recently it has been common to dismiss the possibility of deflation on the grounds that no government or central bank is aiming for it, nor is every likely to -- well, not yet away.
Although not many observers have particular faith in the ECB, there is still a good deal of faith in the powers of central banks in general -- not least among central bankers.38 This is why most people are inclined to be skeptical about the prospects for deflation in Europe and North America. The prevailing view among economists and market operators is that in the end we get the inflation rate not that we deserve, but that the central banks choose. In that case, what really matters is what the central banks (or, where appropriate, their political masters) want, how well they can anticipate the future, and how reliable their tools are for achieving their objectives. The prevailing view is that on all three counts there are grounds for reassurance. But I believe that on all three counts there are grounds for serious doubts.
The widespread market assumption has been that if the economy were ever in danger of sliding into deflationary phase, which is itself the subject of some dispute, central banks would see the danger coming and take appropriate offsetting action. This is naïve. After all, the world's central banks did not see the Great Depression coming, nor the Great Inflation of the 1970s. Nor has the Bank of Japan had a good record, either in foreseeing deflation ahead or taking appropriate action to forestall it. Nor, for that matter, did private-sector forecasters or financial markets see deflation coming in Japan. Japanese long-term bond rates remained as high as 5 percent right up to the beginning of 1995. Moreover, once the danger was upon them, the Japanese authorities still did not seem to appreciate how serious it was. Consequently, they did not act boldly enough and did not take sufficient "insurance" against the special downside risks that deflation poses.19
But do the West's central banks now take the threat of deflation seriously? They certainly take it more seriously than they did when The Death of Inflation first appeared in early 1996. That book met with a barrage of criticism, especially from central bankers, particularly on the subject of deflation, which was widely felt to be beyond the realm of possibilities in the West. Over the succeeding years, though, a number of central bankers have made statements suggesting that they now take the prospect seriously. In particular, the Bank for International Settlements (BIS), the central bankers' bank, said in June 1996: "The forces bearing on the price level are now more balanced than they have been for some decades."
Moreover, the speeches of Alan Greenspan at the Fed have been littered with references to the need to avoid deflation as well as inflation. But all is not sweetness and light. The ECB has also on occasion paid lip service to the deflationary threat, but it doesn't not seem to give it due weight. In his comments at a press conference in June 2003, ECB president Wim Duisenberg said:
There are currently no forecasts indicating any deflationary risks. At a regional level a period of low price increases or indeed declines will improve that region's competitiveness in the monetary area.
Within a monetary union deflation is not a meaningful concept when applied to individual regions, like New Hampshire or Germany.
We are convinced that we don't have to prepare for deflation because we don't see deflation coming. Is the central bank prepared to deal with it if the situation were coming? The answer is yes.20
At the Bank of England, even the arch-hawk, the then deputy governor Mervyn King, said in December 1998: "Central bankers should be as vigilant in counteracting deflation as in preventing inflation." Nevertheless, the doubts remain. After all, Sir Edwards George, then governor of the Bank of England, said as late as August 2002: "I don't at this stage lose much sleep about deflation."21 Let's hope his fellow central bankers don't need a wake-up call.
Not only do central bankers still tend to underrate the deflationary threat, but they are also overconfident of their ability to stop deflation if it appeared. Despite all the propaganda, their policy tools are not so finely honed and effective that central banks can be relied on to take exactly offsetting action when they do see a major shock coming.
In particular, central banks have fallen into the trap of overestimating the power of interest rates to adjust the economy, particularly when the recent history of financial fantasies has left companies with excess capital stock and consumers with weakened "balance sheets" so that they do not respond to reductions in interest rates in the normal way. For real efficacy, monetary policy has always relied on either very large movements in rates or the ability to deliver some sort of shock that affects expectations or the supply of credit. If confidence is lacking, reductions in interest rates of 1/4 or 1/2 percent, which have recently become the norm in the industrialized West, may have next to no effect.
Central bankers think and act as though the economy were some sort of finely tuned engine that only needs the sprockets adjusted and the right buttons pushed for it to leap into life, hover gently, or move up slowly to cruising speed -- whatever the central bankers desire. In fact it is more like a wild animal: unpredictable, awkward, and sometimes downright dangerous.
Accordingly, as and when western central banks are confronted by emerging deflation, even though they might think that their responses are timely and forward looking judged by the potential speed of deterioration of sentiment, they are likely to be slow to reduce interest rates and the pace of reductions is likely to lag behind the pace of deflationary process. This is usually characterized by falling confidence and a collapse of credit availability, against which minor reductions in interest rates may be next to useless. That, of course, means that interest rates will have to fall all the more and stay down that much longer -- assuming, that is, that central banks do want to stop the deflation.
So what do central banks want?
It is not even clear that all central banks would want to stand resolutely against a small rate of deflation. The history of inflation fighting since the battle was first joined in earnest at the end of the 1970s is that the objectives have become more demanding as central banks have enjoyed more success. Far from relaxing as they succeeded in reaching their objectives, central banks have become more ambitious. At the moment across the world, something like an inflation rate of 1-3 percent is presented as the ultimate prize. But it is by no means clear that we have yet reached the final destination.22
Moreover, because of the history of high inflation over the last 30 years and more or less continuous inflation over the last 60, whatever they might think themselves central banks are still paranoid about their reputation as inflation fighters. The US Fed is an honorable exception, but in general, whereas central banks are inclined to see inflationary risks very clearly, they readily play down the chances of deflation. Accordingly, in the event of a major downward shock to demand, prompt offsetting action by central banks could be taken for granted.
Meanwhile, the financial markets are also still supersensitive to the danger of inflation and accordingly tend to set long interest rates at a level that imposes high real interest rates on borrowers. Throughout the steep American economic slowdown in 2001, bond markets continued to fret, and sometimes to panic, about the threat of inflation.
How Could Deflation Happen?
The world is now stuck in a sort of low-inflation equilibrium in which continued downward pressure on a whole raft of prices is offset by continued upward pressure on others. A breakout will require some shock. Nevertheless, there is no mystery about what could cause a breakout downwards: adapting a phrase from President Clinton, it's the stock market, stupid; or perhaps more accurately, the stock market and the housing market.
Moreover, deflation in Japan could spread to North America and Europe. Most commentators and analysts in the West have become complacent about the Japanese slump. It may have lasted for a decade, and for the last several years this has been accompanied by deflation, but so what? Japan is not a substantial borrower from the rest of the world; it is a substantial net creditor. So who cares if the economy keeps contracting and prices keep falling? Anyway, Japanese deflation now seems to be ending.
This complacent view is wrong. The world will find it difficult to prosper when its second-largest economy is in the doldrums. We all need a strong Japan. And there are direct links from Japan's weakness to the rest of the world. In particular, a weak Japan threatens to destabilize the whole of East Asia, which is so critical for the West's prosperity. If Japan slipped back again, one way for it to overcome its deflation might be to export it to the rest of us via an ultra-weak yen.
If that ever happened, it would put pressure on the Chinese renminbi and other Asian currencies. (I briefly discuss Chinese exchange-rate policy in the Finale.) In short, there would be the danger of a mass depreciation of Asian currencies against the dollar and the euro -- exactly the opposite of what is needed. Effectively Asia as a whole would be exporting deflation to the West. Meanwhile, the US would need the recent depreciation of the dollar to be accepted, and probably to go much further. The upshot would be substantial upward pressure on the euro -- and downward pressure on prices in the eurozone.23
Could it be as bad as Japan?
Could countries in the West experience something as serious as the deflationary slump that has engulfed Japan? For most of the 1900s, as Japan floundered, it was common to believe that the answer was a categorical "no." Japan's problems were uniquely Japanese. Nothing like them could or would occur elsewhere.
To be fair, there have been some extraordinary features of the Japanese predicament that it is difficult to see readily repeated throughout the West, and there have been some glaring policy mistakes. Furthermore, the Japanese banking system has been a disaster beyond compare. The Japanese authorities first tried to deal with this problem by not admitting how serious it was, which in the end made matters worse. Only belatedly did they set about trying to clean up bank balance sheets sufficiently to facilitate flows of new credit. At the end of 2002, even official estimates put the ratio of bad debts to GDP at 8 percent, and the reality was surely worse.
By contrast, on the whole, banking systems in the West look relatively strong, even after the financial and economic crisis in America in 2001-2003. Still, complacency is to be avoided. It is in the nature of banking crises that they come and hit you when and where you are least expecting them. Banks have offloaded risks to other parts of the system through the use of derivatives, whose total face value, excluding contracts traded on exchanges such as the International Petroleum Exchange, comes close to $85 trillion.24 It is always possible that major positions in derivatives, perhaps misunderstood or perhaps not even known about by bank managements, could cause some significant bank failures. The sage of Omaha, Warren Buffet, has said: "Derivatives are financial weapons of mass destruction. The dangers are now latent -- but they could be lethal."25 When Buffet is worried the rest of us should quake.
Moreover, not all of the western banking system even looks in tip-top condition. In late 2002 there were five continental European banks whose capital position appeared decidedly weak,26 and at times there have been some strong market rumors that a major continental bank was on the brink of collapse. There are particular concerns about German banks that are suffering the deadly combination of a weak economy and incipient deflation on the one hand and a history of poor profitability on the other.27
Even in Britain, where the banks are highly profitable, generally well capitalized, and well managed, there are potentially serious problems. In December 2002 Standard & Poor's for the first time listed Britain as a country facing potential stress in its banking system if the housing bubble burst. It is right to have done so. When the housing bubble does burst banks will find themselves in much tougher conditions and this is bound to affect their lending policy.
In any case, as the source of all ills Japan's bad-debt crisis is not quite what it is cracked up to be. In the popular perception -- sadly, perhaps in the understanding of the Japanese authorities -- there is a misguided tendency to give these bad debts an exaggerated causative role in the generation of Japan's deflation crisis. The prevailing idea is that they have eroded banks' capital and hobbled their willingness to lend. The implication is that alleviating the burden of bad debts is the solution to the problem. If only the Japanese authorities could muster the will and the funds to wipe the slate clean, all would be well.
Yet if the weakness of Japanese banks is the essential problem, why aren't foreign banks queuing up to make loans to the hoards of frustrated, credit-worthy potential Japanese borrowers? The answer, of course, is that these hoards do not exist. That is the problem. The essential difficulty is the lack of creditworthy borrowers -- and that is partly caused by deflation itself. What is the point of making loans if you know that you will not be repaid? You will be dishing out money for nothing.
In short, the majority of the banking system's travails are the result of weak aggregate demand and deflation rather than the independent source of those problems. Sorting out the banks' bad-debt problems is probably a necessary condition for full Japanese recovery, but it is certainly not sufficient. The same logic applies in Europe and North America. Relatively strong banks help to avoid deflation, but deflation has a way of rapidly turning strong banks into weak ones.
Japan the precursor?
Could we see deflation in Europe and North America? Yes. Japan's deflation is a direct descendant of the bubble economy of the late 1980s. Stock and land prices then soared to extraordinary levels. When the crash came there was a devastating blow to wealth levels, to the balance sheet strength of both lenders and borrowers, and to confidence.
The land bubble has a clear parallel in the West in the spectacular housing bubbles in many countries. And the West's equity bubble in the late 1990s was directly comparable in type to the Japanese bubble a decade earlier. Moreover, there are strong similarities in the surge of investment in Japan in the late 1980s and the surge in the US in the late 1990s. In both cases this left companies with excessive levels of fixed capital, thereby sharply diminishing the subsequent incentive and desire to invest. Furthermore, there are some eerie similarities between Japan and Germany, including the rapid aging of, and then prospective fall in, the population.
Although policy in North America and Europe should be better conducted than it was in Japan, especially by the US Fed, my strong suspicion is that western central banks will easily fall prey to the problems that constrained the Bank of Japan. In particular, the ECB will be slow to see the deflation coming, slow to cut interest rates sufficiently, and very slow to embrace unorthodox monetary policies. Like the Bank of Japan it will be overly concerned with maintaining credibility and normality, inclined to minimize the dangers of deflation and maximize the dangers of the policies designed to overcome it. Again like the Bank of Japan, it will be inclined to see recovery, and an upturn in prices, just around the corner.
But not everything is the same. In some respects the Japanese predicament was much worse. In other respects, though, it was better. The Japanese authorities did not have to deal with the disastrous interaction between company pension provision and stock-market weakness that threatens to dominate economic prospects in several western countries
In mid-2005 most financial market participants were fairly sure that deflation risks were minimal -- as they were sure before the last deflation scare arose. The truth is that with aggregate demand so fragile, price-competitive pressures so intense, and the starting level of inflation so low, deflation is an ever-present danger. And if it materialises it could be deadly.
Central banks like to assure us that they and their policy tools are well up to the job -- even if the job encompasses the overcoming of deflation. They had better be right, or the world economy will face catastrophe.