No. 2 — January 14, 2000

Feature Article


Douglas Ostrom

Issue Index aaaa 2000 Archive Index aaaa Subscriber Area aaaa Home


Tokyo, which maintained a policy of balanced budgets through most of the postwar period, took a dramatically different course in the 1990s. Measured as a share of gross domestic product, Japan's overall government deficit now is the biggest of any major industrialized nation. Moreover, its accumulated debt is huge and growing rapidly. Left unchecked, the rise in the debt burden could accelerate sharply in the early years of this century, partly as a result of rising interest costs but also due to demographic changes that are expanding the ranks of public pension recipients.

Japan's increasing public-sector deficit and debt could have several important effects on the United States. If large enough, the government deficit might lead to a drop in Japan's current account surplus through the interaction of interest rates, currency values and international capital flows. Tokyo assigns considerable weight — too much, according to not a few foreign analysts — to future debt burdens in its formulation of economic policy. Moreover, the existence of a sizable public-sector debt could encourage domestic and foreign savers alike to alter the composition of their holdings, giving Japanese government bonds a more prominent position in financial markets worldwide.

The arithmetic of the debt implies that policymakers can deviate from the long-term goal of stabilizing public indebtedness for years and still incur little cost in terms of the achievement of that objective. This seemingly surprising conclusion arises in part because the scope of the required policy adjustments is great enough to overwhelm the long-run negative impact of short-term policies. Whether the powers-that-be — whose frequent talk about the nature of the debt problem has not always been matched with action — have the will to push through the necessary restructuring remains to be seen, especially since the centers of political power in Tokyo are in flux.


Japan's Position In Global And Chronological Perspective

Deficits - For decades, the term "deficit" in the context of U.S.-Japan relations referred either to this country's recurring shortfall in trade with Japan or to the chronic inability of Washington to balance the budget. Japan had a long postwar history of budgetary discipline. During the two decades following the upheaval in the immediate aftermath of World War II, budget deficits were virtually unknown. Most economic historians credit — or blame, depending on their perspective — Detroit banker Joseph Dodge, who, in the famous "Dodge line" of 1948, made fiscal rectitude part of American occupation policy.

This policy, similar to Mr. Dodge's already successful recommendations for postwar Germany, was followed for years, even after Japan regained full independence in 1952. For example, from FY 1956 (the first year for which comparable data are available) to FY 1973, the budget — inclusive of central government, local government and social security-type imbalances — was in deficit only in FY 1958, when it equaled 0.1 percent of gross domestic product, and in FY 1966, when it was 0.4 percent of GDP. On three occasions during that period, the budget surplus exceeded 2 percent of GDP.1 Moreover, Tokyo was able to avoid borrowing until FY 1965.

The situation changed dramatically after the first oil crisis in 1973. Budget imbalances rose steadily, hitting a high of 4.4 percent of national output in FY 1979. In just 14 years, the amount of national government bonds had surged to 25 percent of GDP.

In the 1980s, Tokyo returned to the ways of Mr. Dodge. As a share of GDP, general government deficits declined each year after 1979 and disappeared in 1987. The peak surplus of 3.5 percent of GDP in 1990 was greater than any comparable number recorded before the first oil shock. However, the years of deficits had left a legacy of an increased volume of outstanding government bonds. Their value did not peak until 1986 at 42.8 percent of GDP.

In the 1990s, the situation reversed again as Tokyo repeatedly resorted to deficit spending to counter Japan's sluggish economic performance. The estimated 1999 financial deficit of 7.6 percent of GDP aggregated across all levels of government was larger than the comparable measure in any of the other 25 Organization for Economic Cooperation and Development members for which the OECD recently developed estimates.2 The weighted average deficit of the 26 countries was 1.2 percent. In 2000, the Paris-based group expects the Japanese deficit to climb to 7.9 percent and the average deficit for OECD members to drop to 1.1 percent.

Figure 1, based on OECD data, summarizes the deficit story since 1982 in Japan, the United States and other OECD member countries. With the benefit of hindsight, Japan's extraordinary surpluses in the 1980s were partly the result of its unsustainable economic performance during the "bubble" period, which lifted tax revenues. Time will tell whether the move toward surplus in the United States during the 1990s was equally a bubble phenomenon.3

During economic downturns, budget deficits grow or surpluses shrink as the government automatically spends more on unemployment insurance and other nondiscretionary items and, at the same time, collects lower-than-expected tax revenues.4 The OECD figures reflect the impact of the 1981-82 and 1990-91 recessions in the United States as well as the late 1980s' bubble in Japan, followed by the weakness of the 1990s, but those influences should not be overestimated. Japan had a recession in the 1985-86 span that apparently did not affect its fiscal balance.

OECD data on structural imbalances, which adjust the deficit or the surplus to what it would be at full employment and divide that result by estimated GDP at full employment, generally show the same trends of deficits and surpluses apparent in Figure 1. For example, the OECD estimated the American structural surplus to be 0.3 percent in 1999 against a structural deficit of 6.7 percent in Japan. The 1990s' trend toward surpluses in the United States and deficits in Japan is the same regardless of whether the financial imbalance measure of Figure 1 is used or the OECD's structural imbalance yardstick.

The U.S.-Japan role reversal is more than just symbolic. Deficits both reflect and influence economic conditions. Japan has a big financial imbalance in part because it spent most of the last decade trying to claw its way out of recession, implementing a series of huge government stimulus packages that combined spending increases and, on occasion, tax cuts. The most recent such initiative was unveiled last fall (see JEI Report No. 45B, December 3, 1999). The United States, by contrast, has not seriously considered using fiscal policy to boost the economy since 1993, when the newly installed Clinton administration made an unsuccessful effort to push a stimulus program through Congress. Indeed, the problem during most of the Clinton years has been an economy that might be growing too rapidly.

Deficits also matter because they affect the supply and demand for savings. Large deficits can bid savings away from businesses that otherwise might use them to build factories or to buy new equipment. This preemption of potential investment funds by bigger government expenditures or by tax cuts is known as crowding out. It is one of the reasons why deficits, while arguably helpful as a countercyclical economic tool, can be dangerous if overused.

Perhaps most significantly for Americans considering Japan's financial deficit, a fiscal imbalance absorbs savings that otherwise might be attracted to overseas projects. Through the combination of a reduced capital flow, higher domestic interest rates and a stronger yen, exports decline and imports expand, reducing Japan's current account surplus. A Japanese deficit weakens the dollar and shrinks the bilateral trade gap. If the U.S. government surplus rises at the same time, both Japan and this country will witness a double-barreled impact on interest rates, currency values and trade imbalances.

During the early 1980s, when the shoe was on the other foot, this process was obvious. The expanding U.S. budget deficit lifted real long-term interest rates here, strengthened the dollar and contributed to a soaring current account deficit. By contrast, in Japan, the budget deficit shrank. Both American and Japanese policies led to a weak yen, a rising aggregate Japanese current account surplus and an exploding trade imbalance with this country. Not coincidentally, American protectionism reached its high-water mark during the decade, exemplified by "voluntary" export restrictions on U.S.-bound Japanese-made cars in 1981 and the tough Omnibus Trade and Competitiveness Act of 1988.

The obvious question is whether the role reversal with regard to fiscal deficits will lead to a similar switch involving trade. That certainly has not happened yet. Over the long run, the answer will depend on a multitude of factors. For example, if growth in this country slowed and the anemic personal savings rate rose, private-sector demand for investment funds could drop. Together with the increasing government surplus, that could lead to reduced reliance on foreign capital and a smaller current account deficit.

Japan's continued high savings rate and slow, or even negative, rate of growth have played a role in perpetuating its current account surplus in the face of large budget deficits. However, economic recovery could lead to a lower savings rate and higher private-sector investment demand, putting downward pressure on capital flows abroad and the current account surplus. In short, the combination of American fiscal surpluses and Japanese deficits may lead to a reversal of the current account imbalances of both countries. However, that result would depend on many other factors besides budget policies.


Debt - Japan's accumulated government debt has soared as a consequence of its continued budget deficits, although those liabilities are offset in part by substantial government assets. In FY 1997, the most recent year for which data are available, government bonds and other liabilities reached ¥470 trillion ($3.9 trillion at ¥120=$1.00) against assets of ¥357 trillion ($3 trillion), implying that net liabilities were ¥113 trillion ($942 billion) (see Figure 2).

Not a few analysts have questioned this net number. If Tokyo, like other Japanese creditors, has more than the normal share of deadbeats among its borrowers, then the asset figure is too large and net liabilities are too small.5 That argument is difficult to assess, however, because the Economic Planning Agency provides few details as to the nature of the government's financial assets.

One particularly mysterious component is the "other assets" category shown in Figure 2. General government assets are divided into categories, one of which is "other." Within that category is a subcategory also labeled "other."6 In FY 1995, the increase in this residual within a residual amounted to more than the total increase in government assets. At ¥25 trillion ($208 billion) in FY 1997, it comprised about 5 percent of total government assets.

But what is this single item that represents an amount more than twice Bill Gates' net worth? An EPA official explained in 1998 that the category includes, among other things, foreign exchange reserves.7 According to published Ministry of Finance data, foreign exchange reserves totaled $223 billion (¥29.7 trillion at FY 1997's yearend exchange rate) in March 1998. This suggests a possible linkage with the subcategory designated "other assets," even though year-to-year variations in the two series do not correlate perfectly.

To the extent that these mysterious assets represent foreign exchange reserves, the debtor is the issuer of the securities in which Japan's foreign exchange reserves primarily are held — none other than the Department of the Treasury, the issuer of U.S. government bonds. So, absent an American government default, this portion of Japanese government assets is safe.

The largest single component of government assets in Japan is the social security-type money entrusted to the Trust Fund Bureau for investment in various quasi-public activities. In FY 1997, the social security and other government financial assets managed by the bureau came to ¥161 trillion ($1.3 trillion), or 37.7 percent of the total.

The Trust Fund Bureau's balance sheet as of March 31, 1998 indicated that these assets were only a small part of the total under its control (see Table). A much larger portion came from the postal savings system, the world's largest depository institution.
e--aaaaa- --aaaaa-

Trust Fund Bureau Balance Sheet (Settlement Basis), March 31, 1998

(in billions of yen)




Long-Term Government Bonds



Short-Term Government Bonds



General Account and Special Accounts



Government Financial Institutions



People's Finance Corp.



Housing Loan Corp.



Agriculture, Forestry and Fisheries Finance Corp.



Japan Finance Corp. for Small Business



Hokkaido-Tohoku Development Corp.



Environmental Sanitation Business Finance Corp.



Okinawa Development Finance Corp.



Japan Development Bank



Export-Import Bank of Japan



Local Public Bodies



Kodan (Public Corporations) and Similar Organizations



Japan Highway Public Corp.



Forest Development Corp.



Metropolitan Expressway Public Corp.



Water Resources Development Corp.



Hanshin Expressway Public Corp.



Japan Railway Construction Public Corp.



New Tokyo International Airport Authority



Japan National Oil Corp.



Honshu-Shikoku Bridge Authority



Japan Agricultural Land Development Agency



Japan Regional Development Corp.



Housing and Urban Development Corp.



Labor Welfare Corp.



Employment Promotion Corp.



Pension Welfare Service Public Corp.



Metal Mining Agency of Japan



Japan Environmental Corp.



Small Business Service Corp.



Social Welfare and Medical Service Corp.



Japanese National Railway Settlement Corp.



Corp. for Advanced Transport and Technology



The Promotion and Mutual Aid Corp. for Private Schools of Japan



Japan Sewerage Works Authority



Teito Rapid Transit Authority



Japan Scholarship Foundation



The Overseas Economic Cooperation Fund



East Japan Railway Co.



Central Japan Railway Co.



West Japan Railway Co.



Japan Freight Railway Co.












Total (excluding cash)



Postal Savings and Postal Book-Transfer Savings



Postal Life Insurance Deposits



Employee Pension Deposits



National Pension Deposits



Other Deposits









Source: Zaisei Kinyu Tokei Geppo (Ministry of Finance Statistics Monthly), July 1999, pp. 58-59. See also Ministry of Finance, Financial Bureau, FILP Report '99. Available at

Although these public and private funds are commingled in a bewildering array of operations, Trust Fund Bureau beneficiaries can be divided into five main groups:

  • The central government itself. The bureau had invested 20.2 percent of its funds in short-term and long-term government securities as of the end of FY 1997 — a figure that rose to 21.7 percent in FY 1998, according to preliminary yearend data.
  • Special accounts of the government and public-sector institutions, which represented 19.9 percent of the total in FY 1997. More than half of those funds were, in effect, loaned back by the Trust Fund Bureau to the original source for so-called self-management. For example, the postal savings system got back ¥45.7 trillion ($380.8 billion) in FY 1998. A large portion of this money also ends up in government securities.
  • Government financial institutions. They, in turn, lend money to the private sector for industrial development projects, small-business finance, housing and other purposes. These loans, which amounted to about 26 percent of the Trust Fund Bureau's FY 1997 assets, are basically commercial loans made through a government intermediary. They may or may not be riskier than loans extended by commercial banks.

    Tokyo is involved because of the perceived failure of the business community to do its job. This factor could mean that borrowers that, for very good reasons, cannot pass muster with private bankers may be able to borrow government funds. The proportion of bad loans in this group can be expected to be higher than among commercial bank borrowers, a depressing thought, given the banking industry's bad-loan problems. However, few of the loans are for land purchases or commercial real estate development, suggesting that the default rate could be lower. Some analysts long have complained that government financial institutions act too much like their tightfisted counterparts in the private sector — a good thing if it has kept potential deadbeats away.

  • Local governments, which accounted for 14 percent of the total.
  • Semipublic agencies. The character of these enterprises is reflected in their Japanese names, which include such suffixes as kodan, jigyo-dan and eidan, each of which is variously translated as corporation, public corporation or authority. They do the work of the government in building and operating highways, bridges, airports and the like. The 18.8 percent of the Trust Fund Bureau total that was lent to public corporations in FY 1997 typically is supposed to be repaid out of user fees. That might help to instill financial discipline, but the tying requirement can be a problem when politicians force these semipublic organizations to embark on dubious projects.

Such projects are the subject of periodic stories in foreign news publications. Articles are legion, for example, about bridges whose tolls approach $100, with obviously devastating effects on demand. Landing and other fees at Kansai International Airport are stratospheric, reflecting the high cost of constructing the offshore facility.


A recent article in The New York Times described how Hamamatsu and Shizuoka, two midsize cities in the same prefecture south of Tokyo and only 40 miles apart, both managed to get world-class performance halls in the space of four years. The Hamamatsu Hall was built with the equivalent of $1.9 billion in public and private money, the Shizuoka facility with the counterpart of $672 million in public money.

Another example cited in the article is a new toll road in Hokkaido, one-fifth of which was built at a yen cost equal to $1.9 billion. It gets few riders, however, because a parallel highway is free.8 The Hokkaido experience is mirrored in an example from the other end of the Japanese archipelago not mentioned in the article. In Kagoshima prefecture, a spectacular, expensive highway runs through much of the region. As in Hokkaido, it gets little traffic and for the same reason: an older, parallel road is free.

A recent article in Bungei Shunju, a popular Japanese monthly, discussed at length the financial plight of the Japan Highway Public Corp.9 In the view of the author, Hideki Kato, a professor at Keio University, JHPC's financial problems are comparable to those of the Japanese National Railways Settlement Corp. In 1998, Tokyo agreed to absorb ¥23.5 trillion ($195.8 billion) of the debt of government-owned JNR Settlement after it became apparent that the organization's revenues from the sale of land and other assets were far from adequate to service the debt it had acquired when the JNR system was privatized in 1987. Another public-sector agency, the Japan Railway Construction Public Corp., will continue to try and collect ¥4.3 trillion ($35.8 billion) from asset sales related to the former JNR (see JEI Report No. 40B, October 23, 1998).

Citing the infamous Tokyo Bay Aqualine expressway, which cost a whopping ¥1.4 trillion ($11.7 billion) to build and carries fewer than 10,000 vehicles a day (less than one-third of the initial projection) and other examples, Mr. Kato argues that JHPC has bad debts of ¥25 trillion ($208 billion), although pooling all the corporation's highways conceals this reality. Despite tolls of around ¥4,000 ($33) for a one-way trip lasting as little as 10 minutes, Aqualine revenues barely cover interest expenses. Nothing is left over for debt retirement or maintenance costs, which are said to be high. By contrast, the Keio University professor notes, one major toll road connecting Tokyo to cities to the south earns the highway corporation annual profits of ¥224.7 billion ($1.9 billion).

Mr. Kato also points out that the financial history of the Tokyo Bay Aqualine is typical in that new highways and bridges are unprofitable and are subsidized by old roads. The clear implication is that eventually, additional road construction will push the JHPC into the red, with consequences for Tokyo's budget.

Of course, as indicated, the highway agency and similar organizations account for less than a fifth of the Trust Fund Bureau's money and a smaller fraction of total government assets. Accordingly, getting a handle on government assets is tricky. In not a few instances, the assets are government bonds, meaning that liabilities can be matched against identical assets. To that extent, the assets offsetting liabilities are more than legitimate — they are the same thing. Assuming that all of the assets in the subcategory "other" are U.S. government bonds and adding to this figure funds invested directly or indirectly through the Trust Fund Bureau in Japanese government bonds point to the conclusion that roughly ¥73.5 trillion ($613 billion) of Japanese government assets, or 17.2 percent, are bonds of the two countries. This number should be somewhat reassuring to those inclined to downplay the role of government assets as an offset to liabilities.

However, the social security-like funds invested through the Trust Fund Bureau end up in a bewildering array of activities as well. To the extent that these assets carry a risk of default, they are not as secure as they might be. The highway funding example suggests that default exposure is not entirely hypothetical. Therefore, some allowance should be made in accounting for these assets to deal with the risk factor.

The biggest gamble associated with these activities — beyond the possibility that the government will lose the relatively modest amounts invested through social security and other government programs — is that Tokyo will be asked to make good on behalf of other claimants if the financed entities run into trouble. This potential liability is even larger than what is indicated by the bottom line of the Trust Fund Bureau's balance sheet. The organizations listed in the Table and others not included have raised money not just through the trust fund but also via other conduits that collectively are part of the Fiscal Investment and Loan Program. At the end of FY 1998, the liabilities of all these operations totaled ¥400.8 trillion ($3.3 trillion).

The postal savings system is an obvious and important example of a FILP creditor. Although the post office is part of the government, depositors' funds invested through the Trust Fund Bureau in the activities listed in the Table are not considered a government liability. Nor, in this case, are the resultant loans to FILP-funded organizations listed as government assets. Technically, the government is just a conduit for money flowing from private parties (depositors in the postal savings system) to borrowers that build bridges and other public infrastructure.

In case of default on the part of a FILP-financed organization, the postal savings system would appear on paper to be vulnerable. The Ministry of Posts and Telecommunications no doubt would argue that the post office, which is under its control, was doing its duty by investing in public projects and that, in any event, the Finance Ministry (never a MPT favorite anyway) misallocated Trust Fund Bureau money and/or failed to keep careful tabs on borrowers.

In this example, MPT would have millions of postal savers, who not coincidentally vote, on its side. In one way or another, the postal savings system would get bailed out. It can be argued, in fact, that the postal savings system routinely gets bailed out in the sense that many of the entities shown in the Table — including the biggest one, the Housing Loan Corp. — require large annual subsidies to balance their books. Absent such payments, the post office would receive less interest, thereby putting its obligations to depositors in greater peril. The point is, Tokyo has contingent liabilities in terms of investments and loans financed by postal savings system money or funds from other public-sector institutions, such as the government-run life insurance system, or kampo, as it is known. The worse the financial health of the organizations funded through the FILP, the greater the likelihood that these liabilities will come due.

But what is the shape of the finances of these government or quasi-public organizations? It is tempting to think that the type of infrastructure projects funded through the FILP are rock-solid. MOF recently added to this perception when it wrote that "[t]here are no nonperforming loans (loans to borrowers in legal bankruptcy, past due loans in arrears by three months or more and restructured loans) in the Trust Fund Bureau."10

This statement is not quite so outrageous as it sounds. Many Trust Fund Bureau loans are to government financial institutions and, as long as these borrowers are current in their obligations, the bureau technically does not have bad loans on its books, even if the government lending institutions do.

However, if their financial statements are to be believed, most FILP-funded organizations operate with very little in the way of a capital cushion. For example, the HLC's balance sheet for FY 1998 showed capital equal to 0.2 percent of assets,11 a tiny fraction of the 8 percent of assets that Tokyo requires of big commercial banks. As a result, under conventional accounting procedures, a few bad loans could bankrupt this FILP-financed organization. Since such an outcome presumably would require a government cash infusion and an admission of failure, the easier course of action is to pretend that few dud loans exist.

One recent headline estimated that bad loans at nine government financial institutions that receive FILP money had hit ¥1.8 trillion ($15 billion) as of the end of FY 1998.12 That would imply a bad-loan ratio of about 1.3 percent. Applying this figure to all FILP activity indicates that the government could be responsible for about ¥5.2 trillion ($43.3 billion) in contingent liabilities. The estimated number may well be too low, but even quadrupling it yields only a modest addition to the government's reported FY 1997 net liabilities of ¥113 trillion ($941.7 billion).

Even assuming that Tokyo has been as irresponsible in its lending as the banking industry — an unrealistic assumption given the different character of the two activities — the public-sector bad-loan figure balloons to only around ¥13 trillion ($108.3 billion). That is just 11.5 percent of reported net government liabilities. In short, even very substantial bad-loan problems in FILP-funded agencies, including public-sector financial institutions, would not seriously alter the overall picture of Japanese government indebtedness.


FILP Reforms

In the future, the FILP might be even less of a problem in creating contingent liabilities for Tokyo. In late 1999, the Finance Ministry developed a proposal that, over time, could substantially alter the FILP as well as the postal savings system and other suppliers of funds to the FILP. These recommended changes would put the government in a less vulnerable position. Legislation incorporating the MOF draft is expected to be enacted by the Diet this year and to become effective April 1, 2001.

The proposal borrows elements from the Federal Credit Reform Act of 1990 in requiring agencies to be forward-looking in their commitments and subsidy requirements. For example, government financial institutions serving Japan's vast number of small businesses run chronic deficits, reflecting their status as agencies created to serve a perceived public need. Until now, however, these organizations have requested subsidies only when required to do so on a cash-flow basis, meaning that commitments to subsidized lending have been made without prior Diet approval. Under the new procedures, public-sector lenders will have to calculate what is called the net present value of their initiatives. When the NPV is less than zero, indicating the need for a future subsidy, Diet authorization will be required.

Another coming reform would sever the mutual requirement that postal savings, social security-type money and other savings-like funds collected by the government be deposited in the Trust Fund Bureau for distribution to FILP-financed agencies and that the bureau essentially accept all these funds. In other words, administrators of the postal savings system would be expected to consider various investment options for deposit money, just as organizations that receive FILP money would have to look for other sources of funding. The Trust Fund Bureau would be shut down. As a result of these changes, bonds issued by the FILP or the agencies that it now finances are likely to assume greater prominence.

Together, the proposed measures have sweeping implications. They are meant to move the system from one in which MOF is responsible for investing in the FILP the pool of funds generated by the postal savings system and some other government-run programs, no matter how large these amounts, toward a framework where organizations that now obtain FILP monies become part of the fiscal policy process, expanding or contracting based on macroeconomic conditions and confirmed needs for public-sector initiatives.

The planned FILP reforms will reduce Tokyo's contingent liabilities in at least three ways. The requirement to be more explicit concerning costs is expected to make public-sector lending institutions and some infrastructure project managers more careful about financing risky schemes and, thus, to ask for less money. At the same time, funds requested to make up shortfalls probably would be more certain of timely repayment.

Moreover, the fact that the operators of the postal savings system and other programs that generate Trust Fund Bureau resources no longer will be able to dump their money in the Finance Ministry's lap means that managers, especially of the postal savings system, will have to increase their understanding of how their agency's expansion affects expected returns. For example, if postal savings surged, then the demand for securities in which to invest these funds also would increase. Greater demand would drive up bond prices, which would lead to lower yields. MOF analysts hope that postal officials will realize that incremental deposits earn a lower return than existing deposits and that they will think twice before offering higher rates to attract additional deposits. On paper at least, postal savings will be smaller than otherwise, thus reducing the government's exposure if something does go wrong.13

Finally, the proposed reforms should encourage organizations that now receive FILP funds to issue bonds with no government guarantee for sale to outsiders. For example, the biggest FILP-affiliated lender, the Housing Loan Corp., has a huge pool of mortgages that easily could become the basis for asset-backed securities to be sold to institutional investors that otherwise never would be interested in a government issue. In some cases, nonguaranteed fund-raising could result over time in the privatization of FILP-financed agencies. Either way, the government's role in these agencies would shrink, further limiting contingent liabilities.

While these arguments are valid, they must be tempered with political reality. In late December, Finance Minister Kiichi Miyazawa, Posts and Telecommunications Minister Eita Yashiro and Health and Welfare Minister Yuya Niwa agreed that the postal savings system and the public pension system would continue to underwrite half of all bonds issued by current FILP-affiliated agencies for seven years, presumably beginning in April 2001. In other words, both sides voluntarily are giving up some reliance on market forces before the FILP reforms are even enacted, let alone implemented. Similarly, Tokyo may guarantee some bonds, reducing or in certain cases eliminating the impact of market discipline.

At best, the impact of the reforms prescribed for the FILP system will be felt only slowly. Existing borrowers from public-sector or quasi-government organizations, including those with mortgages from the HLC, will not be required to alter their arrangements. In the HLC example at least, the changes will occur over a period as extended as the term of the longest mortgage, or roughly 20 years, with most of the adjustment complete in 10 to 15 years.


Why Debt Matters: The Long Run

The biggest problem inherent in the scenario of a shrinking FILP operation and a smaller postal savings system is the long-term trend of the government's debt. The postal savings system, as noted above, might be expected to contract as a result of the planned FILP reforms, but Tokyo's rising debt total could change that. Japanese government bonds are close to the ideal vehicle for the system's risk-averse, domestically oriented depositors. If the supply of these bonds surged, lowering their price and raising the rate of return, the post office would be able to offer attractive rates to its depositors, the adverse effect on yields of the new FILP structure notwithstanding.

Over time, of course, continued government deficits can become part of a vicious cycle where deficits bring more debt, which hikes interest expenses and that, in turn, lifts the deficit still higher. In extreme circumstances, the spiraling debt overwhelms all other measures of economic activity. More typically, however, governments can learn to live with modest deficits that do not get out of control.

In the 1980s and early 1990s, the possibility of out-of-control budget deficits in the United States motivated many members of Congress and three presidents to do what they could to staunch the red ink and, ideally, to eliminate it. Concern that the U.S. deficit would prove impossible to rein in was widespread at the time.

For instance, in a 1993 evaluation of the Clinton administration's budget-deficit package that became law later in the year, the Congressional Budget Office forecast that the initiative would slow the growth of the budget deficit compared to then-existing policies but not by enough to prevent an upward spiral. The national debt was projected to jump by $283 billion in 1994 and by steadily larger amounts each subsequent year. In the nine years between 1994 and 2003, the CBO estimated, the overall debt could climb to 77.4 percent of GDP from 54.9 percent.14

The CBO was careful to note that its projections were subject to a wide margin of error — a caveat that seems particularly important in light of events since 1993. A falling federal budget deficit and the surplus achieved in FY 1999 have trimmed the U.S. net government debt. In mid-1999, it was estimated at about 40 percent of GDP.

Since 1993, of course, virtually every variable that affects the national debt has contributed to the vastly improved outlook. The economy's performance has been stronger than expected, reducing expenditures and raising tax revenues — including those collected on capital gains reaped from the soaring stock market. Moreover, interest rates have stayed low.

Now, more than a few analysts argue that Japan faces the prospect of an out-of-control debt situation — not this year but sometime in the next 10 to 15 years. Based on the better-than-expected American experience, might these experts be as wrong as CBO's 1993 evaluation turned out to be? That is possible, of course, but current debt forecasts for Japan are even grimmer than they were for the United States in the early 1990s. Last summer, for example, the OECD predicted that gross Japanese government debt relative to GDP would soar more than 100 percent between 2000 and 2010.15 The 1993 CBO estimate implied a 45 percent jump in the U.S. debt/GDP ratio over the 1993-2003 period. Moreover, the potential for explosive debt growth in Japan is greater since its debt-to-GDP level was higher at the start of the relevant 10-year period than the respective figure for the United States.

Figure 3 presents three scenarios for the expansion of the net government debt in Japan. Scenario I is the OECD's estimate. It assumed 1 percent real annual economic growth and a 3.5 percent real interest rate (partly because of falling prices that lift real yields above nominal ones). Most critically, the OECD took as given a primary structural deficit (the government deficit less interest payments, calculated at hypothetical potential GDP) that does not change at 5.75 percent of GDP. Under these assumptions, the debt-to-GDP ratio will rise slowly for about a decade but then expand at an increasing rate, reaching 2.87 times national output in 2026.

If anything, this scenario is too optimistic. For example, the OECD separately estimated that public pension outlays could increase to 14.25 percent of GDP in 2020 from 6.5 percent in 1995, a 0.31 percentage point increase per year. If these gains are not fully offset by other spending cuts or tax increases, the debt problem, contrary to the OECD's assumptions, becomes much worse.16

Scenario II makes the same assumptions as the OECD baseline projection except that it allows the primary structural deficit to rise by 0.25 point each year — a rate that still reflects some offsets to rising pension outlays. Under this scenario, net government debt reaches 6.8 times GDP by 2026.

Scenario III incorporates the same gain in the primary structural deficit as Scenario II, but no economic growth also is assumed. The result is a debt that is 8.7 times GDP in 2026. The no-growth hypothesis, while seemingly extreme, is much less so in light of the argument that the potential Japanese labor force already is shrinking at an accelerating pace and predictions that Japan's population may peak in as little as seven years. These labor force considerations imply that, even with modest productivity gains, the economy could shrink.

The simulations suggest that debt levels in Japan eventually would become extraordinary by world standards if current policies continue. The implied Japanese debt-to-GDP ratios are far greater than even those incurred by the United States and Great Britain during World War II. For example, U.S. government debt reached a peak of 114 percent of GDP in 1946 before steadily falling to a low of 24.5 percent in 1974.17 British debt relative to GDP was even higher at three times GDP, a level said to approximate that nation's experience during the First World War and the Napoleonic Wars.18

Both the American and the British experiences suggest that an upward debt spiral can be broken. Moreover, under the OECD's assumptions, Japan will not reach this country's early postwar debt level before 2014. Of course, Tokyo may have a harder time cutting back spending than did Washington and London during their postwar demobilizations. Moreover, neither of the latter had to contend with Japan's current unsettling and inexorable demographic trends.


Implications Of Japan's Growing Government Debt

If Japan's government debt grows at close to the rates shown in Figure 3, several implications stand out. Most obviously, Tokyo will face increasing fiscal stress. If the debt jumps to twice GDP — a level 15 to 22 years away, depending on the assumptions — inflation is zero and the real interest rate averages 3.5 percent, then interest payments will equal roughly 7 percent of national output. Since government expenditures have varied in recent years from just over 30 percent of GDP to slightly under 40 percent, interest charges alone easily could absorb more than a fifth of all outlays. As these payments grew over time, they would add to the pressure for tax hikes.19

This kind of arithmetic puts the onus on Tokyo to bite the bullet and stop the spiraling expansion of government debt. The only question is whether to do it sooner, when the problem is more tractable, or later, when draconian changes would be required. Under the most pessimistic scenario, government debt will grow to 10 times GDP in 2029. In this situation, Tokyo would be paying the equivalent of 30 percent of GDP just in interest payments. Offsetting this expenditure would necessitate either doubling taxes, eliminating the rest of the government or a harsh combination of tax increases and spending cuts.

Even over the short term, the global role of Japan's government debt may gain importance. International investors long have viewed U.S. Treasury securities as the instrument of choice for a wide variety of purposes, in part because of their sheer volume. That could change. Figure 4 shows the amount of American and Japanese government debt. A shrinking debt in the United States and a faster increasing one in Japan have produced near parity in the amounts outstanding. Within a couple of years, Japanese debt will be larger unless the yen weakens against the dollar.

The surge in Japan's government debt has gone almost unnoticed overseas. At first glance, this is surprising, given that Japan is the world's second-largest economy. However, underdeveloped Japanese capital markets make its government bonds an unappealing choice for international investors. Moreover, the country's high personal savings rate and a general aversion to the risk of default or exchange rate changes have made Japanese government bonds easy to sell at home.

These considerations could change if the government debt began to climb along the lines shown in Figure 3. By 2010, under the OECD's assumptions, government debt outstanding will be growing at the rate of 7 percent of GDP a year, double the increase in 2001. Personal savings in Japan in FY 1998 equaled 13.7 percent of disposable income, or 9.6 percent of GDP. This means that if the household sector absorbed all the additional debt, more than 70 percent of personal savings in 2010 would have to be used directly or indirectly (possibly through the postal savings system) to buy government bonds. Most portfolio managers would regard that proportion as a severe overweighting toward one type of security. Japanese savers might agree. Any attempt to diversify portfolios would put upward pressure on interest rates, perhaps to the extent that Japanese government bonds would attract foreign attention.

NEEDS, the forecasting arm of Nihon Keizai Shimbun, the economic daily, predicted in early 2000 that the personal savings rate would fall by 2025 to a mere 3.2 percent of disposable income.20 Such a drop-off might mean that as early as 2010, the household sector would not be in a position to buy many Japanese government bonds. Businesses typically are borrowers, not suppliers of funds to capital markets. That reality implies a lack of domestic purchasers for the government issues that individuals are unwilling or unable to buy.

Under these assumptions and using the OECD's baseline scenario, by 2015, if not earlier, Tokyo, in contrast to current practice, would have to sell a substantial fraction of its bonds abroad. Moreover, companies and consumers would have few funds to invest overseas since they would be putting their money into government bonds. In short, no group in Japan would have the wherewithal to make major investments abroad. At the same time, foreign purchases of government bonds would create a capital inflow. The result — given the equality of capital flows and trade flows — could be a large and growing current account deficit.

Long before this point, however, international investors might realize that their portfolios are underweighted in JGBs. The Finance Ministry would have every reason to encourage this view. In fact, it might be expected to do what it can to make the JGB market work more like the Treasury bond market as a way of minimizing the run-up in interest rates required to whet the interest of foreign investors.21

The longer Japan runs budget deficits, the more forceful are the pressures that make JGBs a factor internationally. At least in this segment of the global financial market, Japan will have achieved one of its cherished goals: internationalizing the yen.22 However, foreign investors legitimately might worry that Tokyo could be tempted to adopt an inflationary monetary policy, thereby reducing the real value of the debt at their expense. As a result, in its formulation of monetary and fiscal policy, Tokyo might be compelled to pay more attention than it does now to the concerns of international investors.


What Should Japan Do Now?

Whatever the issue — the exponential growth of the government debt shown in Figure 3, worries about being in hock to foreigners or severe fiscal dislocations — the Finance Ministry and the government in general long have expressed concern about chronic deficit spending and the accompanying increase in the national debt. This apprehension was part of the motivation for the April 1997 hike in the consumption tax to 5 percent, for example. The argument then was that smaller budget deficits today through higher taxes would limit the rise of the debt tomorrow.

Of course, it has not worked out that way — at least not so far. The budget deficit is larger than ever. The OECD recently predicted that relative to GDP, the gap in FY 1999 would be 88 percent higher than in FY 1996 before the hike in the consumption tax. The economy went into a dive after the consumption tax was boosted and has yet to recover significantly. Tokyo has responded with several deficit-boosting stimulus packages.

Is Japan caught between the rock of stagnant growth and the hard place of out-of-control deficits? Perhaps, especially if Tokyo does not deregulate and otherwise structurally reform the economy. But one aspect of the problem is fairly clear. As Americans offering advice — typically unsolicited — argue, even big deficits are not very relevant to debt-stabilization programs later on.

Adam Posen, an economist at the Institute for International Economics, is among those who have made this argument most forcefully. He describes the quandary of some hypothetical farmers to illustrate policymakers' choices. Should these farmers use the water in the town reservoir to save this year's crop from drought or hold onto it for anticipated troubles 30 years hence? Mr. Posen advocates immediate use because the amount available this year is small relative to the long-term problem, which will require more drastic action. He concludes: "Withholding the money from the appropriate policies to respond to today's crises only harms the Japanese economy without doing anything to truly prepare it for its future burdens. It is a false trade-off."23

The analogy is not exact, however. Higher current budget deficits add to the debt, multiplying the shortfall over time. It is as if the cited farmers in future years would have more difficulty adding water to the reservoir because they had drawn it down at an early date to deal with a short-run problem. Indeed, near-term countercyclical policies pushed Japan's net government debt to an estimated 40 percent of GDP in 1999 compared with Mr. Posen's assumed level of 15.5 percent, the figure in 1996.

Nonetheless, Mr. Posen's point is valid. The effect of Japan devoting a cumulative extra 15 percent of GDP to stimulus initiatives over the first three years of a 25-year period is to increase net debt 25 years later to 3.3 times GDP from 2.9 times. The difference — 40 percent of GDP — between the two scenarios is enormous, but the gap pales in comparison with the huge resultant debt in either case. This calculation admittedly overstates the difference because it assumes that the higher budget deficits in the first three years of the period have no positive impact on the economy's performance or on tax revenues. What the illustration does show, however, is that long-run problems should not seriously constrain solutions to short-term difficulties.



Left unchecked, Japan's increasing government deficit and higher national debt eventually will create enormous problems for the world's number-two economy. Even over the framework of the next few years, rising debt levels will propel Japanese government bonds into a more prominent role in global money markets. While the former is bad news for policymakers in Tokyo and the latter perhaps good news, neither consideration should govern the decisions made to deal with the nation's slumping economy. One policy action of recent years — the boost in the consumption tax — is widely believed to have worsened Japan's economic woes in the short term and possibly in the long run as well. As such, this result serves as a cautionary tale about paying excessive attention to the long run.

In time, of course, Tokyo will have to institute major policy changes to deal with its growing debt crisis. While some observers, such as Mr. Posen, are optimistic that it will do so effectively,24 much will depend on rapidly evolving political arrangements and on the nature of power-sharing between bureaucrats and politicians and within each of these groups. The enormity of the eventual problem will provide a good test of the government's leadership. Success in that regard, either at the beginning of a new century or a couple of decades into it, is not guaranteed, however.

The views expressed in this report are those of the author
and do not necessarily represent those of the Japan Economic Institute

Top aaaa Issue Index aaaa 2000 Archive Index aaaa Subscriber Area aaaa Home


1 aa Keizai Kikaku-cho (Economic Planning Agency), Keizai Hakusho (Economic White Paper) (Tokyo: Okura-sho Insatsu-kyoku, 1999), Appendix pp. 12-13. Return to Text

2 aa Organization for Economic Cooperation and Development, OECD Economic Outlook, Preliminary Edition (Paris: November 1999), p. 68. Return to Text

3 aa The temptation to compare Japan's bubble with recent American experience is hard to resist, particularly for Japanese analysts who believe that they see important parallels. Most of the analogies are either overstated or wrong. See Douglas Ostrom, "Bubble Trouble In America: Lessons From Japan," JEI Report No. 19A, May 14, 1999. Return to Text

4 aa As such, government expenditures and taxes are said to be automatic stabilizers. At least one author has suggested that the role of automatic stabilizers is relatively weak in Japan, especially compared with the United States. See Adam Posen, Restoring Japan's Economic Growth (Washington, D.C.: Institute for International Economics, 1998), pp. 30-34. Weak automatic stabilizers imply that the budget deficit varies comparatively little over the business cycle. Return to Text

5 aa See, for example, David Asher, "Japan Sets Up Its Own Debt Trap," The Asian Wall Street Journal Weekly Edition, September 13-19, 1999, p. 16. Return to Text

6 aa Arthur J. Alexander, "Japan's Fiscal Deficit And Debt: What Happens Next?" JEI Report No. 13A, April 2, 1998, pp. 11-12. Return to Text

7 aa Ibid. Return to Text

8 aa Stephanie Strom, "Public Works and Pointless Waste Go Hand-In-Hand in Japan," The New York Times, November 29, 1999. Return to Text

9 aa Hideki Kato, "Kosoku doro-wa 25-cho no furyo shisanda (High-speed roadways have 25 trillion yen of bad assets)," Bungei Shunju, January 2000, pp. 138-150. Return to Text

10 aa Ministry of Finance, Financial Bureau, FILP Report '99 (Tokyo: 1999), p. 48. Return to Text

11 aa "Zaisei Toyushi Tokushu (Special Issue on Zaito)," Zaisei Kinyu Tokei Geppo, July 1999, p. 68. Return to Text

12 aa "Furyo Saiken Saiko 1-cho 8000 oku-en (Bad loans hit record 1.8 trillion yen)," Nihon Keizai Shimbun, November 22, 1999, p. 1. Return to Text

13 aa Some officials have offered this argument in response to the complaint that the FILP reforms fall short in not privatizing the postal savings system. In effect, they are saying, the postal savings system will have to act like a commercial institution in many respects, even if it remains public. Return to Text

14 aa Congressional Budget Office, Federal Debt and Interest Costs (Washington, D.C.: 1993), pp. 2, 55 and 56. Return to Text

15 aa Organization for Economic Cooperation and Development, OECD Economic Outlook 65 (Paris: June 1999), p. 22. Return to Text

16 aa By contrast, David Asher of the Massachusetts Institute of Technology treats some future social security-type expenditures as unfunded government liabilities. That results in a much larger figure for net government liabilities than conventional estimates. See Asher, op. cit. Return to Text

17 aa CBO, op. cit., p. 92. Return to Text

18 aa E. Victor Morgan and Ann D. Morgan, The Economics of Public Policies (Edinburgh, Scotland: Edinburgh University Press, 1972), p. 116. Return to Text

19 aa In its latest survey of the Japanese economy, the OECD uses the changes required by 2010 as an example of the magnitude of the problem. To eliminate the primary deficit by this date, which, under optimistic assumptions, would stabilize the national debt, Tokyo would have to trim expenditures or raise taxes by about 8 percent of GDP, no easy task. Organization for Economic Cooperation and Development, Economic Surveys: Japan (Paris: November 1999), p. 114. Return to Text

20 aa "Choki Tenbo (Long-term prospects)," Nihon Keizai Shimbun (American Edition), January 1, 2000, p. 15. Return to Text

21 aa See OECD (Economic Outlook 66), op. cit., pp. 170-174, for examples of the ways in which the Japanese government bond market is less sophisticated than its counterpart in the United States and, to a lesser degree, in Europe. Return to Text

22 aa Marc Castellano, "Internationalization Of The Yen: A Ministry of Finance Pipe Dream?" JEI Report No. 23A, June 18, 1999. Return to Text

23 aa Posen, op. cit., p. 80. Return to Text

24 aa Posen, op. cit., pp. 80-82. Return to Text

The views expressed in this report are those of the author
and do not necessarily represent those of the Japan Economic Institute

Top aaaa Issue Index aaaa 2000 Archive Index aaaa Subscriber Area aaaa Home