Japan’s Fiscal Investment and Loan Program (FILP) is a government-operated system for directing public savings into projects that are deemed to have one or more of three public purposes: (i) allocation of societal investments to priorities that might not otherwise be financed by the ordinary operations of the market; (ii)intermediating between the government’s fund-collecting vehicles and various government projects; and (iii) countercyclical financial operations that would tend to stabilize the economy.

Before the reform that was supposed to be implemented on April 1, 2001, the funds intermediated in the FILP process came from many government controlled and administered sources, but principally the Postal Savings System (58%) and the Public Pension System (33%).[1] These were compulsorily deposited with the Trust Fund Bureau of the Ministry of Finance and used in support of FILP’s purposes. In its 2000 Annual Report, FILP noted that the total sum currently invested—apart from the amount invested directly in Japanese Government Bonds—was Y414 trillion, almost entirely in the form of loans. At the current exchange rate (Y=120), this is roughly $3.35 trillion. In recent years, as FILP continued to grow, its outlays reached 60 percent of the size of the Japanese government’s total acknowledged budget, known as the General Account.[2]

Although FILP has a number of stated objectives, as outlined above, it has clearly been a failure in performing the one role that was of singular importance during the 1990s—adjusting the economy through countercyclical spending. Indeed, most commentators now believe that the FILP system has exacerbated Japan’s current financial problems by creating a massive overhang of debt and huge but as yet unrecognized losses in the areas to which FILP funds were directed. If not otherwise accounted for, these debts and losses will show up either as a direct liability of the government to the sources of the funds—the Postal Savings and Pension systems—or eventually as a government obligation to the beneficiaries of both systems if these programs are unable to meet their obligations. There are also other less easily quantifiable losses—discussed below—for which FILP might be held responsible.

The FILP “reform” has a decidedly Japanese cast; it is so incremental that it may not be reform at all. The fund flows in the system are changed, but there are no fundamental changes in this or in the structure of the FILP, at least in the short run. Although funds used to flow as deposits from the Postal Savings and Pension Systems to the Trust Fund Bureau of MOF and then to the FILP corporations, they will now—at least as an interim measure—flow to a new Special Account at MOF in exchange for government guaranteed instruments to be known as FILP Bonds.[3] It remains to be seen whether this “interim measure” does not turn out to be the long term structure, but even if it does not the long term structure the reform envisions—as discussed later in this paper—is not likely to result in any significant changes in the way the FILP system is treated as a fiscal matter by the government or in the way it affects the economy.

Moreover, the reform makes no changes in the structure by which the government operates “off the books” through government owned or controlled corporations. None of these—which were the recipients of FILP funding before reform, and will be the recipients after reform—are to be terminated or privatized, and their roles as instrumentalities of government policy are not shifted elsewhere or assumed by the government directly.

As a result, it is not clear that the FILP “reform” does very much to address either the lack of fiscal or other control that is at the root of the FILP system, or stem the losses that FILP has been causing.

A US Perspective

Anyone familiar with the way the US government conducts its fiscal affairs would have been shocked by a look at FILP in the form it took prior to reform. This was a government subsidy program that was carried on “off the books” in the sense that it was fed by a stream of revenue—compulsory deposits from the Postal Savings and Pension Systems—that was completely separate from regular government revenue, produced government spending and government obligations that were not recorded as such, and made investments that were nominally in the form of loans but were actually thinly disguised subsidies to various sectors of the economy. In addition, the discretion that was granted to the officials in charge of the program was so broad that their performance was essentially nonreviewable.

The language of the documents that describes the reform implicitly recognizes these deficiencies,[4] yet when the reform is carefully examined it has made no substantial changes in the FILP process. Indeed, in some significant ways it reduces the amount of fiscal control that the government may have over the FILP system, and thus opens the way for further damage to the nation’s economy and credit.

In many major respects, the FILP system—before and after reform—violates the precepts on which the United States government operates and keeps its accounts. Indeed, an American observer—knowing nothing about how FILP might have worked in practice—would anticipate fiscal trouble simply from the uncontrolled nature of the FILP program. There is not enough information currently available to be certain about the outcome of the FILP program—and this lack of information is usually an indication that the worst fears are justified—but if its results are as problematic as many observers believe, this would tend to validate the American view.

The fact that reform was deemed necessary suggests that the Japanese government has recognized at least some of these problems, but since the reform as actually implemented makes no substantial changes in how the system is operated, it seems that once again that the government is unable to effect necessary changes in the face of entrenched opposition.

This is not to say that the United States has not made its own major mistakes. The S&L debacle in the late 1980s and the empowerment of Government Sponsored Enterprises (GSEs) today show that the US is no paragon of virtue or prudence. But as this paper will show, both of those departures from good practice in the United States find echoes in FILP on a much larger scale. In this sense, FILP’s failure contains a valuable lesson for the US as well as Japan.

For purposes of this paper, these are the relevant principles on which US government financial policy is based:

• The US government budget is “unified,” so that each spending proposal is forced to compete for priority with all others. There are few if any privileged outlays which are permitted outside the budget’s perimeter. Even where there are special taxes to support certain outlays—such as a gasoline tax supporting a trust fund for highway construction—the funds collected for this purpose may not be spent in any year, or ever.
• US government agencies receive appropriations of funds for each year and are limited to that amount, although in some cases the amount may be expended over several years. Officials of these agencies have no access to significant sources of funds other than through appropriations.
• Where the US government has chosen to operate through corporations, they are funded by the government directly and not permitted to go separately to the capital markets. If necessary, in cases of government deficit, the US government borrows the money itself to provide these instrumentalities with funds for their operations. 

• No US government agency is given the authority to make discretionary outlays for general purposes such as “allocating resources” or “adjusting the economy” through countercyclical spending. To the extent these policies are pursued, they are authorized annually with narrowly focused objectives. For example, an agency will be appropriated additional funds to increase road-building, but those funds cannot not be used for other purposes.
• The US Social Security system is invested only in US government securities. The managers of the system are authorized to make other investments, but these must be made for a financial return, and not for any other purpose. The US government can use the surpluses in the Social Security Trust Fund for any government purpose, but only by borrowing these funds in the same way it would borrow from the public—through the issuance of debt to the Trust Fund.

These principles of government fiscal management would be characterized in the United States as sound public policy. This portion of the paper will discuss the ways in which the FILP system departs from what would be considered orthodox and prudent government fiscal management in the United States.

FILP Before Reform

In general, before reform, the government officials who were the managers of FILP had been granted authority to dispose of substantial portions of the government’s resources outside the government’s regular budget, (i) for government purposes, (ii) without significant controls on the amount of the funds utilized, and (iii) according to criteria so open-ended that it was not possible to evaluate their performance on a regular basis. 

It is all the more astonishing that this authority was given and exercised with respect to funds for which the government was acting, in a sense, as a trustee; these were funds that were compulsorily lent from the Postal Savings or Pension systems in Japan, and will have to be paid back. The idea that they might be used for purposes that would make repayment difficult or impossible seems, to the American eye, quite remarkable—even though in the end the government will make up the losses in some way.

That the FILP system disposes of government resources cannot be doubted. Although it was not using tax revenues, FILP was investing money that the government had collected from the private sector either through borrowing (the Postal Saving system) or through tax-like insurance premia (the Pension system). Both these accounts, in reality, are ultimately liabilities of the government, even if they are not explicitly government-guaranteed. In some way, as a practical matter, the Japanese government will have to make good on these obligations to the public. If the government had directly raised this money through the issuance of bonds to the public, there would be no question that the proceeds were government resources.

Moreover, FILP was disposing of these resources for governmental purposes and in pursuit of government policies. The FILP Report describes three uniquely government roles of which the FILP process is a part:

Allocation of resources. “This function allows the government, acting from the standpoint of the national economy, to provide goods and services that would not be provided at all, or not provided sufficiently, if left entirely to the market mechanism.”

Redistribution of income. “In order to ease excessive inequality in incomes, the government can adopt a progressive tax system on the revenue side and undertake social security measures on the expenditure side.”

Fiscal adjustment of the economy. “By assisting recovery when the economy is declining and by applying the brakes when the economy overheats and the danger of inflation arises, the government can preserve economic stability.” [5]

The Report then continues, to explain how the FILP system meets these government objectives, by asking “What if there were no FILP?”:

Without the FILP, if goals such as the provision of social infrastructure or policies for small- and medium- sized businesses were to be smoothly pursued, tax revenues would be needed to cover the costs. In that case, the tax burden on each citizen would have to be increased.

If such an increase in the tax burden were to be avoided, resources could not be sufficiently allocated. For example, road construction would fall behind and improvement in people’s quality of life would be delayed.

Moreover, if the government did not use interest-bearing funds, relying instead on grant assistance using tax revenues, then in some areas, grant recipients might not exercise the needed self-discipline in undertaking projects and policy gains might not be fully met.

Furthermore, without the FILP, the government would lose a policy tool for flexibly taking economic measures.

Thus, as we can see, without the FILP, there would be a danger that the government could not sufficiently allocate resources and adjust the economy, and the realization of our present convenient and comfortable standard of living would have been greatly delayed. The FILP is an indispensable system that makes a major contribution to the national economy.[6]

The purpose of quoting this statement at length is not to highlight its absurdly self-congratulatory tone in the midst of what many believe is a financial disaster, or its bland assumption that the reader will agree that he or she is enjoying a “convenient and comfortable standard of living” as a result of FILP’s activities, but to establish three points:

• The FILP managers believed themselves to be carrying out uniquely government functions, not an investment activity.
•  These government functions are intended to allocate resources or adjust the economy, not to yield a rate of return.
• The FILP managers have a broad discretionary mandate that make it difficult to review their performance.

In this light, it is instructive to compare FILP’s approach to carrying on a government function with the principles—outlined above—which govern the way the US government carries on similar functions.

The unified budget. The reason for the unified budget is control over expenditure. Thus, funds for highways, bridges, port facilities and the like come out of the same pool of government resources as farm subsidies, law enforcement and education. The value to the nation of each outlay is weighed against the others, and no set of outlays is free of this comparison.

FILP, on the other hand, established a special pathway for outlays made for the broad purposes described above. Expenditures in support of what FILP managers believe is a proper allocation of resources within the economy, or to achieve countercyclical effects, are insulated from competition with the government’s other priorities. It is no wonder, then, that they have grown faster than Japanese government revenue and the Japanese economy as a whole.

Outlays made with borrowed funds. The President’s proposed Budget for 2002 states: “The budget treats borrowing and debt repayment as a means of financing, not as receipts or outlays. If borrowing were defined as receipts and debt repayment as outlays, the budget would be virtually balanced by definition.”[7] This approach assures that the government cannot avoid declaring a deficit by borrowing funds rather than raising them through taxation. Eventually, funds borrowed by the government must be repaid, usually by the taxpayers, but it is generally easier in the short run for governments to borrow rather than tax. The underlying concept of the US government’s treatment of borrowing is that the task of balancing the budget by matching revenues with outlays should not be made easier by allowing the government to borrow without adding to the deficit.

Because it relied primarily on funds borrowed outside the government’s normal process, and its outlays were not recorded as government expenditures, FILP—as it was operated before reform—clearly violates this principle. In effect, although the government was spending money for uniquely government purposes—not simply as an intermediary for the investment of borrowed funds—these outlays were not counted as government expenditures nor included in the government’s budget.  The lack of control this represents accounts for the growth of FILP and the losses it will ultimately produce for the government and the taxpayers of Japan.

Discretionary authority. The US President’s budget proposal contains thousands of line items for specific outlays in each area of government activity. Funds appropriated by Congress also include thousands of line items that narrowly confine the discretionary authority of government agencies.

The managers of FILP were not so restricted. As outlined in the FILP Report, they were authorized to use the borrowed funds for such broad purposes as “allocating resources” or “adjusting the economy.” Granting such broad authority to government officials would be completely foreign to the US governmental system. Even where Congress appropriates funds for the purpose of allocating resources or making countercyclical expenditures, these are narrowly confined to the matters within the jurisdiction of specific agencies. For example, the National Highway Administration will be authorized to spend a certain amount of appropriated funds for highway construction, but not for any other purpose.

Reviewable criteria. When authority is granted to US government agencies, it is generally under conditions and criteria that will permit their performance to be subsequently reviewed. Thus, if an agency is authorized to make loans, the standards on which the loans are to be made are specified—at least in general terms. The same would be true of grants, or for contracting for services. With these criteria and conditions, it is possible for Congress to review annually whether the agency is performing satisfactorily and accordance with the policies Congress set out.

The managers of FILP appear to have had such broad discretion in their use of FILP funds that it would not be possible to evaluate their performance. Some of their investments are presented as loans on which they expect a market rate of return, while others might be for allocating resources or adjusting the economy. There is no way to evaluate their performance in this structure, since a loan that might have been expected to yield a market rate of return could also be a loan that was intended to have countercyclical effects. If it fails the former test, it can always pass the latter. And if it fails both, it can always be cited as allocating resources in some sensible way. This leaves the operations of FILP essentially nonreviewable.

For example, according to the 1999 FILP Report, two FILP companies, the National Life Finance Corporation and the Japan Finance Corporation for Small Business “help small-and medium-sized businesses, lacking in creditworthiness or collateral, to procure funds smoothly…[and] provide policy-based finance, supplementing funds from private institutions.”[8] What are the criteria for making these loans? Are they for the purpose of returning a profit—which would be appropriate for funds borrowed from the Postal Savings or Pension systems—or are they being made for “allocating resources” or “adjusting the economy?” If the loans are made to allocate resources or adjust the economy, it is not difficult to understand  why the government would lend funds to companies that are not creditworthy. But if the purpose of the loan is to receive a return for the Postal Savings system, making a loan to a company that is not creditworthy makes little sense. Despite the difficulty of determining on what basis these loans were made,  The loans outstanding to this sector totaled over Y6 trillion ($520 million) expended by FILP in FY 1999 alone.[9] 

Investment in the private economy.  The US Social Security system and the other trust funds managed by the US government invest only in US government securities. There are a number of reasons for this, but for present purposes one is particularly important: a policy of investing in the private economy would eventually be politicized, making choices based on political considerations rather than value.

Far from eschewing politicization of its investments, the FILP system encouraged decisions on this basis. Statements to the effect that the FILP managers were authorized to “allocate resources and adjust the economy” reflect discretionary authority of such breadth that it would never be possible to tell on what basis they were made. Under these circumstances, it would not be surprising if loans were made to favored political constituencies of the party in power, or were denied to those who opposed the government. This is simply human nature. If this is how loans were in fact made by FILP substantial losses would be expected.

The remarkable thing about FILP, then, is that neither the amount of the funds involved nor the conditions of their use was restricted by such ordinary government devices as a unified government budget—in which each expenditure is evaluated in relation to the others—by annual limits on the funds made available, or by reviewable criteria for success or failure. Thus, the principles of government fiscal management that prevail in the United States—and are intended to prevent losses to the government—were ignored in the growth and operation of FILP prior to reform.

FILP After Reform

A review of the FILP system as “reformed” reveals very little significant change in its fundamental deficiencies—even in those recognized as the reason for the reform. The claims that MOF makes for what was achieved by reform are, upon examination, quite modest. It consists of only three items:

1.                  FILP will be “fundamentally changed from a scheme with compulsory deposit of postal savings and pension reserves into a fund-raising scheme to raise only the necessary amount of funds for FILP agencies’ projects in the market.”

2.                  “The harmonization of the FILP with the market principles, as well as reform and efficient operation of FILP agencies will be promoted.”

3.                  FILP target areas and projects should be reviewed continuously in terms of their complementary role to private business and the financial soundness thereof in the future by making good use of policy (subsidy) cost analysis.[10]

In other words, FILP agencies will only raise what they need, market principles will cause the FILP agencies to function efficiently, and the costs will be better understood. None of these points—which are either obvious goals or fond hopes for any government program—address the real issues associated with FILP. These are the following:

• Before reform, funds used for FILP came in the form of compulsory deposits from the Postal Savings and Pension Systems. They were in this sense implicit liabilities of Japanese government, since the government would have to make up any losses, but they were not recorded in the government’s accounts. After reform, at least on an interim basis, the same entities—the Postal Savings and Pension systems—will be compelled to purchase FILP Bonds and perhaps FILP corporation bonds backed by the government. Eventually, the funds used in the FILP program are supposed to be raised by FILP corporations (which are generally either owned or controlled by the Japanese government) through the issuance of bonds on their own credit. Even if this actually comes to pass, it is naïve to believe that these instruments will not be seen as having the government’s implicit guarantee. These obligations will remain, in other words, liabilities of the government not recorded in the government budget or controlled by the budget process. Realistic estimates of the true indebtedness of the Japanese government thus should include the FILP indebtedness, which could bring the total to more than 200 percent of GDP.[11]
• Similarly, although the total size of FILP investments in any year will be determined in advance, with approval by the Diet, these expenditures will not be treated as outlays of the Japanese government or included in the government’s budget. Their status as privileged expenditures, that do not have to compete with other programs for funding, continues.
• Before reform, the MOF repeatedly argued that FILP loans were intended to earn a return, and that in fact there were no substantial losses in the FILP portfolio. If this were true, “reform” would not have been necessary, and there would not have been references in the reform descriptions to “bloated and inefficient public corporations” enjoying the benefits of FILP funding.[12] Given that FILP loans were avowedly made for purposes of allocating resources or adjusting the economy, and involved projects or borrowers that could not obtain private sector financing, the notion that they would pay for themselves is unsustainable. After reform, the same standards for the use of FILP funds will apply, even though the losses that resulted from the original system were likely to have been part of the reason for reform.

• Before reform, the managers of the FILP program had wide discretion to authorize financing for projects that would “adjust the economy” or “allocate resources.” This discretion still exists, except that now it appears to be shared with the managers of the various government corporations that would actually raise and spend the funds. The basis on which the allocations of funds are made—whether for purposes of receiving a return, allocating resources or adjusting the economy—are still too broad and discretionary to be reviewable.

Taken together, these factors demonstrate that the FILP reform will achieve few of the purposes for which it was originally proposed, and may—as discussed later in this paper—be worse than the system the reforms were intended to replace.

The Harm to the Japanese Economy

Despite the self-congratulatory language of the FILP Report, it can not be seriously argued that FILP significantly helped the Japanese economy to grow out of the slump that engulfed it after the bursting of the real estate bubble in the early 1990s. The data on the economy are quite consistent and quite stark. Real economic growth has been virtually non-existent throughout the 1990s, the banking system may be insolvent, and government debt is at world-historic levels. Although one could argue that things could have been worse without FILP, there are few examples among developed economies of economic downturns of such long duration—especially when the government has spent such enormous amounts to stimulate an economic resurgence.

However, there is an argument that the FILP system—far from simply failing to encourage economic growth—actually contributed to economic stagnation in Japan. In support of this position, the following points could be made.

Government spending.  No matter how they are described, FILP loans to the private sector, or direct outlays by FILP corporations, must be considered government spending. The fact that FILP funds were lent by the MOF’s Trust Fund Bureau prior to reform, and may be borrowed directly in the market by FILP corporations after reform, is not in any way inconsistent with the point that this was and will be simply government spending. If the government of Japan, through ordinary government processes, had borrowed the same amount of funds in the public markets, and then used it in the conventional way to build roads, airports, bridges and port facilities on which it collected fees and tolls, there would be no question that this was government spending. Using government corporations to do the same thing with loans ultimately backed, implicitly or explicitly, by the government is essentially the same thing.

This conclusion might be more difficult to draw if the clear purpose of the loans was to earn a return, but the FILP Report makes clear that the declared purposes of FILP lending are purposes unique to government—allocation of society’s resources and adjustment of the nation’s economy. Earning a return on these loans is clearly a secondary consideration. Although some of the loans were to companies that arguably might be profitable—housing, and small business are examples—substantial outlays were and will be made for such items as education, the environment and infrastructure.[13] That it is at this stage impossible to separate the profitable from the unprofitable investments, or the uniquely governmental from the truly financial, is one of the chief deficiencies of the system. Until there is evidence to the contrary, we must consider that the primary purpose of FILP lending was to do the things governments do for their populations and not to earn a return.

Government spending commands the allocation of society’s resources. If the government decrees that a bridge or a highway is to be built at a particular place, and allocates the resources to accomplish that purpose, the bridge or highway comes into existence, no matter its economic justification. But it is important to note that since society’s resources are finite, the use of its resources in this place necessarily results in the loss of those resources for some other purpose. In other words, in a very real sense every bridge and road built with FILP funds made the cost of building something else—say, a new Toyota factory—that much more expensive.

It is of course true that if the new road opens up an area of the country that would be a particularly favorable place in which to build a Toyota factory, then the road might be thought of as having an offsetting value with respect to the cost of the factory—and in some cases the offsetting value might be greater than the additional cost it imposed on the factory. This is a very complicated computation, and would have to take into account such intangibles as the additional pleasure the road would provide to those who like to take Sunday drives. But in the largest sense—without measuring each individual government investment, or attempting to compare it to the value of some other investment, public or private—if all such investments, taken as a whole, impose significantly more costs on society than they offset, economic growth will suffer.

For this reason, Looking at the data for the growth of the Japanese economy during the 1990s—a period when there has been little economic growth at all in Japan, despite government investment in the economy through FILP—a credible case can be made that FILP may be at least partly responsible for this stagnation because it funneled otherwise valuable capital resources into nonproductive uses.

The enormous size of the FILP program lends some support to this idea. According to the FILP Report, FILP investments in the Japanese economy as of the end of FY 2000 totaled Y414 trillion ($3.35 trillion)—about 80 percent of Japan’s nominal GDP in that year.[14] For comparison purposes, the net stock of US government nondefense investment in the United States at the end of FY 2000 was less than $1.3 trillion,[15] approximately 13 percent of the $10 trillion US GDP.  This suggests that FILP investments constitute a very large part of the Japanese economy and were highly influential in determining the efficiency with which that society’s capital resources were deployed.

In this connection, it is worth noting that annual FILP investments, which were Y27 trillion in 1989, increased to Y45 trillion in 1993 while the economy slid from a growth rate of 7.1 percent to a rate of 1 percent. To be sure, this growth in FILP spending reflects the government’s effort to arrest the economy’s decline, but continued FILP investment in the annual range of Y40-45 trillion over the next six years had no significant effect in stimulating economic growth, which has averaged less than 1 percent annually during this period.[16] The negative correlation between growth in FILP investment and growth in the Japanese economy is a further indication that FILP investment had a negative rather than a positive effect on the economy. If so, the likely reason is the relative inefficiency of FILP in allocating productive resources.

Losses. The discussion above attempts to evaluate the FILP program by looking at the gross results in the Japanese economy. On that test, FILP certainly failed to stimulate growth, and may actually have been a cause of economic stagnation. Another way to assess the FILP system, however, would be to look at whether it was a success as an investment program—i.e., whether it was responsible for the establishment and development of public or private enterprises that enabled FILP to recoup its investment with interest.

A successful investment is one that repays the principal invested, with an additional return such as interest. This is true for a nation as a whole as it is for an individual investor. If the purpose of the investment was to produce or facilitate the production of a good or service, then the willingness of users to pay for that good or service is a measure of whether it is meeting a public need. If a project cannot pay for itself through servicing the debt used to create it, then it has not effectively responded to a public need. From the standpoint of the government that financed it, therefore, the project is an economic failure (although it could have other non-economic benefits).

Normally, government expenditures or investments are not measured by their ability to produce profits, or to enable the government to recoup its investment. Government purchases of parkland, for example, are intended to produce positive externalities that cannot be expressed in monetary terms. However, FILP investments, because they were based on borrowings from the government’s Postal Savings and Pension systems, were at least nominally supposed to pay for themselves in financial terms. If they did, then FILP might well have contributed to Japan’s economic recovery, even if the gross figures for economic growth do not show a positive effect.

However, if these investments were not successful—if they produced losses—then the opposite would be true: the government has wasted valuable capital on nonproductive investments. In this connection, it is important to note that these investments had something of a head start in comparison with private sector investments, since they are made at rates of interest that reflect the Japanese government’s low borrowing costs. If they are not returning even these rates, they were poor investments indeed.

It is in fact highly unlikely that FILP investments were successful in the sense that they have resulted or will result—in the aggregate—in a return of the principal invested, with interest. Although the government has never released detailed figures, market values, or audits of FILP results, MOF has continued to insist that no significant losses have occurred. FILP financial reports have never shown annual deficits, and the government regularly reports that nonperforming loans in the FILP system are less than one percent.[17] If this were correct, reform would not have been considered necessary. The small size of the losses reported is very difficult to credit, however, unless the definition of a nonperforming loan is known. As used by bank regulators in the US, a nonperforming loan is one in which the obligor is so weak financially that the prospect of full repayment is impaired, even if payments of principal and interest on the loan are current. A loan is also counted as nonperforming if it has been restructured in order to meet the financial needs of the obligor.

There are strong indications, however, that—at  least in the past—the definition of nonperforming loan used by MOF in reporting on FILP investments was considerably more lenient—i.e., any loan that is “six months or more past due.”[18] The difficulty with this standard is that it is possible for the lender to turn a nonperforming loan into a performing loan by restructuring it so that the borrower can make at least some payments of principal or interest, or by lending the borrower additional funds sufficient to make regular payments.

Although MOF has not released data that will show the true dimensions of the losses suffered by FILP, it is likely that the ratio of nonperforming loans in the FILP portfolio is quite high. This is true because of the nature of FILP. Despite the fact that they were using funds borrowed from the Postal Savings and Pension systems, the FILP managers were apparently not required to invest those funds in such a way as to earn a market or any other rate of return. As noted above, FILP was seen as a fund for carrying out government functions—allocating resources and adjusting the economy—and not simply as a way of finding suitable investments for the savings and pension funds of Japanese citizens.[19] Under these circumstances, it would be logical to expect that the losses in FILP would be at least as great as the losses suffered on loans by Japanese commercial banks, which were presumably lending with only one standard in mind—the ability of the borrower to repay.

Accordingly, to determine whether the FILP loans are recovering principal and interest, one would have to separate out the loans that were made on a commercial basis from those that were made for governmental purposes such as allocating resources. As noted above, it is not possible from the available data to make such distinctions; the FILP managers were given such wide discretion that their performance in this or any other area can not be independently evaluated.

Nevertheless, one way of getting at the likely performance of the FILP loan portfolio is to consider conditions at other institutions that were lending for a profitable returns during the same period. Here the results suggest that FILP has probably suffered large losses. The losses at Japanese commercial banks that have gradually come to light are extremely large—larger as a percentage of assets than the losses of the US S&L industry when that industry collapsed in the late 1980s. For example, at the end of 1997, MOF announced that Japanese banks then held over Y76 trillion in loans that might not be recoverable. That amounted to 12 percent of all loans and loan guarantees of these institutions.[20]

There is reason to doubt whether even this enormous number fully states the losses of the Japanese private banking system. For one thing, Japanese banks and their customers do not use accounting systems that are sophisticated enough to yield an accurate estimate of which loans are in trouble. Audited financial statements, or financial statements that follow a set of consistent principles, are not always required by Japanese lenders. For another, the definitions of nonperforming loans are lax, as discussed above. Finally, when Japanese banks, life insurance companies and corporations have failed, the actual losses have been many times the accounting losses reported before failure.[21]

If the losses of Japanese banks are as high as 12 percent, and perhaps higher, then the losses on FILP loans—which, for example, are described in the FILP report as made to small businesses that are not otherwise considered “creditworthy”—are likely to be substantially higher. For one thing, there is no reason to believe that the managers of FILP corporations were better at selecting creditworthy borrowers than the managers at commercial banks—especially when creditworthiness as a criterion for lending was not a central objective of the FILP program.

If in fact FILP losses are even larger than those of Japanese commercial banks, the capital involved would have been more productively employed in the economy by funneling it through the private commercial banks than by deploying it through FILP. Again, this suggests that FILP had a net harmful effect on Japan’s effort to recover from economic stagnation.

Diversion of financing resources. Yet another way to look at FILP’s effect on the Japanese economy is to consider how the resources of the Postal Savings and Pension systems would have been deployed without FILP. If these entities had invested in Japanese Government Bonds from the outset, the government’s claim on the capital markets would have been reduced and more funds would have been available for productive investment. In this case, the government would have been required to make loans to the FILP corporations, incurring an enlarged deficit but presenting a more accurate fiscal picture.

On the other hand, if the Postal Saving and Pension Systems had invested directly in the markets—instead of depositing their funds with MOF’s Trust Fund Bureau for use in FILP—these funds might have become a source of growth capital for the private economy. Under these circumstances, again the government would have been required to borrow the funds—and increase its reported deficit—if it intended to continue to operate through the FILP corporations.

As it turned out, the use of the resources of the Postal Savings and Pension systems by FILP may have been the least productive use for these funds in terms of the needs of the Japanese economy.

Price Keeping Operations. Other commentators have noted the adverse effects of the price keeping operations pursued by FILP through its investments in the equity markets.[22] The purpose of these investments was to prop up the banking system by assuring that the share values included in their capital computations remained high. However, the artificially high equity price levels that these operations produced made it difficult for other companies to raise equity capital at rates that would enable them to operate profitably, and again retarded the growth of the Japanese economy.

Effectiveness of government efforts to stimulate consumer spending. Another area where FILP may have had an adverse effect on a Japanese economic recovery is its effect on consumer spending. One of the puzzles of the Japanese slump has been its apparent resistance to most measures that other governments have used with success to stimulate their economies. Unlike many other countries that have been afflicted with serious economic reverses, Japan in the early 1990s had many resources with which to stimulate economic growth, including an enormous stock of hard currency reserves, a strong credit rating and a highly productive private sector economy. Nevertheless, repeated borrowing and spending by the government—in part through FILP and in part through other government mechanisms—has failed to re-ignite significant economic growth.

Many observers of this phenomenon have related it to a very high propensity of the Japanese public to save rather than consume. Thus, government spending on infrastructure, housing, and other construction—although it clearly pumped money into the private economy—did not result in substantial growth in consumer spending. The same was true of tax reductions that were supposed to put more money into consumers’ pockets. In both cases, the Japanese public seemed to prefer to add to its savings rather than to spend the additional resources.

Although this reaction could have many causes, including a deflationary economy in which goods may cost less at a later time than they do now—making it rational to put off spending even when additional money is available—another possible explanation is that the public is saving because it is uncertain about its prospects for retirement. The Japanese pension system is known to be underfunded, and the public may well sense that the investments being made with pension contributions through FILP will eventually prove to be losses. Thus, they may believe that unless they save significant additional amounts they may be forced to live on very limited government pension payments—if any—when they ultimately retire. If so, an unwillingness on the part of the public to spend and consume—and its consequence in prolonging the period of economic stagnation—can be counted as another result of the FILP system.

The Ineffectiveness of FILP Reform

The government’s reform of the FILP system went into effect on April 1, 2001. Its most salient feature is the requirement that the government-owned and controlled corporations which had previously been the recipients of FILP support eventually obtain their financing by issuing their own debt in the capital markets.[23] The assumption seems to be that these corporations, if they are not efficient and profitable, will not be funded and will eventually disappear.[24]

This assumption seems naïve. Many if not all of these government corporations are engaged in activities that are in support of government policies. Some of the larger ones are the Government Housing Loan Corporation, Japan Highway Public Corporation, Japan Finance Corporation for Small Business, National Life Finance Corporation, Japan Development Bank and the Japan Bank for International Cooperation. Many of the top officials of these companies came from MOF or other ministries connected with their purposes, and they have developed powerful constituencies among those they currently finance. In other words, it is very unlikely that the government will allow them to fail or otherwise disappear. Indeed, the MOF’s description of the reform contemplates that  there will be a transition period to full privatization of the FILP corporations, during which the government will guarantee their obligations.[25] This will tend to cement in the minds of investors the notion that these corporations are performing a government function and will not be allowed to fail.

As it happens, there is an analogy in the United States. The Federal National Mortgage Association, better known as Fannie Mae, was begun as a government agency in 1934 for the purpose of encouraging home ownership through purchasing mortgages from banks and other residential mortgage lenders. By creating a market for the mortgages that banks were then holding in portfolio, Fannie Mae enabled banks to liquidate their existing loans and finance new mortgages. In 1970, Fannie Mae was “privatized,” in the sense that it was permitted to sell equity securities to the public, and thereafter—like the FILP corporations—was to raise its financing in the private capital markets. The privatization legislation stated explicitly that Fannie Mae securities were not guaranteed by the United States government.

Despite this statement, however, investors in the United States and elsewhere do not believe that the United States government will allow Fannie Mae to fail. There are many reasons for this view, but one of them is certainly that the company is carrying out a government policy—helping to keep the residential mortgage markets liquid. Accordingly, even though Fannie Mae operates with capital levels that are a fraction of those required of banks, it is treated in the capital markets as a better than AAA credit, and pays an interest rate that is just slightly higher than the US Treasury itself. In fact, investors are not wrong. When other similarly situated government corporations in the United States—known as Government Sponsored Enterprises (GSEs)—have encountered financial difficulties Congress has stepped in with a financial rescue plan.

Given these facts, it is very difficult to believe that the FILP corporations, which are controlled by the Japanese government and are pursuing government policies, will not be seen in the same light by Japanese and other investors.

If that happens, even if eventually the FILP corporations are able to issue bonds themselves, there is essentially no difference between the FILP system as it existed at the end of FY 2000 and the new system that is to replace it. Under the reform, after an interim period in which they will continue to invest directly in FILP, the Postal Savings and Pension systems are supposed to invest their funds by purchasing Japanese Government Bonds and investing directly in the capital markets. If this happens, the purchases of JGBs will reduce the market supply and send investors in search of equivalent credits. The former FILP-financed corporations, owned or controlled by the Japanese government, are likely to be seen as enjoying the implicit—if not the explicit—backing of the government.

If this follows the US pattern, these corporations will pay somewhat more in interest than they paid to FILP—since there will be some small risk on their bonds that the market will price as greater than direct government risk—but they may have more freedom to grow. Unless Japanese accounting systems improve, it will be difficult for the government as well as investors to determine whether these corporations are profitable or not. And since the government will not be financing them directly, it will have less authority to control their investments or limit their growth. In a sense, this system could turn out to be worse than FILP itself, without relieving the government responsibility for their debts.

What’s to be Done?

If this diagnosis is correct, the FILP “reform” does not go remotely far enough. It still leaves the activities of a very large number of corporations—at last count somewhat over 80—with the authority to pursue government policies without any clear mechanism for control. The US experience with Fannie Mae other similar GSEs indicates that these entities—far from disappearing because of inefficiency or lack of profitability—will acquire a life of their own. The situation after reform will be more complicated but not substantially different from what prevailed before reform, and it could even be worse if these corporations—like the GSEs in the United States—gain the political power to evade control by the Japanese government.

Thus, the correct prescription for reform of FILP would be to adopt a fiscal management script that is generally closer to that followed in the US:

• All government outlays intended to allocate resources or adjust the economy should be made through a regular government budget process in which each allocation must compete with others for funding.
• If the government chooses to operate through controlled corporations, these should be funded out of a unified budget, supported by tax revenues, or government borrowing if there is a government deficit. Government corporations, carrying out government functions, should not be given direct access to the capital markets.
• Government instrumentalities that are authorized to invest in the private sector should be given a single criterion on which their success will be evaluated. That criterion—depending on the government’s policies for how it will carry out its program—can be such things as earning a market rate of return, allocating resources, or adopting countercyclical measures, but only one of these. That way, the performance of the managers can be evaluated in light of their results.
• All government agencies should have limited budgetary allocations with which to carry on their programs, renewable each year. No government agency should have access to unlimited amounts of funds outside the government budget.
• Government operated savings and pension programs can invest in the private markets, but only for market returns. If the government wants to use the resources of these entities for other purposes—such as building roads or airports—it should borrow the funds through the issuance to them of government bonds and then allocate the borrowed funds within the context of the unified budget. Only in this way can the true cost of government policies be determined.

Thoroughgoing reforms that sweep away existing structures have been rare in Japan, but the situation is urgent enough today that more radical steps are necessary than the government has thus far been willing to contemplate.

One of these steps is a comprehensive review of the FILP corporation, privatizing those which can be profitable, and terminating those that cannot. Where necessary, the government should assume the assets and liabilities of the terminated corporations, since their obligations will turn out to be obligations of the government eventually in any event.

FILP corporations that can survive without government support are meeting a legitimate public need—that’s why they are able to survive—while those that cannot must be justified on some other basis. If they are performing functions that—in the government’s judgment—are necessary for the welfare or convenience of the public, then they should be financed out of the government’s general account, and weighed against other priorities. Until these principles are adopted, it seems likely that the Japanese government and the Japanese economy will continue to pile up debt without economic recovery.