What Went Wrong

What Went Wrong
Aggregate Demand, Structural Reform, and the Politics of 1990s Japan
Robert A. Madsen
BRIE Working Paper 162
September 2004
© Copyright 2004, by BRIE
Robert A. Madsen
Senior Fellow
Center for International Studies
Massachusetts Institute of Technology
2
Abstract
What Went Wrong
Aggregate Demand, Structural Reform, and the Politics of 1990s Japan
Robert A. Madsen
This paper argues that the cause of Japan’s stagnation in the 1990s was not inefficient
corporations, a failure to implement adequate reforms, or bad macroeconomic policy.
The problem was more fundamental: a structural inadequacy of aggregate demand. By
the middle 1980s Japan was approaching economic maturity, and its savings rate should
have fallen as consumption replaced private non-residential investment as a source of
new demand. Demographic factors, however, prevented this adjustment from taking
place. Much of the population was now entering middle age, the stage of life in which
people everywhere increase their savings in order to prepare for the exigencies of
retirement. The behavior of these older people kept the savings rate elevated long past
the point where such elevation was helpful; and the country consequently suffered from a
surfeit of capital which, if not absorbed by some sector of the economy, might well have
pushed it into a prolonged recession or even a depression.
The most obvious way to resolve this imbalance would have been to ship the excess
funds abroad through a much larger current account surplus. The government, however,
could not weaken the yen in order to produce this effect because Japan’s trading partners
were complaining that it was already exporting too much and, perhaps paradoxically,
because important domestic interest groups were also opposed to a policy of aggressive
depreciation. Political considerations likewise prevented the government from enacting
structural reforms that might have lowered the savings rate, as is evident in this paper’s
review of conditions both in the overall economy and in the automobile, retail, banking,
and construction industries. By default, then, Japan was forced to rely on a combination
of excessive corporate investment and ever-larger government budget deficits as a means
of employing capital and forestalling recession. This strategy cannot be adjudged a
complete failure, for it bolstered demand and enabled the country to achieve GDP growth
of some 1% per annum. But it was certainly sub optimal, leaving the economy highly
inefficient and causing the national debt to increase dramatically. Japan would therefore
have fewer resources with which to remedy its profound structural distortions when it
finally attempted to do so in the 2000s and 2010s.
3
Acknowledgements
Among those who provided information and advice in the writing of this paper are Robert
Aliber, Jennifer Amyx, David Atkinson, Brendan Brown, James Fiorillo, Hidetaka
Fukuda, Richard Jerram, Paul Hewitt, Richard Katz, Akira Kojima, Ronald MacKinnon,
Kathy Matsui, Yukihiro Moroe, Tadashi Nakamae, Marcus Noland, Daniel Okimoto,
Hugh Patrick, Adam Posen, Takamitsu Saito, Tomoko Saito, Hiromichi Shirakawa, Yves
Tiberghian, Maria Toyoda, Robert Ward, Brian Waterhouse, and Tetsufumi Yamakawa.
I am grateful for their assistance.
4
What Went Wrong
Aggregate Demand, Structural Reform, and the Politics of 1990s Japan
Robert A. Madsen
In the late 1980s many observers regarded Japan's economy as "miraculous", capable of
sustaining high levels of economic growth and employment over many decades and
perhaps indefinitely. During these years Western universities offered courses on
Japanese business practices and pundits published countless books purporting to reveal
the wellsprings of the country's prosperity. Soon, however, the tenor of the discourse
changed. During the 1990s Japan’s growth rate fell to an average of only one percent per
annum and the country experienced two outright recessions. The costs of this slowdown
were immense, including the forgoing of considerable potential wealth, damage to the
country’s social fabric, erosion of public confidence in the government, and years of
intense international criticism. Japan thus entered the 21st century considerably poorer
and weaker than had seemed likely just a decade before.
The purpose of this paper is to explain how this reversal of fortunes occurred. Namely,
what forces caused Japan's economic performance to deteriorate so markedly in the
1990s, and why did those forces remain virulent for such a long time? To answer these
questions, the paper examines the country’s fundamental problem of inadequate
aggregate demand, the relationship between that phenomenon and the "bubble" which
inflated in Japan’s various asset markets during the late 1980s, and the effects of these
two phenomena on the industrial and banking systems. It then evaluates the
government's efforts to reinvigorate the economy. The conclusion to which the analysis
leads is that Japan's most powerful economic actors—political leaders, civil servants,
bank managers, and corporate executives—never sincerely tried to cure the country of its
malady. Fearing the implications of the necessary reforms for their own parochial
interests, those people effectively opted to sustain the status quo no matter what the cost
5
to the nation as a whole. So while it is true that political factors did not cause Japan’s
malaise, they definitely impeded the country’s recovery.
Part One: The Burdens of History
Every economic system is unique, embodying the universal mechanisms that create jobs
and produce wealth but also any number of unique cultural characteristics and historical
influences. To the extent that these idiosyncratic traits depart from the principles of
textbook economics, they usually introduce industrial and financial inefficiencies that
retard a country’s material progress. In the late 1980s and early 1990s Japan was widely
perceived as an exception to this rule; during these years many commentators believed
that the country's peculiarities represented an empirical breakthrough toward greater
productivity, profitability, and growth. In its extreme form this theory posited that Japan
had evolved a superior form of capitalism that would gain universal respect once some
theorist—an Asian Adam Smith, it was said—had elucidated its workings.1 From today's
perspective, however, these conjectures appear largely fallacious. Many of Japan's
distinctive attributes were in fact symptomatic of an ailment that ultimately transformed
the country from an object of international emulation into a cautionary example of
deflation, official profligacy, and political dysfunction.
That severe distortions were present in the economy was not obvious until quite recently.
From the 1950s through the early 1970s Japan evolved much as other parts of Asia would
later do. In these decades a large number of people migrated from the countryside to the
cities, providing a more or less constant stream of workers to support industrial
expansion.2 The supply of credit likewise increased fast during this period, for
government policy reinforced a popular predisposition toward frugality and kept interest
1 Interviews with officials in the Economic Planning Agency by the author, February 1992. Eisuke
Sakakibara, Beyond Capitalism: The Japanese Model of Market Economics (New York: Economic
Strategy Institute, 1993), p. 2.
2 Kent Calder, Crisis and Compensation (Princeton, NJ: Princeton University Press, 1988), p. 172.
6
rates lower than they would otherwise have been. This, in turn, facilitated a great deal of
capital spending. Gross capital formation—meaning the sum of government, corporate,
and housing investment—rose in real terms from 13.1% of GDP in 1955 to a peak of
32.8% in 1973. This was a very high level, to be sure, but it was appropriate for a
developing country with an insufficient stock of plant and equipment.3 During this
period Japan also enjoyed easy access to advanced Western technologies through trade,
licensing agreements, and other media of exchange. The combination of these
sophisticated technologies with the country's abundant savings and labor proved a potent
mix: from 1955 to 1973, its GDP grew at an average real rate of just over 9% per annum.
South Korea, Taiwan, and other East Asian countries would later demonstrate
comparably rapid modernization, but Japan remained the single big economy to achieve
such a feat.
At some point, though, the Japanese economy should have gone through a structural
change in which the balance of commercial activity shifted away from investment and
towards consumption. There were two reasons for this. First, by the early 1970s rural
Japan had shed most of its surplus labor, and wages in industrial centers were beginning
to rise.4 As corporations’ payroll obligations grew more onerous, the number of potential
investments that would yield more revenue than they cost decreased. Firms’ investment
budgets should accordingly have begun to shrink. The second reason that Japanese
corporations should have become more conservative was the country's attainment,
sometime in the late 1970s or early 1980s, of the global technological frontier as defined
by the United States, Germany, and other mature economies.5 Having adopted most of
the world's best commercial ideas and production techniques, Japan could no longer
3 This paragraph relies on the national accounts in 68SNA, published by the Cabinet Office’s Economic
and Social Research Institute (ESRI), which employs a price index based on 1990 prices and extends back
to 1955. The figures used in the rest of this paper are from 93SNA, which is indexed to 1995 prices and
reaches back to 1980.
4 Richard Katz, Japan: The System that Soured (Armonk, NY: ME Sharpe, 1998), pp. 206-10. Brendan
Brown, The Yo-Yo Yen and the Future of the Japanese Economy (New York: Palgrave, 2002), pp. 63-5.
5 Compare: Porter et al., pp. 16,118; Mikitani, “The Facts”, in Ryoichi Mikitani and Adam S. Posen, ed.,
Japan’s Financial Crisis and Its Parallels to US Experience (Washington, DC: Institute for International
Economics, 2000), pp. 29,31; Katz, pp. 131-8; and Brown, pp. 64-5.
7
create new industries through mere imitation: its enterprises must now devote more time
and money to the incremental improvement of existing goods and services and to the
invention of entirely new products. Technological progress would consequently become
slower and more expensive and, again, the scope for profitable investment would narrow.
Confronted by these two trends, Japanese corporations should have curtailed their capital
expenditures and started returning more of their earnings to workers through better
salaries, and to shareholders through more generous dividends. The household sector’s
share of national income would then rise, and people would be able to reach their
financial goals even as they spent more money on material comforts. Consumption
would thus come to replace capital spending as a source of marginal aggregate demand,
and GDP would continue growing at a respectable, if somewhat slower, pace.
In the event, politics interrupted this process of structural maturation. By the early 1970s
agriculture, small businesses, and some of the other groups on which the ruling Liberal
Democratic Party (LDP) had historically depended for donations and electoral assistance
were beginning to decline.6 The Oil Crisis of 1973-1974 exacerbated this difficulty by
causing energy prices to surge in a manner that threatened broad swathes of corporate
Japan. Rather than allow these forces to thin the ranks of its supporters, the LDP pursued
a dual strategy of shielding its existing friends from market forces even as it extended its
patronage to a range of new industries. Among the tactics employed in these efforts were
subsidized loans through the official budget; generous grants for research and
development which, in effect, sheltered companies from competition; and the
organization of formal and informal cartels in parts of the economy that suffered from
excess capacity or especially intense competition.7 Yet another tactic was increasing the
size of the Fiscal Investment and Loan Program (FILP), a sort of unofficial “second
budget” over which the elected authorities had more control than the regular budget and
6 Calder, pp. 104-9,174-5.
7 C. Fred Bergsten, Takatoshi Ito, and Marcus Noland, No More Bashing (Washington, DC: IIE, 2002), pp.
131,142,144-9. Akiyoshi Horiuchi, "The Big Bang: Idea and Reality," in Takeo Hoshi and Hugh Patrick,
ed., Crisis and Change in the Japanese Financial System (Boston, MA: Kluwer Academic Publishers,
2000), p. 236. Yoshinori Shimizu, "Convoy Regulation, Bank Management, and the Financial Crisis in
Japan", in Mikitani and Posen, pp. 58-60. Calder, pp. 155-7.
8
which they routinely used to reward favored companies and industrial groups.8 During
the 1970s and 1980s, furthermore, the Ministry of Finance (MOF) continued to depress
interest rates and obliged commercial banks to channel credit into sectors that tended to
sympathize with the ruling party. So even as some industries were undergoing
deregulation and reform, in many other areas a "hidden . . . safety net" was being "built
into the structure of the private political economy” in order to protect the Liberal
Democrats and their corporate allies.9
Whether or not this was a nefarious arrangement depends on one’s point of view. On the
one hand the cost of subsidizing weak enterprises was tolerable as long as the economy
was growing fast, and few people were upset by marginal sacrifices made in the name of
social stability. Indeed, those many Japanese citizens who appreciated political
quiescence and high levels of employment might well have approved of the new
emphasis even if they clearly recognized its costs. In this sense the widening of the
LDP’s electoral base to include ever more parts of the economy could be portrayed as a
reasonable distributive outcome rather than just a cynical political maneuver.10 On the
other hand, over time Japan would pay a steep price for its system of disguised
protectionism and the benefits it conferred. This is evident in two related phenomena:
the country’s worsening industrial inefficiency, and a savings rate that stayed elevated
long past the stage at which such elevation was helpful.
[Chart 1: Japan's Dual Economy]
In the industrial economy, the effect of intimate cooperation between the LDP, the
bureaucracy, and much of the business community was consistently to vitiate the market
forces that might have redirected resources from feckless enterprises to more profitable
8 Maria Toyoda, Why Thinking Globally Makes It Difficult to Act Locally (and Vice Versa): The Political
Economics of Financial Liberalization and National Elections (Ph.D. dissertation, Georgetown University,
1998), pp. 82-4,86,94,165.
9 Katz. p. 105; also pp. 45,82-3,86-96,157-60,166-9,170-7,179-95. Calder, pp. 112-4,124,156-7,160-
1,163,166,172,175-9,231. Horiuchi, p. 236, likewise describes the financial system as a “safety net”.
10 Calder, pp. 155-7. Also, Sakakibara, Beyond Capitalism, p. 2; and Adam S. Posen, Restoring Japan's
Economic Growth, (Washington, DC: Institute for International Economics, 1998), pp. 33-4.
9
and expansive ones. The result was the gradual emergence of what Michael Porter and
others have termed "two Japans”. The more impressive of these comprised those
manufacturers and exporters that competed in global markets and hence were compelled
to meet, or exceed, international standards of excellence.11 Many of these still number
among the world’s most renowned corporations. The weaker “Japan”, by contrast,
included such domestic industries as "retailing, wholesaling, transportation and logistics,
construction, energy, health care services, and food preparation" as well as the banking
and securities sectors. Reclining upon the “hidden safety net”, these predominantly “nontradable”
industries were able to survive, and in many cases prosper, even though they
were often extremely irresponsible in their use of labor and capital. This misallocation of
resources was in fact a grave problem. Not only did it undermine the profitability of the
backward reaches of the economy, but it also harmed Japan’s most successful firms. For
while the country’s best enterprises were largely global in orientation and competence,
geography compelled them to procure many of their supplies from local companies at
prices that were well above than those which prevailed abroad.12 Thus the inefficient
“Japan” impaired the performance of the efficient one and dragged down the productivity
of the entire economy.
Coincident with, and essential to, this progressive distortion of the Japanese industrial
system was the country’s stubbornly high private savings rate. This feature of the
economy arose from decisions made independently by both corporations and households.
Among the many factors that companies consider in deciding how much of their earnings
to save are the range of available investment opportunities, the size and variability of
their own income flows, and the costs of raising capital from external sources. But the
mechanisms of corporate governance are also important, for if they function poorly firms
will feel less pressure to disgorge their surplus funds to shareholders. Instead, they may
11 Michael Porter, Hirotaka Takeuchi, and Mariko Sakakibara, Can Japan Compete? (Cambridge, MA:
Perseus, 2000), p. 6; see also pp. 23-6,34-65,11-6. Compare McKinsey Global Institute (MGI),
Manufacturing Productivity, (Washington, DC: MGI, 1993), Executive Summary, pp. 1-2; and Why the
Japanese Economy Is Not Growing: Micro Barriers to Productivity Growth, (MGI: July 2000), Executive
Summary, p. 1, and Synthesis, p. 30.
12 Katz, pp. 47-8,189,196. Bergsten et al., pp. 116-7. Porter et al., p. 12. This remained true through the
early 2000s. Nikkei Weekly, 13 August 2001.
10
squander their cash on wasteful internal projects, deposit it in bank accounts, or use it to
speculate in equities, bonds, and real estate. In the United States and some European
countries, firms with large stocks of such assets would be subject to intense shareholder
criticism and perhaps to hostile takeovers by outside investors who reckoned that the
money could be used to greater advantage elsewhere. In Japan, however, the "main
bank" system, ubiquitous webs of interlocking shareholdings, and the government’s
predisposition to protect wasteful corporations and their managers meant that market
discipline was not very strict.13 There was nothing, in short, to stop companies from
hoarding their earnings. So rather than falling as the Japanese economy matured and the
expected profitability of potential investments decreased, the corporate rate of savings
remained buoyant. Indeed, in many years businesses were basically self-financing and
did not need to borrow from the household sector at all.
[Chart 2: Sources of Savings]
A different set of variables determines a country’s household savings rate. Nationality
may be significant in this regard, for in certain periods some peoples exhibit more
parsimony than others. A society’s level of economic development also plays a role. For
whereas the citizens of less developed countries often save a high proportion of their
income in order to exploit the rich investment opportunities they enjoy, workers and
investors in fully industrialized economies have fewer such opportunities and so tend to
save less. In Japan’s case, though, the critical factors were the unique attributes
described above: namely, demographics and the insidious effects of the “hidden safety
net”. According to the “lifecycle hypothesis” of financial behavior, which is empirically
valid across most countries, people’s savings rates vary with their ages.14 Young people
and new families generally spend more on food, clothing, housing, and education than
13 Bruce Stokes, A New Beginning: Recasting the US-Japan Economic Relationship (New York: Council
on Foreign Relations, 2000), pp. 63-4. Also, Takeo Hoshi and Hugh Patrick, “The Japanese Financial
System: An Introductory Overview”, in Hoshi and Patrick, p. 7; and Mikitani, “The Facts”, p. 29.
14 MGI, The Global Capital Market: Supply, Demand, Pricing, and Allocation (Washington, DC: MGI,
1994), Chapter 3, pp. 4-8.
11
they earn, so their savings rate is negative. The Middle-aged, meanwhile, typically earn
more as a group than they immediately require and so are able to put aside considerable
sums of money in preparation for the exigencies of retirement; this group is usually a
country’s most sedulous savers. Once those workers quit their jobs and join the ranks of
the retired, however, their income drops off precipitously and they begin “dissaving”, or
running down their accumulated wealth. All other things being equal, then, a country’s
household savings rate will be low if its population is disproportionately young or old,
and high if a large percentage of its citizenry is middle-aged.
[Chart 3: Comparative Gross Savings Rates]
This analysis explains much of what happened in Japan. In the 1970s the household
savings rate started to decline, as one would expect of a largely industrialized country
that was approaching the technological frontier. But then demographic realities
intervened. Over the next two decades the number of people in the most provident age
bracket—those who were 45 to 60 years old—surged by an enormous 42.1%.15 This
remarkable change implied that the Japanese household sector would persist in saving
more, and consuming less, than would normally be the case for so fully developed an
economy. The demographic trend will reverse early in the 21st century, when those many
middle-aged workers leave the labor force and begin spending the wealth which they
have so assiduously acquired, but the effect of their actions during the period discussed in
this paper was to keep Japan awash in private-sector capital.16
The “hidden safety net” underscored this phenomenon. People save money for specific
purposes: to purchase homes, for instance, to finance children’s educations, or to
facilitate comfortable retirement. Explicitly or implicitly, individuals have targets for
how much money they will need to attain these goals. It follows, therefore, that anything
15 More precisely, the change occurred between 1980 and 2000. UN Population Division, World
Population Trends, 2000 Revision.
16 The ranks of people above age 65 will swell from 28% of the population in 2000 to 38% in 2010, and
then to 50% in 2020. Ibid., and The Economist, 5 July 2003, p. 67.
12
that retards the speed with which households grow more affluent will compel them to
save more of their income over a longer period of time than they would otherwise choose
to do. The “hidden safety net” had precisely this effect. In part because of official
regulation, the real interest rate earned on bank deposits from the 1950s through the
1980s was exceedingly low—as, incidentally, were the returns offered by the postal
savings system, which was itself essentially a huge public bank.17 In addition, the
government’s policy of cosseting troubled companies through the provision of cheap
credit and other forms of assistance enabled those firms to stay in business, where they
monopolized labor and capital that should ideally have been redeployed to more
productive uses. So corporate profitability worsened, and the returns households earned
on their stock and bond portfolios were pushed down to levels much lower than in
comparably sophisticated economies.18 The situation improved slightly during the
collective transport of the late 1980s, but then deteriorated precipitously when the asset
bubble burst in the early 1990s.19 The losses which households subsequently incurred
forced them to forswear anything more than a modest, gradual decrease in their savings
rate through the end of the century.
However rational this frugality was for individual households, the very high private
savings rate which it entailed was a bane for the nation as a whole. This was true not
only because the surfeit funds lowered the cost of capital and facilitated the government’s
subsidization of value-destroying corporations, but also in the more fundamental sense
that surplus savings are equivalent to inadequate aggregate demand. It is a basic
principle of economics that when too many actors postpone spending for too long,
corporations cannot sell their output and react by shutting down plants and offices, laying
17 Robert Aliber, Forward, in Brown, pp. xxviii,xxix,xxxiv; and Brown, pp. 66-7. Aliber calculates that
that the expected real rate was only 0.1% in the 1960s, -2.9% in the 1970s, and 0.6% in the 1980s. Letter
to the author, January 2002.
18 While acknowledging that the data are too sparse to admit firm conclusions, Albert Andoh surmises that
individual investors may actually have lost money on their securities holdings between 1970 and 1985, and
again in the 1990s. Albert Ando, "On the Japanese Economy and Japanese National Accounts" (National
Bureau of Economic Research, Working Paper 8033), December 2000, Appendix Table 2B. MGI, 2000,
Synthesis, p. 2.
19 In the 1990s households lost some ¥500 trillion, or roughly a year’s GDP, in wealth. Bergsten et al., p.
67.
13
off workers, and cutting back on investment in a way which perforce causes GDP to
shrink. This recessionary pressure then encourages households and companies to reduce
their spending still more, which further attenuates aggregate demand. To prevent this
contractile cycle from causing a recession or depression, it is necessary for some sector of
the economy to borrow the excess funds and expend them on actual goods and services.
There are, however, only a few possible destinations for this money: it might flow
abroad to finance a trade surplus, be used by the government to support deficit spending,
or be wasted by companies in the form of unwise capital expenditures. In the late 1980s
and 1990s Japan would employ all three of these alternatives, albeit with decidedly mixed
results.
Part Two: The Bubble and Its Implications
The bubble that inflated in Japanese asset prices in the late 1980s has been described in
many publications, and a detailed treatment of that misadventure lies beyond the scope of
this paper. Nevertheless some elements of the subject are crucial to an understanding of
Japan's attempts at financial and industrial restructuring during the 1990s and so merit
brief description here. Most importantly, the bubble encouraged companies to enlarge
their investment budgets and this, along with a sizeable trade surplus, enabled GDP to
keep growing swiftly and the country to avoid protracted recession. Less desirably, the
passage of so much credit through an industrial system that was already deeply flawed
had the collateral effect of aggravating Japan’s structural inefficiencies. This unpleasant
fact manifested when the bubble imploded in the early 1990s, initiating an extended
slowdown in corporate investment and exposing anew the country’s underlying
macroeconomic imbalance. The ultimate import of the bubble, therefore, was to buy
Japan several years’ prosperity at the price of mortgaging its future.
In the process of maturation, as noted above, a country’s potential growth rate gradually
slows as it converges with those of the other advanced, industrialized economies. While
researchers differ as to when exactly Japan arrived at this watershed, there is fairly
14
unanimous agreement that it had done so by the middle 1980s.20 From that point forward
its economic performance should have been dictated by roughly the same forces as were
operating in the United States and Germany. The experience of those countries, in turn,
suggests that a mature nation should invest some 10-12% of its GDP in private
productive capacity each year. Spending below this rate leaves the country with too little
plant and equipment and stops it from growing as fast as it could whereas spending over
it for any significant length of time eventuates in chronic oversupply, poor corporate
profitability, and recession as firms belatedly decrease the scale of their operations to a
level commensurate with demand.21 But if private capital expenditures stayed in the
target range, they usually add enough to the productive stock to enable GDP to grow at 2-
3% per annum without engendering any major imbalances in the industrial and financial
systems. It consequently made perfect sense that Japan's rate of private non-residential
investment would decline to 13.0% of national output in the first half of the 1980s and
that the pace of the economy’s expansion would slow to 3.0% per annum. Such
deceleration was just what one would have expected from a country at that stage of
industrialization.22 Less understandable, however, was the fact that in the latter half of
the decade Japanese corporations dramatically increased their capital expenditures and
the rate of GDP growth rose to a yearly average of 4.7%. This sudden acceleration of
commercial activity contradicted the empirical precedents and raised serious questions
about how Japan was allocating its resources.
20 Porter et al., pp. 16,118. Mikitani, “The Facts”, pp. 29,31. Katz, pp. 131-8. Brown, pp. 64-5.
21 The United States, for instance, exceeded that level from 1996 through 2000 and subsequently suffered a
long economic slowdown as corporations reduced their capacity and waited for demand to take up the
slack: hence the “jobless recovery” of 2003-2004. See Chart 4.
22 93SNA, Government Cabinet Office, ESRI.
15
[Chart 4: The “Bubble” and Asset Prices]
The solution to this puzzle lies in Japan's capital markets. In the middle and late 1980s
the other G7 nations—and particularly the United States—were demanding that Japan
stimulate its economy so as to suck in more imports and reduce its current account
surplus. Conservative officials in the MOF, ever concerned about the government’s
finances, argued that this should be done not through deficit spending but rather by
loosening monetary policy. So the Bank of Japan (BOJ), which was then statutorily
subordinate to that ministry, increased the money supply somewhat more quickly than its
internal rules would normally have permitted. Abetted by careless regulators, the
commercial lenders amplified the impact of this change in monetary policy by extending
new loans to good borrowers and bad ones alike. Vast growth in liquidity ensued, and
much of this poured immediately into Japan’s various asset markets. From March 1985
to their respective peaks, the price of commercial real estate in Tokyo rose by 155% and
the value of the Nikkei Average more than doubled. Such was the genesis of Japan's
infamous asset bubble.
Besides enriching those who owned corporate equities and property, the sudden growth
in the money supply also worked a change in firms’ investment decisions. Since the
prices of goods and services were relatively stable at the time, the flood of new money
caused real interest rates to fall.23 This translated into a lower cost of capital, which then
motivated companies to amass more plant and equipment than they would otherwise have
done. The once-celebrated propensity of Japanese companies to prioritize the acquisition
of market share over short-term profitability was thus reinforced, and the pace of capital
formation quickened.24 By 1988 the ratio of private non-residential investment to GDP
had risen to 16.5%, and in 1990 and 1991 it would surpass 19.0%. This was almost half
again what the country had expended on productive facilities a decade earlier, and it
23 Olivier Blanchard, "Bubbles, Liquidity Traps, and Monetary Policy" in Mikitani and Posen, pp. 186-7.
24 Ibid. Also, Shimizu, p. 60; Katz, pp. 215-6; and Porter et al., pp. 75-7.
16
greatly exceeded the ratios that obtained contemporaneously in Germany and the United
States.
[Chart 5: Corporate Investment]
In immediate macroeconomic terms, this was an auspicious turn of events. If Japan had
followed the normal trajectory for an industrializing economy, corporate investment
would have continued to decrease through the late 1980s and there would soon have been
insufficient demand to exhaust all of the country’s domestic savings. The textbook
solution to this problem would have been for the unneeded money to flow abroad so that
foreigners could procure more of Japan’s output. Unfortunately, Japan’s trading partners
were already complaining vociferously that it was exporting too much and hinting at the
possibility of countervailing protectionism; they would doubtless have reacted harshly if
net exports grew much further.25 There was also some question, as a practical matter,
whether the world could have adapted to what must, given the size of the Japanese
economy, be an immense increase in the global supply of funds. Accommodating such a
change would have forced many countries’ current accounts into deficit, pushed up their
unemployment rates, and in some cases provoked serious domestic strife. So Japan faced
a daunting task in the middle 1990s: assuming that it could not channel much more of its
savings abroad, it must find a domestic destination for its capital or watch as the
weakness of aggregate demand pulled the country into a long and possibly deep
recession. The loose monetary policy and lax banking practices of the bubble period
resolved this dilemma by stimulating rapid growth in corporate investment, which rose
far above the efficient level and thereby added considerably to aggregate demand and to
GDP. Thus the same dynamics that inflated the asset bubble also produced the economic
efflorescence that characterized the years after 1985.
Needless to say, few objected to this faster growth. Corporations benefited from Japan’s
new vitality insofar as their revenues increased and pressure to restructure abated; banks
25 Bergsten et al., pp. 34-6,100-3.
17
gained from strong loan demand and from appreciation in their stock portfolios; and
households prospered because employment and wages remained strong even as the
returns generated by their stock and real estate investments improved.26 Their spirits
buoyed by this upturn, the Japanese people curtailed their savings somewhat and
consumed a bit more of their income, thus further augmenting aggregate demand. The
euphoria of the bubble years also redounded to the advantage of the Liberal Democratic
Party, which took credit for the economy's resilience and managed to stay in power
despite a series of corruption scandals that might have proved disastrous had they
coincided with a deceleration in GDP growth.27 There were occasional signs that not all
was well; some commentators warned, for instance, that Japan’s reliance on copious
capital investment was dangerous and that the financial system was foundering. But
these were voices in the wilderness. Almost all of the country’s businessmen,
bureaucrats, and political leaders were preoccupied with their quotidian responsibilities
and saw no need to focus on what looked like trivial defects in the industrial and banking
sectors.
The passage of time, however, would soon dispel this insouciance. In 1989 the
mandarins in the BOJ started hiking interest rates in order to dampen the market’s
enthusiasm and stop the economy from overheating. Stock prices accordingly peaked in
the summer of that year and then started to fall. By 1993 the Nikkei average had lost
55% of its value and the price of commercial property in Tokyo had plummeted 47%.
This decline in asset prices would in fact last through the end of the decade, eventually
engendering a bout of comparably destructive price deflation.
The significance of these events was not entirely obvious in the early 1990s. It is always
difficult to identify a bubble during its inception or indeed for months after it begins to
26 Andoh, pp. Appendix Table 2B. Martin Wolf, in Financial Times, 9 May 2001.
27 The Recruit Scandal occurred in 1989 (Schlesinger, pp. 233-7), and soon thereafter the LDP lost control
of the House of Councilors ( ibid., p. 248). The early 1990s brought the Sagawa Kyubin Scandal (ibid., pp.
245-7,251); a mass defection from the LDP (ibid., pp. 264-70); and the Liberal Democrats’ 1993 loss of
control of the House of Representatives, and hence of the government. The LDP returned to power several
months later (ibid., pp. 271-7), but their sojourn in opposition had persuaded the party’s leaders that their
power was tenuous and that controversial policies should be eschewed.
18
collapse—witness Allen Greenspan’s ebullient public comments in October 2000, long
after the American technology bubble had begun to implode.28 So too, a decade earlier,
did Japanese executives, regulators, and politicians feel confident that their economy
would soon resume its erstwhile fast growth. The mood in Tokyo was still sanguine as
late as 1993, when Eisuke Sakakibara, the prominent MOF official, published the English
version of his triumphal Beyond Capitalism, which contended that Japan’s economic
system was superior to Western capitalism and should be imitated by the rest of the
world.29 To such commentators it seemed that the necessary adjustment was well
underway, for not only had asset prices dropped sharply but corporations were also
shrinking their investment budgets and promising to implement ambitious cost-cutting
schemes. If one assumed—as these optimists generally did—that Japan’s potential
growth rate remained as high as those of many developing nations, logic suggested that
the proposed reforms would soon restore the economy to health.
This optimism, though, failed to account for several facts. One was the sheer magnitude
of the Japanese bubble, which dwarfed those experienced by many other countries and
would have entailed a lengthier period of recuperation even in the absence of other
complications. Another was the deep involvement of the commercial and retail banks,
for this implicated the mechanisms of credit creation and impaired the efficacy of
monetary policy, thus intensifying Japan’s deflationary tendencies. Still another was the
nexus of interests shared by Japan's older and less efficient corporations and banks, the
civil service, and the LDP inasmuch as cooperation between these groups inhibited
institutional change and discouraged the reallocation of human and material resources
that would have helped produce a lasting recovery.30 But most important was the
inadequacy of aggregate demand. What Sakakibara and so many others failed to perceive
28 Allan Greenspan, “Challenges for Economic Policymakers”, a speech at the 18th Annual Monetary
Conference, Washington, DC, 19 October 2000.
29 The Japanese edition is more blatant in this regard, but the sense of superiority is still evident in the
English. Sakakibara, Beyond Capitalism.
30 Mikitani, “The Facts”, p. 29. Porter et al., pp. 4-14,75. Sakakibara admits the nexus, if not its
disadvantages. Sakakibara, Beyond Capitalism, pp. 5,28,45,62-3.
Footnote continued on the next page.
19
in the early 1990s was that the bubble had concealed a structural surplus of savings and
that its collapse was both uncovering and exacerbating that vulnerability. For as
households and corporations grew poorer, they reacted by curtailing their expenditures
and expanding their savings and this caused the gap between aggregate supply and
demand to widen.31 As urgently as at any time in the past, Japan needed to find an outlet
for its surplus capital lest its economy stagnate or start to contract.
Part Three: Industries beyond Capitalism
In the early 1990s Japan faced an immense challenge. Over the longer term the best way
to lower the savings rate was to restructure the country’s more backward industries so
that they wasted less of their earnings on useless capital projects and disgorged more
wealth to workers and individual investors, who could then attain their financial goals
while also spending more on goods and services. In the short term, though, such
restructuring would bring bankruptcies, layoffs, anxiety, a diminution in consumption
and investment, and very possibly a recession. To make a reform program work,
therefore, the government must loosen fiscal or monetary policy to protect households
until the restructuring process reached fruition in a higher rate of GDP growth. Both of
these stimulatory options, however, were foreclosed by political factors. Foreign
governments, as explained above, objected strenuously to the prospect of aggressive
monetary easing whereas officials in the powerful MOF adamantly opposed the prospect
of big budget deficits. Japan was thus left with no macroeconomic means of facilitating
industrial and financial reorganization. Further complicating this predicament was the
belief, both at home and abroad, that the economy was basically sound and required
neither fundamental reconstruction nor bold stimulus. So Tokyo was inclined to
temporize, adopting marginal reforms and modestly expansionary fiscal policies while
31 The year-over year rate of growth in consumption averaged 5.3% in 1990-1992, then slowed to 2.2% in
1993-5 and to 1.1% in 1997-1999. The propensity to consume declined from 75.3% in 1990, to 72.5% in
1995, and then to 71.3% in 1998. Tokei Geppo, November 1997, pp. 2,10; and November 2000, pp. 4,14.
20
eschewing comprehensive change of the sort, which might have threatened the country’s
political and social institutions.
It is easy to demonstrate, on a macroeconomic level, that Japan undertook little
restructuring during its “lost decade”. The most conspicuous symptom of its illness
during the late 1980s had been inordinate spending on plant and equipment, and this
profligacy continued throughout the following decade. It is true that the ratio of private
non-residential investment to GDP fell from 19.1% in 1991 to only 14.4% in 1994, but
this was still well over the rate appropriate to a country of Japan’s maturity. Put simply,
many industries were still amassing capacity that they could never realistically hope to
employ. The situation worsened further in 1995-1997, when a series of anomalous events
seemed to portend a return to the torrid growth of the bubble period and persuaded the
corporate sector to increase its capital expenditures to over 16% of GDP in 1997 and
1998.32 But then, unfortunately, the pace of economic activity decelerated again and
Japanese and foreign economists realized that much of the new productive capacity was
entirely superfluous. As one writer would later quip, corporate investment in these years
was "so inefficient that it might as well be considered consumption".33
[Chart 6: Capacity and Utilization]
There are many ways to gauge the effect of this rampant capital spending. For the
industrial economy as a whole, data compiled by the Ministry of International Trade and
Industry (MITI, subsequently METI) reveal that Japan was exceedingly slow in adapting
when aggregate demand slackened in the early 1990s.34 The ministry’s utilization index
shows that, as one would expect, the bursting of the bubble depressed consumption and
investment and hence lowered the operating rate for existing facilities quite sharply.
Manufacturing activity would remain subdued for the rest of the decade. Yet despite this
prolonged weakness in demand, corporations kept adding to Japan’s stock of plant and
32 On the unsustainability of this apparent recovery, see note 101.
33 Martin Wolf, Financial Times, 16 February 2000.
21
equipment, as reflected in MITI’s capacity index, until November 1997. Moreover, the
subsequent cuts in investment budgets were so small that the total volume of plant and
equipment at the end of the decade was exactly the same as it had been in late 1992. On
a net basis industrial Japan had managed to achieve no restructuring.35
[Chart 7: Capital Productivity]
[Chart 8: Profitability of the Japanese Economy]
The failure to restrict capacity at a time of flagging demand further vitiated the economy's
efficiency and profitability. This is apparent in the ratio of GDP to productive capital,
which measures how much output a country can generate per unit of plant and
equipment. This indicator demonstrates that Japan's operational efficiency started to
erode in the 1970s, when the LDP became more magnanimous in its dealings with
powerful vested interests, and that the erosion persisted through the end of the century.
The pace of the decline slowed somewhat during the bubble years, when pigs could and
did fly, but then accelerated again in the early 1990s. Studies of Total Factor Productivity
(TFP) confirm this general pattern, as do a wide variety of anecdotal evidence.36 The
same is true of the Return on Equity (ROE) earned by Japan’s big non-financial firms,
which fell from 8.2% in 1988 to an average of 3.1% between 1992 and 1999. Estimates
by Goldman Sachs suggest that at no point during this period did Japanese industry earn
enough to cover its cost of capital, which means that the average company was actually
34 The name of this agency was subsequently changed to the Ministry of Economics, Trade and Industry, or
METI.
35 For an even more pessimistic conclusion, see: Robert Feldman, "ROA Whodunit: Every Nook and
Cranny", in Morgan Stanley, Global Economic Forum, 31 May 2000, pp. 3-6.
36 Compare: Bart van Ark and Dirk Pilat, "Productivity Levels in Germany, Japan, and the United States:
Differences and Causes", Brookings Papers on Economic Activity: Microeconomics No. 2 (Washington,
DC: Brookings Institute, 1993), pp. 56-73; Okubo, Sumiye, Prospects for Growth in Japan in the 21st
Century (Washington, DC: Department of Commerce, 1996), p. 36; and OECD, Economic Surveys: Japan,
Volume 2002, Supplement No. 2-January 2003, p. 105. Anecdotally, studies by McKinsey & Company
estimated that Japan’s capital productivity declined marginally from 63% of the US average in 1990-1993
to 61% in 1999 while its labor productivity fell from 83% of the American standard in 1990 to only 69% in
1999. MGI, Manufacturing Productivity (Washington, DC: MGI, 1993), Executive Summary, pp-2,
Exhibit S-1; and Conclusion, pp. 1-3. MGI, Capital Productivity (Washington, DC: MGI, 1996), Synthesis,
pp. 1-5 and Exhibit 1. MGI, 2000, Aggregate Analysis, p. 7 and Synthesis, pp. 1,30. See also IHT, 11
August 2000.
22
destroying value.37 The macroeconomic record of the 1990s was thus one of chronic
over investment and overcapacity and consequently of little corporate contribution to
sustained GDP growth.
[Chart 9: Automobile Industry: Capacity and Output]
Microeconomic analysis of particular industries—automobiles as an example from the
world-class export sector, and retail stores as typical of the backward domestic sector—
not only vindicates these macroeconomic conclusions but also elucidates their causes.
Some of Japan’s finest enterprises, and the object of much international emulation, were
its car and truck manufacturers. Productivity among these firms was consistently
superior to that of their competitors in Germany and the United States.38 By the middle
1980s, however, the industry was already showing some signs of overcapacity. This
problem was manageable as long as the bubble inflated demand for vehicles at home and
a stable yen promoted exports abroad; auto production accordingly peaked at 13,487,000
units in 1990 and slowed only slightly to 13,245,000 in 1991. Assuming that this
represented the textbook definition of "full employment", or 85% of total capacity, then
the industry was probably able to produce almost 16 million cars per annum during this
period. It was therefore inevitable that the manufacturers would be badly hurt when the
asset deflation of the early 1990s sapped domestic demand, causing yearly sales to fall
sharply and never again to surpass 10,196,000 units.
That demand would decrease ought not to have been a surprise. With very little
population growth and a comparatively equitable distribution of income and wealth,
Japanese households already had almost all the cars they wanted; the scope for more
domestic sales was quite limited. Meanwhile, between 1993 and 1998 the manufacturers
37 For the ROE data, see: Goldman Sachs, Japan Strategy Flash, 14 May 2001, p. 3. On the value
destruction, see: Goldman Sachs Global Strategy Research, Who Are the True Value Creators?, 13
February 2002, pp. 1-3. See also, Porter, pp. 3,101.
38 MGI, 1993, “Executive Summary”, Exhibit S-1; MGI, 1996, “Synthesis”, Exhibit 3-1b,3-1c,3-2; and
MGI, 2000, “Executive Summary”, p. 1.
23
had doubled their overseas productive capacity from 3 million to 6 million autos per
annum because, in the words of one CEO, “we [wanted] to build cars in the markets
where we sell them".39 This, after all, was the best way to avoid the imposition of trade
restraints. But since these new facilities could easily accommodate increases in overseas
demand, they had the collateral effect of weakening demand for vehicles produced by
Japan’s domestic factories. Constrained at home and abroad, the domestic industry was
laboring under an immense burden of excess plant and equipment.
A simple concern for profitability should have compelled these corporations to close
some of their Japanese factories and reconcile themselves to a permanently lower level of
output. But rather than consolidating their facilities in this way, the companies
maintained their domestic production capacity at roughly its 1992 level throughout the
decade, leaving over a third of their plant idle for a very long time.40
Some insight into this paradoxical conduct may be gleaned from the history of Nissan
Motor, a company that in the 2000s would come to symbolize the feasibility and promise
of enterprise reform. Nissan's first essay in restructuring came in 1993. By this time the
company had seen its rate of capacity utilization decline to 60% and was steadily losing
money.41 In response to this adverse development management decided to close a plant
in Zama City, on the outskirts of Tokyo, and to redeploy its personnel from that site to
other factories around Japan. This was a rather timid proposal: even Nissan's labor
union endorsed it as the best way to ensure the survival of the company and at least some
of its employees’ jobs. The plan suffered a setback, however, when Fuji Bank and
Industrial Bank of Japan (IBJ), Nissan's two largest creditors as well as major
shareholders, concluded that the closure might offend the Tokyo city government and the
big political parties and hence did not fully commit themselves to it. Much of the public
39 Honda Motor's president, Hiroyuki Yoshino, is quoted in: Wall Street Journal (WSJ), 7 February 2002.
40 Compare the Japan Automobile Manufacturers Association data (JAMA) with Nikkei Business, 9 March
1999.
41 The following discussion relies heavily on Chapter Nine of Tiberghien, Yves, Political Mediation of
Global Economic Forces: The Politics of Corporate Restructuring in Japan, France, and South Korea,
Ph.D. Dissertation, Stanford University, 2002.
24
did in fact disapprove, for Nissan was patently violating Japanese commercial norms by
inconveniencing its employees and interrupting business relationships with suppliers and
distributors. This widespread chagrin galvanized the Liberal Democratic Minister of
Labor, Masakuni Murakami, who visited the Zama facility and publicly condemned the
restructuring scheme. Not to be outdone by this populist appeal, the Socialist Party,
which was on the verge of gaining control of the government, launched a full
parliamentary investigation in which it openly excoriated the company and its managers.
Then, after the factory was shut down, Zama City refused to permit the sale of the
property on which it sat, effectively forcing Nissan to keep the useless asset on its books
indefinitely.42 The company and its individual executives thus paid a high price for their
rather modest cost-cutting effort.
Nissan's later and more successful attempt at corporate reorganization took place in a
very different political climate. Much of corporate Japan was shocked in the autumn of
1997, when Tokyo allowed Yamaichi Securities, Sanyo Securities, and Hokkaido
Takushoku Bank—all prominent financial institutions—to fail. The reaction was most
extreme at the troubled Fuji and Daiichi Kangyo Banks, which feared that the authorities
might likewise allow them to succumb to bankruptcy. Since they had lent large sums to
Nissan, these creditors demanded that the manufacturer do whatever was necessary to
restore itself to profitability, a goal it had attained in only one year since the closure of
the Zama plant in 1993. Nissan could not reinvigorate itself without outside capital,
though, so the two banks enlisted the government’s help in searching for a merger partner
that might be willing to play the white knight. Everyone would have preferred that this
be a domestic investor, but as the months passed it became clear that no Japanese
company was prepared to rescue the embattled auto company. This forced the
government to look farther field. Germany's Daimler was intrigued enough to make
some initial inquiries but dropped out of the negotiations after a few months. The
balance of opinion in Tokyo then shifted in favor of France's Renault; and in early 1999
Prime Minister Obuchi, MITI, the Ministry of Foreign Affairs, and several other official
42 Interviews with a confidential source who was a ranking government official during the 1993-4 period
Footnote continued on the next page.
25
agencies publicly stated that they approved of that company's bid for a controlling stake
in Nissan.43 This, they all realized, was the only way to prevent the corporation from
collapsing in a manner which would not only cost its employees their jobs but also do
grave damage the financial system.
The almost universal agreement that foreign ownership and reform were the only way to
save Nissan was crucial to that company’s transformation. In October 1999 the new Chief
Executive Officer, Carlos Ghosn, announced a bold "revival plan" that included
discharging 21,000 workers over three years, closing five out of eleven domestic
factories, the liquidation of equity stakes in over 1,000 affiliated corporations, and the
severance of commercial relations with the vast majority of Nissan's subcontractors.44
Although ostensibly revolutionary, these specific initiatives had all been formulated by
mid-level managers in the years after 1993 and were widely recognized as essential to the
company's future viability. The reason they had not been implemented previously was
because Nissan's executives feared another political backlash. Ghosn, however, had more
freedom. As a foreigner with firm support from the Japanese government he could safely
assail the lifetime employment system, the customary relationships between corporations
and local communities, and the manufacturer’s financial connections to other companies.
The revival plan was still criticized—in a famous article Toyota’s chairman, Hiroshi
Okuda, intimated that Nissan's executives should consider committing suicide—but these
conservatives now wielded much less influence than they had during the Zama episode.
When Nissan returned to profitability in fiscal 2000, most commentators agreed that the
firm had been correct in undertaking its reforms.45 But Nissan was an anomalous case;
for the other auto companies the political and social barriers to restructuring were as
formidable as ever. In fact, as late as the summer of 2001 the strident Mr. Okuda could
still be heard asserting that profit maximization was not a proper goal for a Japanese
and subsequently became a member of Nissan’s board of directors. January and March 2002.
43 Asahi Shimbun, 14 March 1999; and Kyodo News, 19 October 1999. Financial Times, 13 July 1999.
44 Chigusa, pp. 130-1. Confidential Nissan source, March 2002.
45 NikkeiNet, 27 November 2000.
46 Economist Intelligence Unit (EIU), Japan Country Report, June 2000, p. 30.
26
corporation to pursue.46 Such stubbornness was quite common in the automobile sector,
and there was no countervailing sense of crisis such that the other producers might follow
Nissan's lead. The vehicle manufacturers consequently assumed a cautious stance,
announcing some relatively minor changes but retaining most of their existing capacity
into the early 2000s.47 Some executives still hoped that the ever-elusive return of strong
GDP growth would enable them to put their excess plant to use, but most had come to
view overcapacity as a permanent feature of their industry. Nissan was thus the exception
proving the rule that in the Japanese context restructuring was still exceedingly difficult.
If the automobile industry was one of Japan’s most competitive, the retail sector was one
of its least. Productivity in this part of the economy was in the 1990s less than half that
of its counterpart in the United States, so the case for reform was compelling.48 Yet as
was true of the vehicle manufacturers, retailers confronted barriers to change, which
virtually guaranteed that their financial performance would steadily deteriorate.
Prominent among these obstacles were the predominance of inefficiently small
companies, a lack of competition between large corporations, and an insistence in some
segments of the industry on increasing capacity almost regardless of the level of
consumer demand. Again like the auto companies, much of this intransigence should be
ascribed to political and social values that informed not just the retail market but also the
entire economy.
[Chart 10: Retail Industry Structure]
During the last few decades the nature of the retailing business in many industrialized
countries has evolved dramatically, with discount outlets, specialty stores, and
supermarkets displacing inefficient "mom-and-pop" operations. In Japan, however, such
47 For Toyota’s timid restructuring scheme, see: Nikkei Net, 21 December 2000. On the industry’s
overcapacity, see: Nikkei Business, 9 March 1999; and Asian Wall Street Journal, 7 November 2001.
48 MGI, 2000, “Synthesis”, p. 32. Also, James Kondo, William W. Lewis, Vincent Palmade, and Yoshinori
Yokoyama, “Reviving Japan’s Economy”, McKinsey Quarterly, 2000, No. 4, p.22.
27
evolution was impeded by an array of cultural and legal institutions that protected
feckless little enterprises. A combination of low property duties and high taxes on capital
gains, for instance, penalized those owners who closed their shops and sold off the
underlying real estate. Obviously, this discouraged industry consolidation. So too did
the Large Scale Retail Law, which until its revision in April 1999 gave local communities
and shopkeepers the right to block the entry of big stores into their neighborhoods.49 In
addition, beginning in 1998 the government provided small firms with some ¥4 trillion in
loan guarantees and sizeable sums in rent assistance, grants for new investment, and other
subsidies.50 The motive for this largesse was blatantly political—tiny retailers were
among the more stalwart allies of the LDP and its coalition partner, New Komeito—and
its effect was to relieve the pressures on poorly performing enterprises to exit the market.
In the late 1990s, therefore, “mom-and-pop” stores still accounted for over half of all
retail employment despite the fact that they were the least competitive companies in a
very unproductive industry.51
A separate dynamic eviscerated market forces at the upper end of the market. The degree
of continuity here is evident if one compares a list of Japan's ten largest retailers in 1983
with the corresponding list for 1998, for over the intervening fifteen years only one
corporation had dropped from that august rank.52 Among the reasons for this remarkable
stasis was the fact that the government and the banks deemed some of the retailers "too
big to fail". In the 1970s and 1980s many department stores, supermarket chains, and
other corporations had borrowed against their real estate holdings and branched out into
such unrelated fields as hospitality, transportation, tourism, and real estate.53 By the
49 MGI, 2000, “Synthesis”, pp. 11,16-9. OECD, Economic Surveys: Japan, Volume 2001-December 2001,
pp. 121-2. Bergsten et al., p. 137.
50 MGI, 2000, “Executive Summary”, pp. 3,9; and “Synthesis”, p. 19.
51 Kondo et al., p. 24.
52 By contrast, only five US retailers managed to stay in the top 10 over this period. MGI, 2000,
“Synthesis”, p. 39. The degree of concentration in the Japanese industry was also unusually low: in 1998
large outlets there still accounted for only 20% of total sales as opposed to 39% in the United States.
Furthermore, only one foreign company—Toys ‘R’ Us—had achieved a 1% market share. Ibid., pp.
8,10,41.
53 Mycal was typical in this regard. Nikkei Net, 17 September 2001. For another example, see: Thomas R.
H. Havens, Architects of Influence (Cambridge, MA: Harvard University Press, 1994), pp. 203-4.
28
1990s a number of these conglomerates had become so massive that their failure would
do real harm to other companies, banks, bondholders, and shareholders. Rather than let
this happen, Japanese financial institutions persisted in extending credit to all of the retail
giants irrespective of their operational health. Timorous regulators, meanwhile, feared
the impact that big bankruptcies might have on the labor market and accordingly
refrained from demanding that the banks desist from this self-destructive behavior. So
more funds flowed into marginal and insolvent companies, innovation was discouraged,
and this part of the market underwent very little structural change.
[Chart 11: Retail Capacity and Sales]
The third peculiarity in the retail sector was a determination among many managers to
increase the size of their businesses more aggressively than commercial conditions
warranted. The 1990s were not good years for Japan’s most prominent supermarket and
department store operators; total revenues for these “large-scale retailers” rose by only
10.5% between 1990 and 1999, from ¥20.9 trillion to ¥23.1 trillion. Given that the
industry was already overcapitalized, it could probably have accommodated this modest
growth in demand without having to increase its rate of investment very much. Some
shops should surely have been relocated to better sites and their design and management
improved, but there was little need for more overall capacity. Nevertheless the retailers
spent huge sums over the decade, expanding their total floor space from 17.3 million
square feet at the end of 1990 to 27.2 million square feet at the close of 1999.54 The
number of people working in the industry likewise rose significantly. Thus the big
retailers acted in a fashion precisely the opposite of what economic theory would have
dictated.
To understand this insistence on obtaining ever-greater scale, one need only note the
barriers to exit that confronted inefficient enterprises and the indulgent attitude evinced
by their creditors. It was in fact very expensive to shut down large stores. One estimate
54 NikkeiNet, 27 June 2002.
29
put the cost of closing a single such outlet—including penalties for the premature
cancellation of leases, generous severance packages for workers, and negotiations with
community leaders—at approximately ¥500 million in the late 1990s.55 Instead of
incurring these losses, which they deemed prohibitive, many retailers chose to leave their
older, loss-producing facilities open even as they availed themselves of falling land prices
to establish new shops in prime urban districts throughout the country. By the end of the
decade, therefore, several conglomerates were managing numerous modern boutiques in
addition to their older networks of often deserted big stores.56 The banks acquiesced in
the new investment because they had already lent too much money to the major retailers
and could not afford to bear the losses that would ensue if any of them declared
bankruptcy. The upshot was a bizarre situation in which creditors offered more and more
money, on progressively better terms, to companies that were marching implacably
towards insolvency. Daiei and Sogo corporations exemplified this problem: their
respective debts rose by 62% and 204% between 1992 and 1999, and yet the interest rates
they paid on their obligations eventually fell well below those charged to Japan’s best
retailers.57 Sustained by these funds Sogo was able to survive until the controversial
Shinsei Bank withdrew its loans in 2000, and Daiei staggered on even longer.58 Along
with the prevalence of inefficient small companies and the lack of competition between
big enterprises, this tendency to over invest ensured that the retail industry remained
overcapitalized, generated a constant stream of non-performing loans (NPLs) for the
financial system to absorb, and provided very little impetus towards stronger GDP
growth.
55 This was roughly $3.8 million at the current exchange rate. Nakamae International Economic Research,
Quarterly Report, June 2000, pp. 10-11. Goldman Sachs's retail specialist, Yukihiro Moroe, believes this
estimate is too low. Interview, March 2002.
56 For example, The Oriental Economist, October 2001, p. 10.
57 MGI, Synthesis, pp. 19,49. Mikuni & Company, “Can Japanese Economic Policy Escape the Bond
Market’s Discipline?”, (unpublished) 23 September 2002.
58 Sogo failed in 2000, and Mycal collapsed in 2001. Nikkei Net, 15 September, 17 September 2001, 31
October 2001, and 7 November 2001. Daiei was bailed out several times in 2000-2002 but remained
fragile. The Economist, 9 February 2002, pp. 61-2.
30
[Chart 12: Cross-shareholdings]
As illustrated by these specific cases, Japanese firms in a wide range of industries had an
incentive to invest ambitiously even in adverse economic circumstances. Of course,
countervailing influences were also important from time to time. In 1996 and 1997
Prime Minister Ryutaro Hashimoto oversaw the deregulation of the airline industry,
partially liberalized the telecommunications sector, and initiated a "Big Bang” program
of reforms that was intended to bring Japan's financial and accounting practices up to
global standards by the year 2000.59 Another sign of progress was the unraveling of the
web of equity stakes that tied companies and banks together and enabled them to
influence each other’s business decisions. One measure of these “cross-shareholdings”
shows them decreasing from 52% of the stock market’s total value in 1988 to 49%
thereof in 1995 and then, somewhat more quickly, to 39% in 1999. Since these financial
relationships sheltered firms from the rigors of the market, their attenuation both reflected
and intensified the process of restructuring. The same was true of corporate
bankruptcies, which proliferated in the latter half of the decade both numerically and in
terms of the failing enterprises’ net liabilities.60 Yet as signified by MITI’s index of
overall industrial capacity and by the falling ratio of GDP to the nation’s productive
stock, these changes were marginal—certainly insufficient to produce a quantifiably
better macroeconomic performance.
In summary, then, Japanese industry never really adjusted to the implosion of the bubble
in 1989-1991 or to the resulting diminution in aggregate demand. Corporations
decreased their investment budgets somewhat; but the volume of such spending was still
59 On the Big Bang and its checkered history, see: EIU, Japan Country Report, 1st Qtr 1998, p. 21.
Horiuchi, pp. 237-247. Shoichi Royama, "The Big Bang in Japanese Securities Markets", in Hoshi and
Patrick, pp. 258,260-9. On the more general deregulatory efforts, see: Bergsten et al., pp. 137-9.
60 Tokei Geppo, November 1997, p. 6; and November 2000, p. 7.
31
too high relative to household consumption and net exports, so large swathes of plant lay
idle throughout the decade. Many factors informed this recalcitrance, including cultural
norms, laws, and the preferences of employees and local and national politicians. Also
essential was the absence of a sense of crisis, which might have inspired the Japanese
people, and their leaders to overcome their conservative proclivities and enact the
reforms necessary to render the industrial economy more profitable, enrich households,
and bring down the national savings rate. But no crisis befell the country until the bank
failures of 1997-1998, and that challenge was quickly surmounted through government
bailouts and a burst of fiscal spending. The process of restructuring accordingly made
little headway, aggregate demand continued weak, and the economy remained susceptible
to recessionary pressures.
Part Four: The Complicity of Financiers
The financial system both facilitated the corporate sector’s reckless investment and was
harmed by it. Economic deceleration of the magnitude that occurred in Japan between
the late 1980s and the early 1990s would have harmed any country’s banks inasmuch as
it both impaired the quality of existing assets and eviscerated demand for new loans. But
in the Japanese instance these more general difficulties were aggravated by several
unique traits, including an unusually heavy reliance by industrial firms on bank
intermediation, an ill-conceived program of deregulation which left overcapitalized
lenders chasing ever worse customers, and the aforementioned inclination of elected
leaders and civil service to use the banks as a tool for the achievement of social and
political goals as well as to pursue strictly economic desiderata. Never did the balance of
political interests favor the deconstruction of this system. There were occasional forays
into reformist territory—witness, again, the elements of the Big Bang program that were
actually adopted—but for the most part the banks’ lending practices did not change
during the period in question. As before, those institutions insisted on supplying funds to
both creditworthy and uncreditworthy borrowers, and many of the latter invested
unwisely and hence were later unable to repay their debts. To this extent the costs of
32
sustaining aggregate demand through inefficient capital expenditures were borne by
Japan’s financial system.
[Chart 13: Distribution of Household Assets]
[Chart 14: Source of Corporate Funds]
Banks have long played a central role in the Japanese economy. As late as 1999 more
than half of Japan's household savings were deposited in such institutions, or in the postal
savings system, whereas the comparable figure for the United States was just 10.7%.
These intermediaries, in turn, channeled a large proportion of the funds they raised into
industrial enterprises: at the end of the decade the average non-financial enterprise still
raised 41% of its funds from bank loans as opposed to 13% for US companies.
Bolstering these lending relationships were strong ownership ties because, as stated
above, banks and their clients often owned much of each other’s equity.61 These crossshareholdings
were the cement that bound the "main bank" groups together, with several
affiliated companies clustering around a single lender to constitute an informal
association that often influenced its members’ investment, operational and marketing
decisions. By the middle 1990s this system had begun to break down, but financiers and
borrowers still routinely interfered in one another’s internal affairs.
The prominence of banks in the Japanese economy was partly an historical phenomenon,
as is the case in some European countries, but it additionally reflected the preference of
bureaucrats and corporate executives to work through institutions that were more
susceptible to political suasion than, say, anonymous stock and bond markets.62 The
tractability of the banking system—or, more accurately, the political sensitivity of its
61 In 1997 Ken Courtis wrote that since an estimated 85% of banks' equity is owned by large corporations,
"in effect, Japan's financial institutions are owned by their largest borrowers." Kenneth Courtis, "Japan:
Big Bang or Wee Whimper," Deutsche Morgan Grenfell, August 1997, pp. 6-7. Generally, David Asher
and Andrew Smithers, "Japan's Key Challenges for the 21st Century", in Karl Jackson, ed., Asian Contagion
(Boulder, CO: Westview Press, 1999), p. 31.
62 Thomas F. Cargill, “What Caused Japan’s Banking Crisis?”, in Hoshi and Patrick, pp. 44,46. Benjamin
M. Friedman, “Japan Now and the United States Then: Lessons from the Parallels”, in Posen and Mikitani,
Footnote continued on the next page.
33
managers—was useful in many ways. Most immediately, it enabled the government to
contain financial shocks. When an individual lender encountered serious trouble the
regulators would cajole healthier firms into refinancing it or, perhaps, merging with it.
During the era of this "convoy system", which lasted from the middle 1950s to the middle
1990s, these centrally organized rescue campaigns worked so well that no big bank ever
collapsed. But the utility of biddable financial intermediaries was not limited to
addressing financial crises: it was also conducive to the resolution of problems in the
industrial sector. In the event of a recession or market downturn, regulators and
corporate leaders could enlist the banks’ help in organizing cartels, mergers, and other
mechanisms for managing excess capacity and preventing inconveniently big
bankruptcies. Such negotiations were the first resort when Nissan neared insolvency in
1999; it was only when that company's main banks demurred and no other domestic firms
could be pressed into a salvage operation that the government finally endorsed Renault’s
bid for control. Public spectacles of this sort, though, seldom took place. The ability of
the regulators to deflect minatory economic forces before they triggered such disruptions
was in fact one of the defining elements of Japanese dirigisme.
Another salient attribute of Japan's banking industry was the changing composition of its
clientele. As financial deregulation proceeded in the late 1970s and 1980s, Japan's best
manufacturers and exporters reduced their dependence on bank loans and went instead to
capital markets to raise most of their funds. By the end of the century this rather small
category of enterprises would be financing itself much like the best Western
corporations.63 Unfortunately for Japan, these internationally competitive borrowers
turned away from the banking system just as it was undergoing tremendous growth.
When interest rates were deregulated in 1985 they started to rise, enticing households to
put much more of their money into savings and demand deposits. The banks’ total
pp. 41,47. Nobuhiro Hiwatari, "Crisis and Change in the Japanese Financial System", in ibid., pp. 113-
6,133. Shimizu, pp. 58-9. Hoshi and Patrick, “Introduction”, pp. 2,6-8.
63 Ibid., pp. 8-10. Shimizu, pp. 60-2,66,94. Takeo Hoshi and Anil Kashyap, “The Japanese Banking Crisis:
Where Did It Come From and How Will It End?”, National Bureau of Economic Research, Working Paper
7250 (July 1999), pp. 8,14,11-3,40-1.
34
liabilities consequently doubled, rising from ¥237 trillion in January 1985 to ¥495 trillion
in December 1990. Such prodigious expansion put the financial institutions in a
precarious situation. They needed to increase their lending markedly but could only do
so by lowering their standards and providing credit to a weaker and less profitable
clientele—property developers, for example, and small and medium-sized enterprises,
individuals, overseas corporations, and the services industry.64 The only big customers
that stayed loyal to the banks during this transitional period were those which were too
fragile to appeal to Japan’s relatively discriminating debt and equity investors: those, in
other words, which had no other choice.
As the quality of their loan portfolios deteriorated, the banks sought to hedge their
exposure to dubious companies by relying more heavily on real estate and corporate
securities as collateral. Such liens had historically been very important in Japan. In the
most sophisticated Western economies, banks base their lending decisions on the capacity
of borrowers to earn enough money to repay the interest and principal of their debts.
Collateral is frequently evaluated in individual decisions, but the foremost concern is
whether a firm can consistently fulfill its contractual obligations. In Japan and other
Asian countries, by contrast, banks often ignore a potential borrower's earning power and
focus instead on the value of its tangible assets. This worked well as long as property
prices were flat or rising, as was the case in Japan from the early 1950s through the late
1980s, because the value of the underlying real estate was never in jeopardy. It therefore
seemed natural to those bankers who were, during the bubble period, extending loans to
so many marginal enterprises to rely more heavily on collateralization. While the ratio of
property-backed loans to all secured loans remained fairly constant through these years,
the amount of money lent against such collateral rose from some 70% of its market value
to, in many instances, well over 100%.65 Where real estate was unavailable, furthermore,
64 Hoshi and Kashyap, pp. 4,6,7-10,16-7,16-7,18. Shimizu, pp. 58-9,63,65,66-70,75-6,81-2,95. Kazuo
Ogawa and Shin-ichi Kitasaka, "Bank Lending in Japan: Its Determinants and Macroeconomic
Implications" in Patrick, p. 163. Kazuo Ueda, "Causes of Japan's Banking Problems in the 1990s" in Hoshi
and Patrick, pp. 64,66,76-7. Friedman, pp. 45-6.
65 Shimizu, pp. 60,62,70-1,76,81-2,84-5. Mikitani, “The Facts”, p. 32.
35
the banks assumed liens on equity- and bond-holdings to secure lines of credit to new
clients. But their speculation did not end there. Eager to participate in the stock market’s
apparently endless appreciation, the lenders also purchased a great deal of corporate
paper for their own accounts.66 In this way Japan’s major financial institutions placed
multiple bets on the proposition that asset prices could move in but one direction.
[Chart 15: Bank Deposits and Lending]
Because of this reckless gamble, the market crash that began in 1989 did far more
damage to the financial system than would otherwise have been the case. According to
the government's statistics, which would time and again prove too optimistic, the volume
of NPLs in the banking industry increased from a negligible percentage of its total assets
at the beginning of the decade to 6% thereof in 1997.67 This was a prodigious sum,
representing over 4% of annual GDP. Households recognized early that the banks were
compromised and responded by entrusting less of their money to them. After growing
very swiftly in the late 1980s, deposits at such institutions shrank from ¥495 trillion at the
end of 1990 to ¥445 trillion in January 1994 even as the assets in the postal savings
system, which enjoyed a government guarantee and hence was perceived as a safer
alternative, ballooned from ¥134 trillion to ¥182 trillion. As their capital bases
contracted the banks should logically have begun to reduce their outstanding loans or, at
the very least, have refrained from extending new ones. They did not, however, do this.
To the contrary, the banks expanded their outstanding loans and discounts from ¥443
trillion at the end of 1990 to over ¥480 trillion in late 1995 and early 1996. Even more
strikingly, they supplied more capital to the real estate industry, which had been
devastated by falling property values and plainly deserved no additional credit.68 It was
as if the commercial lenders had not noticed that the bubble had burst.
66 Also, Ueda, p. 77. In the negotiations that produced the Basle Accords, Japan had successfully argued
for the inclusion of a provision letting its banks count 45% of their unrealized capital gains on equities as
Tier Two capital. EIU, Japan Country Report, 1998 2nd Qtr, p. 21.
67 Management and Coordination Agency, Statistics Bureau, Japan Statistical Yearbook, 2000, p. 445.
68 Nobuhiko Hibara, "What Happens in Banking Crises? Moral Hazard vs. Credit Crunch", 5 November
2001, unpublished, pp. 20-21,29.
36
[Chart 16: Household Deposits]
The explanation for this curious behavior, once again, is that none of the major economic
actors had an incentive to admit that the financial system was malfunctioning or to
implement the reforms that would have repaired it. Most obviously, the lenders did not
want to write off their bad loans because doing so would be embarrassing and, if the
process depleted too much reserve capital, perhaps fatal.69 Underscoring this reluctance
to foreclose on insolvent borrowers were the government's explicit and implicit guarantee
of deposits and the low interest rates that obtained throughout the 1990s, for these
minimized the cost of holding questionable assets on one's books. In effect, the monetary
authorities had given the banks an inexpensive "call option" on any improvement in their
customers' businesses, for if an eventual economic recovery revitalized the delinquent
borrowers their debts might become collectable or even profitable. It was to preserve the
value of this option that financial institutions, especially those banks that had already
incurred substantial damage, kept rolling over loans to some of their worst borrowers,
including real estate and construction companies and such embattled retailers as Sogo,
Mycal, and Daiei.70 Large parts of corporate Japan, in the meantime, were also reluctant
to espouse a program of thorough financial restructuring. On the one hand the managers
of unprofitable companies realized that they would be among the first victims of a
contraction in credit whereas, on the other, senior executives at many stronger enterprises
worried about the fate of their suppliers, distributors, and affiliated companies. To such
businessmen it seemed better to let sleeping dogs lie.
The civil service, too, was inclined to dither. In the early 1990s such leaders as Kiichi
Miyazawa, who would serve as both finance minister and prime minister, occasionally
69 Ibid. Posen, 1998, pp. 127-9.
70 Cargill, p. 50. The Oriental Economist, October 2001, p. 10. Mikuni, p. 7.
37
bruited the notion that the banks were undercapitalized and needed an injection of public
money, but the bureaucratic and popular response to these statements was uniformly
negative.71 Voters had in fact long resented the financiers’ cozy relations with big
business, the regulators, and the ruling party, and had lately come to feel that collusion
between these groups was to blame for the asset bubble and its painful aftermath. There
was, in short, virtually no political support for the use of state funds to bail out the big
lenders. The dangers of challenging this sentiment became evident in 1995, when the
Ministry of Finance (MOF) admitted that the Jusen mortgage companies were
overwhelmed by NPLs, could not recover on their own, and must be dissolved.72 The
ministry initially tried to spread the costs of these closures among the Jusen's owners, but
some of the creditor banks refused to participate in a traditional "convoy" exercise and
the powerful agricultural cooperatives enlisted the aid of their allies in the LDP to
circumvent the requirement that they pay their full share. This forced the government to
absorb ¥685 billion in losses. Popular fury over this subsidy tarnished the MOF’s prestige
and contributed to the Diet's decision in 1997 to scrutinize that ministry’s internal
operations, deprive it of control over the Bank of Japan, and entrust the supervision of the
commercial banks to a newly formed Financial Supervisory Agency (FSA).73 Thus
chastened, the bureaucrats stuck their heads back in the sand; issuing frequent assurances
that the financial system was secure and hoping that the voters’ hostility would soon
dissipate.
[Chart 17: Distribution of NPLs by Industry]
Japan’s political leadership—meaning effectively the LDP, which controlled the
government for almost all of the 1990s—shared the widespread predisposition to
procrastinate. Lacking its own expertise in financial matters, the party depended on the
71 Shimizu, p. 77. Cargill, pp. 51-2,62.
72 Ibid. Hiwatari, pp. 113-6. Bergsten et al., pp. 71-3.
73 EIU, Japan Country Report, 2nd Quarter 1998, pp. 14-5; and 3rd Quarter 1998, pp 15-6. Jennifer Amyx,
Japan’s Financial Crisis: Institutional Rigidity and Reluctant Change (Princeton, NJ: Princeton University
Press, forthcoming 2004), Chapter 8. Adam Posen, “Introduction: Financial Similarities and Monetary
Differences”, in Posen and Mikitani, pp. 8,16.
38
inadequate information emanating from the MOF and consequently could not easily
formulate appropriate remedial policies.74 But the Liberal Democrats may have failed to
act even if they fully comprehended the fragility of the big financial institutions for the
simple reason that most of the NPLs were concentrated in the real estate, construction,
retail and wholesale, and services industries from which they and their coalition partners
derived critical campaign donations and logistical assistance. Fear of the electorate
reinforced this reticence, for the LDP had not escaped the Jusen imbroglio entirely
unscathed and therefore did not want to brave the public’s wrath by using government
money to help another category of unpopular financiers. Thus the balance of interests
between banks, businesses, bureaucrats, and parliamentarians favored a policy of
inaction, letting more capital flow to debtors in the hope that they might be resuscitated
by an eventual resurgence of GDP growth.
This conservative consensus foundered in November 1997, when Hokkaido Takushoku
Bank, Yamaichi Securities, and Sanyo Securities failed. These bankruptcies engendered
real anxiety among the citizenry and in financial and political circles, where calls for
reform grew more vociferous. The opposition Democratic Party capitalized on the mood
of discontent by proposing the establishment of a new regulatory scheme that would
stiffen disclosure standards, mandate higher reserve ratios against bad loans, create a
Financial Reconstruction Commission to process NPLs, and allow the government to
nationalize insolvent banks.75 At first the LDP objected to this proposal, but a setback in
the Diet election of July 1998 promptly persuaded it to reverse course. In October the
government duly agreed to enact the controversial legislation, set aside ¥60 trillion for
possible use in strengthening the banking system, and pledged to install reformist
officials in the relevant regulatory agencies. Over the next two years much changed.
Long-Term Credit Bank (LTCB) and Nippon Credit Bank (NCB) were nationalized, and
fourteen other City Banks (large commercial lenders) promised to restructure themselves
in exchange for ¥7.5 trillion in state loans. It was also during this period that most of the
74 Amyx, Chapters 8 and 9. Hiwatori, pp. 113-4.
75 EIU, Japan Country Report, 4th Quarter 1998, pp. 16-7,22-5. Amyx, Chapter 9. Hoshi and Patrick,
“Introductory Overview", pp. 17-8. Hiwatari, pp. 116-22. Bergsten, et al., p. 78.
39
big lenders declared their intentions to merge into four main groups—Sumitomo Mitsui,
Bank of Tokyo Mitsubishi, Mizuho, and United Financial of Japan—and to write down a
much larger proportion of their non-performing assets.76 Even the smaller regional banks
and credit cooperatives now felt the need to consider consolidation.
Yet these changes were more cosmetic than substantive. Though beyond the temporal
scope of this paper, LTCB’s post-nationalization experience sheds light on the political
topography in which the reform process played out. In particular, the story demonstrates
how equivocal a role the Financial Supervisory Agency—now renamed the Financial
Services Agency (FSA)—played in the enforcement of Japan's new laws and regulations.
The drama began in March 2000, when a consortium of Western investors purchased the
defunct LTCB from the government and changed its name to Shinsei, or "Rebirth". As
part of the purchase agreement Shinsei received an injection of cash and the right to sell
loans from the original portfolio back to the government if they lost more than 20% of
their face value within a contractually stipulated period of time.77 The new company
conversely made several pledges, including an agreement to supply a certain volume of
credit to the small and medium-sized enterprises that were so important to the LDP. On
the basis of this agreement Shinsei became the first City Bank to fall into foreign hands,
providing what many observers reckoned was the strongest indication to date that Tokyo
had become serious about financial reconstruction.
Soon, though, relations between the FSA and Shinsei soured. It was to be expected that
the two parties would bicker over whether various assets met the conditions necessary to
be "put" back to the government, but other conflicts were more disturbing. One such
dispute concerned the bank’s NPLs. In preparing its accounting statements for fiscal
2000 Shinsei informed the regulators that it intended to classify 19.6% of its portfolio as
non-performing.78 The FSA objected to this proposal because it knew that all the City
76 NikkeiNet, 30 November 2000 and 4 January 2001. Japan Times, 6 September 2000 and 25 November
2000. Reuters, 26 August 2000. Asian Wall Street Journal, 18 December 2000.
77 NikkeiNet, 19 October 2001, 7 November 2001, and 5 March 2002.
78 Confidential interviews with two senior managers at Shinsei Bank, October 2001 and January 2002.
40
Banks held similar portfolios, and if Shinsei pronounced such a high proportion of its
assets to be risky the market would expect the rest of the major lenders to do the same.
Those banks would then face a real dilemma. If they followed Shinsei’s precedent they
would reveal themselves to be insufficiently capitalized, but if they refused to do so they
would cast doubt on the veracity of their accounting statements and on the competence of
the regulatory authorities. To forestall such discomfiture, the big banks and senior
regulators reportedly urged Shinsei to lower its NPL estimate towards the industry norm
of about 7%. Shinsei, however, would not budge. Rather than conform to what it saw as
the industry’s obfuscatory norms, it went ahead and declared that a fifth of its loans were
partially or fully impaired.79
A related quarrel arose when Shinsei decided that some of its largest customers were in
default and accordingly decided to call in its loans to them.80 This confronted the other
City Banks, several of which had lent sizeable sums to the same corporations, with
another awkward choice. They could either withdraw their funds from the weak firms,
booking huge losses in the process, or purchase Shinsei's positions at full face value in an
attempt to maintain the fiction that the borrowers were still solvent. The effect of this
latter approach, of course, would be to transfer bad assets from Shinsei to the other
lenders, essentially transforming them into repositories for the industry’s NPLs.
Unwilling to accept either of these fates, the City Banks and the FSA privately requested
that Shinsei refrain from foreclosing on those large enterprises whose bankruptcy would
harm the broader financial system. Shinsei acceded to these pleas in some instances; but
in others the renegade bank went its own way and thereby precipitated some spectacular
corporate failures, including that of Sogo Department Stores. In the process, needless to
say, Shinsei’s managers infuriated executives at the other City Banks as well as numerous
conservative officials and parliamentarians.
79 NikkeiNet, 10 December 2001.
80 Asian Wall Street Journal, 26 September and 5 October 2001. The Oriental Economist, October 2001,
pp. 11,14. Shinsei interview, January 2002.
41
It was a third dispute, though, that brought the feud between Shinsei and the FSA to
national attention. At the end of fiscal 2001 the regulators found that the bank was not
allocating enough money to small and medium-sized corporations and demanded that it
do more to honor its contractual commitments. Shinsei acknowledged its obligation in
this regard, but contended that the class of borrowers did not deserve more credit and that
the government should not force it to make uncollectible loans.81 The FSA replied by
issuing a "business improvement order" which ironically demanded that the bank achieve
its lending targets irrespective of the cost of doing so, and by quietly pressing Shinsei’s
Japanese employees to break ranks with their Western colleagues and support the
government’s position. In the most egregious of the FSA’s communications, a senior
official went so far as to instruct the bank to lend more money to four specific
companies that happened to enjoy close ties to the LDP. Shinsei thought this an
unwarranted intrusion into its operational decisions and counterattacked by briefing the
media on the FSA's tactics—even letting reporters read some of the agency’s confidential
memoranda.82 The ensuing articles provoked an uproar. Seizing upon them, the
opposition parties convened hearings in the Diet and pointedly questioned the new
Koizumi cabinet and the relevant FSA personnel about their oversight of the financial
sector. This brought Shinsei a short respite, during which the government concentrated
on fending off parliamentary criticism, but the bank’s relations with the bureaucracy
remained tense for years thereafter.
The proposition for which the Shinsei interlude stands is that the reforms of the late
1990s neither fundamentally altered the way in which the big banks did business nor
severed the sometimes improper ties between them and the civil service. The lenders had
enlarged their capacity enormously during the bubble period, and there was subsequently
very little consolidation; in Anil Kashyap’s sardonic words, "the major Japanese banks
were simultaneously among the largest banks in the world, and the least profitable".83
81 Ibid., and NikkeiNet, 29 October 2001.
82 The Wall Street Journal articles cited in this section are the result of Shinsei’s leaks.
83 Kashyap, “Discussion” in Mikitani and Posen, p. 109. In 1999 Hoshi and Kashyap, pp. 40-2, estimated
that at least 25% of the industry’s capacity, and a third of Japan’s 140 banks, should be eliminated.
Bergsten et al., pp 83-4.
42
The Obuchi reforms of 1998-1999 brought a much needed re-capitalization of these
institutions, but the promise of cost reductions, better management, and greater
profitability went largely unfulfilled.84 Indeed, once the immediate crisis had passed the
FSA’s lack of enthusiasm for corporate restructuring manifested anew. Instead of forcing
the City Banks to foreclose on insolvent enterprises and write off the concomitant NPLs,
that agency encouraged the lenders to forgive delinquent obligations and to extend still
more credit to vulnerable borrowers.85 To facilitate this, the regulators tacitly permitted
the banks to miscategorize their outstanding loans. Most of Sogo's and Mycal's debts, for
instance, were classified as either fully performing or as "requiring monitoring", rather
than as risky or uncollectible, until after those corporations filed for bankruptcy
protection.86 Clearly the lenders and the regulators were determined to conceal the truth
as long as they possibly could. The LDP acquiesced in this deception partly because it
did not comprehend the depth of the rot in the financial system but also in deference to its
corporate allies, many of which would have been hurt by the imposition of tougher credit
standards.
[Chart 18: Commercial Bank Profitability]
Hence the erosion of the lenders’ balance sheets that had begun in the late 1980s
persisted through the end of the century. The return on assets (ROA) earned by Japan’s
commercial banks on their core operations, including losses on the disposal of NPLs, was
negative every year from 1993 onwards. These enterprises could, and did, garner some
additional funds by selling the stocks and bonds that they had acquired before the
inflation of the bubble and realizing the associated capital gains. But this became more
difficult as the stock market relentlessly declined through the 1990s, and by the year 2000
84 Cargill, p. 49. Asian Wall Street Journal, 18 December 2000. NikkeiNet, 4 January 2001. HSBC's
Brian Waterhouse, quoted in the Japan Times, 25 November 2000. Kashyap, “Discussion”, pp. 109-10.
85 Cargill, pp. 42,44-5. Ueda, p. 78. Horiuchi, p. 235. Asian Wall Street Journal, 6 March 2002.
86 NikkeiNet, 15 September 2001. The Oriental Economist, October 2001, p. 10. A study of 605 companies
that went bankrupt in the first half of 2001 found that 70% of the loans to them were significantly
overvalued on their creditors’ balance sheets. Financial Times, 20 April 2001.
Footnote continued on the next page.
43
the banks’ equity portfolios had virtually no unrealized gains left to exploit.87 With scant
operating income and no more securities profits to reap, the lenders were increasingly
forced to run down their reserves in order to dispose of bad assets. Yet since they were
still supplying funds to insolvent borrowers, their non-performing loans continued
inevitably to mount. Even the disingenuous accounts published by the FSA registered
this fact: they showed that the volume of NPLs at the City Banks increased every year
from 1998 through the end of the decade and into the early 2000s. Conditions were direr
still, of course, if one accepted the relatively credible estimates emanating from privatesector
analysts, who estimated that the total of such dubious assets at the turn of the
century was somewhere between ¥100 and ¥150 trillion.88 By this point it seemed that
the industry, and especially the some of the more troubled regional banks, might
eventually need another infusion of public capital if they were ever to return to health.
[Chart 19: Officially Recognized NPLs]
At a higher level of abstraction, it is clear that Japan's banks were caught between
macroeconomic exigency on the one side and political realities on the other. The country
needed more aggregate demand in order not to fall into prolonged recession.
Deregulation and restructuring might have caused consumption to strengthen in a manner
that lowered the savings rate somewhat and, to that extent, reduced the need for
extraordinary stimulus. But thorough reform never occurred because too many corporate
executives, bank managers, regulators, and politicians feared the ramifications of such
change for the economy as a whole and for their own particular interests. The requisite
incremental demand might also have stemmed from growth in the current account
surplus, yet this too failed to materialize. By default, then, the country remained
dependent on inordinate corporate investment. It was only possible to engage in this
87 By late 2001 they would be sitting on some ¥5 trillion in unrealized losses. The Economist, 9 February
2002, pp. 62-3.
88 For example, The Economist, 9 February 2002, pp. 62-3. In 2001, Goldman Sachs would put the total at
a vertiginous ¥237 trillion. Goldman Sachs Global Equity Research, Totally Rethinking Japanese Asset
Quality, July 2001. Even the normally generous IMF rejected the FSA's figures as incredibly low. Japan
Economic Newswire, 20 August 2001.
44
copious capital spending, however, as long as the banks kept supplying capital to the
weaker industries that comprised so much of Japan’s surfeit productive capacity. When
those industries subsequently performed poorly—as they almost invariably did—they
transferred a large proportion of their losses to the financial system in the form of new
NPLs. Some regulators and political leaders recognized this destructive dynamic and
professed a desire to obviate it, but these people were always in a distinct minority and
could not marshal much political support for their views. A dysfunctional financial
system consequently became part of the complex of problems that the leaders of 1990s
Japan left for their successors to resolve.
Part Five: Raiding the Treasury
Although corporate investment, financed in the main through bank loans, helped sustain
aggregate demand throughout the 1990s, its ability to serve this function diminished over
time. There was a modest fall in household savings during the decade; but corporations
were conversely becoming more parsimonious, and this meant that the combined savings
rate stayed very high. In the meantime, though, the ratio of capital expenditures to GDP
fell from just over 19% in 1990 and 1991 towards 15% in 1999. The gap between
aggregate supply and aggregate demand accordingly widened, portending commercial
stagnation at best and economic contraction at worst. In order to avoid such an outcome,
Japan must find yet another source of supplemental demand. Ultimately it would be the
government that filled this gap, increasing its deficit spending so ambitiously that it not
only counteracted the decrease in corporate investment but also produced some impetus
toward economic expansion. This, incidentally, is how Japan managed to grow during
the 1990s rather than stagnating or falling into depression. But the progressively greater
reliance on fiscal stimulus was by no means a cost-free proposition: by the turn of the
decade the national debt, too, would have begun to cast a shadow over Japan’s future.
Before describing Tokyo’s fiscal policy and its shortcomings, it may be useful to explain
in more detail why the external alternative was never earnestly pursued. The excess of
45
savings over investment that afflicted Japan from the 1980s through the end of the
century meant that a weaker yen and a bigger current account surplus were the most
appropriate means of bolstering aggregate demand. Put simply, the exchange rate should
have fallen so as to enable the country to export its surplus capital. The price deflation
which characterized the decade of the 1990s underscored the potential advantages of
depreciation, for a weaker yen might also have engendered inflationary pressures and
have slowed the pace at which NPLs were appearing on the banks’ balance sheets.
Depreciation and greater capital outflows would also have helped some members of the
international community; for these were the years in which the newly independent
peoples of East Central Europe and the former Soviet Union were competing with other
poor nations for scarce investment funds, and anything which lowered global interest
rates would have facilitated their economic modernization.89 East Asia and the world
would also have benefited if a new exchange rate had stabilized the Japanese economy
and transformed it from a source of deflation and financial risk into an engine of growth.
Yet however persuasive the arguments in favor of a cheaper yen, there were equally
convincing reasons to believe that such an expedient was impracticable. The first of
these was the sheer size of the necessary change. Through much of the 1990s potential
net savings in the private sector, calculated as the residual that would have resulted if
corporations had invested at the same rate as their US and German counterparts and the
government budget had been balanced, totaled somewhere near a tenth of GDP. The
current account surplus, though, was only 3.1% of national output in 1993.90 In order to
employ all of the country's extra capital, then, Japan’s external surplus might conceivably
have needed to triple in size. Singapore and Switzerland have exported a tenth of their
GDP in excess funds each year over long stretches of time, but they represented such
small parts of the international economy that this posed little challenge to other countries.
Japan, by contrast, was so massive that an attempt to reconcile its aggregate supply and
demand through the current account would have fundamentally altered the international
89 The IMF’s argument to this effect is summarized in Brown, pp. 70,132.
90 Statistical Yearbook, pp. 138,423. Martin Wolf, in Financial Times, 9 May 2001. Brown, pp. 75-
79,164,169.
46
pattern of trade and finance and have imposed wrenching change on other countries’
domestic institutions.
It was understandable, therefore, that few foreign governments were willing to
contemplate anything more than a small adjustment in the yen’s value. To adduce the
most salient example, in the early 1990s Washington shared Tokyo’s view that Japan’s
structural flaws were not severe and could soon, and easily, be overcome. Confident in
this respect, the United States concentrated its diplomacy on trying to prize open Japan’s
various markets as a means of reducing the size of its own trade deficit.91 During these
years the Clinton administration also enacted several concrete policies which had the
effect of inhibiting growth in Japan’s net exports. The tightening of fiscal policy during
the 1990s was one such measure, for it lowered interest rates within the United States and
stopped the dollar from rising against the yen as much as it would otherwise have done.92
That these initiatives might push Japan into a depression was a possibility, which, at this
stage, worried almost no one in the US capital. Thus the Americans initially welcomed
the steep appreciation of the yen between 1991 and 1995, when the currency briefly
touched ¥80 per dollar, before belatedly realizing that this was crippling the Japanese
economy and retarding world growth.
Japan’s neighbors in East and Southeast Asia shared the US opposition to a weaker yen,
for such a development would have constrained their ability to export goods and services
while also harming those domestic enterprises that competed with Japanese imports.
Concern on this score intensified during the Asian Financial Crisis of 1997 and reached a
climax when the yen plummeted to ¥147 per dollar in 1998. Much of East Asia, the
United States, and a few European countries complained vociferously about this
depreciation, claiming with some justification that it was exacerbating the region’s
travails.93 The costs of this international opprobrium were significant. By the summer of
91 Bergsten et al., pp. 15-8,100. Brown, pp. 59,131,157,166-7,186,194-6,200.
92 Ibid., pp. 76,169. Bergsten et al., p. 46.
93 Ibid., pp. 97,100,199. EIU, Japan Country Report, 2nd Quarter 1998, pp. 15-6; and 3rd Quarter 1998, pp.
17-8.
47
1998 many officials in Washington had concluded that Tokyo was no longer a reliable
diplomatic ally and that the United States should start treating Beijing as its foremost
“strategic partner” in the western Pacific.94 President Bill Clinton’s decision to fly to
China for a state visit without making the customary courtesy call on the Japanese
government was just the most public sign of the White House’s disillusionment with its
old ally. In these circumstances, there was little chance of Japan’s making progress
towards such cherished goals as a permanent seat on the United Nations Security
Council.
Some economists and Japanese politicians have contended on these grounds that
outsiders were responsible for the strength of the yen in the 1990s—and even for the two
recessions that befell the country during that decade. 95 This argument, though, is
spurious. If the Ministry of Finance and the Bank of Japan had acted assertively they
could almost certainly have devalued the yen no matter what other countries thought.
The monetary authorities, after all, controlled their own printing presses and could issue
as much yen, and purchase as much foreign currency, as they wished. The MOF would
in fact use this power over exchange rates, which it retained after the BOJ gained
statutory independence in April 1998, quite effectively in the first half of 2003. During
those months the US dollar weakened considerably, and the ministry reacted by
intervening to drive the yen down in tandem, thereby keeping the yen-dollar exchange
rate fairly stable but improving Japan’s terms of trade relative to the European Union and
many other economies.96 The success of this endeavor belied the notion that in the 1990s
Japan had sat impotently by as foreign recalcitrance forced the yen to move ineluctably
upward. Also discredited in early 2003 was the erroneous proposition that monetary
94 Ibid., 1st Quarter 1999, pp. 16-7.
95 Eisuke Sakakibara, "US-Japanese Economic Policy Conflicts and Coordination during the 1990s", in
Posen and Mikitani, pp 167-185. Perhaps the most rigorous exposition of this theory is Ronald I.
McKinnon and Kenichi Ohno, Dollar and Yen (Boston, Massachusetts: Massachusetts Institute of
Technology, 1997).
96 Whereas the Euro appreciated by almost 30% against the dollar, the yen gained only about 10%. The net
effect on Japan’s overall terms of trade was basically neutral. Financial Times, 24 June 2003. NikkeiNet,
13 May 2003, 26 May 2003, and 9 June 2003. The scale of this intervention diminished in late 2003 and
stopped, at least temporarily, in March 2004. This change reflected the fact that the dollar was now
strengthening against the yen. Goldman Sachs, Japan Economics Analyst, 14 May 2004.
48
policy becomes irrelevant once a country’s interest rates have fallen to zero, for the
higher cost of imports entailed by the MOF’s currency intervention played some role in
the deceleration of Japan’s price deflation which ensued towards the end of that year.
The truth is that like any other country Japan was perfectly capable of debasing its own
money.
Hence the question is not whether Japan could have weakened the yen enough to make its
current account surplus expand but rather why it chose not to do so. Surprisingly,
perhaps, for most of the period from 1985 through 2000 the balance of opinion in Tokyo
favored the status quo or a slight cheapening of the yen but never dramatic depreciation
of the sort that would have contributed substantially to aggregate demand. Both BOJ
Governor Yasushi Mieno in the early 1990s and, after a brief interim, Governor Masaru
Hayami in the late 1990s felt that a robust currency was beneficial to Japan insofar as it
compelled companies to cut costs and achieve greater efficiencies. The record is replete
with their statements in this regard, and the governors generally acted in a manner
consonant with their convictions.97 From time to time other parts of the civil service
expressed a similar preference for tight monetary policy and a comparatively high
exchange rate. The Ministry of Foreign Affairs was one such institution; it was happy to
see Japanese corporations take advantage of the strong yen to invest in Asia because that
investment enhanced Tokyo’s influence in the region. Meanwhile, ranking officials in
both MITI and the MOF occasionally advocated appreciation as a means of placating the
importunate Americans without having to liberalize Japan’s domestic economy.98 The
97 For example, Nihon Keizai Shimbun, 30 December 1989; Reuters, 27 November 1991; Jiji Newswire, 27
April 1992; Reuters, 2 November 1994; Nikkei Weekly, 23 March 1998; Wall Street Journal, 6 April 1998;
Nihon Keizai Shimbun, 11 December 1998; Financial Times, 10 December 1999. See also, Paul Krugman,
“Delusions of Respectability”, February 1999; ibid., “Further Notes on Japan's Liquidity Trap”, June 1998;
and Jinushi, et al., pp. 117,121-2,139-40,144.
98 See, in particular, press statements by the MOF’s Sakakibara, who espoused yen strength in a way that
belies his more recent contention that the United States is to blame for Japan’s “lost decade”. Jiji
Newswire, 8 May 1997; and Le Monde, 23 May 1997. More generally, Brown, pp. 3,71,76,149,150,152-
5,165-6,195. Such statements were important not only because they endorsed the BOJ’s policies but also
because they altered the balance of risks as perceived by market participants. At many times during the
1990s investors were tempted to short the currency but feared that the government might later intervene to
push its value up. To avoid the losses this would entail, they refrained from assuming strongly bearish
positions; and this helped maintain the yen’s strength. Ibid., pp. 4,73-6,118-9,145,155,162,180-3.
49
MOF’s behavior changed somewhat after it ceded control of monetary policy to the BOJ
in 1998. Now the ministry more consistently advocated marginal depreciation as a means
of assisting the country’s exporters, and yet it still acted as if engineering the yen’s fall
were the central bank’s duty rather than using its own statutory powers to affect the
exchange market. Even worse, by this point the MOF’s calls for looser monetary policy
had become counterproductive because Governor Hayami was determined to defend the
central bank’s new autonomy from political interference.99 Thus there was no extended
period of time during the 1990s during which the relevant parts of the civil service
uniformly favored an exchange rate low enough to enable Japan to reach macroeconomic
balance.
Strong political leadership might have overcome this bureaucratic impasse. The Diet
could at any time have passed a law dictating that the MOF and the BOJ to devalue the
yen in order to enlarge the current account or, alternatively, to combat deflation. The
Liberal Democrats, however, were never sufficiently dissatisfied with the status quo to
insist on such a radical departure. For although the prospect of a moderately weaker
currency and somewhat faster GDP growth doubtless appealed to these politicians, their
highest priority was protecting those interest groups that were essential to their own
political power. If they thought about the matter at all, therefore, they would probably
have objected to precipitous depreciation to the degree that it hamstrung their friends in
import-dependent industries, such as energy production and retail and wholesale
distribution, and because it might provoke Japan’s trading partners into imposing
restraints on the country’s exports.100 Many LDP parliamentarians presumably also
worried about the opinions of voters, many of whom would have objected if a lower
exchange rate diminished the purchasing power of their savings or further tarnished
Japan’s international prestige. For these reasons the path of least resistance for the ruling
party was to let the MOF and the BOJ pursue what was in reality a strong yen policy.
99 Kashyap, “Discussion”, pp. 112-3.
100 Brown, pp. 127-8,131,157,166-7,184,186,192-6,200.
50
[Chart 20: Absorbing Surplus Capital]
Since the external surplus was never allowed to widen enough to alleviate Japan’s
deflationary tendencies, fiscal policy became the key to the country’s economic stability.
The extent to which the government supplemented private demand may be seen in the
trajectory of the national budget, which moved from an average annual surplus of 1.8%
of GDP in the years 1989-1991, to a deficit of 2.4% in 1993, and thence gradually to a
deficit of 7.1% in 1999.101 As these figures imply, fiscal spending added almost a tenth
to yearly economic output over the course of the decade. This more than compensated
for the decline in private non-residential investment and helped Japan eke out GDP
growth of some 1.0% per annum rather than succumbing to a lengthy recession or
depression. Such was no mean accomplishment given the deflationary forces that
prevailed in the aftermath of the bubble’s implosion.
Tokyo’s use of its budget to fend off economic contraction was in fact so effective as to
cast doubt on one of the three major criticisms leveled against its fiscal policy during the
1990s. That first argument—that the government might have revivified the economy at
almost any time had it opted for a sharp burst of fiscal stimulus, or “pump priming”,
rather than slowly increasing the size of the deficit—emanated regularly from officials in
the US government and from some corporate analysts and academic researchers.102
These people believed that conditions in the Japanese economy were fundamentally
sound in the late 1980s and that it was cyclical factors that, at the turn of the decade,
pushed aggregate demand temporarily below aggregate supply. To close this gap the
government should have immediately adopted a program of generous tax cuts and wellconceived
public works projects in order to stimulate household consumption and
101 EIU, CountryData, 2001.
102 For example, Posen, 1998, pp. 2-6,17-21,29-30,37-8,44,49; and Posen, "Nothing to Fear but (Fear of)
Inflation Itself”, International Economic Policy Briefs, IIE, October 1999.
51
corporate investment. The economy would then have resumed the high rate of growth
that it had attained before the bubble collapsed. An ordinary ailment, in other words,
called for an ordinary cure.
The errors informing this view are numerous. Most conspicuously, Japan’s problems
were not cyclical but structural. If the economy had been fundamentally healthy, its
productivity and profitability would not have declined so inexorably from the 1970s
through the 1990s. Nor would the “dualistic” bias in the country’s industrial system,
whereby an extremely competitive export sector indirectly subsidized a much larger and
less efficient domestic sector, have grown more deeply entrenched with the passage of
time. But these deleterious trends were demonstrably, measurably, real. Their presence
attested to the existence of flaws, which were both profound and chronic.
Equally undeniable was the related phenomenon of a very high private savings rate, a
feature which was more typical of a developing country with a low standard of living
than of a mature economy and which caused such immense trouble during the 1990s.
What Japan required was to find a new equilibrium in which some combination of
households, government, and foreigners consistently provided more aggregate demand,
leaving fewer surplus funds to discourage commerce and push prices down. Isolated
doses of deficit spending, no matter how voluminous, could not produce such a result.
They could absorb excess capital and add to aggregate demand in the short term, yielding
some GDP growth in any one- or two-year period, but as soon as the stimulus had
dissipated the weight of superfluous savings would again cause commercial activity to
recede. This is why fiscal schemes like those enacted in 1995-1996 and 1998-1999,
which many of the cyclical enthusiasts initially declared sufficient to trigger a selfsustaining
recovery, failed in the event to do so.103 It was only when many such packages
103 For claims that Japan had achieved a sustainable recovery in 1995-1996 see, for instance, Posen, 1998,
pp. 29,37,41,46-51; and Posen, “Introduction”, p. 9. This contention, however, is contradicted by the
details underlining the period’s impressive GDP figures. First, the Kobe Earthquake of January 1995 had
destroyed a vast quantity of public and private infrastructure, and replacing these facilities contributed
greatly to the economic growth of the following years. By the first quarter of 1997, though, most of the
necessary reconstruction was underway and construction starts and machinery orders, both leading
indicators of investment, had resumed the downward trajectories that obtained before the earthquake. A
Footnote continued on the next page.
52
were superimposed one upon another, steadily expanding the budget deficit and replacing
the incremental loss of corporate investment, that even negligible economic growth
became possible over longer periods of time.
A related flaw in the cyclical theory is its declared expectation that fiscal stimulus would
revive the economy by engendering greater capital spending as well as more
consumption. For although stronger consumption would definitely have helped Japan,
the last thing the country needed in the 1990s was for companies to amass more plant and
equipment. Isolated industries such as high technology or healthcare might ideally have
added to their productive capacity, but this should have occurred in the context of
decreasing overall investment as Japan finally converged toward the pattern that
characterized the United States, Germany, and other mature economies. In the meantime,
the establishment of another domestic automobile factory, the opening of another
unprofitable retail outlet, or the incorporation of one more construction firm could not
possibly have improved the efficiency of the economy. Quite the opposite: such efforts
would have lowered the rates of return available to households, pension funds, and other
investors, and hence have fostered a higher savings rate over the medium and long term.
It is hard to imagine a development more inimical to the well being of a country that was
already suffering from inadequate aggregate demand.
slowdown in that sector was consequently inevitable. Tokei Geppo, February 2000, p. 17. Second, while it
is true that the Hashimoto cabinet’s imposition of new taxes in April 1997 caused households to reduce
their spending after that date, it also motivated them to increase their expenditures beforehand.
Consumption was thus falling through the third quarter of 1996, then improved sharply in the next two
quarters as households rushed to purchase goods and services that they would otherwise have bought later.
Ibid., January 1998, p. 2. This suggests that if there had been no change in the tax laws, household
spending would have followed a much more linear path: consumption would have been weaker throughout
the 1996 recovery, and the end of the construction boom in 1997 may well have pushed Japan back into
recession. It is also worth noting that during the 1996 recovery Japan was still operating at least 30%
below full employment—meaning that the fiscal stimulus had not closed the output gap and hence that
Japan had not escaped its deflationary circumstances. AWSJ, 5 January 2001; and Nikkei Business, 9
March 1999. This, too, suggests that the recovery would not have lasted much beyond the middle of 1997.
Posen is also one of the economists who claim that the fiscal expansion of 1998-1999 presaged a
sustainable recovery. Posen, “Nothing to Fear”. But as in 1996, the output gap remained very large
throughout this period, and the economy slipped back into stagnation as soon as the fiscal stimulus
dissipated. In fact, neither of these recoveries was sustainable.
53
The second major objection to Japan’s fiscal policy was that it did not promote
restructuring of the sort which might have shifted the balance of economic activity away
from corporate investment and towards consumption. This argument appears more
cogent. There were limits to the efficacy of such reforms, for a nation with so many
middle-aged people was likely to persist in accumulating wealth under almost any
conditions. But as witnessed by the modest fall in the household savings rate during the
bubble period, workers and individual investors were not altogether insensitive to
changes in company profitability. It should accordingly have been possible to make
some progress if, for instance, the government had thoroughly reformed the banks, shut
off the supply of credit to poorly performing enterprises, and raised the standard of
corporate governance so that labor and capital were allocated more appropriately. Such a
process would assuredly have been traumatic in the short term, as marginal corporations
failed and the ranks of the unemployed swelled; but the government could alleviate some
of this stress through tax cuts, the extension of better welfare benefits, and moderately
inflationary monetary policy. When the reforms reached fruition in a higher rate of GDP
growth and better returns on investments, households would presumably feel richer and
curtail their savings just as they had done in the heady days of the late 1980s. The surfeit
funds that must be exported through a larger current account surplus or absorbed by
government budget deficits would then diminish, and Japan’s future would look
commensurately brighter.
Regrettably, more often than not the government tried to preclude meaningful
restructuring rather than to encourage it. This was evident in the pressures applied to
Nissan Motor Corporation in 1993 and 1994, when politicians proved more reactionary
than organized labor, and in the regulators' attitude toward the country’s dysfunctional
banks throughout the decade. The same conservative emphasis informed the Obuchi
cabinet's provision of almost ¥1 trillion in consumption vouchers to households in the
years 1998-1999, supposedly the peak of reformist fervor, for this largesse was intended
in part to ease the financial stresses under which inefficient "mom-and-pop" operations
were laboring so that they would continue to assist the LDP and New Komeito in national
54
elections.104 Prime Minister Obuchi's simultaneous extension of ¥30 trillion in loan
guarantees to small and medium-sized corporations was likewise designed to prevent the
collapse of uncompetitive enterprises that were sympathetic to the ruling coalition. The
import of these initiatives, however, was to leave the state in the unenviable position of
underwriting loans to companies that even Japan’s reckless bankers considered unduly
risky. Still another policy designed to slow the reallocation of resources was the apparent
decision, in the aftermath of Shinsei Bank’s privatization, not to let foreigners gain
control of another big lender. In keeping with this new policy, in 2000 the government
sold NCB to a predominantly Japanese consortium despite the fact that that group had
much less expertise in corporate rehabilitation and was offering $540 million less than
the highest Western bidder.105 In all these cases the authorities acted to shield vested
interests from the winds of change rather than letting market forces rationalize the
economy.
Yet the most egregious example of sub optimal fiscal spending—judged, again, from an
economic vantage point as opposed to a political perspective—was the bounty bestowed
on the construction industry. By the late 1980s Japan had a very large stock of factories,
office buildings, houses, and apartments, so the private sector’s appetite for new
construction would probably have waned even if asset prices had not subsequently
collapsed. But collapse they did, and so consecutively did industry demand. Private
orders to the 50 largest builders decreased from ¥19.2 trillion in 1990 to ¥9.6 trillion in
1999, and this drove their ROE down from 26.2% to 11.3%.106 The only way to reverse
this financial decline would have been to shut down superfluous companies and lay off
redundant workers until total capacity fell to a level consistent with market demand. But
rather than shrink, the sector actually expanded. The percentage of the workforce
employed there rose from 9.1% at the beginning of the decade to 10.1% at its end, and
104 EIU, Japan Country Report, 1st Quarter 1999, pp. 17-8. Kondo, et al., pp. 28-30.
105 Stokes, p. 66.
106 The construction orders: Tokei Geppo, January 1998, p. 17; and May 2001, p. 28. The ROE figures:
Goldman Sachs, Outlook for the Global and Japanese Economy/Financial Markets, 2001-2002, January
2002, pp. 34-5,65. Goldman Sachs, “Will Koizumi's Structural Reforms Fail Japan?” 15 May 2001, p. 2.
55
the number of firms in the industry rose even faster.107 By the latter date some experts
would reckon that half of the builders’ productive capacity was excess.
[Chart 21: "Official" vs "Unofficial" Budgets]
Such prodigious overcapacity could not have been maintained at a time of flagging
household and corporate demand without significant assistance from the Japanese
government. Given the importance of the building companies to the LDP’s political
fortunes, however, there was never much doubt that such help would be forthcoming.
Each year the Liberal Democrats ensured that the government’s regular budgets included
big dollops of cash for public works projects. When public scrutiny and opposition from
fiscally conservative officials in the MOF prevented the ruling party from being as
generous in this regard as it wished, it adopted less transparent expedients. The
aforementioned Fiscal Investment and Loan Program was one such mechanism. It grew
enormously during the period in question, from ¥20.5 trillion in 1985, to ¥35.8 trillion in
1990, and then to ¥52.9 trillion in 1999.108 Since approximately one-third of the FILP’s
yearly expenditures were devoted to infrastructure development, this expansion entailed a
vast increase in official support for construction companies.109
Still another channel for public subsidies was the government’s supplementary budgets.
At several points during the decade it became clear that economic activity was slowing,
and the political leadership in Tokyo reacted by formulating stimulus packages that were
expected primarily to precipitate GDP growth but secondarily to protect Japan’s myriad
construction firms. This latter motive is reflected in the composition of the fiscal
schemes.110 Whereas well-designed tax cuts might have fostered corporate restructuring
and hence have strengthened consumption, such items accounted for only 13% of the
107 Ministry of Health, Welfare and Labor, cited in Goldman Sachs, Outlook, p. 35. The Economist, 9 June
2001, p. 63. Wall Street Journal, 5 March 2002. Also, Japan Times, 21 December 2000.
108 Ministry of Finance, Research and Statistics Bureau, Keizai Tokei Nenkan, various issues.
109 The one-third figure understates the true level of government support, for roughly an additional tenth of
the budgets was designated for small and medium-sized enterprises—many of which also happened to be
construction firms. Toyoda, pp. 82-3.
110 EIU, Japan Country Report, December 2001, p. 7. Bergsten et al., p. 63. Posen, 1998, pp. 8,118-9.
56
fiscal stimulus as compared to 42% for expenditures on infrastructure. Over the course
of the decade these supplementary packages added more than ¥25 trillion to the public
works provisions of the original annual budgets.111 Finally, the ruling party also used
commercial banking system to sustain the builders. Small, and burdened by considerable
excess capacity, many of these corporations were constantly losing money and generating
NPLs for their creditors to absorb and should, therefore, have been forced into
bankruptcy. But the regulators and senior Liberal Democrats did not want this to happen,
so they quietly encouraged Japan’s lenders to continue supplying capital to them.112
What the authorities feared most, after all, was not destabilizing the financial system or
wrecking the government’s finances but losing the pecuniary and logistical advantages on
which their power was predicated.
[Chart 22: Fiscal Stimulus Packages]
The third and most persuasive criticism of Japanese fiscal policy during the 1990s was
that it endangered the country’s long-term financial stability. The government began the
decade with what was by global standards a low level of national debt and a sizeable
budget surplus. This was an important fact; it meant that Japan had the wherewithal to
use aggressive tax cuts and sagacious spending programs to neutralize the contractile
effects of the industrial restructuring that was essential to lowering the corporate and
household savings rates. But Tokyo opted to use its resources to preserve, rather than to
change, the country’s economic system. This strategy was not without macroeconomic
merit—once again, the gradual enlargement of the government deficit soaked up surplus
funds and forestalled a protracted fall in GDP—but it left the economy largely
111 These data are “mamizu,” or actual spending. Reading the stimulus numbers is a fraught business
because the headline figures were deceptive, suggesting a much greater volume of new spending than
actually occurred. Taking the government’s claims at face value would lead to the conclusion that outlays
on infrastructure totaled almost ¥60 trillion. OECD, December 2001, p. 52. A more careful reading, by
contrast, produces the ¥25 trillion figure. Goldman, Japan Economics Analyst, 30 November 2001, p. 3.
Generally, Bergsten, et al., pp. 62-5. It should also be noted that some of the money used to finance these
expenditures came from the FILP program, which means that adding together the infrastructure allocations
in the original, supplementary, and FILP budgets would result in an overstatement of the government’s
total construction activity.
112 NikkeiNet, 30 November 2001. Wall Street Journal, 5 March 2002.
57
unreconstructed, constrained by inadequate aggregate demand, and needing still more
extraordinary stimulus. As a consequence the budget deficit grew ever larger and the
national debt, measured in the more appropriate gross terms, climbed from a respectable
65% of GDP in 1990, to 124% of that standard in 2000, and to even greater heights in the
following years.113 Such indebtedness did not pose any insuperable difficulties during
113 EIU, CountryData, 2001, 2004. The superiority of the gross data stems from problems with the net data.
The method used to calculate a country’s net debt is to subtract the value of certain categories of stateowned
or state-controlled assets from the gross debt in order to arrive at figures that more accurately reflect
that nation’s financial balance. After making such adjustments, the IMF put Japan’s net indebtedness at
80% of GDP in 2003, as compared to 166% in gross terms. IMF, World Economic Outlook, 2004 p. 12.
The assets used to make such adjustments are generally debits in various accounts at the central and local
governments, though in at least one case—Christian Broda and David E. Weinstein, “Happy News from the
Dismal Science: Reassessing Japanese Fiscal Policy and Sustainability”, June 2004,
http://nyfedeconomists.org/broda/pub.html; summarized in The Economist, 26 June 2004, p. 78—the
analysts also subtracted the value of the JGBs owed to the central bank on the grounds that the government
would suffer no adverse consequences if it defaulted on that portion of its borrowings. This led to an
estimate of 62% of GDP at the end of 2002.
While it is certainly true that the Japanese state possesses many valuable assets and that these
could be used to offset large portions of the gross national debt, this approach suffers from three important
shortcomings. First, between 1990 and 2003 the government’s total indebtedness increased from ¥300.1
trillion to ¥771 trillion. EIU, CountryData, 2004. It follows from this that both its gross and its net
positions deteriorated significantly over that period. The incremental increase in Japan’s borrowing—
nearly a year’s GDP—is in fact bigger than the entire net debt as reckoned by the IMF, Broda and
Weinstein, and many other economists.
Second, until the government redeems its outstanding JGBs it must pay the interest on them and
remains liable for their principle. In this sense the concept of net debt is irrelevant. The government could
not, for instance, default on the 41% of the outstanding JGBs owned by private institutions without
harming those institutions, damaging the economy, and infuriating the electorate (BOJ, Flow of Funds
data). Nor could the authorities renege on the 29% of the JGBS that are owned by the postal savings,
postal insurance, and public pension systems, for these agencies likewise derive their funds from individual
and household depositors. There may in theory be more flexibility with regard to the 13% of the debt held
by public financial institutions and the 15% owned by the Bank of Japan, but the private sector would
presumably react negatively to any indication that the government intended to cancel some or all of its
financial obligations. Such an announcement could in fact trigger a sell-off that pushed up yields and
render it more difficult for the country to service its remaining debt. In practical terms, therefore, Japan
will remain saddled with considerable financial obligations, and hence vulnerable to changes in interest
rates, until it actually retires a meaningful volume of the outstanding JGBs.
Third, most calculations of Japan’s net debt are not really “net” at all. The statisticians diligently
identify and subtract the value of concealed and off-balance sheet assets from the gross national debt, but
they do not ask whether those assets are liquid: whether, that is, they can be sold for all, most, or any of
their supposed worth. This omission leads to a substantial overstatement of the value of the government’s
portfolio and hence of its net worth. Equally problematic is those analysts’ failure, after reducing the gross
debt by subtracting the government’s assets, to add back the government’s hidden liabilities. One category
of ignored losses is those in the FILP system, which one recent estimate put at nearly 20% of GDP. Takero
Doi and Takeo Hoshi, “Paying for the FILP”, NBER Working Paper #9385, 2002. Other government
accounts also contain significant unacknowledged losses, which tend to be overlooked because they are
difficult to quantify. Advocates of net figures also tend to neglect the probability that the government will
be called upon to absorb losses incurred by Japan’s regional banks and it’s many troubled insurance
companies. Less obvious still are the contingent liabilities in Japan’s various pension schemes. Purists
Footnote continued on the next page.
58
this period, when interest rates were exceptionally low and the government could easily
float new bond issues. But if the country’s debt mounted much further the cost of
financing it might eventually exceed the revenues that the state could derive from even a
reformed and relatively vigorous economy.
Though subject to the caveat that economic prognostication is always an imprecise art,
medium-term simulations undertaken by the Economist Intelligence Unit (EIU) convey
some sense of the impending challenge.114 In that group’s baseline scenario the pace of
growth in Japanese GDP peaks well above 3% in 2004 and then recedes to the country’s
medium-term potential rate of some 1% from 2007 onwards.115 Because the prices of
goods and services are now beginning to rise, the average real interest rate on outstanding
Japanese Government Bonds (JGBs) falls from 1.0% to 0.4% over the forecast period,
which implies that the cost of financing the national debt declines even though the
government is borrowing ever more money. Also beneficial is a big increase in annual
sales of state assets, such as stakes in railroads and other corporations, in order to garner
observe that these shortfalls should not be considered current debt because they have not yet accrued. This
view is certainly valid with respect to the present, but it becomes less compelling if one is trying to predict
the government’s future financial condition. The deficiency in the public pension system is projected to
total some ¥450 trillion (Wall Street Journal, 24 May 2004. NikkeiNet, 23 April 2004), and that of the
private schemes is expected to reach ¥530 trillion (IMF calculations, summarized in EIU, Japan Country
Report, March 2004, p. 18). The government could renege on its commitments, and allow the underfunded
private pension systems to do the same, but this would have a profoundly negative impact on household
confidence, economic growth, and public support for the elected authorities. For these reasons it seems
logical to assume that the political leadership will ultimately strike a compromise, providing some
proportion of the payouts promised to retirees by both public and private institutions. That compromise, in
turn, will suddenly push the net debt figures substantially upwards. Accurately predicting Japan’s financial
trajectory over the few decades therefore requires that one estimate the timing and magnitude of the
pension burden’s materialization and then add that number—which could easily be as much as a year’s
GDP—to the net debt projections. If this is deemed too complex a task to accomplish with any precision,
then a second-best solution is to use the gross national debt as a crude, ballpark means of estimating
Japan’s future financial position.
114 The reform scenario may be found in EIU, Japan Country Forecast, June 2003. The rest of the results are
unpublished.
115 The IMF alternatively estimates that the potential rate in 2004-2005 is about 1% and assumes that
comprehensive restructuring will overcome the negative effect of a shrinking labor force and produce a
moderately higher trend growth rate in the latter half of this decade. IMF, World Economic Outlook, 2004,
p. 28; see also, pp. 3,10,12. The Japan Center for Economic Research (JCER) believes that the potential
rate could rise well above 1%, and perhaps towards 2%, by 2009; but only if comprehensive reforms are
enacted and large numbers of women and retirees join the workforce, thereby relaxing the demographic
constraint. JCER, 30th Medium-Term Forecast of the Japanese Economy: Fiscal Years 2003-2010,
Footnote continued on the next page.
59
more revenue. Yet despite this combination of economic growth, lower interest
expenses, and enhanced revenues, the budget deficit widens from 7.4% of GDP in 2003
to 8.5% thereof in 2008. Even worse, the gross national debt grows from 155% to 208%
of GDP, leaving Japan much more vulnerable to domestic and international shocks than it
is today.
The alternative scenario posits a moderate degree of fiscal tightening and yields a slightly
less tenebrous conclusion. In this simulation the government raises the consumption tax
from its present level of 5% to 7% in 2007, when it likewise begins imposing a series of
annual increases in the income tax. These assumptions are in fact quite generous.
Cognizant that a similar rise in the consumption duty in 1997 so offended voters that they
forced then-Prime Minister Ryutaro Hashimoto to resign from office, today’s Prime
Minister Junichiro Koizumi has pledged that the LDP will refrain from enacting any
changes in the relevant laws until 2007 at the earliest.116 Underscoring this cautious
attitude is the fact that national elections to both the upper and lower houses of the Diet
are scheduled for later that year. Since it would be impolitic to start detailed discussions
of tax reform until those polls are safely in the past, the earliest that new duties could
realistically be levied is probably 2008 or 2009. But if one ignores these political
considerations and initiates the retrenchment process in early 2007, as in the EIU
exercise, the result is that the budget deficit stops widening and stabilizes at about 7.3%
of GDP. This represents an improvement over the baseline scenario but, with the debt-to-
GDP ratio surpassing 200% in 2008 and still growing at roughly the same speed as in
2003-2004, it is hardly cause for celebration. Thus the reforms proposed in the EIU
analysis prove to be but a modest first step in what must perforce be a long and arduous
program of financial consolidation.
http://www.jcer.or.jp/eng/. If any of these conditions fails to materialize, the JCER projections would fall
back towards the EIU forecast.
116 For Koizumi’s pledge, see: NikkeiNet, 19 January 2004, 22 January 2004, and 9 April 2004.
60
[Chart 23: Budget Deficits Trajectory]
[Chart 24: National Debt Trajectory (found on same worksheet as 23)]
The situation is by no means hopeless. By international standards Japan is taxed lightly:
the state took an average of only 31% of GDP each year between 2000 and 2004, as
opposed to almost 40% for the members of the OECD and even more for several of the
EU nations.117 This means that there should be plenty of room to expand government
revenues in order to eliminate the budget deficit and begin paying down the national debt.
This desirable outcome would be jeopardized, however, if Tokyo did not act decisively
and at the right moment. On the one hand it must not raise taxes before private-sector
demand has strengthened and the output gap closed, for doing so would push the country
back into recession and prolong the economy’s dependence on state stimulus. The
recovery of 2002-2004 has surely reduced that gap, but it will probably be another two or
three years before the government may safely begin the process of aggressive fiscal
retrenchment. On the other hand, Tokyo cannot delay painful tax hikes far beyond the
end of the present decade without letting the debt grow so large that the average interest
rate on JGBs starts to rise and the pressures on the government budget intensify. The
temporal window within which to make the necessary adjustment may thus be fairly
narrow.
Some commentators believe that Japan will easily negotiate these problems of timing and
political judgment because “it seems hard to imagine the Japanese government” failing to
act in the country’s best interests.118 Such optimism, however, is at variance with the
history of political and regulatory behavior retailed in this paper; a history which
unfortunately suggests that the authorities very well could postpone fiscal tightening until
it is too late to contain the national debt. The temptation to dither may in fact be growing
more compelling as the electorate ages, for older and more conservative voters generally
tend to oppose dramatic change of the sort Japan so conspicuously needs. An eventual
117 EIU, CountryData, 2004.
118 Broda, et al., p. 32.
61
default on the national debt, either overtly or through a bout of rapid inflation, can
therefore not be ruled out.119
With this in mind Japan’s fiscal policy in the 1990s was plainly sub optimal. It is true
that the increase in deficit spending neutralized the contraction in corporate investment,
bolstered aggregate demand, and generated a modicum of GDP growth. Put simply, the
government succeeded in fending off depression. But Tokyo exhibited its largesse in
ways that preserved a severely distorted economic system in which corporations
transferred too little wealth to workers and investors, the private-sector savings rate
remained elevated, and the NPLs in the financial sector proliferated rapidly. The price
deflation which began in the latter half of the decade was both a symptom of this
distortion and an aggravating factor in it, putting additional stress on the banks and
necessitating still more deficit spending.120 Even worse, perhaps, this deflation depressed
the interest rates that Tokyo was obliged to pay in order to borrow money from investors
and hence rendered it easier for profligate Liberal Democrats and civil servants to persist
in wasting valuable public resources. Japan accordingly entered the 2000s largely
unreformed, as reliant on fiscal stimulus as ever, and with the possibility of a debt crisis
looming on the distant horizon.
Conclusion
In the late 1980s and 1990s Japan suffered from what one writer has termed “the
dilemma of the one-party state”.121 The LDP had dominated the nation’s politics for so
long that it had become identified with a particular industrial and financial structure: it
119 A normal default occurs when a country cannot pay its foreign creditors and is forced to reschedule, or
renege upon, its commitments. But if the debt is held domestically and its interest rate is fixed, then a
government need not admit that it is defaulting and can simply inflate away some of the debt’s real value.
This option is available to Tokyo because the proportion of the JGBs held by Japanese entities is very high:
92.6% in 1999, and more thereafter. OECD, Volume 2002, Supplement, No. 2, p. 62.
120 Krugman, “ Respectability”, February 1999; and “The Land of the Rising Yen”, September 1999. Also,
Asian Wall Street Journal, 26 December 2000.
121 Mikitani, “The Facts”, p. 29.
62
was quintessentially a party of the status quo. The government was thus ill prepared to
redress the imbalances that emerged in the middle 1980s. By this point the economy had
matured, and household consumption should have begun to replace corporate investment
as a source of new aggregate demand. This, however, did not happen. Demographic
factors, poor corporate governance, official intransigence, and other forces prevented the
private-sector savings rate from falling and thereby left Japan saddled with much more
capital than it could reasonably expect to employ within its own borders. This problem
of excess savings could have been resolved through a much larger current account
surplus, which would have shipped Japan’s excess funds abroad; through corporate
restructuring, which would have lowered the savings rate; or through some combination
of both. The economy would then have moved towards a new and more stable
equilibrium, and respectable GDP growth would have ensued.
Yet these remedial possibilities were never seriously explored. Misunderstanding played
a role in this failure in the early 1990s, for these were years in which very few people in
Japan or abroad appreciated the virulence of the country’s illness or perceived the
urgency of the need for reform. Later in the decade, however, it was largely self-interest
on the part of the main economic actors which precluded a return to health. The BOJ
actually preferred a strong yen, which put it in de facto alliance with the United States
and those other countries that adamantly opposed an increase in Japan’s current account
surplus. The MOF and the country’s elected governments occasionally expressed
dissatisfaction with this state of affairs, but they never used their own powers to drive the
value of the yen down far enough to reconcile the national savings rate with the
requirements of full employment. So the burden of adjustment fell squarely on Japan’s
domestic economy. But here, too, the major actors refused to act. The Liberal
Democrats, the civil service, business executives and bank managers all feared the
ramifications of structural change and consequently refused to endorse a program of
comprehensive reform. To the people who mattered, in short, the cure seemed worse
than the disease.
63
By default, then, Japan was compelled to find other, less appropriate, ways to instill vigor
in its economy. Part of the answer was excessive corporate investment. During the
bubble years companies had greatly increased their capital expenditures, and this
absorbed much of the surplus capital and gave the economy substantial positive
momentum. In the early 1990s, however, firms began to curtail their investment and this
forced the government to step in, expanding its budget deficit in order to compensate for
the new corporate parsimony. This gambit succeeded in forestalling depression,
produced some GDP growth, and preserved the bastions of LDP power. But it also
pushed Japan closer to the shoals of financial trouble, for by the turn of the century the
national debt was swelling so rapidly that an eventual financial crisis was no longer
inconceivable. In the meantime, the disappointments of the 1990s had eroded public
confidence in the nation’s political leadership, the bureaucracy, and the business
community—precisely the institutions that would be called upon to implement the
corporate and fiscal reforms necessary to contain the national debt. In this sense the
legacy of the "lost decade" was to leave the country with significantly fewer resources
with which to manage the challenges it will encounter in the late 2000s and 2010s.