The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
FRBSF WEEKLY LEIlER May 12,1989 The FSLIC Bailout and the Economy The Bush Administration's proposal for raising the funds to cover deposit liabilities at hundreds of insolvent thrifts is gathering momentum in the Congress. One provision would raise $50 billion "off budget" through bonds issued by a federally-sponsored agency. (That debt would add to the nearly $40 billion in liabilities already amassed by the Federal Savings and Loan Insurance Corporation-FSLlC-in handling insolvent institutions prior to this year.) The principal on the bonds issued by the federal agency would be covered by zero coupon bonds that would be purchased with resources from the thrift industry. However, the Treasury would be responsible for most of the interest. (See Letter of March 31, 1989.) This raises a question: if the cost of servicing this debt really is the Treasury's obligation, shouldn't the liability be "booked:' that is, put on the federal budget and added to the federal budget deficit? On this question, Martin Feldstein recently argued that this debt legitimately belongs off budget since debt financing of FSLlC expenses will not affect aggregate demand, raise interest rates, nor crowd out private investment. This Letter shows that this argument is true only in a narrow context. The actual issuance of agency bonds merely would "book" government debt that, in effect, already has been incurred. However, in a broader context, because the method of financing government spending seems to affect saving and spending decisions, government debt incurred in connection with deposit insurance liabilities does have macroeconomic consequences. Accordingly, steps should be taken to ensure that in the future changes in the expected expenses of the deposit insurance system are not ignored in the federal budget process. Redistribution of losses Thrifts become insolvent when the value of their assets falls below that of their liabilities. To society as a whole, such a decline iii value represents a loss of wealth, which in itself should have a negative effect on aggregate demand. Deposit insurance cannot diminish the size of this loss. But the existence of deposit insurance and the method used to finance it may affect the incidence of the loss, by protecting depositors of failed institutions and placing the burden on others in the economy. Payouts by the deposit insurance fund, then, can be viewed as transfer payments to depositors. When the reserves of the deposit insurance fund are inadequate to cover these payouts, as in the case of the FSLlC currently, funds have to be raised from other sources. To the extent that the solvent portion of the industry cannot raise the needed funds, taxpayers may be called on to bear a share of the losses in order to honor the deposit guarantees. Raising taxes Such a. transfer would mean that depositors as a group would be better off than if theywere not insured, but currenttaxpayers generally would be worse off. With no net change in wealth, this transfer should not have any additional effect on aggregate spending and interest rates, assuming depositors' tastes regarding spending and saving broadly reflected those of society as a whole. This argument suggests that although the actual loss in wealth has an effect on the economy, shifting the burden of that loss from one group (depositors) to another (taxpayers) should not have an impact. A simple numerical example can help illustrate the economic consequences of financing FSLlC expenses through taxation. Consider first the case where no deposit insurance is provided. Starting with a healthy thrift industry that holds $100 million in loans funded by $90 million in deposits and $10 million in capital, household wealth is $100 million (since deposits and thrift stock are both assets). If an economic catastrophe were to reduce the value of the thrift industry's loans by $30 million, the industry would be insolvent and household wealth would decline by $30 million. Without deposit insurance, FABSF depositors would absorb $20 million of losses and equity holders would absorb $10 million. When deposit insurance costs are financed through current taxes, the outcome in terms of private credit and household wealth is identical to that for the case.qf no deposit insurance. To make depositors whole, the government would. payout $2Qmillion, increasing taxes by an equal alJlountto cover the liability. thus,.household wealthwould decline by $30 million: $10 million from the loss in the value ofthe thrift industry's stock and $20 million associated with the increase in taxes. The thrift industry would hold $70 million in loans and total deposits would equal $70 million. Thus, the volume ofprivate credit and household wealth would be unaffected by deposit insurance when it .is financed through current taxes. As a consequence, an FSLlC bailout financed through taxes should not affect the economy. Debt financing How does this scenario change when the FSLlC shortfall is covered by issuing government debt? Some have argued that budget deficit~ hav~ the same economic consequences as. raising taxes. When faced with a futuretax liability associated with an increase in governmentborrowing today, so this. argument goes, rational households will increase their current savi ng to· generate sufficient resources to cover the h igherfuturetax liability. As a result, spending will be curtailed by the same amount as if current taxes had been raised. Thebond~ held by households would not increasewealth since there would be an offsetting increase in future tax liabilities. Thus,using debt toJinance the FSLlC bailout would not be stimulative. An extension of the earl.ier example will help to illustrate this point. Instead of raising taxes by $20millionto cover.the loss to depositors, assumethat the governmentissues $20 million in bonds and gives these bonds directly to the thrifts. In this case,deposit liabilities would total $90 million and assets would total $90 million~ $70 million in. loans and $20 million in government bonds. As in the case of tax-financed deposit insurance, household net wealth would -'Cdecline by the $10 million capital loss and by the $20 million future tax fiabilitythat the bonds represent. Changing the example so that the government sells the bonds to the public does not alter these results. For example, households could purchase the $20 million in government bonds and the government then would inject cash into the thrift industry to make up the loss to depositors. Assuming households draw down their holdings of deposits to purchase the bonds, the net result would be that the thrift industry would have $70 million in loans and the public would have $70 million in deposits. The $20 million in bonds held as assets by households would be balanced by a future tax liability of $20 million. Thus, household wealth still would decline by $30 million. However, other economists have argued that households may not perceive their net worth as declining by the full $30 million, particularly if some of this future tax burden falls on future generations and current households place a higher value on their own spending than they do on the spending of future generations. In this case, households probably will not increase their saving to compensate fully for higher future taxes. As a result, government budget deficits would transfer wealth from future generations to the current generation. In the two debt-financing examples above, if current households ignore entirely the liability created by the government debt, the effective decline in wealth for current households would be only $10 million, the value of the thrift stock, rather than $30 million. The $20 million in government bonds would be perceived as adding to current household wealth since the expected rise in taxes in the future would not be perceived as a liability. In the near term, this smaller decline in household wealth would mean higher aggregate demand and interest rates than if current taxes had been increased. The examples show that depositors and thrift stockholders are not any better or any worse off when insurance costs are financed through government debt rather than through taxes. The difference between the two methods of financing deposit insurance lies in the way current taxpayers view the future tax liability connected with the $20 million in government debt. On the one hand, if current taxpayers treat the future taxes as a current liability, there will be no difference between tax-financed and debtfinanced deposit insurance. On the other, if current taxpayers do not view themselves as liable for the future tax burden, debt financing can affect current wealth, inducing more current spending and less future spending. In the near term, this would be stimulative and would boost interest rates relative to the levels that would have prevailed if deposit insurance were financed through current taxes. The experience in the 1980s suggests that households do not adjust saving fully to compensate for an increase in future tax liabilities. Since 1982, the federal budget deficit has increased very sharply, but the U.s. saving rate has continued to decline. This evidence suggests that an increase in government borrowing does affect aggregate demand over and above the effects on demand from an increase in government spending. Booking the debt Thus, government debt incurred in covering the deposits of insolvent thrifts does have economic consequences. This does not mean, however, that the act of issuing the $50 billion in bonds to assist the FSLlC will have an impact. Since most of the problem thrifts have been insolvent for some time, the expenses and government debt already have been incurred as far as their economic consequences are concerned. Executing the plan to assist the FSLlC merely would book the debt, and should have no additional impact unless the amount of debt issued is materially different from that expected. Treasury or agency? Part of the debate in the Congress concerns using a federal agency rather than the Treasury to issue the $50 billion in bonds. By using a federal agency to issue the debt, the proceeds from the bonds would be treated as revenue that offsets FSLlC's expenses. Although federal agency debt is somewhat more costly than direct Treasury debt, the practical appeal of this approach is that it would be easier to meet the Gramm-RudmanHollings targets for the federal budget deficit. The use of agency, or off-budget, debt also has been rationalized on the grounds that the "new" debt has no economic consequences. Although this argument technically is true in the sense that booking debt that already has been incurred should have no further consequences, it fails to acknowledge the real economic impact of this previously unbooked liability. Thus, if there is an economic rationale for issuing the bonds on an off-budget basis, it is merely that two wrongs do not make a right. That is, it was wrong to exclude the unbooked expenses of the FSLlC from past budgets. That wrong cannot be corrected by increasing budget deficits in the immediate future to reflect past losses. Fundamental problem The more fundamental problem with the argument that using debt to finance FSLlC expenses does not have economic consequences is that it provides erroneous guidance concerning how future "unbooked" deposit insurance liabilities should be handled. This argument implies that since debt incurred by the government (or the FSLlC) does not have economic consequences, it is okay in the budget process to ignore unbooked expenses incurred by the FSLlC in the future. The analysis in this Letter suggests the opposite; debt financing (whether implicit or explicit) does affect the economy and should be taken into cons ideration. In principle, then, any increase in the expected cost to the deposit insurance system of resolving problems in the future should be included in the federal budget as a current expense, in accordance with generally accepted accounting practices. Alternatively, regular deposit insurance premia should not be counted as future federal revenues since, in theory, such premia merely reflect the change in the value of the future unbooked liabilities of the deposit insurance system. However, in its estimates of the budget effects of the Administration's proposal to assist the FSLlC, the Office of Management and Budget includes future deposit insurance premia from both banks and thrifts as revenue that offsets FSLlC expenses in resolving current problem thrift cases. Such treatment of these insurance premia means that the true macroeconomic impact of changes in deposit insurance liabilities will not be reflected in future federal budgets. Fred Furlong Research Officer Opinions expressed in this newsletter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco, or of the Board of Governors of the Federal Reserve System. Editorial comments may be adllressed to the editor (Barbara Bennett) or to the author.... Free copies of Federal Reserve publications can be obtained from the Public Information Department, Federal Reserve Bank of San Francisco, P.O. Box 7702, San Francisco 94120. Phone (415) 974-2246. ¥ Research Department Federal Reserve Bank of San Francisco Alaska nevada Arizona Oregon California Hawaii Utah Idaho Washington Address Correction Requested