Japan 2001: Decisive Action or Financial Panic
by Adam S. Posen, Peterson Institute for International Economics
© Institute for International Economics. All rights reserved.
The author is a senior fellow at the Institute for International Economics and coeditor of Japan's Financial Crisis and its Parallels to U.S. Experience (2000) and author of Restoring Japan's Economic Growth (1998).
Japan's economic tumult during March 2001 has prompted renewed fear of an impending crisis. Mixed macroeconomic data of the past months indicate that at least a brief recession is likely. To add to the panic, Japanese Finance Minister Kiichi Miyazawa announced on 8 March that Japanese government finances are "very close to collapsing." The Bush-Mori summit of 19 March, and the Bank of Japan (BOJ) rate cut earlier that day, raised both hopes and skepticism that Japanese officials were preparing to address their fragile banking system. Transpacific pundits have once again begun to chant, "Ah! At last true crisis will come to Japan and force change."
It is true that the Japanese economy could benefit from structural reform and the Japanese banking system specifically needs to finally write off accumulated nonperforming loans (NPLs). The Mori government and the Liberal Democratic Party (LDP)-led coalition in the Diet have been unwilling to bear the financial consequences that writing off NPLs will have for small businesses and thus have discouraged the banks from doing so. Statements like Miyazawa's, or the reluctance of the BOJ to actually ease the problem could provoke an outflow of capital from Japan, which would raise long-term interest rates, and thus cause widespread bankruptcies. And until the week of 19 March, it appeared that both Miyazawa and BOJ Governor Masaru Hayami were intentionally engaging in brinkmanship with the Diet, the markets, and with each other to scare their respective counterparts into the desired actions. All true, and all truly mistaken policies with likely bad results for Japan, the United States, and the world economy.
The resolution of the Japanese bad loan problem is not only overdue but also remains highly doable, given Japan's rising potential growth rate and already huge private savings. What is new is that the partial reforms of the banking system undertaken to date have finally brought it to the point of no return: if write-offs of the NPLs are not undertaken by the time Japanese banks reveal their mid-FY2001 results at the end of September 2001 under the new accounting standards, or if the new accounting standards are not adhered to, money will simply leave the Japanese financial system, if not Japan altogether. If the NPLs are directly addressed before then, the Japanese economy could rebound quickly from a finite deflationary period of asset reallocation—in fact, there would be room for expansionary monetary policy to offset the contraction. Given that the present choice therefore is between a controlled implosion of the Japanese banking system (with efficiency benefits) and an uncontrolled financial crisis (with international spillovers), which may leave the banking problems unaddressed or worsened, the US government has a direct interest in using its financial leverage to encourage the Japanese government to quickly take the first option.
Making a Crisis Deadline Out of Progress
Prior to the Nikkei sell-off of March 2001, Japan's mixed macroeconomic data and declining stock market since mid-2000 are indicative of structural reform in progress:
In short, a greater share of Japanese investment capital is being allocated according to market signals than ever before in the postwar era, raising returns and efficiency through the urban-industrial parts of the economy. The signs of oncoming recession are not the final verdict on the economy but a reflection of both the costs of restructuring and the effects of a US slowdown. Japanese potential growth is actually rising and macroeconomic policy should take advantage of this by compensating expansion on the monetary side and by slow contraction fiscally-not by trying to deepen a recession.1
As mentioned above, the one structural barrier to Japan's return to growth has been the ongoing accumulation of NPLs and the resulting continued undercapitalization of the banks.2 Properly capitalized banks do not rollover bad loans because they have enough of their own capital at risk and can bear the costs of writing off bad loans. The costs to Japan of its dysfunctional financial system are enormous in terms of lost growth, missed investments in new firms and projects (due to rollover of bad loans), and low returns on savings. Still, even the Financial Services Agency (FSA) under reform-minded Minister Hakuo Yanagisawa (also chairman of the Financial Reconstruction Commission), let alone the Diet, has been unwilling to take the last step and force the writing-off of NPLs.
The reluctance to pull the trigger appears to be based on the fact that such write-offs would leave most Japanese banks under water, requiring the unpopular steps of another injection of public funds and/or round of closures. The banks' own reports indicate there are 82 trillion yen (16 percent of a year's GDP) worth of NPLs for which they have been partially provisioning. Reliable private sector estimates add another 40 trillion to 50 trillion yen to this number based on the facts that lending has continued into the decadent construction and retail sectors since the last bank recapitalization of April 1999 and that even loans classified as Class I and good Class II (performing) usually show some defaults.3 Meanwhile, the entire banking system has only about 60 trillion yen in capital and losses are expected on their nonfinancial equity holdings.
Ironically, what makes a crisis almost inevitable in Japan this year are the banking reforms undertaken, however grudgingly, to date. Tighter supervision and greater transparency, requiring Japanese banks to mark-to-market value their securities in FY2001, have made the banking system stronger than it was in 1997-98, the last time a financial crisis loomed. The belief that the new accounting standards will actually be enforced is reflected in the fact that the Japanese banks have been selling off their cross-shareholdings and investment stock (that sell-off was the initial force behind the Nikkei's decline in recent months and was also a useful reallocation of capital). Yet, these regulations also mean that come 30 September 2001, when the Japanese banks will have to make their first half-year report under the new accounting standard, all will be revealed. Unlike prior false deadlines announced by Japanese regulators that were finessed, put off, or ignored, this one will be enforced by the markets.
Implementing the write-off of the bad loans should follow what is now a reasonably standardized sequence from other countries' banking crises. First, supervisors truly count and publicly recognize the extent of the bad loans, using market signals about the worth of various assets—those banks whose losses exceed their capital are shut down, with bank shareholders losing all their equity and depositors being reimbursed up to deposit insurance limits. Those banks that have positive but too little (i.e., below prudential standards) capital after taking loan losses, are either sold or merged with other more viable banks or are recapitalized by an infusion of public funds, which are conditional on certain performance requirements—shareholders again take a loss, with some equity shifting at least temporarily to the public sector.4 Also, foreclosed collateral (land) is sold off.
On purely economic terms, the undercapitalization, the NPLs, and the long-term downward trend in demand for banking services should prompt the FSA to close or merge half the banks in Japan. Though the political reality will prevent this, the FSA must make clear to the public that the Japanese banking sector must shrink. The surviving banks, including those with public funds, will be forced to pay off the loan losses over several years, absorbing most profits. Even with proper attempts to hold bank shareholders and management accountable for past mistakes, there will be a large cost to the rest of the Japanese economy. Numerous small- and medium-businesses that cease to have their credit rolled over will go bankrupt and still more will suffer from the effects of declining demand and rising unemployment. Japanese taxpayers will in the end have to pay the lion's share of capital injections into the banking system and the deposit insurance payouts, though they will get back some money from the sale of bank assets.
None of this is open to dispute. The bottom line is the same for all countries facing banking crises. It is based on the same immutable logic about the incentives of undercapitalized banks and the effect of rolled-over NPLs. If the FSA forces the write-offs without capital injections and bank consolidation, the undercapitalized Japanese banking system will return to the worst days of 1998 and cease to function as either a safe repository of savings or as an allocator of investment. If the FSA allows the banks to "waive" or forgive the loans, instead of writing them off, the Japanese government simply gives these failed businesses a cost-free transfer of funds5; this would add to the bill Japanese taxpayers will pay, thus extending the mistake of the last 10 years. The latest proposals to prop up the Japanese stock market with public purchases of bank-owned stock will not address the problem either. Such measures do not recapitalize the banks or end the wasteful rollovers of loans to firms that should close, and thus behave the same as loan forgiveness (but they give bank stockholders a taxpayer-paid bonus for their poor investments).
Making a Macroeconomic Crisis Out of a Banking Implosion?
The Japanese government confronting its banking system today is in the position of a city planner looking at a multistoried building in danger of collapse on a busy downtown corner. Some time in the coming months, the building will fall down; until then, fewer and fewer people will occupy or go near it, lest they be hit by tumbling debris. The usual response to such a situation is to plant explosives for a controlled implosion, and on a set day, after warning bystanders and occupants, collapse the building in on itself (and only itself). Then rebuilding on that spot can begin. Were the Japanese government to pull the trigger on resolving the bad loan problem—with all its attendant damage to bankrupt small- and medium-enterprises on life-support, and to artificially solvent bank shareholders—it would be performing such a controlled implosion. There would be damage but the bottom would be found in both the Japanese stock and land markets, and expansionary monetary policy could offset the deflationary pressure until the banking system was rebuilt. Capital and labor freed to seek higher returns by these corporate shutdowns would raise Japanese growth within a few quarters.
Alternatively, leaving the NPL problem unaddressed while everyone learns about the Japanese banking system's solvency, and while the world economy slows, would be the equivalent of letting the building collapse abruptly. The bystanders in this analogy who avoid this dangerous though central downtown corner, and thus deplete the city of business, would be akin to the Japanese savers putting their money under their mattresses, into zero-return postal savings accounts, or into capital flight abroad, depleting Japan of investment. Prompting panic in the Japanese government bond (JGB) market by currently unnecessary and ill-timed calls for fiscal consolidation and/or by continued countenance of deflation by the BOJ would be even worse. Such policies would be just as wantonly destructive a way of addressing the NPLs as it would be for that city planner to not only remove the collapsing building by blowing up the entire city block, but by placing the explosives under the sound buildings on the opposite side in the hope that they would collapse onto the unsound one.
Thus, the type of crisis at risk in macroeconomic brinkmanship—be it passive failure to offset the costs of the banking problem or active attempts to scare others into changing course—is very different from the costs of resolving the banking problem. For one thing, such escalation is unnecessary. Characterizations of Japanese public finances as being on the verge of collapse are misplaced. The real problem with outstanding Japanese government debt is not its large size.6 Until the signs of panic during March 2001, Japan had a relatively flat yield curve, low long-term nominal bond yields, a stable yen in the Y112-Y116/dollar range, and deflationary expectations among price and wage setters, all of which indicated a near complete lack of market concern about the solvency of the Japanese government. This calm was justified because Japanese net national savings remain enormous, net Japanese government debt is far lower than the gross debt to GDP ratio—less than 6 percent of JGBs outstanding are held by foreigners—and all of the debt is yen-denominated.
Instead, the real problem with Japanese government debt is one of illiquidity. As long as the banking system problems remain unresolved (i.e. growth is constrained by poor capital allocation and as a result banks are over-invested in JGBs), the Japanese bond market is vulnerable to destructive panics (i.e., abrupt, large, sell-offs in the JGB market). This is just what occurred in January 1999, when long JGB interest rates spiked 175 basis points following an unjustified Moody's downgrade of Japanese public debt and a doom-predicting pronouncement about the debt out of the Ministry of Finance's (MOF's) Trust Fund Bureau. An inadequately quick countering expansion of liquidity from the BOJ, let alone a tightening of policy, would effectively validate and prolong such an interest rate rise. This is precisely what happened in February 1999, after the BOJ was slow to respond to the spike, and long-bond rates have remained since then 50-75 basis points higher than in December 1998.
It is possible that the BOJ's announcement of 19 March that it would not raise rates again until consumer price index (CPI) inflation is greater than zero indicates that a cooperative response is coming from the Japanese government. Yet all the players in this game—BOJ, FSA, MOF, Ministry of Economy, Trade, and Industry (METI), and Diet members—have reversed their stated positions many times without any real change in actions. The risk is that Japanese officials are so intent on playing chicken with each other and with the Diet that this JGB panic scenario is likely to repeat. Some budget hawks are willing to risk a bond sell-off in hopes of forcing a BOJ loosening and budgetary restraint from the Diet; Hayami remains unwilling to actively counter deflation in hopes of scaring the government into carrying through the bank clean-up. From whatever source, such an induced JGB panic raises the probability of outright financial crisis so long as the NPLs are outstanding. Even if Hayami, Miyazawa, and Yanagisawa are simply slow to come to agreement, they may well find that their prior brinkmanship has led to the crisis that so many in the cheap seats have asked for.
Such a crisis of self-destructive macroeconomic policy, however, would not lead to the resolution of the banking crisis and the NPLs problem, and will likely make them worse. Provoking a panic in the JGB market, through calls for either deflation or austerity, will lead to a combination of capital flight from Japan and sharply rising long-term interest rates. Such a crisis will hit the good or restructuring Japanese firms along with those inefficient firms on financial life-support. In fact, given the econometric evidence that information imperfections in financial markets increase in recessions, the resulting financial contraction will likely disproportionately hit new investment projects from good firms and the creation of new businesses.
Some cynics retort that such macroeconomic gambits were and are justified because there is no other way to get action on banking reform. Yet, these cynics are themselves naïve about political economy if they believe that the onset of a crisis is more likely to prompt constructive action than a negotiated deal among the players. As seen not only in the case of the construction firms and small retailers in Japan from 1992 to 1998 but also in the examples of the chaebol in Korea in 1998-99 and the natural resource monopolies in Russia since Gorbachev, especially during hard times, it is the already politically protected businesses that are spared, not the ones that suffer as public spending is cut. Meanwhile, based on pure economics, contractionary macroeconomic policy in recessionary times is not creative but destructive because it actually puts more of the wrong firms out of business—you gain nothing from causing additional national suffering.7 That is why the structural improvements underway in Japan described above took hold during the weak expansion since the fall of 1998 and not before.
Most important, unlike directly addressing the banking system problems, such an economy-wide crisis once set in motion has no clear ending point. Once capital starts leaving Japan, and once sell-offs of JGBs begin, most policy responses (such as raising short-term interest rates) will only make matters worse. Such a crisis would arise out of the willful acts of frustrated Japanese policymakers— policymakers apparently incapable of resolving the banking crisis that the United States, France, Sweden, and even numerous developing economies managed to resolve in much less time—and only accelerate the collapse of confidence.
So what should the Japanese government do? The alternative policy agenda is not new but nonetheless is more likely to be successful than policies promoted solely for the sake of their radical nature:8
1. Japanese officials should stop calling attention to government debt until the banking recapitalization is complete. Talking only makes an imminent problem out of something that can and will have to grow before being slowly resolved.
2. The FSA should force the banks to write-off the NPLs and be recapitalized (with public funds) or shut down. Now that the more transparent accounting rules have forced sales of shares and fuller disclosure, this is inevitable. It is both necessary and sufficient to return Japan toward sustained growth.
3. The BOJ should offset the transitional deflationary force of the bankruptcies caused by the NPL write-offs with expansionary monetary policy through purchases of JGBs, dollars, and euros—but not stocks—with printed yen.
4. Whether the BOJ (or US industry) wants it to or not, the yen is going to depreciate until the banking problem is resolved, either through a responsible monetary expansion (as in point 3 above) or through capital flight when a larger crisis hits. Better for the BOJ to anchor the long-term expectations with a multi-year inflation or exchange rate target and try to claim credit for the yen's decline than to fail in an attempt to fight the depreciation.
5. The Japanese government budget can engage in essentially revenue-neutral expansion by shifting spending from wasteful public works to tax cuts aimed at salarymen.9 An enormous difference between the multipliers on the two types of policies was seen in the 1990s, with (temporary) tax cuts and public works spending largely saved or ignored.
6. The remaining financial burden on Japan of illiquid real estate markets can be partially alleviated if the tax cuts in point 5 above emphasize housing. One option would be to offer money saved on cancelled public construction projects to home sellers as rebates for their mortgage losses incurred. Such a measure would be both structurally sound and expansionary.
What a Crisis Would Mean
In blunt terms then, the choice for Japan in 2001 is between the political crisis for the LDP leadership, which would follow writing off the NPLs/recapitalizing the banks, and the outright financial crisis, which will occur if 30 September is reached without addressing the bad loans problem—sooner if the brinkmanship of the MOF or the BOJ or an external shock causes panic in the vulnerable system before then. If such a wider crisis were to occur in Japan, it would entail disintermediation from the Japanese banking system—that is a sharp rise in nominal interest rates, widening interest rate spreads, a breakdown in corporate investment, deep cuts in bank lending, further declines in equities and land prices, possible bank runs, and an outright contraction of real activity. Its effects would not be transmitted solely through the financial system, however, but would also directly erode consumer confidence in the government and the economy, and would lead to an outright outflow of investors from yen-denominated assets.10 Such a sequence of events began to unfold in Japan in the first half of 1998, but was halted by the reform efforts (including partial bank recapitalization) of the government of the late Prime Minister Keizo Obuchi.
The international repercussions of such an abrupt collapse would include the frightening litany made familiar by the Asian economic crisis of 1997-98:
The direct impact upon the now weakening US economy would be large. Obviously, there would be import surges driven by a declining yen, particularly in steel, autos, machinery, and electronics, broadening our current account deficit to unprecedented levels and increasing congressional demands for trade protection.11 Slower Japanese economic growth and declining Japanese asset prices mean declining demand for American goods and declining wealth for American investors in general.
On the capital flows side, we can expect sales of Japanese-owned American assets to rise for the next couple of months, through the start of the new fiscal year in April and a bit beyond. Japanese businesses will want cash on hand, and even after recent developments in US markets, their American assets will be easier to sell and more valuable than their illiquid Japanese land assets and depreciating Japanese securities. After this initial capital outflow, however, Japanese savers will join foreigners in selling yen-denominated assets and money will flow back into the US and European markets from Japan, perhaps for many months. This is not all bad, primarily because a declining yen relieves some of the inflationary pressure, which might make the Federal Reserve hesitant to cut interest rates. Yet such portfolio flows would be as much "hot money" for the United States as they would be for an emerging market and they might well push artificially and unsustainably both asset markets and the current account away from their needed corrections.
Whether the impact on the world economy from a crisis in Japan would be less than in 1997-98 is an open question. Now that many economies in the region have adapted to a stagnating Japan and many investors have gotten out of Japanese assets, the impact could be lesser than last time, or the impact could be even greater because a significantly slower US economy and a less-than-fully recovered Asian region would be less able to offset the demand effects of Japanese contraction. If we prefer not to find out the answer to that question from experience—and if the Japanese government remains more afraid of taking the political hit for pulling the trigger on the bad loans and bank recapitalization than of an economic crisis—the United States government must put pressure on Japan now for a direct resolution of the NPL problem.
Just as the Clinton administration found, the Bush administration will soon find that when the government that rules the world's second largest economy does not live up to its economic responsibilities, it is everybody's immediate problem. Even those priests of purer high politics-based foreign policy will find that a national security alliance with a nation rests on that nation's economy growing on average at its potential. This will be especially true of Japan—an ally whose primary contribution comes from being wealthy rather than from being combat ready. If we believe that the long brewing financial crisis is finally about to hit Japan due to its incomplete banking reform and its self-eroding monetary policy, there is little question the United States has a national interest in getting those reforms completed and reversing that monetary policy. The real question is whether the US government can do anything to make the Japanese government implement its promised policies.
In fact, a big stick remains within easy reach of the US government and luckily it is one that works best when wielded quietly. US financial regulators not only have the capability but the legal responsibility to insulate the American banking system from foreign banks that represent a significant payment risk. At the far outside end of this competence, a subset or all of a specific country's banks can be excluded from being parties with American banks—such a radical measure is tantamount to economic warfare and would never even be considered as an option for Japanese banks or those of any democratic ally. There are many much smaller steps, however, which can usefully be taken or threatened to be taken by American regulators to increase the pressure on the Japanese government to clean up the NPLs before a financial crisis breaks out.
The Fed and other regulators can, for example, increase their scrutiny of Japanese bank subsidiaries in the United States or of American banks doing extensive business with Japanese counterparts, as they have done in the past with Latin American financial institutions during times of financial fragility there. Increased scrutiny on either set of institutions represents a small tax that is discouraging at the margin transpacific flows; more importantly, as word gets out to the markets that supervisors are taking such care with Japanese banks, interbank overnight rates will increase for Japanese banks (the Japan premium will return). Requirements that American subsidiaries of Japanese banks have enough liquid assets on hand in the United States to settle overnight balances would have similar effects on the markets and would put direct pressure on the BOJ (which in practice would probably provide most of the short-term liquidity to its domestic banks). While not often remarked upon, such limited but powerful measures were taken by US regulators in 1998, prior to the Obuchi government's partial bank reforms.
Of course, as word gets out into the markets, such measures run the risk of prompting the crisis the US government is hoping to get the Japanese government to prevent. For the United States, facing Japan in coming months would be much like the International Monetary Fund (IMF) confronting the Thai government in spring 1997: if the IMF did not express to the markets its belief that Thai policies were leading to a financial crisis, there would be no credible threat to provoke a change in policy. Yet too great a disclosure too soon could bring on the crisis before policy could change. Most observers of the Asian crisis have since concluded that it would have been better for the world economy had the IMF put more pressure on Thailand and disclosed more earlier.
The early warning indicators of financial crisis have been going off in Japan for years now, and only Japan's accumulated wealth, the reluctance of its savers to invest abroad, and the partial policy reversal of the Obuchi government in late 1998 have let Japan get by until now. If the NPLs are not resolved sometime between now and 30 September 2001 that breathing space will run out and a financial crisis will hit Japan. A financial crisis in Japan will assuredly be a "systemic risk." Because Japan is a surplus country, the IMF has no real traction on its behavior; the G-7 process, demonstrating its latest failure to sufficiently exercise peer pressure and surveillance, has appeared to have had no effect on Japanese behavior except in September 1998, when direct American and financial market pressure on Japan were highest. If the Bush administration is sincere about its commitment to crisis prevention instead of bailouts, it will use its bank supervisory powers—and their effects on market sentiment-to carefully and quietly but firmly pressure the Japanese government to address its accumulating bad loan problem.
In the end, it is up to the elected representatives of the Japanese people and their cabinet, not foreigners, to implement thorough banking reform. If those leaders take the decisive actions of which they have so often spoken, there will be no crisis (and no need for US pressure). But right now, those Japanese politicians seem unconcerned, if not eager, to provoke an avoidable macroeconomic crisis in Japan and around the world rather than face the political crisis that they believe bad loan foreclosure and banking recapitalization would cause. All those with some say about Japanese economic policy—which includes the US government as well as the Japanese private sector—must encourage the government that succeeds Prime Minister Mori to recognize that the collapse is imminent and that it is better to set off a controlled implosion than have the building fall down on you.
1. Adam Posen, "Recognizing Japan's Rising Potential Growth," NIRA Review, Winter 2001, Vol. 8, No. 1, pp. 48-53.
2. See the chapters by Friedman, Glauber, Kashyap, and Shimizu in Japan's Financial Crisis and Its Parallels with U.S. Experience, Ryoichi Mikitani and Adam S. Posen, eds. 2000. Institute for International Economics, Washington DC.
3. Prices for distressed loans in secondary or securitized credit markets that exist in Japan today are consistent with these estimates of the worth of outstanding NPLs.
4. Some measure of supervisory discretion will have to be used to determine which banks to shut, both because market signals are noisy during periods of economic contraction and uncertainty, real estate markets for collateral will remain illiquid, and because there is a limit to how many accountants and supervisors the FSA can mobilize in the time remaining. Nevertheless market indicators should be taken seriously and general principles for assessing capital adequacy should be announced ahead of time.
5. Apparently, Yanagisawa is now proposing such half measures (see David Ibison, "Japan set to duck key banking reform," Financial Times, 21 March 2001, p. 13). Assuming that whatever the Japanese government does will be a consensus policy and that Yanagisawa would come to the negotiations as the representative of the tough reformer position, this augurs poorly, summit agreements notwithstanding.
6. See Adam S. Posen, "The Real Trouble With Japanese Debt," The International Economy (July/August 2000).
7. As argued in Adam S. Posen, Restoring Japan's Economic Growth, chapter 6. 1998. Institute for International Economics, Washington DC; and in Adam S. Posen, "Little creative benefit in recession," Financial Times, August 2, 2000, p. 11.
8. The details of how to implement these or similar policies have been set out in previous publications, including Restoring Japan's Economic Growth (chapter 5). 1998. Institute for International Economics, Washington DC; Adam S. Posen, "Implementing Japanese Recovery," Policy Brief No. 99-1, Institute for International Economics, Washington DC; Adam S. Posen, "Nothing to Fear but Fear (of Inflation) Itself", Policy Brief No. 99-9, Institute for International Economics, Washington DC; and Japan's Financial Crisis and Its Parallels with U.S. Experience, Ryoichi Mikitani and Adam S. Posen, eds. 2000. Institute for International Economics, Washington DC.
9. Strict Keynesian algebra, of course, tells us that a balanced budget multiplier is less than one, whereas all public works spending has a multiplier of at least one. In the case of Japanese public works, however, much of the public works spending takes on the attributes of transfers, and construction firm owners, seeing the end of the gravy train coming, seem to display Ricardian behavior. This is under research.
10. Adam S. Posen, Restoring Japan's Economic Growth, chapter 4, 1998, Institute for International Economics, Washington DC, spells out in more detail such a financial crisis scenario, its economic underpinnings, and the international implications given below.
11. See the analysis in Marcus Noland, et al, "The Continuing Asian Financial Crisis: Global Adjustment and Trade," Working Paper No. 99-4, Institute for International Economics, Washington DC, which examined what was likely to happen as a result of the crisis of 1998-98.