Japan Emerges from the Shadow of Deflation

President and Chief Research Officer, RIETI

In the fourth quarter of 2003 the growth rate of consumer prices (excluding fresh foods) came to precisely 0% over a year earlier, and the growth rate of real gross domestic product reached a solid 3.6%. Can we safely say that the Japanese economy is finally pulling out of its deflationary phase of falling prices? To answer this question, we need to look into the specific factors that have been turning deflation around.

One special feature of the deflation that has plagued Japan since the 1990s is that it has been moderate and cyclical. The moderation of the decline in consumer prices is evident in the fact that the maximum speed of their fall, registered from the first quarter of 2001 through the third quarter of 2002, was 1%. As for their cyclical nature, just as there have been three identifiable business cycles over the past 10-plus years, there have also been three similar price cycles. During the business recovery in 1996 and 1997, prices were ascending at a 0.8% rate. The first actual decline in prices occurred subsequent to the jitters in the financial system that broke out in November 1997. Then, in the first half of 1999, there was a phase of no change in the price level. During 2000 the economy entered another downswing, and that set the stage for the phase of 1% deflation that I just mentioned. Business began to recover during 2002, and prices returned to zero change in the last quarter of 2003.


Let us now consider the decisive differences between mild and sharp deflation and also the significance of the small cyclical price movements. When talk turns to deflation, what comes first to mind for many economists is the Great Depression of the 1930s in the United States. The deflation of that period was marked by the high speed of the fall in prices, which plummeted at an annual rate close to 10% from 1929 through 1933, and by the continuity of the descent, which lasted for almost four years without notable cyclical movement. This happened because the real economy (real GDP) was rapidly contracting, losing about 25% of its weight altogether. When you add the plunging prices to the shrinking real GDP, you find that the economy suddenly shrank to about half of its former size when measured in nominal GDP.

The differences between moderate and extreme deflation are of considerable import in economic theory. One of the foremost problems deflation causes is that when prices go down, the essential value of debts grows larger by a corresponding amount. This makes it difficult for debtors, which often are companies, to pay back what they owe. In this situation, the pace of debt expansion is the same as the real interest rate (nominal interest rates minus price changes). When nominal interest rates are about zero (as has been the case in Japan), the real interest rate is a positive figure matching the deflation rate. In the United States during the depression, accordingly, real interest rates were 10% at the minimum. When interest rates are at such a high level, companies are in no position to make capital investments. Investment in fact plummeted in the United States, and in 1933 it was insufficient even to match depreciation-a most unusual situation.

What about Japan? With nominal interest rates at zero and deflation proceeding at a 1% pace, real interest rates were 1%. These are short-term rates. The real long-term interest rate was about 2.5%, since the nominal rate for 10-year government bonds was about 1.5%. Interest rates at that level are by no means high.

In theory one derives a figure for the anticipated real rate of interest by subtracting the expected inflation rate from the nominal interest rate. The question one asks is how much inflation people expect, or in this case, how much deflation. According to a regular survey of 2,400 households conducted by the Cabinet Office, from the second quarter of 2001 to the second quarter of 2003, an interval during which the rate of deflation at times ran as high as 1%, people were expecting 0.3% deflation at the worst. As far as expectations are concerned, accordingly, the real interest rate was between 0.3% (for short-term rates) and 1.8% (for long-term ones).

If even in that case interests rates seemed high to business managers, this can only have been because the profits they could expect to realize from investment were too meager. In short, because the deflation Japan experienced was a very moderate variety, the problems that excessively steep interest rates can cause did not occur.


What should we make of the cyclical movement observed in our country's deflation rate? The upswing in the business cycle that got underway in 2002 is actually the third one since the collapse of the speculative bubbles early in the 1990s, following earlier recovery phases in 1995-96 and 1999-2000. Three features not evident in the earlier phases stood out in the most recent one. (1) The upturn was self-propelling instead of being induced by Keynesian stimulus measures. (2) The level of corporate profits was higher than in the two earlier phases thanks to declining wages and progress in straightening out balance sheets. (3) Exports to other Asian countries were enjoying a vigorous expansion, since the Asian region as a whole has become the factory of the world and Japan has found a role for itself within this new division of labor.

The first feature was the result of the recovery's having started at a point after Japan had more or less exhausted what it could do in the way of Keynesian-style pump priming to get the economy going. We can call it a recovery that got going without the aid of a macroeconomic stimulus policy. Among the fiscal Keynesian tools the government has put to use, stepping up public investment is typical. As the result of Japan's having made liberal use of such government spending, the general government deficit has reached 7.4% of GDP, and the combined public debt of the national and local governments has mounted to 155% of GDP. Starting in 2000, each quarter has seen a 5%-6% contraction (yearly rates) in public investment (public fixed capital formation). This is the context in which the latest upturn took hold. In both 1996 and 1999, by contrast, public investment was being expanded, thereby giving a boost to demand in these two earlier recovery phases.

Because Japan also ran out of monetary Keynesian medicine it could administer, we should not imagine that monetary policy has added momentum to the recovery. When deflation was getting worse in 2001 and 2002 (even though it only reached a rate of 1%), advocates of inflation targeting stepped forward. They proposed that the government embark on a radical course of quantitative monetary relaxation in order to quell deflation, which means they wanted it to greatly expand the base money supply.

Actually the Bank of Japan has taken steps to keep the market flush with funds by swiftly lifting the volume of the current deposits held with it by commercial banks. In 2002 it had ¥15 trillion in such deposits, but it increased this to ¥25 trillion shortly before Fukui Toshihiko took over as the new BOJ governor, and now the current deposits are in excess of ¥30 trillion.

The question, though, is whether the BOJ's policy of quantitative monetary relaxation, which is represented by the huge balance of current deposits and which aims at expanding the monetary base with interest rates close to zero, in fact made any contribution to the latest recovery. The only possible answer that can be made is just about as close to a flat no as one can get. What is plainly obvious is that when short-term interest rates are already at zero and cannot go any lower, the tool of cutting interest rates cannot have any effect on them, no matter how much the volume of base money grows.

But what about the impact of quantitative relaxation on long-term interest rates? In the spring of 2003 the rate for 10-year government bonds slid all the way to 0.5%. This was the result of the fear of a bubble in the market for government bonds stirred up by alarmist cries from the deflation pessimists. When signs of an economic recovery became more readily visible after that, though, the rate settled down at just under 1.5%.

Back in March 2001 the BOJ let it be known in no uncertain terms that until the rate of increase in consumer prices (excluding fresh foods) stabilized at or above zero, it would stick to its stance of zero interest rates and quantitative relaxation. This announcement three years ago came to be believed so completely by market participants everywhere that even today, when consumer prices have stopped falling, nobody is anticipating any rise in short-term rates. Long-term rates have, accordingly, stayed at or under 1.5%. But this is exactly where they were three years ago; they have not become noticeably lower only recently.


In that event, what is the channel or mechanism by which the policy of quantitative monetary expansion (which is increasing the supply of base money at an annual rate in the 15%-20% range) is supposed to be stimulating business activity? The fact of the matter is that economic theory has not elucidated any channel that could accomplish this. Suda Miyako, who serves as a member on the BOJ Policy Board, made this very point in the January 2004 issue of the BOJ's monthly bulletin, Nippon Ginko Chosa Geppo: "No answer has been provided to the question of the effect of an expansion in quantity under zero interest rates, either in the world of theory or in the world of empirical evidence."

The measure of money that bears the closest relationship to actual economic activities is known as M2 + CDs (certificates of deposit). During 2003 this money supply was increasing at only a 1.5% rate over year-earlier levels. This was, moreover, even below the 3% rate of 2002. Here we can see that the relationship between base money and the money supply has been severed. After all, bank credit is continuing to contract.

Since 2003 the central bank has been intervening in the foreign exchange market on an unprecedented scale in a bid to avert appreciation of the yen, and Japan's foreign exchange reserves are snowballing. The selling of yen and buying of dollars in the foreign exchange market expands the supply of base money. Massive intervention thus means a massive increase in base money. What was it that made intervention on this scale possible? The answer is the adoption of the quantitative relaxation policy, which made it appropriate for the BOJ to allow a huge increase in base money. If not for that, the central bank would have been compelled to "sterilize" its interventions by selling enough short-term government paper to absorb the increase in base money. Another factor was that no matter how greatly base money expanded, no threat of inflation would emerge, since as I noted, this expansion was not producing an increase in the money supply. If other countries were doing the same thing Japan was doing, though, we would see inflation take off in no time at all.

Two factors lie behind the contraction in bank credit I mentioned. One is that commercial banks are fearful of a fall in their capital adequacy ratios, which has caused them to hold back from lending to clients seen as presenting a high credit risk. The other is that companies have continued to focus on reducing their excessive debt load even after their cash flows have improved. In the corporate sector, as a result, an extraordinary saving-investment balance has emerged, with saving (retained earnings) running in excess of investment in plant and equipment and in inventories. The usual situation is the opposite one of internal reserves falling short of investment. Companies then make use of borrowing from banks and other forms of external finance to cover that part of their investment they cannot fund from retained earnings.


This brings us to the question of the forces that have engendered an autonomous recovery in the private sector, which has had to do without help from stimulus of the Keynesian type. There are two key factors at work here: One is the progress businesses have achieved in improving their balance sheets, and the other is the rise in corporate profit rates (capital efficiency).

Balance sheets began to look better when firms finally managed to rid themselves of the excessive debts they accumulated in the 1980s and when the decline of asset prices, especially of land and stocks, came to an end. Thanks to the exhaustive use of Keynesian measures, Japan's real GDP actually expanded by 10% over the period people have been calling the 10 years of stagnation. And now, at last, firms are reaching the end of their struggle to regain financial health. The ratio of corporate debts to sales has now returned to where it was before Japan experienced its bubble economy, and so has the ratio of land prices to nominal GDP.

The improvement in profitability constitutes an even more positive source of the forces involved in the self-propelling recovery. As we have seen, anticipated real interest rates during the period of mild deflation were very low, ranging from 0.3% for short-term rates to 1.8% for long-term ones. By no means can they be considered excessive. The true problem was the levels of corporate profit rates, which were so low that they could not support borrowing even at this low level of interest.

Plain evidence of this can be seen in an abnormally high share of national income that goes to labor. As of 2002 labor's share had reached 75% in national income statistics, surpassing labor's share in the United States and becoming the largest in the world. In the 1980s, by contrast, the average share in Japan was 65%. Against this, if we define the "capital earning rate" as the ratio of operating profits to tangible fixed assets, we find that Japanese companies in the 1980s managed to sustain an average rate of 12%, far above the real interest rate of 5% at the time. By the end of the 1990s, though, the capital earning rate had dropped all the way down to just under 5%.

A major factor depressing corporate profits in periods of deflation is downward rigidity in wages. It is hard to push through pay cuts. Since deflation means a downward trend in the prices of the products that companies sell, profits inevitably become slimmer unless there is a corresponding decline in nominal wages. Labor's share of national income moves up, and the capital earning rate moves down. Until 1997 nominal employees' compensation continued to rise. Wages eventually began to come down after the banking crisis of November 1997, but the descent only really picked up speed over the recent three-year period from 2001 to 2003. According to preliminary national income statistics for 2003, the rate of wage deflation reached 2.9% last year.

An increase in productivity (total factor productivity) brought about by technological innovation has contributed substantially to the better-looking corporate profit picture. One example is of course the often-cited field of consumer electronics, where digitization and other advances are allowing companies to earn big profits on popular new products, such as single-lens reflex digital cameras, flat-panel television sets, and DVD recorders. But this is by no means the only area in which profits are improving.

When I visited Chongqing in China at the end of 2003, I saw work underway on the large-scale project to build a monorail system, which Japan is helping to finance. Chongqing, which has a population of 10 million, is located in a hilly region where conventional streetcars and buses cannot easily be used and the construction of a subway system is difficult. The city instead is looking to the monorail to move people around. Tens of thousands of workers have been brought in to handle the foundation work. They are good at things like digging holes and tunnels, and they work for daily pay of only about ¥300-¥400. The Y-shaped pillars on which the trains will run are made of domestically supplied reinforced cement. In short, these elements of the project are mainly Chinese in content, combining a very labor-intensive approach to construction with a somewhat technology-intensive approach to the supply of reinforced cement.

But 60% of the project's expenses go for things like huge units of construction machinery and monorail cars. These represent an area where Japan has a comparative advantage and sophisticated abilities. The cars, for instance, depend on multiple high-tech systems arrayed in overlapping layers, all of which come within the compass of Japanese technology. They must have built-in signal systems embodying operation, control, and communication technologies; they require an intricate, precision-made electrical system including parts like spark plugs and batteries; and they need a power plant with components like engines, motors, and generators.

The elements required for the development of humanoid robots, another area where Japan is on the leading edge in the world, are similar. As Araoka Takuya points out in an article that appeared in the August 5, 2003, issue of Ekonomisuto, robots also depend on the integration into a single system of overlapping high-tech layers (for control and communication components, finely designed and precisely functioning parts, and a power source).

In this way we find that the profitability of Japanese companies is improving as a combined result of efforts in the areas of restructuring, where wage deflation is one representative trend, and of technological innovation, which is giving birth to digital consumer electronics, monorail cars, and robots. Profit rates are now easily high enough to clear the low hurdle of real interest rates in a deflationary setting, and they are supplying the motive power of the self-propelling recovery.


With the rapid rise of the Chinese economy over recent years, many people presume that it is becoming the factory of the world. This view is, however, too narrow. Actually it is Asia as a whole that is emerging as the world's factory. And Japanese companies, through direct investment and other approaches, are drawing China into the production processes of a fast-evolving vertical division of labor.

In finer detail, we can see that Japan and Asia's newly industrialized economies (Singapore, South Korea, and Taiwan) are exporting capital goods as well as intermediate goods (materials, parts, and components) to China and also to Southeast Asia. China and the Southeast Asian nations then manufacture finished goods (mainly consumer goods, including digital appliances), making use of their low-wage labor supply and midlevel technologies for the processing and assembly processes, and export these goods to the developed nations, Japan included.

The production networks that have formed in Asia divide the manufacturing for, say, an electronic product, into segments, each of which is sited in a country with a comparative advantage for handling it. The result is a deftly organized division of labor within each industry. China would be unable to export were it not being supplied with intermediate goods and other inputs by Japan, the NIEs, and some of the Southeast Asian countries.

China's exports give it a huge surplus in its bilateral trade with the United States, one in excess of $100 billion in 2003. The surplus arises because while China sells large quantities of finished goods to the United States, it does not buy that many intermediate goods from it. But China's overall (multilateral) trade surplus amounts to only some $25 billion. This tells us that in its trade with some other countries and regions, China is running a deficit. And a substantial portion of the deficit is the result of the import of intermediate goods.

The Japanese economy is playing a key part in the construction and operation of the new Asian factory of the world with its vertical division of labor. Japan's role entails the export of capital and intermediate goods embodying advanced technologies. This explains why the Japanese these days are toning down talk about how China presents a threat and is the source of deflation and are instead looking with interest at the enormous Chinese market. And in fact we can see here one of the final features of the recovery now underway in Japan. That is, Japanese exports are on the roll again because they are being put to use all across Asia in the new factory of the world. These exports are generating momentum in this third attempt to get a recovery on track because they are expanding quite briskly, not in the direction of the United States but toward China and other parts of Asia.

>> Original text in Japanese

* The article was reprinted from Japan Echo August 2004 edition. The original article appeared in the Ekonomisuto, April 12, 2004. (Translation by Japan Echo) No reproduction or republication without written permission of the author, the Mainichi Newspapers, and Japan Echo.

August 2004 Japan Echo

August 5, 2004