The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Federal Reserve Bank of NewYork Quarterly Review W in te r 1980-81 V o lu m e 5 No. 4 1 In M e m o ria m : J o h n H e n ry W illia m s 1887-1980 3 In fla tio n and S to c k V a lue s: Is O u r T ax S tru c tu re th e V illa in ? 14 24 27 C u ttin g th e F ed era l B u d g e t: A n a ly z in g H ow Fast E x p e n d itu re , G ro w th Can Be R ed uce d C u rre n t d e v e lo p m e n ts T h e b u s in e s s s itu a tio n T he fin a n c ia l m a rk e ts 30 G lo b a l P a ym en ts P ro b le m s : The O u tlo o k fo r 1981 36 O il P ric e D e c o n tro l and B e yon d 43 S o c ia l S e c u rity and S a ving s B e h a v io r 50 T re a s u ry and F ederal R eserve F o re ig n E x ch a n g e O p e ra tio n s The Quarterly Review is published by the Research and Statistics Function of the Federal Reserve Bank of New York. An “ In M em oriam ” honoring JOHN HENRY WILLIAMS (1887-1980) begins on page 1. Among the mem bers of the function who contributed to this issue are MARCELLE ARAK (on inflation and stock values, page 3); JAMES R. CAPRA (on analyzing how fast Federal budgetary expenditures can be reduced, page 14); WILLIAM J. GASSER (on the outlook for global payments problem s in 1981, page 30); PAUL BENNETT, HAROLD COLE, and STEVEN DYM (on o il price decontrol and beyond, page 36); and PAUL WACHTEL (on social security and savings behavior, page 43). An interim report of Treasury and Federal Reserve foreign exchange operations for the period August through October 1980 starts on page 50. In Memoriam John Henry Williams 1887-1980 John H. W illiam s was the rare com bination of the scholar, outstanding in academic pursuits, and the active practitioner of the art of central banking. Fam iliar with the evolution of econom ics, the bulk of his career was concerned with the application of that discipline to public policy. Born in Ystrad-gynlais, Wales, his parents emi grated to the United States when he was an infant. The fam ily settled in North Adams, Mas sachusetts, where John grew up. After earning his bachelor’s degree in 1912 from Brown Uni versity, he taught English there until 1915. In that year he married Jessie Isabelle Monroe, by whom he was to have two daughters and, already in his late twenties, began the study of economics at Harvard. There he won a Ph.D. and the Wells Prize fo r his classic study on Argentine trade. After teaching at Princeton and Northwestern, he returned in 1921 to the faculty of Harvard, where he remained until his retirem ent in 1957. A distinguished academ ic career soon brought him national recognition. In 1932, at the nadir of the Great Depression, he was appointed a mem ber of the United States delegation to the Prep aratory Committee for the World Monetary and Economic Conference. In the spring of the fo l lowing year, when the conference was about to open, he joined this Bank as Assistant Federal Reserve Agent and immediately became involved in the efforts to stabilize the exchanges. Deeply concerned with this objective, he remained at the Bank full time until October 1934. Thereafter, he divided his time fo r more than twenty years be tween this institution and Harvard where he be came, in 1937, the first Dean of the Graduate School of Public Adm inistration. At this Bank, he was appointed in 1936 Vice President in charge of the Research Function to which he attracted many able economists. From 1947 until he reached retirem ent age in 1952, he served as Economic Adviser, continuing thereafter as con sultant to the Bank fo r another decade. Among posts and honors too numerous to list was his election as President of the Am erican Economic Association in 1951. Two years after the death in 1960 of his first wife, he married Katherine R. M cKinstry who survives him. Although it was his m ajor professional con cern, advising on policy never came easily to John Williams. He wrote that he always tried “ to look all round a problem rather than to plunge forthw ith for the bold solution” . His circum spection reflected a deep understanding of the com plexity of the problems confronting the authorities. In dealing with these problems, theory was certainly essential. He liked to quote Keynes’ view that w ithout theory we are “ lost in the w oods” . But, by its very nature, theory was a sim plification of reality. Moreover, the most influential theories were products of unique circum stances and, in deed, had their origins in views about policy growing from those circum stances. In effect, the ory was often a rationalization for policy. Since circum stances were constantly changing, he warned that those who drew prescriptions glibly from theory were dangerous as policymakers. Understanding both the value and the lim ita tions of theory, John W illiam s was constantly testing hypotheses against the realities of the market. In doing so, he found much to justify his skepticism. He particularly questioned conven tional views about the gold standard. The classi cal specie flow mechanism was a beautiful in tellectual construct which, however, failed to FRBNY Quarterly R eview /W inter 1980-81 m irror the realities. The international monetary system, which it purported to describe, was in fact one in which Britain maintained a gold stan dard, w hile most other countries based their currencies on sterling. He found related faults in classical trade theory which neglected both the dynamic relationship between the center and pe ripheral countries and also the adjustment diffi culties that the spread of m anufacturing in the periphery caused for traditional industries in the centers themselves. His view about the key role played by the in dustrial and financial centers shaped Dr. W il liam s’ advice about the handling of international monetary problems. In the thirties and forties, much of the w o rld ’s econom ic activity was cen tered in the United States and Britain. Their cur rencies were the media in which trade and finance were conducted. The problem of exchange insta bility, which bedeviled the discussions of those years, boiled down to negotiating a m utually acceptable relationship between the dollar and sterling and then m aintaining that relation— stable but not immutable— through appropriate domestic policies in the two center countries. Such views clearly influenced the United States Government in the negotiation of the T ripartite Agreement of September 1936. They also were the basis for John W illiam s’ reservations about the Bretton Woods agreements. These reservations focused prim arily on the International Monetary Fund. Dr. W illiam s c riti cized numerous aspects of its articles but his m ajor concern was that the Fund, which was designed to help correct relatively modest and tem porary international imbalances, would be incapable of perform ing this function in the very d ifficult circum stances expected at the end of hostilities. B ritain’s external difficulties would be p articularly severe. Unless “ heroic measures” — a continuation of Lend-Lease or a large low-cost loan— were granted by the United States, Britain would not be in a position to cooperate in the FRBNY Q uarterly R eview /W inter 1980-81 reestablishm ent of a m ultilateral trade and pay ments system. Yet, such “ heroic measures” were beyond the capacities of the Fund; in th eir ab sence, the trade and exchange restrictions that had been erected during depression and w ar would almost certainly be extended long into the postwar period. Thus, establishm ent of the Fund would create only a facade of cooperation w ith out the substance. As events developed, mea sures even more heroic than Dr. W illiam s had advocated were adopted in the troubled years follow ing the war— the Anglo-Am erican loan, aid to Greece and Turkey, and the Marshall Plan. These, combined with the cooperative efforts of Western Europe, eventually built an international environment in w h ic h the Fund could effectively function. In fu lfilling its role throughout this disturbed period, the Federal Reserve benefited greatly from the broad experience and wisdom of John W illiam s. In 1956, as he approached his seven tieth birthday and accepted the need to lighten his professional reponsibilities, this Bank’s board of directors expressed its appreciation, stating that His w ide-ranging knowledge and ex perience in econom ic affairs, his sound judgment, and his whole-hearted dedi cation to the public interest have marked Dr. W illiam s’ contributions to the work of the Federal Reserve System during years of depression, war, and inflation. In addi tion to the wise counsel he has brought to deliberations, he has been a constant source of encouragement and inspiration to others on the Bank’s staff, always generous of his time and wisdom, thus carrying some of his prim ary vocation into his work at the Bank to its enduring benefit. If all this were not enough, he will remain long in the memory of his asso ciates at the Bank who treasure him as a true and steadfast friend. Inflation and Stock Values Is Our Tax Structure the Villain? At one time, investors regarded common stocks as a good inflation hedge. Because stocks represented the ownership of real capital, people thought that their value would rise roughly in proportion to the general price level, at least over periods of several years. For the last decade or so, however, stock prices have not kept pace with inflation. The Standard and Poor’s index of stock prices, for example, stood at 133 in the fourth quarter of 1980, up only 26 percent from its 1968 fourth-quarter level. Yet, the price level more than doubled in that same period. This meant that the real value of equity fell almost 50 percent. Why did this tremendous drop in real value of equity occur? Some observers have suggested that inflation itself may account for this phenomenon. One theory is that the tax structure in the United States, particu larly that applicable to corporations, becomes more burdensome when the price level rises. As a conse quence, a change in inflation can reduce a corpora tion’s real aftertax earnings. This could, in turn, lower the value of owning equity. This article explores the question of whether the tax system— along with the acceleration in inflation— could account for the poor performance of stock prices. Overall, the analysis indicates that the tax structure may well have played a sizable role in reThis is a revised version of an article that is part of a forthcoming Federal Reserve System study of the Federal tax structure. I would like to express my appreciation to Patrick Corcoran, Patric H. Hendershott, Patrick Lawlor, Martha Scanlon, Thomas Simpson, and Helmut Wendel for useful comments and suggestions, and to Joseph Snailer for statistical assistance. ducing real stock prices. At the same time, the analy sis indicates that the tax structure cannot account for the whole decline. A closer look at real stock prices Stock price averages such as the Standard and Poor’s index of 500 common stock prices moved up sharply in the early 1960s and then more slowly from 1966 to 1973 (Chart 1). Then, in 1974, prices plunged. Although they recovered somewhat thereafter, stock prices un til very recently remained below their 1973 peak. In constant dollars, stock price performance was much worse, falling dramatically since 1968 (Chart 2). Real stock prices peaked in the 1965-68 period and then declined through 1970. Although there was some recovery from 1971 through 1973, real stock prices did not regain their previous peak. Then, in late 1973 and 1974, real stock prices dropped precipitously back to their 1954-55 level. They have not since recovered substantially. How can one explain this phenomenal drop in real stock values? One simple hypothesis is that stock holders were paid dividends in excess of aftertax cor porate earnings. In this case, corporations would not have had sufficient funds to replace equipment or structures as they depreciated unless they borrowed. Whether corporations ran down their stock of fixed capital or borrowed to maintain it, the amount of fixed capital owned free and clear by stockholders would decline. The data, however, do not support this hypoth esis: in every year from 1967 to 1979, corporations paid dividends smaller than their aftertax “true” FRBNY Quarterly Review/Winter 1980-81 3 Chart 1 Standard & Poor’s Stock Price Index of 5 0 0 Stocks Chart 2 1941-43=10 Standard & Poor’s index deflated by the GNP price deflator Index 1 3 0 ---------------------------------------------------------------------------------------- " R e a l” Stock Prices Index 1 3 0 -------------------------------------------------------------- Sources: Standard & Poor’s index of 500 stocks: Standard & Poor’s Corporation; gross national product implicit price deflator: United States Department of Commerce, Bureau of Economic Analysis. Source: Standard & Poor’s Corporation. profits (see glossary). Thus, the stock price per dollar of equity investment, w hich includes retained earn ings, declined even more sharply than the real stock prices shown in Chart 2. A second hypothesis is that inflation was responsible fo r the decline in equity values. Here the data do lend support. For example, the acceleration of inflation in the seventies (Chart 3) does coincide roughly with the deterioration of real stock values. Moreover, statistical analyses over long periods of tim e indicate that stock prices were negatively correlated with the rate of inflation.1 O ther statistical studies show that the re turns to equity— which may have been reflected in equity values— were also negatively affected by in 1 See Franco M odigliani and Richard A. Cohn, “ Inflation, Rational Valuation and the M arket” , Financial Analysts Journal (M a rc h /A p ril 1979); Bruno Oudet, “ The Variation of the Return on Stocks in Periods of Infla tion” , Journal o f Financial and Quantitative Analysis (M arch 1973); and John Lintner, “ Inflation and Security Returns” , Journal of Finance (M ay 1975). 4 forFRBNY Digitized FRASERQ uarterly R eview /W inter 1980-81 flation.2 All this evidence suggests a negative correla tion between inflation and stock values. However, it does not explain the linkage. One explanation of the linkage is that the structure of the tax system reduces equity returns when inflation accelerates. Tax nonneutrality as an explanation of stock prices A tax is “ neutral” with respect to inflation if it collects the same tax monies, in real terms, from a given amount of real income regardless of the price level. That is, the taxation ratio associated with a given real income does not change w ith inflation. Both the per sonal income tax and the corporate income tax codes in the United States contain features that are not neu tral. For example, the marginal tax rate brackets of 2 See Eugene F. Fama, “ Stock Returns, Real Activity, Inflation and Money” , Graduate School of Business, University of Chicago W orking Paper (1979). the personal income tax are based upon dollar income rather than real income. If tax rates are unchanged, a proportional rise in prices and nominal incomes will put taxpayers in higher marginal tax brackets and their taxes w ill rise more than in proportion to prices. As a result, a larger percentage of their income w ill be paid in taxes even though their real income is no higher. Also, the dollar value of realized capital gains is taxed even if the asset did not appreciate in real terms, i.e., no additional purchasing power was achieved. At the corporate level, the Federal tax code has two main features that cause an increase in the tax bur den when prices accelerate: (1) “ nom inal” inventory profits are taxable3 and (2) allowable depreciation is based upon the original, rather than the replacement, cost of equipment and structures. Inventory profits Corporations are taxed on total nom inal inventory profits. Like capital gains, inventory profits are taxed even if the goods do not appreciate in real terms. The value of inventories is typically computed by using one of two accounting methods: “ first in-first o u t” (FIFO) or “ last in-first o ut” (LIFO). For a corporation using FIFO, the oldest item in inventory is assumed to be the first sold. The value of a fixed volume of raw ma terials, say, w ill rise as “ o ld ” items are taken from inventory and new higher priced ones are added. In contrast, fo r corporations using the LIFO procedure, the item inventoried most recently is the one assumed to be removed from inventory and replaced with a newly produced item. The inventory profit calculated by this method is typically small, unless a firm liquidates an extensive portion of its inventory. As a consequence, firms have an incentive to switch to LIFO and some of them did switch, particularly in 1973-74. Many more, however, were reluctant to do so, perhaps because of costs entailed in making the switch or because they feared that their stock price would decline if they imple mented an accounting change which reduced reported profits even though increasing true aftertax profits. On balance, only a small proportion of the inventory profits are computed on a LIFO basis and, in aggregate, inventory profits are therefore substantial in an infla tionary period. For example, inventory profits soared in 1973-74 and again in 1979 when inflation acceler ated (Chart 4). As a consequence of this link between inventory profits and inflation, the tax burden associ ated with inventories increases in real terms when inflation accelerates. Depreciation allowances Corporations are perm itted to deduct allowances fo r depreciation of th e ir fixed capital— structures and equipment— in com puting th eir taxable income. These allowances are based upon the “ service life ” of the capital good, as specified by the Internal Revenue Service (IRS), and the o rigin al cost of the capital good. The service lives set out by the IRS are gen erally shorter than the useful service lives of capital goods. Thus, capital goods can be depreciated faster than they wear out. When prices are rising, however, the depreciation allowances that are permitted, based upon original cost, w ill understate the true cost of replacing capital goods. And the more rapidly the price level is projected to increase, the smaller is the anticipated present value of the depreciation allow ances on a new capital good. For example, when the inflation rate is 8 percent, a corporation is permitted to deduct only 53 percent of the “ true” depreciation on a thirty-year structure (Table 1). Debt While the Federal code taxes nominal capital gains, which may not represent an increase in the general purchasing power of the asset, some im plicit real Glossary Cash flow is de fin ed as profits b e fo re ta xes plus c a p ita l co nsu m p tio n a llo w a n c e s plus net in terest paid. A neutral tax (in an in flatio n a ry sense) co lle c ts the sa m e m onies, in real te rm s, fro m a given am ou nt of real in co m e re g a rd le s s o f th e p ric e level. Reported profits (a fte r ta x e s ) are c o rp o ra te ta x a b le in co m e less c o rp o ra te ta x lia b ility. Adjusted profits a re re p o rte d profits m inus (a) inventory profits and (b) a co rre c tio n fa c to r to put d e p re c ia tio n on a re p la c e m e n t-c o s t basis. True profits are ad ju s ted profits plus th e redu ctio n of th e real v a lu e of net o u tstand ing fin a n c ia l d ebt du e to inflation. True profitability is th e ratio of tru e profits to ca p ita l, va lu ed at re p la c e m e n t cost, less th e m a rk e t va lu e of net debt. The ratio rate of return on total capital of to tal a d ju s te d c a p ita l is c a lc u la te d as the in com e— in terest plus a fte rta x profits, a d ju s te d to e lim in a te in ven tory profits and to re fle c t d e p re c ia tio n on a re p la c e m e n t-c o s t basis — to th e re p la c e m e n t co st of ca p ita l. 3 There is no easy way to calculate true inventory profits. FRBNY Q uarterly R eview/W inter 1980-81 5 capital gains are not taxed. Consider, fo r example, the real value of a corporation’s financial debt. When ever the price level increases unexpectedly, the real value of the corporation’s outstanding debt declines and the shareholders’ real wealth increases. Yet there is no tax on this real gain. (Unexpected inflation would cause some wealth shift toward debtors even if part of it were taxed.) Second, a change in the anticipated rate of infla tion that affects nominal rates of interest may also benefit shareholders in a firm which has net debt outstanding.4 Suppose, fo r example, that the ex pected rate of inflation rose by 1 percentage point. To earn (or pay) the same real rate of interest, the aftertax nom inal yield would have to rise by 1 percent age point in order to offset the inflation increase. A cred itor in a 25 percent marginal tax bracket would re quire an interest rate increase of 1 y3 percentage points to net 1 percent more after taxes [(1 - .25) (1 1/3) = 1]. The corporation in a 46 percent tax bracket, in con trast, would require a 1.85 percentage point increase in the nominal bond rate to pay 1 percentage point more after taxes [(1 - .46) (1.85) = 1]. Any sm aller increase in the nominal rate of interest would improve its real income. Therefore, if the interest rate increased by 1 V3 percentage points, just enough to maintain the real aftertax earnings of the recipient of interest, the corporation’s real aftertax cost would decline. Chart 3 Growth Rate of GNP Price D eflator From four quarters earlier Percent 1 1953 55 60 65 70 75 4 There are two parts to this argument. The first concerns the tax treatm ent of interest and the second the difference between the tax rates of the corporation that pays interest and the individual who receives it. In general, the real cost of borrowing after taxes and inflation is: r — p — T, where r is the nominal interest rate, T is the reduction of taxes permitted because of the interest payment, and p is the expected annual percentage decline in the real value of the principal that is owed. A tax w hich is neutral with respect to the rate of in flation would allow a deduction of the real interest cost (r — p) per do lla r of debt. The aftertax cost would therefore be (I — tc) (r — p), where tc is the corporate tax rate on marginal income. One way of looking at this neutral tax system is that it allows all interest to be deducted but counts the reduction of the real value of the debt as taxable corporate income. (That is, the aftertax real cost could be written as: r — rtc — p + tcp, w hich is identical to the neutral tax form ula shown a few lines above.) In the United States tax system, however, nominal interest payments, rather than real interest payments are tax deductible. The aftertax real cost of a d o lla r of debt to the corporation is therefore: (I — tc) r — p. From the view point of the interest recipient, a neutral tax system would apply the marginal tax rate to the real interest earnings. The recipient, under a neutral tax, would therefore be left with (I — tp) (r — p) after taxes and inflation, where tp is the personal tax rate on marginal income. But, under the United States Federal tax code, nominal interest is fully taxed, so that after taxes and inflation the earnings per do lla r of principal are: (I — tp) r — p. If the inflation rate w ent up by 1 percentage point, the interest recipient would be at least as well off providing the nominal rate of interest increased by more than I / ( I — tp) w hile the corporation would be at least as well off providing the interest rate increased by less than I / O — t c). Digitized 6for FRASER FRBNY Quarterly R eview /W inter 1980-81 Source: GNP price deflator from United States Department of Commerce, Bureau of Economic Analysis. To summarize, inflation influences the aftertax real income of stockholders, reducing it through the gener ation of taxable nominal capital gains and nominal inventory profits, as well as through the reduction of the real value of depreciation allowances, and increas ing it through the tax treatm ent of debt and debt ser vicing. Can we say on balance how large an effect inflation has had on the value of stockownership? First, let us define precisely what we mean by “ infla tion ” . For purposes of computing the impacts on real stock values, three different cases must be distinguished: • the occurrence of inflation that was expected, • the occurrence of more inflation than was ex pected, and • an increase in the rate of inflation expected to prevail in the future. Table 1 The Present Value of Statutory Depreciation Allowances Relative to the Present Value of Price-Level-Adjusted Depreciation Allowances In percent Ten-year equipment* Sum-of- Straightyears d ig its line Inflation rate Thirty-year structure* S traightline 0 102 108 111 2 95 100 88 88 93 73 83 87 61 77 82 53 4 v * r e >, 6 8 ... ...................... Statutory lifetim es. Statutory depreciation allowances are based on the sum -ofyears d ig its form ula for equipm ent and the 150 percent de clining-balance form ula for structures. (For structures, a switch is made to the straight-line form ula in the eleventh year, so that the present value of statutory allowances is as large as possible.) The statutory allow ances for both eq u ip ment and structures use the stated lifetimes. The alternative sum -of-years digits and stra ight-lin e allow ances for eq uip ment and the straight-line allowances for structures are based on price-level-adjusted depreciation form ulas extending over lifetim es 25 percent longer than the statutory lifetimes. The entries in the table are ratios of the present value of the statutory allowances and their price-level-adjusted alternatives. The real aftertax discount rate is 3 percent. Source: Taken from Richard Kopcke, "A re Stocks a B argain?” , New England Econom ic Review (M a y /J u n e 1979). Each of these events should in principle have a differ ent effect on stock prices. When expected inflation occurs, the real valuation of the firm should not be affected; any effect on anticipated real earnings should have altered equity valuation when the anticipation was form ulated.5 Unexpected inflation, in contrast, can alter the real value of the firm ’s equity when it occurs since its impact on real tax liab ility was not anticipated. For example, 5 The real value of equity equals the present discounted value of expected future real earnings. To the extent that actual dividends are less than the perm anent level of dividends (where permanent dividends are defined as that constant level w hich has the same present value as the stream of aftertax corporate p ro fits), the real value of the firm w ill rise over time. In the case where dividends are equal to perm anent aftertax profits, the real value of the firm should remain constant. this inflation would give rise to a once-and-for-all nominal inventory profit on which corporate tax must be paid. In addition, it would cause a loss in the real value of the depreciation allowance on capital pur chased prior to the unexpected price rise. Tending to offset these negative effects is the unexpected reduc tion of the real value of the firm ’s outstanding debt. A change in the expected rate of inflation affects real tax liabilities in ways sim ilar to those from unex pected inflation— through the creation of inventory gain and the understatement of depreciation. However, in this case, both of these effects are ongoing. (Note that, in the case of an unexpected price rise, there is a one-time loss on existing fixed capital only. New equipment, purchased at the higher price level, would have a depreciation allowance that is the same per centage of replacement cost as was typical prior to the unexpected price level rise.) In addition, stockholders can anticipate that the accrued nominal capital gain between any two future points of time w ill be larger if the price level is expected to rise more rapidly. Should they sell, the realized capital gain and th eir personal tax liability would be larger in the higher inflation case. It is possible to obtain a rough idea of the maximum effect of a change in the expected rate of inflation by examining the form ula for the rate of return and figuring how much it would be affected by inflation w orking through each tax feature.6 For example, the present value of depreciation allowances can be ex pressed as a function of the rate of inflation. How much a change in the rate of inflation impacts the present value of depreciation allowances can therefore be calculated. The effect on depreciation allowances can then be translated into the effect on taxes and into the effect on aftertax income. The percentage im pact on stockholder returns is an upper lim it of the possible percentage im pact on real stock prices. If there are other assets whose real returns are unaffected and these assets were available in unlim ited supply, then stock prices would have to fall enough to produce the same real return on equity as prevailed before the inflation increase. That is, stock prices would have to fall as much as the real return. Suppose, on the other hand, there were few alternative assets. At the same time, the public wanted to maintain the same stock of accumulated wealth despite the lower returns. In this case, there could be no attempted shift out of equities and the public would simply end up accepting a lower return on stocks. In addition, my estimates overstate the impact because: 6 These calculations assume no change in the capital intensity of production and no change in the firm ’s d e b t/e q u ity ratio. FRBNY Quarterly R eview /W inter 1980-81 7 (1) The investment tax credit, which has been greatly increased since its inception, is not figured into my calculations. This would offset part of the negative effects on stock values. (2) Taxes have been reduced on average partly in response to inflation-caused rises in reve nues. Therefore, figuring the impact while holding the tax structure constant w ill over state the net effect. (3) There has been a shift away from straightline depreciation to accelerated depreciation, a reduction of perm issible service lives for the calculation of depreciation deductions, and a shift from FIFO and LIFO. All these changes tend to reduce the impact of infla tion on stock values. The results of the calculations for a change in the expected rate of inflation are displayed in Table 2, first column. My estimates show that the prescribed rules for depreciation allowances are the tax element with the largest impact. Indeed, a 4 percentage point rise in the expected rate of inflation could lower stock values by 11 percent through this one tax feature. The taxation £of inventory profits and the taxation of capital gains at the individual level each account for about a 5 percent fall. W orking in the opposite direction, the real interest rate effect could raise the return by about 5 percent, offsetting about one quarter of the negative effects of the other three tax features. The effects of a once-and-for-all bout of unexpected inflation are shown in Table 2, last column. Because unexpected inflation is not reflected in the interest rate, the gain to the firm from the reduction of the real value of outstanding debt is not offset by higher interest payments on that old debt. (In the case of a change in inflationary expectations, the interest rates would be higher, lim iting the gain to the firm.) This large posi tive benefit from inflation washes out almost all nega tive effects of inflation on inventory profits and the understatement of depreciation allowances. Altogether, a 4 percentage point increase in the expected rate of inflation could lower real stock prices by as much as 17 percent. The expected rate of infla tion has probably risen by 6 percent over the past decade. According to my calculations, the increase in the expected rate of inflation coupled with our tax system could have caused a 25 percent decline in real stock prices. Therefore, of the 50 percent decline in real stock prices in the past decade or so, the tax structure could account fo r as much as half. Although this suggests that the tax structure may have had a significant effect on stock values, clearly it is not a full explanation. Indeed, at least half of the decline in stock values remains to be explained by other factors. Kopcke and Feldstein, Green, and Sheshinski (FGS) also evaluated the impact of inflation on stockholders’ returns.7 Kopcke calculated the effect of the same four tax elements that I examined, obtaining estimates 7 Richard Kopcke, “ Are Stocks a B argain?” , New England Econom ic Review (M a y/Ju n e 1979); M artin Feldstein, Jerry Green, and Eytan Sheshinski, “ Inflation and Taxes in a Growing Economy w ith Debt and Equity Finance” , Journal of P olitical Economy (A pril 1978), Part 2. Table 2 Inflation’s Effect via the Tax System Com ponent of tax system Percentage change in equity value due to a 4 percentage point rise in the expected inflation rate* Percentage change in equity value due to an unexpected once-andfor-all rise in the price level of 4 percent Tax on inventory profits ........................................................... - 5.4 -0 .6 Tax on understated depreciation allowances .................... -1 0 .9 -0 .9 Effect on nominal debt and debt s e r v ic in g ........................ 4 .8 f 5.3 Capital gains tax (in personal incom e tax code) ........... - Total — 16.8 .............................................................................................. * Upper lim its of the im pacts. t Assumes that real rate of interest earned by bondholders remains constant, the corporation reaping the entire gain from the tax treatm ent of interest payments. (R efer to discussion in the text.) Source: M arcelle Arak, “ Can the Performance of the Stock Market Be Explained by Inflation C oupled with Our Tax System ?", Federal Reserve Bank of New York Research Paper Number 7820. Digitized8for FRBNY FRASER Quarterly R eview /W inter 1980-81 1.1 0 -0 .4 about 50 percent larger than mine. In a different ap proach, FGS compared two situations with different rates of inflation. According to their model, a 6 per cent inflation differential leads to a 21 percent differ ential in the rate of return on equity, a bit less than my calculations indicate. All in all, the different method ologies indicate that the tax system could be an im portant factor in the perform ance of the stock market but it cannot explain the entire decline in real stock prices. Criticism of the corporate taxation argument Although taxes appear to be a plausible explanation of at least part of the stock price decline, several re searchers have argued that the historical data are in consistent with this explanation. One piece of evidence cited is the ratio of taxes to before-tax cash flow (see glossary). This tax ratio declined from the fifties to the sixties to the seventies, whereas the tax structure hypothesis suggests an in crease in the ratio of taxes to capital income.8 Although the movement of the ratio of taxes to cash flow is suggestive, it is not necessarily an accurate measure of the tax burden on stockholders. First, it uses all capital income rather than income earned by stockholders. If a larger fraction of funds is raised through debt, the relative tax burden w ill fall because interest is deductible in com puting taxable corporate income. Second, the ratio of taxes to corporate in come reflects current taxes. But a change in the ex pected inflation rate w ill affect anticipated future taxes and their ratio to cash flow. The ratio of current taxes to current cash flow could be affected very little. Another piece of evidence cited is the rate of return on total capital (see glossary). This rate of return shows no trend in the postwar period as a whole, although it was somewhat lower in the midseventies than in the m idsixties, when it was particularly high. In this case also, it is not accurate to interpret the total return to capital as a measure of the return to stockholders. From the sixties to the seventies, there was a shift toward debt finance which has a more ad vantageous tax treatment. Because interest payments create a tax deduction for the corporation while divi dend payments do not, the increased use of debt w ill raise total capital income, other things being constant. (Of course, it also raises leverage and riskiness.) For example, a corporation which raised the proportion of capital financed with debt by 10 percentage points 8 A ccording to Fama (1979), the decline in the tax ratio resulted from improved depreciation allow ances— shorter service lives and acce l erated depreciation— and the de du c tib ility of interest payments. In the seventies, the larger investment tax credit was important. could raise its total return on capital by about V2 percentage point.9 Let us look more closely at the income of stock holders and th eir return on capital. To obtain the in come of stockholders, reported aftertax corporate profits (see glossary) must be adjusted to eliminate inventory profits and to reflect depreciation on a replacem ent-cost basis; both of these adjustments reduce aftertax profits. Then, to this adjusted profits (see glossary) figure must be added the gain to stock holders from the reduction of the real value of their net financial liabilities. Inflation lowers stockholders’ real debt to bondholders, banks, etc., so that the cor poration could issue more nominal debt w ithout raising the future real burden of its debt; the funds from the new bond issues could be used to increase stock holder dividends w ithout reducing the corporation’s 9 Let K be the capital stock, D the corpora tion’s debt, r the interest rate, and G gross earnings after labor and depreciation costs. Total capital incom e is aftertax corporate profits (I — t) (G — rD) plus interest payments rD. If the fraction “ b” of capital is financed by debt, incom e per d o lla r of capital is (I - t) (G - rbK) + rbK „ x G --------------^ -------------------------- or (I - t) — + trb. A change in “ b ” alters the return by t r ( A b ) . If " t ” is 0.46 and r is 0.12, then A b of 0.1 produces a change in the rate of return of 0.55 percent. FRBNY Q uarterly R eview /W inter 1980-81 9 Table 3 Views on Inflation and Stock Values M ajor reason why inflation harms stock value Author Is the corporate tax structure relevant? Are other tax elem ents im portant? ..................................................... Taxation of equity a partial explanation Yes Yes, capital gains F a m a ..................................................... No true connection No No Hendershott ........................................ Favored tax treatment of housing No (E quity values should be helped by inflatio n) Yes, treatm ent of housing Kopcke ................................................. Taxation of equity ex plains a large portion Yes Yes, capital gains ............................... Use of a nom inal interest rate to discount profit streams, plus error in calculating profits No No Arak M odigliani-Cohn Sources: See text. ability to maintain the same level of future real d ivi dends. Thus, according to standard econom ic defini tions of “ incom e” , such gains on outstanding liabilities should be included in income. Reported profits and true profits have been very different in recent years (Chart 5). The divergence between the measures in the fifties and early sixties reflected prim arily the relatively long service lives specified by the IRS. These kept depreciation allow ances below true depreciation. As service lives were liberalized, this situation changed. When inflation ac celerated in 1973, however, it became the predominant influence on the relationship between profit measures. True profits began to fall very far short of the standard profits. For example, in the fourth quarter of 1979, true profits were running at a $90 billion annual rate, 23 percent below reported profits. The adjusted profits measure— used by many ana lysts— fell even more relative to standard profits. But it is apparent that this measure substantially overstates the effect of inflation on stockholder income. The adjusted profits measure involves subtractions from reported corporate profits fo r inventory profits and true depreciation but does not add in the gain to stock holders from their reduced bond obligations. The true profits figures can be used to calculate the tax rate of, and rate of return to, stockholders. The tax burden on stockholders (as measured by taxes Digitized10 for FRASER FRBNY Quarterly R eview /W inter 1980-81 relative to before-tax true profits) declined from the fifties to the sixties (Chart 6). Since the 1960s, how ever, the tax burden on profits increased, in contrast to the tax burden on total capital income cited above. The rate of return to capital owned by stockholders — the stockholder analogue to the rate of return to total capital— was computed using true profits in the numerator. The denom inator was the replacement cost of capital minus the market value of (net) financial debt, as calculated by George Von Furstenberg.10 The decline in the stockholder returns from the high levels of the sixties to the seventies was enormous (Chart 7), whereas the total capital return did not decline much. The data therefore support the view that the tax burden on stockholders increased since the sixties. The data also suggest that there was a very substan tial decline in the aftertax return to equity capital, a decline only partly attributable to the higher effective tax rate. Alternative explanations of the fall in real stock prices Economists have put forth several alternative explana tions of the decline in real stock prices (Table 3). One cogent argum ent begins with the observation that our 10 George Von Furstenberg, “ Corporate Investment: Does Market Valuation Matter in the A ggregate?” , Brookings Papers on Economic A ctivity (1972:2). tax system treats owner-occupied dwellings in a spe cial way. In an inflationary environment, homeowners expect the value of th eir houses to appreciate; at the same time, interest rates w ill be high, reflecting the expectation of price rise. Homeowners can deduct their interest payments in figuring their taxable income. However, the services rendered by owner-occupied dwellings, that is, the im p licit rental value, is not taxed, and the capital gains are taxed only when a home is sold and then only in some circum stances.11 In effect, if an owner lives in his own house, the “ dividends” — the current rental services— are not taxed as they would be if provided by a third party. Also, the capi tal gains on owner-occupied housing are effectively taxed less heavily than capital gains on other assets because home-sale capital gains taxes often can be postponed by reinvestment or com pletely avoided by selling after age 55. When inflation accelerates, both interest costs and expected capital gains increase and the asymmetry in tax treatm ent becomes more valu able. This asymmetry in the tax treatm ent of owneroccupied housing has caused the user cost of housing to decline substantially. For example, if a person is in a 45 percent tax bracket, the decline has been about 4 percentage points according to Hendershott (1979).12 What effect would the reduction of the cost of hous ing have on stock prices? Lower housing costs w ill influence people to buy rather than rent and to buy larger a n d /o r higher quality houses. The shift of funds toward housing and away from other investments would tend to push down equity prices. Profits rela tive to stock prices would then be higher, comparable to the attractive yield on homeownership. This argu ment is both logical and consistent with most of the facts including the rapid increases in the prices of homes. The one fact that does not quite fit is that bond yields have increased about as much as the rate of inflation, so that the real return on bonds has not risen along with the return on houses and corporate equity. A different argument is that inflation causes people to make mistakes in evaluating investment opportu nities. Modigliani and Cohn, for example, hypothe size that investors use a nominal interest rate in calculations which should be done with a real interest 11 For those under age 55, gains from sale of a principal residence w hich are reinvested in a new principal residence are not taxed at the tim e of receipt. For those over 55, $100,000 of the capital gain may be excluded from taxation, subject to certain conditions. i 2 Patric H. Hendershott, “ The D ecline in Aggregate Share Values: Inflation, Taxation, Risk and P rofita bility", Conference on the Taxation of Capital (N ovem ber 16-18, 1979). rate. During an inflationary period when the nominal rate is substantially higher than the real rate, this error means that they are discounting future earnings too heavily and therefore undervaluating equity ownership. Suppose, fo r example, that current dividends per share of a particular corporation are $2, the real return on risky investments is 7 percent, and the expected in flation rate is 8 percent. The nominal return to risky investments is therefore 15 percent (= 7 + 8 ). With an inflation rate of 8 percent, dividends w ill probably be 2(1.08) next year, 2(1.08)2 the follow ing year, etc. The value of a share of stock is the present discounted value of that flow of dividends. Discounting this stream of nominal earnings by the nominal rate of interest, Chart 5 Alternative M easures of Aftertax Corporate Profits of Nonfinancial Corporations Billions of dollars Source: Reported and adjusted profits: United States Department of Commerce, Bureau of Economic Analysis. True profits: calculated by the author as described in the text. FRBNY Quarterly Review/Winter 1980-81 11 C h a rt 6 Taxation of A lternative Measures of C orporate Profits of N onfinancial Corporations Chart 7 Aftertax Profitability of Corporate Capital Rate of return F o ur-quarter moving average Ratio .6 5 0 ------------------------------------------------------------------------------- S ource: Tax payments and reported profits: United States D epartm ent of Comm erce, Bureau of Econom ic A nalysis. True profits: calculated by the author as described in the text. the value of the share of stock is: (a) 2 + 2(1.08)/1.15 + 2(1.08)7(1.15)2 + or roughly (b) 2 + 2/1.07 + 2 /(1 ,07)2 + .. . + which amounts to about $30. Note that, according to (b), the current dividend should be discounted at the real rate of interest, not the nominal rate of interest. (This is true for other returns and inflation rates as well.) If the current dividends were discounted by the nominal return of 15 percent, the stock would be m istakenly valued at only $13! In addition, Modigliani and Cohn hypothesize that investors make a second mistake: they fail to include the reduction of the real value of outstanding debt caused by price increases as part of profits. They test these hypotheses by analyzing the factors that influenced share prices in the past. Specifically, the authors estimate an equation for share prices which includes among other items (a) the nominal rate of interest and (b) a weighted average of past in flation rates that was assumed to represent expected inflation. Since the real rate of interest can be repre sented as a nominal interest rate less the expected Digitized12 for FRASER FRBNY Q uarterly R eview/W inter 1980-81 Num erator is true p ro fits ; denom inator is capital valued at re placem ent cost less the m arket value of net financial debt. See te x t fo r a description o f the ca lc u la tio n s and the data sources. rate of inflation, (b) ought to get a coefficient of oppo site sign to (a). As it turns out, however, both the inter est rate and the inflation rate variable get negative coefficients! The negative coefficient on the price vari able is not significantly different from zero in a statisti cal sense. However, even zero is much too low a coefficient.13 The authors interpret this result as evidence that investors are making two valuation errors— misusing a nominal rate as a real rate and failing to include the fall in the real value of outstanding debt as part of equity earnings. How strong is their argument? Hendershott pointed out that it is difficult to reconcile such a misvaluation with the fact that the nominal bond rates have risen about one for one with the increase in inflation. By his model, investor shifts from stocks into bonds cause the real aftertax returns, adjusted for risk, to be equal. Therefore, if investors did not properly account for inflation, bond returns would have stayed low, in tan dem with real returns on stocks. 13 Expected inflation should have an equal and opposite sign from the nominal rate of interest— to convert the nominal rate to a real rate— plus a coefficient reflecting the anticipated future inflationproduced capital gains on the outstanding debt. Moreover, there are other ways to explain the empiri cal results obtained by Modigliani and Cohn. For ex ample, a weighted average of past inflation rates could be a poor estimate of the inflation rate expected to prevail over the long term. On the other hand, because nominal bond rates incorporate price expectations, changes in bond rates could be a good proxy for changes in expected inflation. Indeed, if variations in the real rate of interest tend to be small, then most of the changes in the bond yield will reflect changes in price expectations. In this case, the bond rate would be proxying for expected inflation and its coefficient would represent the effect of expected inflation on equity values rather than the effect of real interest rates on equity values. By this interpretation, the co efficient of -0 .0 5 9 obtained in one of their regres sions indicates that each 1 percentage point increase in the expected rate of inflation would reduce stock values by 5.9 percent; a 6 percentage point increase in the rate of inflation would therefore reduce real stock prices by about 35 percent. Interestingly enough, this is within the range of the Arak-Kopcke stock price impact calculated from the tax structure. While many explanations of stock price behavior are related to inflation in some way, others are not. For example, some economists argue that equity prices have declined for the simple reason that cor porate profitability before taxes has dropped sharply. Charts 6 and 7 lend support to this view; they show that stockholders’ (aftertax) return dropped substan tially while the tax rate on stockholders increased only moderately. Another factor may be that the growth prospects during the 1960s were much brighter than during the 1970s. Since stock values are based upon expected dividend growth, the outlook could well be an important element. Conclusions and implications There is no single factor that can plausibly explain the substantial fall in real stock values over the past ten to fifteen years. However, the tax system— the corporate and capital gains tax as well as the tax treat ment of housing— probably has played a significant role. Besides lowering real stock values, the current tax system may impair productivity by lowering desired capital investment and encouraging shorter lived capital than is optimal from an economy-wide vantage point. Moreover, the tax system gives firms a large incentive to leverage themselves. Taken together, there would be important gains from reforming the corporate tax system to get rid of the features which cause non neutrality with respect to inflation. Of the features considered above, the depreciationallowance rules are the single most important in terms of the impact on real stock values. Moreover, the de preciation allowances probably were important in in ducing business to build less durable capital than is desirable from society’s viewpoint.14 The ideal solution is to base allowances on replacement cost, rather than on original cost, while using write-off schemes that approximate the true depreciation of each piece of capital. Ad hoc schemes to improve depreciation allow ances, such as shortening the permissible service lives or widening the scope for use of accelerated deprecia tion, work imperfectly. Only at one particular inflation rate and with one particular technological mix will they exactly offset the shortfall in the true depreciation generated by the use of original cost. If the inflation rate were to fall, such schemes would lead to higher profits and longer lived equipment than is economi cally efficient. According to the Bureau of Economic Analysis, Department of Commerce, the understate ment of depreciation was about $17 billion in 1979. If this were added to the depreciation write-offs cur rently allowed, it would have cost the United States Treasury less than $8 billion in 1979, far less than some of the other schemes that have been proposed to im prove depreciation write-offs. Another issue is whether the United States wants to retain tax provisions that allow the full deduc tion of nominal interest payments by both business and homeowners, and the full taxability of interest receipts. For the corporation, the deduction of nominal interest payments about offsets the taxability of nominal inventory profits. However, for the homeowner there is no similar offset; the homeowner clearly benefits. Although this country wants to encourage homeownership, inflation undoubtedly has widened the encour agement far beyond the original plan. Some tax change that would alter this situation without greatly hurting current homeowners would be desirable. 14 Patrick Corcoran, “ Inflation, Taxes, and the Composition of Business Investment”, this Q u a rte rly R e v ie w (Autumn 1979), pages 13-24. Marcelle Arak FRBNY Quarterly Review/Winter 1980-81 13 Cutting the Federal Budget Analyzing How Fast Expenditure Growth Can Be Reduced Federal outlays in fiscal year 1981 threaten to exceed $660 billion, well above the second budget resolution ceiling of $632.4 billion and the goal of $635 billion contained in the Stockman-Kemp memorandum to the then President-elect Reagan on “Avoiding a GOP Eco nomic Dunkirk”.1 Federal spending as a percentage of gross national product (GNP) could exceed the postwar high of 23.1 percent, and the unified budget deficit could be $60 billion or greater. When combined with an off-budget deficit of about $23 billion, this would result in new Treasury marketable financing of over $80 billion. At this point, it is highly unlikely that projected unified budget outlays for 1981 can be re duced. Various changes, some of which are cosmetic2 and do not affect the size of Government, may be proposed. However, a major push during the next few months to cut spending for 1981 could very well end up a wasted effort and at the same time use up “po litical capital” necessary for meaningful cuts in 1982 and 1983. The outlook for Federal outlays in 1982 and 1983 under current policies is for continued high rates James R. Capra, Senior Economist in charge of the Fiscal Analysis Staff, Monetary Research Department, is the author of this article. He was formerly the Chief of Budget Projections for the Congressional Budget Office. The views expressed are those of the author and do not necessarily reflect those of the Federal Reserve Bank of New York or the Congressional Budget Office. 1 This memorandum was written by David Stockman and Jack Kemp in late November 1980. More recent policy statements indicate that the administration may ultimately set a target of $645-650 billion for 1981 Federal outlays. 2 Among the cosmetic changes are asset sales from the Farmers Home Administration to the Federal Financing Bank (F FB )— an off-budget agency— and changes in the timing of offshore oil sales. Digitized14 for FRASER FRBNY Quarterly Review/Winter 1980-81 of spending. With the start of fiscal year 1982 only eight months away, a legislative calendar devoted primarily to the control of Federal spending could reduce projected outlays for fiscal year 1982 by about $10-15 billion. A significantly larger reduction would take an extraordinary effort on the part of the new administration and a degree of cooperation by the Congress that is rarely seen. Realistically, however, the earliest target date for a full offset (through spend ing cuts) to the $30-35 billion per year revenue loss from a 1981 individual income tax cut probably would be fiscal year 1983. The 1981 problem There appears to be a consensus that Federal spend ing programs ought to be cut, or at least the rate of growth reduced. However, there is a misconception that this can be done rapidly— in 1981, for example. The two largest components of spending are national defense and benefit payments for individuals (Table 1). Defense clearly will not be reduced in the near future; rather there appears to be widespread support for in creases. Almost all benefit payments programs are en titlements, which means that eligible beneficiaries have a legal claim on the Federal Government. Changes require substantive legislation that would take months to formulate, negotiate, and implement. For example, the recently enacted budget reconciliation bill— an omnibus bill that changed current law for many pro grams— was formulated in the spring of 1980, negoti ated in part in the first budget resolution conference in the early summer and in part by conferences bn various components of the bill during the remainder of the summer and in the fall. Some of the new pro visions of current law that are the result of this bill will take months to implement, making the total time from Estimated Federal Outlays for Fiscal Year 1981 ■ .' ■ Benefit payments to in d iv id u a ls ..................................... 331.4 Other grants to state and local go vernm ents! .............. 58.3 Net interest ........................................................................... 67.4 Other Federal operations ................................................ 56.2 T o t a l........................................................................................ 660.0 ---- -------------------------------------------- ----- -----------* Includes Department of Defense m ilitary and defense-related activities of the Department of'E nergy but not m ilitary retired pay which is included under benefit payments. t Includes those grants that are not for benefit payments Table 2 Possible Proposals for 1981 Outlay Cuts and Savings that Might be Claimed In billions of dollars Proposal Amount Type of cut -----------------------------------------------------------Small Business Adm inistration disaster lo a n s ..................................... 1.2 (O n e tim e ) Medicare and m e d ic a id .................... 1.0 (P erm ane nt)* Strategic petroleum re s e rv e ............. 0.8 (C o s m e tic )t Solvent refined coal dem onstration plants I and I I ..................................... 0.2 (Perm anent) Public service e m p lo y m e n t............. 0.4 (P erm anent) ............... 0.8 (Perm anent) Trade adjustm ent a s s is ta n c e ........... 0.7 (Perm anent) Economic support fund .................... 0.4 (D elay) Unem ploym ent insurance social security in c re a s e .................... ...... 4.5 (O ne tim e) Asset sales ................................................ 1.5 (C osm etic) Outer continental shelf le a s e s ___ ___ 1.8 (C osm etic) Travel, pay, and c o n s u ltin g ............. ...... 1.7 (C osm etic) Tea, ..................................................... m o --------------------------------- . * For the permanent items the savings to 1982 and 1983 outlays would be different from the 1981 savings. t This reduction is listed as cosm etic since the change probably would not reduce the long-run costs to the Federal Government and may ultim ately result in an increase. proposal to implem entation almost a year.3 After defense and benefit programs, two sm aller categories— grants and other Federal operations— re main for consideration. For grants, outlays during the remainder of fiscal year 1981 largely represent pay ments fo r obligations that have already been incurred or contracts already signed. M ajor programs in this category include aid for elementary and secondary education, grants for the construction of wastewater treatm ent plants, and the Federal-aid highways pro gram. For the education programs, obligations for the 1981 school year were made in the summer of 1980. For the construction programs, 1981 outlays prim arily represent the execution of contracts signed in 1980 and prior years. Breaking these contracts would be very difficult and very expensive. The final category— other Federal operations— is comprised of many differ ent Federal programs, ranging from the strategic pe troleum reserve to farm price supports and pay fo r the nondefense Federal work force. For the m ajor pro grams, the problems with reducing 1981 outlays are sim ilar to those for national defense, benefit payments, and grants. The strategic petroleum reserve is a high p riority item. The new adm inistration and a clear Con gressional m ajority favor increases rather than a scalingdown of the program. The farm price supports program is an entitlem ent and changes probably w ill not be forthcom ing until a new farm bill is considered this spring. Even m ajor changes in the bill w ill probably not significantly affect 1981 outlays. Federal pay, on the other hand, could be reduced by attrition or even by layoffs. However, even a 10 percent reduction of Federal civilian agency employment would save less than $1 billion in fiscal year 1981. The bleak prospects for changes that would reduce 1981 outlays need to be emphasized. If the prim ary focus of the upcoming debate over control of spending becomes fiscal year 1981, the prospects for meaningful reductions in 1982 and 1983 may be jeopardized, with near-term savings being achieved through delays or even exchanged for subsequent program expansions. The recently enacted reconciliation bill provides a good example of the potential problems, with over emphasis on near-term savings. In the House-passed bill, 1981 savings in medicare and m edicaid were ex changed for program expansions in 1982-85. Another example is child nutrition where an immediate one tim e cut was finally agreed to in exchange for no re ductions in 1982-85. Some reductions of 1981 outlays 3 Another example of problem s with making cuts in benefit payments quickly is the food stamp program. In each of the past two years reforms have been enacted, but it now appears that in both cases the changes will take more than one-year longer to im plem ent than anticipated at the tim e of passage. FRBNY Quarterly R eview /W inter 1980-81 15 are, of course, not impossible, but large cuts are highly unlikely. Nevertheless, the new administration is likely to pro pose budget cuts for 1981. The following list represents some of the major components that have been dis cussed recently, together with the savings in the fiscal year ending September 1981 which might be claimed for them. The savings listed are highly dependent on early enactment. • Change the newly enacted Small Business Ad ministration (SBA) authorization to make farm ers who were victims of the 1980 springsummer drought ineligible for SBA disaster loans. This change could be assumed to result in a one-time saving of $1.2 billion in 1981. (The new authorization made victims of future droughts ineligible for SBA loans.) • Make miscellaneous changes in the laws gov erning medicare and medicaid, including caps on certain fees for services. By assuming al most immediate enactment, about $1 billion in permanent savings could be claimed. • Fund the strategic petroleum reserve by having the Federal Government sell shares in the stored oil or by issuing bonds to defray the cost of oil purchases. If early enactment of this complex proposal were assumed, an in crease of $0.8 billion in offsetting receipts and a decrease in net Federal outlays would be claimed. The future-year effects are unclear, depending on whether the oil reserve is used and on the assumed rate of return to share holders or bondholders. In all likelihood, the proposal would increase the long-term costs to the Federal Government. • Delay, or possibly terminate, construction of the two solvent refined coal demonstration plants (SRC I and II) at a savings of $0.2 billion. • Terminate all funding for countercyclical public service employment, including a rescission of funds already appropriated. Savings of $0.4-0.6 billion in fiscal year 1981 might be claimed, al though action would have to take place quickly. A significant portion of 1981 funds have already been obligated. As discussed later, this cut would have a larger effect on projected out lays for 1982 and 1983. • Make miscellaneous changes in the unemploy ment insurance laws, which if enacted quickly would save $0.8 billion. Trade adjustment as sistance changes, saving about $0.7 billion, also might be proposed. • Change the method of disbursement of credits Digitized for 16 FRASER FRBNY Quarterly Review/Winter 1980-81 • • • • to Israel through the Economic Support Fund back to the pre-1979 approach. By assuming early enactment of this change, the new ad ministration could claim the delay of $0.4 bil lion in outlays until 1982. Postpone the July 1, 1981 social security in crease until October 1, 1981. This proposal, which would affect recipients of social security, railroad retirement, supplemental security in come, and veterans’ pensions, would be for a one-time postponement. Savings of $4.5 billion in 1981 could be claimed, but there would be no lasting effect on Government spending levels.4 Increase asset sales of Federally held mort gages and insurance by the Farmers Home Ad ministration to the FFB. These sales would shift about $1-2 billion in outlays off-budget. The change would be cosmetic since the FFB purchase would then become part of the offbudget deficit and would still be financed through the issuance of Treasury debt. Move a scheduled sale of outer continental shelf leases from September to August 1981 so that all the receipts would offset outlays in 1981 rather than in 1982. This could reduce the 1981 budget totals by $1.5-2.0 billion but would increase the 1982 totals unless the 1982 sched ule is revised. Finally, the new administration may propose miscellaneous rescissions of already enacted appropriations for travel, pay, and consulting services of about $1-2 billion. The size of the cut may be somewhat dependent on how much is needed to reach the announced goal for total outlay reductions. Outlay savings in 1981 would be difficult to achieve because, even if the Congress enacted a rescission of budget au thority, agencies would probably absorb the cut in slow spending programs rather than in travel and pay where outlays flow quickly from budget authority. Thus, the reduction would be to 1982 outlays. These possible proposals which total as much as $15 billion are summarized in Table 2. Each of the proposals for 1981 reductions would require a major effort on the part of the new adminis tration. Even proposed delays and one-time reductions could encounter stiff resistance that might either ulti4 In theory, this proposal would lead to savings of about $400-600 million in interest on the public debt in 1982 and later years if the temporary reduction of 1981 outlays were not used to fund a larger tax cut or a larger defense program. mately defeat the proposed changes or at least could stall enactment until the savings would be significantly reduced. In the long run, the Congress and the Pres ident can control virtually every dollar spent by the Federal Government by making changes in the numer ous laws governing Federal expenditures. However, forging a consensus on just which laws ought to be changed— and how they ought to be changed— takes time. Also, after that consensus is reached and laws are changed, implementation is anything but im mediate. Only eight months will remain in fiscal year 1981 after the new administration takes office and probably only four months will remain by the time a third budget resolution for 1981 is passed by the Congress. The most realistic (and possibly the best) strategy may be to forget about large budget cuts for fiscal year 1981 and to work out proposals carefully that will affect 1982 and more importantly 1983. Unfortunately, it is already getting late to do as much as might be desired about 1982. As will be shown later, more than a 2 per cent reduction of projected outlays for 1982 as a result of Congressional action over the next eight months is hard to visualize. A reduction of even that size will not be possible if the 97th Congress and the ad ministration spend the next several months on quickfix options designed to reduce the 1981 budget totals. Spending reductions for 1982 and 1983 At present, with no new program initiatives (except for defense), Federal outlays for 1982 and 1983 may ba projected at $760 billion and $850 billion, respectively. The projection assumes 5 percent real growth of de fense, 2 percent real growth of benefit payments, pri marily due to demographic and case-load changes, and no real growth of grants and other Federal operations. The estimates in Table 3 provide a useful baseline from which spending cuts can be considered. In evaluating potential budget reductions, the follow ing factors are important. All cuts require joint action by the administration and the Congress. With the pass age of the Congressional Budget Act, the President can no longer impound funds. The proposals that follow have been restricted to those that are con sidered to be politically and technically feasible in the sense that they have a reasonable chance of being proposed, adopted, and implemented in time to achieve the savings listed for 1982 and 1983. Finally, although outlay savings proposals for national defense exist, they are not likely to result in a smaller spending total. The entire national defense discussion is now being framed in terms of real growth, with 5 percent real growth being the minimum figure under active consideration. Cuts in low priority or unnecessary de fense expenditures are likely to be offset by increases in order to sustain the real growth target that emerges from the debate over the next few months. The remainder of this article will review in some detail the ways in which projected Federal spend ing for fiscal years 1982 and 1983 could be cut. The reductions of benefit payments, grants, and other Federal operations that will be discussed and are sum marized in Table 4 cut across many different Federal programs and represent an ambitious agenda for the first session of the 97th Congress that would require numerous changes in current law. The options do not include all possible budget-cutting alternatives that have been or will be proposed, but in general they are the ones that have been most prominently discussed during the past year. Some have been included in President Carter’s budget proposal. Additional budget-reduction alternatives may be put forward, and some that are not considered in this article could ultimately be enacted. However, since each change requires sepa rate consideration and negotiation, it is doubtful that a program significantly larger than the one outlined in this article could be formulated, negotiated, and implemented within the next eight months.5 Benefit payments fo r individuals The largest benefit payments program is social security with projected outlays in fiscal year 1982 of over $160 billion. The upcoming July 1981 benefit-level increase, estimated at over 12 percent, will raise the annualized cost of social security by over $17 billion. Proposals to make major changes in social security indexing stand little chance of being enacted and probably will not even be proposed since they affect over 35 million recipients, most of whom are eligible voters. As a general rule, to stand any significant chance of passage, proposed cuts in social security should be designed to affect either subsets of existing recipients or future recipients.6 For example, the following cuts could be proposed (Table 5). • Under current law, dependent benefits are paid to unmarried students between the ages of 18 and 21. This benefit, which is not based upon need and costs about $2 billion annually, could 5 A longer list of budget-reduction options can be found in R e d u c in g th e F e d e ra l B u d g e t: S tra te g ie s a n d E x a m p le s (Congressional Budget Office, February 1980). The Congressional Budget Office plans to update this report in March 1981. ‘ This assumption may turn out to be wrong. However, the chances of large near-term social security reductions that affect all current beneficiaries appear to be so remote that it would be unproductive to formulate a budget policy that depends to a significant degree on their enactment. FRBNY Quarterly Review/Winter 1980-81 17 Table 3 Table 5 Projected Outlays Assuming 5 Percent Real Growth for Defense and No New Nondefense Initiatives* Social Security Savings By fiscal year; in m illions of dollars Item By fiscal year; in billions of dollars Spending category 1982 1983 203 National defense ............................... Benefit p a y m e n ts ................................. ............... 375 418 Other g r a n ts .......................................... ............... 64 69 .......................................... ............... 86 93 Other Federal operations .................. ............... 63 67 ....................................................... ............... 760 850 Net interest Total 1982 1983 Phase out student b e n e fits ............................. 200 Elim inate minim um b e n e fit............................. 65 135 Elim inate lump sum death b e n e fit............... 165 190 Phase out survivor benefits for high school children, age 16, 17, and 18 ........... 300 2,000 Total 730 3,125 .................................................................... 800 * Also assumed is $16 billio n in off-budget spending for 1982 and 1983. Table 4 Reductions of Federal Spending for Fiscal Year 1982 and the Effects on 1983 By fiscal year 1982 Billions of dollars Spending category 1982 Percent* 1983 Billions of dollars 1983 Percent 2.0 Benefit p a y m e n ts .................. ......................................................... — 5.9 1.6 — 8.5 Other g r a n t s ........................... ......................................................... - 3.0 4.6 - 4.4 6.3 Other Federal operations .. ........... ............................................. — 2.9 4.6 - 6.5 9.7 ........................................ ......................................................... -1 1 ,8 1.6 — 19.4 2.3 Total For each spending category, the percentage represents the cut as a fraction of projected spending for the category. For the total, the percentage represents the sum of the reductions expressed as a fraction of projected total Federal outlays. Table 6 Table 7 Unemployment Compensation Savings Income Support Savings By fiscal year; in m illions of dollars By fiscal year; in m illions of dollars Item 1982 1983 Item Elim inate the national trigger for extended benefits ............................................ 1,000 1,000 Reduce trade adjustm ent assistance b e n e fits ................................................................ 1,400 300 Total 2,400 1,300 .................................................................... FRBNY Q uarterly R eview /W inter 1980-81 Digitized 18 for FRASER 1982 1983 500 Monthly incom e r e p o rtin g ............................... 400 Child nutrition ................................................... 200 200 Food stamps .......................... .......................... 700 1,000 1,300 1,700 Total .......................................... .......................... be phased out starting in 1982 by stipulating that payments would not be made to students who reach their eighteenth birthday after October 1, 1981. • A minimum social security benefit of $122 is currently provided to insured workers who re tire at age 65, regardless of the level of their past earnings. Many of those who receive the benefit have earned pensions under other pro grams, typically civil service retirement. Elimi nation of this benefit and coverage of those actually in need through supplemental security income (SSI) would save about $100 million per year. • All surviving families receive a lump sum death benefit of $255. The benefit is out of date, not having increased significantly since 1954. Cov erage of those families in need could be pro vided through SSI. • Currently, families with children under 18 are entitled to survivor benefits for each child and for the spouse under the assumption that the parent cannot work away from home while a child is in his or her care. A phasing-out of benefits for high school children age 16, 17, and 18— where the rationale for the benefit is probably not applicable— would save up to $2 billion by 1983. In the area of health, there are numerous proposals to restrict the growth of medicare and medicaid. Pro jected costs in 1982 for hospital insurance, supple mentary medical insurance, and medicaid total more than $67 billion. Like social security, however, medi care and medicaid benefits are paid to millions of recipients (more than 45 million). Proposals with a reasonable chance of enactment would save about $1 billion in 1982 and $1.8 billion in 1983. Mandatory hospital cost containment might save more. However, the new administration and the Congress may wait another year or two to evaluate whether hospitals have voluntarily moderated their price increases. The earliest consideration of a new cost containment proposal probably will be in connection with the 1983 budget. Unemployment compensation, with an estimated fiscal year 1982 cost of over $20 billion, is another area where reductions might be feasible. One of the largest cuts would be achieved by eliminating the na tional trigger for extended benefits (Table 6). Currently, an extra thirteen weeks of benefits are paid to all re cipients when the national insured unemployment rate exceeds 4.5 percent even though a state’s rate may be below 4.5 percent. This proposal was included in the Senate’s version of the reconciliation bill but was re moved in the House-Senate conference agreement on the bill. Other reductions of unemployment benefits could be achieved by implementing the Government Accounting Office recommendations that trade adjust ment assistance benefits be paid only to those who have exhausted their unemployment insurance benefits and be payable at the same level as unemployment in surance benefits. The rationale for this reduction is that under current law it is possible for trade adjustment assistance recipients to receive benefits (when com bined with other income transfer payments) which exceed their take home pay prior to becoming unem ployed. Clearly, this is likely to create a disincsntive for trade adjustment assistance recipients to start looking for work. Miscellaneous income support or welfare programs such as aid to families with dependent children ($9 billion), food stamps ($12 billion), supplemental secu rity income ($8 billion), and child nutrition ($4 billion) are potential targets for reductions (Table 7). A nation wide monthly income-reporting system, together with one-month retrospective accounting (that is, basing each month’s benefits on the previous month’s income), would eliminate some of the current welfare abuses. Other reductions include making permanent the change from a semiannual to an annual cost-of-living adjust ment for child nutrition and certain miscellaneous food stamp cuts. Finally, veterans programs are a potential target for reductions, although it has been extremely difficult to obtain passage of legislation that would cut the $20 bil lion of benefits and administrative expenses. The larg est cut that has been discussed recently would require private insurance companies to reimburse the Veter ans Administration for insured persons treated in vet erans hospitals, so-called “third party reimbursement” (Table 8). The House and Senate Veterans Committees have been very reluctant, however, to report this leg islation. Other changes for veterans, such as reducing burial benefits by the amount of the other Federal burial benefits received by veterans, would have a smaller effect on outlays. Grants to state and loca l governments Over the last thirty years the largest growth of Federal spending, on a percentage basis, has occurred in grants to state and local governments for other than benefit payments. The following are some possible cuts (Table 9). • Federal spending on highways is growing very rapidly. It is not clear that this growth is desir able in light of the need to cut back on energy consumption. Reimposition of a tight obligation FRBNY Quarterly Review/Winter 1980-81 19 Table 8 Veterans Savings By fiscal year; in m illions of dollars Item 1982 1983 Third party re im b u rs e m e n t...................... 350 400 Veterans' com pensation, pensions, and burial benefits ................................... 100 100 Gl b ill c h a n g e s .......................................... 60 80 510 580 Total .............................................................. Table 9 Cuts in Grants to State and Local Governments By fiscal year; in m illions of dollars Grants Highways ..................................................... Environmental Protection Agency low priority c o n s tru c tio n ........................... Public service jobs ................................... . . . 1982 1983 700 1,500 50 350 1,000 1,100 Other Comprehensive Employment and Training Act p ro g ra m s ...................... 700 750 Im pact aid for school d is tr ic t s ................ 250 350 D iscretionary health p ro g ra m s ............... 300 300 3,000 4,350 Item 1982 1983 Strategic petroleum reserve oil purchases . . 1,000 3,000 Term ination of solvent refined coal dem onstration plants I and II construction . Total .............................................................. . . . Table 10 Cuts in Other Federal Operations By fiscal year; in m illions of dollars 500 1,000 C om m odity Credit C orporation price support reductions ............................... 100 1,000 Railroad cuts in low priority ro u te s ................ 300 350 Federal payment to postal s e r v ic e ............... 250 250 Wage board pay raises (nondefense) ___ 60 60 R eduction of outm oded soil and water conservation p r o je c ts ........................................ 100 100 Reduction of civilian agency em ploym ent . . 440 480 Lim it nondefense travel and tra n s p o rta tio n ..................................................... 100 200 2,850 6,440 Total .................................................................... FRBNY Q uarterly R eview /W inter 1980-81 Digitized20 for FRASER ceiling on the Federal-aid highways program could hold 1982 and 1983 outlays to about $8 billion per year. • The Environmental Protection Agency (EPA) makes grants for the construction of waste water treatment plants. Because of various re quirements specified in the law that emphasize “ ready to go” rather than high priority pro jects, approximately 25 percent of the funds have been used fo r low priority projects. Un fortunately, the savings from a change to elim inate these projects build slowly. Nevertheless, the 1985 savings would be about $1 billion, out of a projected cost of $4 billion. • Expenditures for countercyclical public service jobs w ill total about $1 billion in 1981. This program has demonstrated a marked procy clical pattern. It probably should (and in all likelihood will) be terminated in 1982. • Many other Comprehensive Employment and Training Act (CETA) programs do not appear to be effective or duplicate private-sector pro grams. A 10 percent cut in the $7 billion in projected outlays probably could be achieved without impairing the effectiveness of the pro grams. The cuts could be across the board or targeted toward specific programs like sum mer youth or the public service jobs program for the structurally unemployed (Title II of CETA). • The impact aid program compensates school districts for children whose parents live or work on Federal property. Annual funding is about $800 m illion. The purpose is to compensate school districts partially for educating pupils where the local tax base is reduced because of Federal property ownership or where enroll ments are raised because of a Federal em ployer. Parts of this program clearly do not meet the intended needs. Past adm inistrations have unsuccessfully proposed cuts, but last year the Congress came closer to approving re ductions. • During the past year, several proposals were made to reduce discretionary health grants which overlap with other programs. These in clude drug and alcohol abuse, mental health, family planning, and health planning. A pro posal submitted in 1980 by Republican mem bers of the House and Senate would save about $300 m illion (or slightly over 10 percent of the $2.5-3 billion spent on these programs). The reductions to state and local government grants summarized in Table 9 are for categorical grant pro grams, that is, grants where the Federal Government specifies the precise purpose or use of the funding. There appears to be little, if any, support in the Con gress or in the new administration for cuts in commu nity development block grants or in general revenue sharing.7 In fact, most Republicans advocate the res toration of the $2.3 billion state share of general rev enue sharing in 1982. One way to offset such an in crease, however, might be to require states to forfeit funds for categorical grants on a dollar-for-dollar basis in exchange for revenue-sharing funds. This pro vision was written into the recently enacted general revenue-sharing reauthorization; however, a plan for implementing what could become a complex system of credits and debits does not yet exist. Other Federal operations This category contains numerous Federal programs. Some are not particularly effective, but few are large when compared with defense and the major benefit payments programs. • Federal outlays for the purchase of oil for the strategic petroleum reserve are projected at about $3.5 for 1982 and $4.5 billion for 1983.8 All these outlays would not neces sarily be offset by the proposals for public capitalization or debt financing of the reserve discussed earlier. Various factors, ranging from the marketability of the certificates of owner ship to the coupon rate (if any), would affect the size of the offset. The savings for 1982 would also be affected by the timing of a change in the law and lags associated with implementation. For the purposes of this analysis, savings (offsets) of $1 billion in 1982 and $3 billion in 1983 are assumed. The esti mates reflect gradual implementation of a change enacted late in fiscal year 1981 or early 1982 and a program that includes an annual interest payment on the debt out standing.9 7 Over the next few months proposals may be made to cut Urban Development Action Grants, a program to help cities revitalize their economic bases and reclaim deteriorated neighborhoods. Reductions of appropriations, however, will not significantly affect outlays until after 1983. 8 This projection assumes a fill rate in excess of 100,000 barrels per day and further increases of world oil prices. 9 It should be noted that, in the long run, this proposal may cost more than current policies, depending on whether the oil purchases are debt financed or equity financed and on the relationship between oil price increases and interest rates. • As discussed earlier, the termination of con struction at the two solvent refined coal demon stration plants (SRC I and II) might be a poten tial budget reduction option since the current program has already shown the feasibility of the technology. • Several changes could be made in the farm price support program administered by the Commodity Credit Corporation (CCC). The CCC spent about $3 billion in fiscal year 1980. Although several aspects of the various CCC programs could be changed so that outlays would be reduced, the Congress will be under considerable pressure for program expan sions. The disaster payments program prob ably can be eliminated since it largely dupli cates the newly enacted Federal crop insurance program. Also, cuts in dairy price supports, such as indexing support levels to annual rather than semiannual changes in prices, ap pear to be justified. • Federal support of railroads totals $1.9 billion, including funds for construction and operating subsidies. Subsidies for low priority routes could be reduced, saving about $300 million per year. • The $1.2 billion annual Federal payment to the Postal Service could be reduced. The cut need not specify elimination of Saturday mail de livery, as was proposed in March 1980. The Postal Service probably should be allowed to decide how to absorb the cut— either by rais ing rates or by new efficiency initiatives. • The current procedure for computing pay raises for Federal blue-collar workers (wage board employees) overstates the percentage increase needed to maintain comparability with the private sector. The savings from reform would be over $500 million by 1983. However, three fourths of the $10 billion in pay is an expense of the Department of Defense. Most of the cut probably would be offset by other defense increases in order to maintain a 5 per cent (or greater) real growth target for defense outlays. • Miscellaneous soil and water conservation projects that have outlived their usefulness and actually are in direct opposition to wildlife con servation projects could be reduced. The bud get for the Soil and Water Conservation Service is about $400 million. • The Army Corps of Engineers currently spends about $1 billion per year for construction and operating costs on the nation’s network of FRBNY Quarterly Review/Winter 1980-81 21 inland waterways and to help maintain deepdraft ports. An increase in the per gallon fuel tax paid by inland waterway users could de fray some of these expenses. (The receipts are treated as offsets to Federal outlays.) An in crease of over 5 cents per gallon would be required for each $100 million in offsets. It is likely that such a proposal would encounter stiff opposition. The most recent increase in the tax was debated several years prior to enactment. • A reduction of Federal civilian agency employ ment through attrition (for example, a two for one attrition-replacement policy) would reduce employment by 2 percent. Assuming a 1982 payroll of $24 billion, savings would be $400-500 million. However, it is not clear that such a policy is desirable, compared with more targeted cutbacks. The savings of $400-500 million could be achieved through attrition, major cutbacks in certain departments like the Department of Energy, or by a 10 percent cut back in the $5 billion spent by Federal agencies to formulate and enforce Federal regulations. • Federal travel and transportation cost about $9 billion annually, with over 75 percent attrib utable to the Department of Defense. The $1.7 billion for civilian agencies is embedded within programs presented throughout the bud get. Although data are maintained on travel and transportation expenses, budget and ap propriations review is generally done on a programmatic basis rather than on an object class or input basis. Pending detailed review of travel and transportation policies, a general provision limiting 1982 expenses to 1981 levels could be attached to each nondefense appro priation bill, saving between $100-200 million. (It is not clear that such a policy is appropri ate for the Department of Defense since most funds are used to transfer military personnel and move equipment.) The reductions summarized in Table 10 do not include some across-the-board cuts that are expected to be proposed by the new administration. In particular, reductions of expenditures for consulting services may be proposed. However, unlike travel, little data is avail able on where or how money is spent or on how to make the reductions. The new administration may include a cut of about $700 million for such services in its bud get, but it would be extremely difficult to implement the reductions. Digitized22 for FRASER FRBNY Quarterly Review/Winter 1980-81 Net interest No reductions of interest on the public debt have been included since it is unclear whether the spend ing cuts will lower the deficit or be used for larger defense increases or for larger tax cuts. If spend ing cuts were used to lower the deficit, interest costs in 1982 would drop by about $0.6 billion for each $10 billion reduction of the deficit because of a lower level of outstanding interest-bearing debt; by 1983 the savings would be $1.8 billion, if the $10 billion deficit reduction were continued. Summary If all the reductions outlined in this article were en acted, the savings would be $12 billion in 1982 and about $20 billion in 1983. These savings would repre sent about 2.0 percent of nondefense spending in 1982, 3.0 percent in 1983, and approximately 2 per cent of total Federal spending in each year. Between 80 and 90 percent of the reductions could be accom plished only by rewriting existing laws rather than through regular appropriations action. The process of changing laws generally requires extended and drawn-out negotiations and is subject to greater delays than appropriations. Because of the time required to negotiate and implement the various changes in cur rent laws, the savings totals in this article are prob ably an upper limit on what can be achieved through action during the remainder of this year. By com parison, the push for reductions in the fiscal year 1981 budget that started last March probably resulted in nondefense legislated savings of only $4-6 billion, despite the fact that the effort had administration and bipartisan Congressional support for achieving a bal anced budget to resist inflationary pressures. However, many of the reductions discussed in this article may be opposed by the Democratic leadership in the House. Also, the target date for a balanced budget appears to be slipping further into the future. Conse quently, advocates of spending cuts cannot use the balanced budget argument to defeat amendments that exempt various programs from budget cuts. An alternative to the goal of a balanced budget is expenditure cuts that offset the revenue loss from a tax cut similar to the first instalment of the Roth-Kemp proposal and from a business tax cut. However, the revenue loss in 1982 from a 10 percent across-theboard cut in individual income tax rates— about $30-35 billion10— is by itself in excess of what reasonably can be expected in the way of outlay cuts. A package of 10 This estimate assumes enactment of a bill by about July 1,1981 and changes in withholding tables by September 1, 1981. The cut would not be retroactive to January 1, 1981. cuts that yields significantly more than $10-15 billion for 1982 may not be possible. In general, a larger reduc tion of total outlays probably would require making ad ditional separate program changes rather than making each of the changes listed in this article more drastic. The program of changes that has already been outlined could occupy most of the time of the first session of the 97th Congress with debate and decisions on Fed eral spending and consequently could leave little time for consideration of major changes in taxes, Federal regulatory policies, or energy policy. (The tax-writing committees of the Congress are also the committees responsible for social security, medicare, unemploy ment compensation, and welfare.) Active consideration of more proposals in all likelihood would either be post poned or add to the overall confusion, making it more difficult to achieve any reductions. The Congress and the administration may have to settle for a longer range goal of offsetting the revenue losses from a tax cut by 1983. For 1983, the revenue loss from a one-time 10 percent cut in rates enacted in 1981 would be about $35 billion. The expenditure savings of $20 billion for 1983 in Table 4 represent estimates of the second-year effects of making per manent the program changes that reduce 1982 outlays. Additional changes that either start in 1983 or begin in 1982 but have no outlay effect in 1982 could probably reduce 1983 outlays by another $10-15 billion, yielding total reductions in 1983 of $30-35 billion. These could include cuts in contributions to international financial institutions, reductions of (or elimination of) future funding for the space shuttle, additional cuts in entitlements, and further reductions of Federal employ ment. These changes, together with those outlined in this article, could come close to offsetting the income tax cut by 1983 but would probably still fall short of off setting the $50-55 billion revenue loss from both a 1981 business tax cut and an individual income tax cut. The possibilities for offsetting the revenue loss from an individual income tax cut earlier than 1983 appear to be limited. A reduction of the defense real growth target might make a difference, but it would entail a major shift in policy. Swift enactment of comprehensive social security and medicare changes that affect all beneficiaries could contribute to an earlier income tax cut, but such changes do not appear likely because of various political considerations and also would be at variance with earlier policy pronouncements. A flood of changes in existing laws well beyond that envisioned in this article could possibly offset an income tax cut by 1982, although both political and time constraints make this extremely difficult. What seems to be clear, however, is that the process of reconsidering basic legislation would need to begin right away, regardless of whether enough changes can be enacted to affect total Federal outlays significantly in the immediate term. Otherwise, unless the Congress begins immediately to consider and to act on numerous changes in current laws, the chances for any spending cuts for 1982 may slip away and the opportunities for reducing spending growth in 1983 and beyond could be severely cir cumscribed. James R. Capra FRBNY Quarterly Review/Winter 1980-81 23 The business situation Current developments Chart 1 The 1980 recession is unique by postwar standards.* The advance in economic activity during the year before the downturn was very weak . . . Percent Percentage increase in real GNP in the year the cyclical peak “before — I I I IBBS IHi I JE 3 I 1948-4 919 53-54 1957-58 1960-61 1969-70 1973-75 1980 . . . and the decline started off sharply . . . Percent I Annual rates of decline in real GNP in the first §~ quarters of rece s s io n s --------------------------------------------- . . . but the economy seemed to turn quickly around. Percent || Annual rates o f change in real GNP two quarters 2 - a fte r the p e ak-------------------------------------------------------- ■ —4 --------------- --------— - . I It ------------------------------------------------------ I ~ ~ I I J ______ I......... 1948-49 1953-54 1957-58 1960-61 1969-70 1973-75 1980 *D a ta for periods preceding 1969 do not reflect latest revisions. Source: United States Department of Commerce, Bureau of Economic Analysis. 24 FRBNY Quarterly Review/Winter 1980-81 Economic activity advanced in both the third and fourth quarters follow ing the sharp second-quarter decline. The recovery surprised many observers, be cause it was earlier and more vigorous than had been expected. Industrial production rose more than 7 per cent from July to December, reversing about 80 per cent of the January-July decline. Construction activity also picked up substantially, and sizable gains were recorded in payroll employment. At the same tim e, however, there were indications that further substan tial growth in early 1981 is unlikely. Retail sales, in constant dollars, were essentially flat from July to November and then tailed off. Domestic auto sales were sluggish, and permits for housing construction began to decline at the end of the year, as interest rates reached the very high levels attained last spring. While the business indicators were mixed, strong infla tionary pressures persisted. The current business cycle has been very different from earlier ones (Chart 1). The 1980 recession was preceded by a year of weak econom ic growth, and the ensuing downturn in econom ic activity was very rapid. The 9.9 percent annual rate of decline of real gross national product (GNP) during the second quar ter— led by rapid declines in the auto and housing sectors— was the steepest on record in the postwar era and by far the largest decline at the start of a reces sion. A fter this abrupt slowdown, the pace of business activity picked up again in the third quarter. By most conventional measures, the 1980 recession should be one of the shortest on record. In terms of its over all magnitude, however, its peak-to-trough decline would be close to the average for previous postwar recessions. The pattern of the current business cycle has been reflected in the movements of both industrial produc- Chart 2 Chart 3 Although there have been significant gains in em ploym ent in recent months . . . Consumers have m aintained a higher savings rate . . . Millions of persons Percent P ayroll err ploym ent I I I I I J FMAMJ I I I I I I J A S O N D J 1-J 1 1 1 1 1 1 1 I I F M A M J J A S O N D 1 . . . and initial claim s for unemployment benefits have declined . . . Th ousands . . . and have not increased purchases. Index . . . the unemployment rate has rem ained well above 1979 levels. Percent Source: United States Department of Commerce, Bureau of Economic Analysis. Source: United States Department of Labor, Bureau of Labor Statistics. tion and payroll employment. Industrial output rose more than 7 percent from July to December, the larg est five-month gain since late 1975, when the economy was recovering from the last recession. The recent strength has been broadly based, stemming from ad vances in the production of consumer durables, inter mediate products, and basic materials. Reflecting the pickup in econom ic activity, employm ent has been rising in recent months, recovering about 95 percent of the February-July decline. But, with the total labor force expanding vigorously through November, the un employment rate has been in the 7.4 to 7.6 percent range for eight consecutive months (Chart 2). The domestic auto industry had a central role in the first months of the upturn. Sales were sluggish going into the recession, dropped further during the second quarter, and rebounded— although to still fairly low levels— in the third quarter. By and large, auto mobile sales fo r the new-model year have not met producers’ expectations. Dom estically produced autos sold at an annual rate of 6.5 m illion in the fourth quar ter, just m atching the third quarter’s sales pace. The new fuel-efficient models were expected to be very popular, but high prices and tight credit conditions still cloud the sales outlook. Uncertain sales prospects and high financing costs have encouraged dealers to maintain low inventories, and this could tem porarily delay sales even if demand strengthens. By historical standards, the housing recovery also has been less than vigorous. Nevertheless, the ad vance in housing starts— from a 900,000 unit annual rate in May to more than 1.5 m illion in October, No vember, and December— contributed substantially to econom ic growth in the fourth quarter. With interest FRBNY Quarterly R eview/W inter 1980-81 25 rates markedly higher at the year-end, however, build ing activity showed signs of turning down once more. Single-family housing starts and building permits reg istered modest declines, but the pace of multifamily starts increased, sustained by work on new apartment buildings to be occupied under Federal rent subsidies. Nevertheless, the total issuance of single- and multi family permits in December was more than 300,000 below September’s rate, signaling a likely reduction of housing activity in coming months. Weakening demands for new homes and autos fit into a broader picture of stagnant consumer spending. In constant dollars, retail sales were essentially un changed from July to November after rising by almost 4 percent from May to July. A number of factors make the outlook for consumer spending uncertain. The savings rate remains high in comparison to its level in late 1979 (Chart 3), suggesting that consumers have remained cautious. (Recent data revisions have raised the level of the savings rate but without changing its pattern in the 1979-80 period.) On the other hand, there is evidence that consumers are beginning to borrow once again after the precipitous decline that occurred during the credit restraint program. Outstanding con sumer credit grew by $3 billion from July to November, offsetting 40 percent of the March-July decline. 26 FRBNY Quarterly Review/Winter 1980-81 Despite the legacy of the sharp recession in the first half of 1980, inflation continued at a rapid rate. The consumer price index increased at a 12 percent annual rate in the September-November period, boosted by the volatile mortgage rate and food price compo nents. The latest statistics on wage increases also suggest strong inflationary pressure. Average hourly earnings in manufacturing, adjusted for interindustry shifts, rose at a strong 9 percent annual rate in the fourth quarter. However, this was considerably slower than the previous three quarters, suggesting that in flationary pressure, while still strong, may be easing somewhat. With the uncertain outlook for auto sales, and high interest rates causing housing to slow, it is unlikely that the recent advance in business activity— if it lasts — will have the vigor of the early stages of past expan sions. Given current economic conditions, it would take remarkable strength in several sectors of the economy to achieve a robust expansion, despite the prospects for some additional stimulus in the coming year from tax cuts and increased defense spending. Moreover, these sources of strength are likely to be offset, at least in part, by weakening export demand as the economies of other major industrial nations experience slowdowns. The financial markets Current developments Chart 1 After rising sharply for several months, interest rates declined in the final weeks of the year. Percent Percent Long-term rates F ive -ye a r G ove rnm ent s e c u ritie s * 6 A aa-rated c o rp o ra te bonds G overnm ent s e c u r i t i e s * -----(T he Bond B uyer in d e x ) ----- 1 J u j u J 11 1 1 1 i n 11 1 1 1 1 1 1 1 11 i i l n u l i 1 1 1 i . j l i l u u l j u l i - i . i i l F M A M J J A S O N D 1980 * These yields are adjusted to five-year and twenty-year maturities and exclude bonds with special estate tax privileges. Sources: Federal Reserve Bank of New York, Board of Governors of the Federal Reserve System, Moody’s Investors Service, The Bond Buyer, and Donoghue’s Money Fund Report of Holliston, Massachusetts. Financial markets tightened throughout most of the fourth quarter as interest rates approached and in many cases exceeded the record highs of last spring. But, in the final weeks of the year, interest rates de clined (Chart 1) as the m arket reacted to slow er money supply growth and indications that the economy would be weakening. For most of the fourth quarter, however, the financial markets were responding to news of a much stronger than expected economy and very rapid money supply growth. As a result, the three-m onth bill rate rose from about 11 percent at the end of September to 16.7 percent in the week of December 17. To keep the discount rate in line with other market rates and to restrain the rapid growth of money and credit, the Federal Reserve raised the discount rate tw ice during the fourth quarter to a level of 13 percent and imposed a surcharge on frequent borrowers from the discount w indow .1 Partly as a result of the strengthening in econom ic activity in the second half of the year, M-1B rose at a 10.5 percent annual rate, compared with 2.3 percent in the first half of the year. But the rapid growth of M-1B did not stem only from the strengthening in econom ic activity. A considerable volume of funds were shifted from savings deposits, which are not part of M-1B, into automatic transfer accounts (ATS) which are a component of M-1B; ATS accounts, negotiable order of withdrawal accounts (NOWs), and credit union share drafts com prise the “ other checkable deposit” component of M-1B. The inflow of funds from savings 1 The surcharge above the basic discount rate, which was set at 2 percentage points on November 17 and increased to 3 percentage points on Decem ber 5, applies only to borrow ings fo r "ad justm e nt purposes” by institutions with deposits of $500 m illion or more and is charged when such borrowings occur in two or more successive weeks in a calendar quarter or when borrow ings take place in more than four weeks in a calendar quarter. FRBNY Quarterly R eview /W inter 1980-81 27 accounts into ATS accounts during the latter half of 1980 was concentrated at comm ercial banks outside the New York, New Jersey, and New England geo graphic area. Apparently, these banks were prom oting ATS accounts aggressively in advance of the intro duction of nationwide NOW accounts fo r all financial institutions as of December 31 to solidify in advance market shares fo r interest-bearing checkable deposits. Furthermore, to the extent that these ATS accounts replace prospective NOW accounts, reserve require- Chart 3 While the demand for short-term business credit has increased sharply since July . . . Jan Chart 2 With short-term interest rates rising, the spread between m arket rates and the return on money m arket funds becam e positive after July . . . Percent 2------------------------------------------------------ . . . leading to a decline in the assets of money m arket funds . . . Billions of dollars 90 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1980 . . . the spread between the prime rate and the com m ercial paper rate has narrowed . . . Percent 8 ----------- Jan Feb Mar Apr May Jun Jul 1980 Aug Sep Oct Nov Dec . . . and a larger share of short-term financing has been provided by banks. Percent Percentage of short-term 1980 . . . and encouraging growth of six-month certificates. Billions of dollars nonfinancial commercial paper Sources: Federal Reserve Bank of New York, Board of Governors of the Federal Reserve System, and Donoghue’s Money Fund Report of Holliston, Massachusetts. 28for FRASER FRBNY Quarterly Review/Winter 1980-81 Digitized Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1980 Sources: Federal Reserve Bank of New York, Board of Governors of the Federal Reserve System. ments of these banks will be reduced for a while, since requirements on ATS accounts will be phased in from the previous ratio required against savings deposits (3 percent) to the higher level required against trans action accounts (12 percent for deposits in excess of $25 million). On the other hand, reserve requirements on NOW accounts outside New York, New Jersey, and New England will be at the level for transaction ac counts immediately. The rapid growth of other checkable deposits dur ing the second half of 1980 makes it difficult to inter pret the growth of the narrow monetary aggregates — M-1A and M-1B— relative to the annual targets for 1980 that were set by the Federal Open Market Com mittee (FOMC) last February. From the fourth quarter of 1979 to the fourth quarter of 1980, M-1B increased at an annual rate of 7.3 percent, about % percentage point above the 6.5 percent upper limit set for 1980, while M-1A advanced at a 5.0 percent rate, well within the 6.0 percent upper limit of its annual range. The upper limits of the M-1A and M-1B targets were set with a 1/2 percentage point spread, assuming only negligible effects from ATS and NOW accounts. As it turned out, however, the spread between the annual growth rates of M-1A and M-1B was 2.3 percentage points as the public shifted a larger than expected volume of funds from demand deposits and savings deposits into other checkable deposits. After allowing for these unexpected shifts, both M-1A and M-1B increased at rates roughly equal to the upper limits of their respective annual ranges. The broad aggre gates— M-2 and M-3— also showed strong growth during the second half of the year and recorded growth rates of 9.9 percent and 10.0 percent, respec tively, slightly above the 1980 targets. The introduction of nationwide NOW accounts on December 31, 1980 will continue to complicate the interpretation of the narrow monetary aggregates dur ing 1981. As funds shift from demand deposits into NOW accounts, the growth of M-1A will continue to be reduced. This particular shift will not affect M-1B growth because both components are included in the definition of M-1B. However, other funds, primarily savings deposits, will also be moved into NOW ac counts, which will add to M-1B growth in the same way that shifts of savings deposits into ATS accounts added to M-1B growth in 1980. As a result of funds shifting from demand and savings deposits into NOW accounts, M-1B growth is likely to be considerably stronger than M-1A growth during 1981. Reflecting the strong demand for money during the fourth quarter, short-term interest rates increased sharply. The rate on six-month Treasury bills rose from about 11.5 percent at the end of September to 15.7 percent in mid-December. As a result of the higher yields which were consequently available on six-month money market certificates, the public invested a large volume of funds in these certificates at banks and thrift institutions during the latter part of the year (Chart 2). At the same time, the average yield on the assets of money market mutual funds increased from a low of about 8 percent during the summer months to 16.6 percent at the year-end. However, since the yield on money market mutual funds rises only gradually as portions of the existing portfolio of money market in struments mature and are reinvested, the return tends to lag behind other market rates. Consequently, the assets of money market funds declined during the fourth quarter. In contrast, during a period of declining inter est rates, money market mutual funds should be more attractive than other market instruments because the yield would not fall so quickly. The rapid increase in interest rates occurred during a period of strong demand for short-term business credit. Commercial and industrial loans exclusive of bankers’ acceptances plus commercial paper issued by nonfinancial corporations expanded at an 18.7 percent annual rate between July and December (Chart 3). In part, this overall strength stems from the increased cost of borrowing in the long-term debt mar ket. During the fourth quarter, rates on long-term corporate bonds were as much as 3 1A percentage points above the trough level recorded last summer. Many corporations became reluctant to incur such high borrowing costs on a long-term basis and turned in stead to short-term sources of credit to meet their financing needs. As a result, new corporate bond offerings averaged only about $2.8 billion per month during the fourth quarter, compared with $5.6 billion during the spring and summer months. As corporations turned to the short-term market, the interest rate spread between commercial paper and the prime rate was an important factor contributing to the composition of their short-term borrowing. Even though the prime rate rose to a record 2 1 1/2 percent by the end of December, commercial paper rates rose by a greater amount and, as a result, corporations raised a larger proportion of their short-term credit needs at banks rather than in the commercial paper market. Another factor contributing to greater reliance on bank loans was the downgrading of several large issuers of commercial paper. FRBNY Quarterly Review/Winter 1980-81 29 Global Payments Problems The Outlook for 1981 Since late in 1978, oil prices have risen sharply and the major oil-exporting countries have again amassed large financial surpluses. Correspondingly large deficits have been contracted by oil-consuming nations. With political tensions in the world’s major oil-producing region at a new high, the questions of how large these deficits and surpluses may become and how long they can persist have taken on renewed urgency. This article reviews recent developments and considers the 1981 outlook for international payments of the Organi zation of Petroleum Exporting Countries (OPEC) mem bers and the non-OPEC developing countries.1 The in dustrial country members of the Organization for Eco nomic Cooperation and Development (OECD) also face serious problems. But, except for a couple of the least developed OECD members, the central problem is to reduce oil demand with minimum adverse effects on employment and inflation rather than how to finance the oil imports. Less developed countries (LDCs), too, must adjust to higher oil prices, but these adjustments at best take time. With their often limited capacity to adjust and limited sources of external funds, many LDCs find that their external financing constraint quickly binds. The forced adjustment that then results tends to be more costly than necessary. The combined current account of members of OPEC grew from near balance in 1978 to a surplus of over $110 billion by 1980 (Chart 1). Most of this increase i The definition of non-OPEC developing countries used here excludes southern Europe, China, and South Africa. In discussions of gross bank finance and gross oil trade, data for offshore financial centers and offshore refining centers are excluded as well. 30 forFRBNY Digitized FRASERQuarterly Review/Winter 1980-81 was against the OECD member countries. The OECD accounts for nearly 90 percent of the world’s oil im ports, and by 1980 its aggregate current account had deteriorated to an estimated deficit of over $75 billion from a surplus of under $10 billion in 1978. Meanwhile, the combined deficit of non-OPEC de veloping countries widened from about $25 billion to over $50 billion. The outlook for 1981 is critically dependent on very uncertain oil prices, so that two different price sce narios are considered. If the recent Iran-lraq supply interruptions are overcome early in 1981, an in crease in the oil price at least in line with inflation in industrial countries is likely. A 12 percent OPEC oil price increase on average from 1980, along with a continued rapid rise in their imports would reduce the OPEC current account surplus to about $80 billion. The economic slowdown in major industrial countries is expected to continue, and this would lower the com bined OECD deficit some $30 billion to around $45 billion. The deficit of the non-OPEC developing coun tries, on the other hand, would be expected to rise nearly $10 billion to around $60 billion, as prices of their primary commodity exports stagnate in the face of weaker demand in the industrial world. Even this scenario assumes that developing countries maintain a tight check on real import growth as their deficits are constrained by the availability of finance. A higher price of oil in 1981 would result in a larger OPEC surplus and a larger OECD deficit. For instance, if the average oil price received by OPEC rises 25 percent to $40 per barrel for the year, their surplus would again exceed $100 billion. Most of this increase Chart 1 Global Current Account Balances Billions of dollars 120 -20 -4 0 Non-OPEC developing countries t 40 60 80 1973 1974 1975 1976 1977 1978 1979 1 9 80 1981a 1981b Estimate Projections § Members of the Organization of Petroleum Exporting Countries (OPEC). M em bers o f the O rg anizatio n fo r E conom ic C oop era tion and D evelo pm e nt (OECD). Excludes eastern and southern European countries, China, and South Africa. Projection 1981a assumes a $ 3 5 per barrel average OPEC contract price for oil, while the 1981b projection assumes a $ 4 0 per barrel average price. Sources: OECD, International Monetary Fund (IMF); Federal Reserve Bank of New York estim ates, projections, and adjustments for consistent country coverage. again would be reflected in the deficit of the OECD area. As a group, these countries would not have a serious financial constraint so long as they continue to attract the bulk of OPEC investments. Non-OPEC developing countries, a few of which would gain from an oil price rise, would face a further $5 billion erosion in their current account, bringing it to near $65 billion. The oil-im porting countries in this group would have to cover a $10 billion higher oil bill. Be cause oil imports are concentrated in a few countries, the higher oil price could present serious financing problems for individual countries even though it pro duces only a small increase in the combined deficit. In this context, it has to be kept in mind that payments interruptions by one or more of the m ajor debtor coun tries could raise the cost of borrowing for all and com pound the adjustm ent problem fo r others. OPEC The combined OPEC current account surplus is now estimated at about $110 billion in 1980, up from only about $5 billion two years earlier (Chart 2). The group’s annual export receipts more than doubled to over $300 billion during this period, as the 140 percent surge in oil prices dominated a 10 percent decline in oil production and export volume. But, by 1980, more than a third of the $150 billion increase in export revenue was being spent on current im port and trans fer payments abroad. These payments responded slow ly at first to the rising oil receipts. Most OPEC members entered 1979 with relatively austere plans for econom ic developm ent and imports. Their emerg ing fiscal and balance-of-payments deficits between 1976 and 1978 led most OPEC countries to cut back their import-intensive government spending plans. How ever, by early 1980, OPEC real im port growth once again appears to have been in excess of 20 per cent per year. As a result, merchandise imports are estimated to have risen to about $140 billion in 1980, nearly $40 billion above their 1978 level. Moreover, the OPEC deficit on net services and transfer payments has risen about $10 billion since 1978 to more than $50 billion despite growing earnings on OPEC invest ments abroad. Real OPEC imports are likely to remain strong as Iraq and Iran reconstruct war damage, or at least rearm. Moreover, the heightened political tensions w ill likely increase arms purchases elsewhere in the region. If oil prices remain about constant in real terms (a 12 percent nominal year-over-year growth), this continued rise in imports would reduce the 1981 OPEC surplus to around $80 billion. On the other hand, a 25 percent increase in oil prices to around $40 per barrel for the year average would again produce a surplus in excess of $100 billion. The conditions under which a higher oil price might occur are not implausible. The Iraq-lran w ar has, as of this w riting, driven spot prices to the $40 per bar rel range and led OPEC to announce increases in th eir posted prices to an average of about $35 per barrel. However, the OPEC price structure remains split. Under announced plans, Saudi Arabian prices FRBNY Q uarterly R eview /W inter 1980-81 31 still remain below those of equivalent grade oil from other Middle Eastern producers w hile prices for higher quality African crude oils remain well above their traditional premia. These price differentials seem un likely to be sustained, but it is unclear whether market forces w ill dictate cuts in the premium prices being charged by Algeria, Libya, and Nigeria or induce Saudi Arabia and other price moderates to recon sider their discounts. The lower price scenario is consistent with an early return of Iran and Iraq pro duction and exports to near th eir prewar levels, and no additional supply disruptions elsewhere. With slug gish activity in industrial countries, this would return the world oil m arket to the oversupply situation that was apparent in the third quarter of 1980. Then, inventories had reached record levels and spot prices were falling, even though OPEC production had declined more than 10 percent from its level a year earlier. This outcome would allow a consolidation of OPEC prices around the $35 per barrel level. Production cutbacks by the high surplus Arabian Gulf producers in line with their longer term plans would prevent a further price ero sion. On the other hand, a prolongation of hostilities, a spread of the war, or new political disruptions in Chart 3 Chart 2 Disposition of OPEC S u rp lu s* OPEC Current Account Billions of dollars Billions of dollars 35 0 Merchandise exports 300 -5 0 — 1 00 - M e rc h a n d is e im ports -150 I 200 Trade balance 150 100 50 ■ M ~50 - n * The total surplus available for disposition equals the OPEC balance on goods, services, and private transfers plus borrowings by OPEC members plus adjustments for leads and lags of oil-export receipts. Net services and transfers -1 0 0 1 150 t Increase in liabilities of banks in industrial countries as reported to the Bank for International Settlements (BIS). Current account 100 50 QI— 1973 f Includes direct investment, loans, portfolio investment, and unrecorded items. 1974 1975 1976 1977 1978 1979 1980 Estimate Sources: IMF; Federal Reserve Bank of New York estim ates and adjustments for country coverage. 32 forFRBNY Digitized FRASERQuarterly R eview /W inter 1980-81 § Excludes purchases of World Bank bonds in international capital markets. Sources: IMF, OECD, BIS; Federal Reserve Bank of New York estimates and adjustments for country coverage. other m ajor oil-producing countries could tighten the oil market substantially and produce another run-up in oil prices. While one can easily imagine even higher oil price projections based on worsening political scenarios fo r the M iddle East, the $40 per barrel oil price assumption provides the flavor of th eir impact. In investing its surplus, OPEC continues to favor low -risk investments, particula rly government securi ties of m ajor countries and deposits in large inter national banks (Chart 3). The effect has been to shift the job of lending to most oil-im porting countries— including developing countries— over to banks and other participants in the w orld capital markets. At least three quarters of the available OPEC surplus in 1979 and 1980 was invested in industrial countries or in Eurocurrency deposits of banks from these coun tries, and the banks alone have taken about half the surplus. The remaining quarter includes direct credits to developing countries, indirect funding through m ulti national organizations, and unrecorded items. Direct and indirect assistance to other developing countries has not grown in real terms since 1974 and has fallen far short of the growth of the OPEC surplus in the last two years. Under the lower oil price scenario for 1981, the level of OPEC lending to LDCs would increase little from the $10-12 billion estimated for 1980. In the past during periods of declining surplus, such lend ing has fallen back although with a lag. Also, as in the past, much of this lending would be in the form of concessional loans and would follow the political ties of the high surplus OPEC members with Middle Eastern and North African countries. Increased OPEC investments at market-related terms may be a ntici pated. However, these investments probably would compete with bank lending in those few more ad vanced developing countries that are adjusting well to the oil shock, rather than complement bank lending in countries where adjustm ent proves more difficult. The main difference under the alternative higher surplus scenario would be increased bank placements. Some increase in direct LDC assistance and the fund ing of m ultilateral institutions might also be possible, but the past growth and direction of these flows suggest they would not compensate fo r the additional oil cost to many LDCs w ithout a serious effort to augment official recycling. Non-OPEC developing countries The non-OPEC developing country current account d eficit mounted to over $50 billion in 1980, more than double its level two years earlier. This deterioration was nearly equal to the $35 billion growth of the annual oil-im port bill of the group over the period (Chart 4). Chart 4 Non-OPEC Developing Countries’ Trade Account Billions of dollars 2 5 0 ------------------------------------------------------------------------Merchandise exports 200 -------------------------------------------------------------------Oil 150 100 50 H 0 -5 0 -100 -1 50 -------------------------------------- ■200 ---- ---------------------------—25 0 - M erchandise imports -3 0 0 50 Trade balance 0 m I -5 0 1973 I 1974 I 1975 I 1976 1977 " 1978 l 1979 1980 Estimate Sources: IMF, OECD; Federal Reserve Bank of New York estimates and adjustments for country coverage. But the direct im pact of higher oil prices on the developing countries is very uneven. Four coun tries— Brazil, India, Korea, and Taiwan— account for nearly half of the group’s oil-im port bill. These four also accounted for about half of the deterioration of the deficit. Many sm aller countries with less export or borrowing potential have been even more seriously affected in proportion to their own income and output. At the other extreme, those developing country oil exporters that are not members of OPEC showed about a $15 billion increase in net oil receipts over the 1978-80 period.2 These countries have expanded their oil production nearly 25 percent since 1978, but their dom estic oil consumption and nonoil imports have also grown. As a result, they showed only a modest $2 b illion improvement in th eir current account deficit by 1980. 2 The major non-OPEC developing country oil exporters are Mexico, Oman, Trinidad and Tobago, Egypt, Malaysia, Angola, Bahrain, Peru, Syria, and Tunisia. FRBNY Quarterly R eview /W inter 1980-81 33 On top of their higher oil-im port bill, developing country exports have suffered from weakening de mand in th e ir markets in industrial countries. The slowdown in real gross national product (GNP) growth of the industrial countries during 1979-80 cut 1980 developing country exports $10 billion to $15 billion below what they would have been. Moreover, the full im pact of this slowdown has not yet been felt. Pri mary comm odities prices were relatively strong for LDC exporters until just recently and rose about 35 percent over the past two years. But the increases were concentrated in a few products— sugar, copper, tin, and rubber— and benefited only some countries. Many developing countries also import prim ary com modities, p articularly foods, and have been hurt by the nearly 40 percent rise in grain prices. For 1981, the non-OPEC developing country cur rent account d eficit is projected to widen to about $60 billion, if oil prices remain constant in real terms. The further slowing in industrial country growth and the weaker com m odities prices w ill further re duce the growth of export receipts. Thus, most of the deterioration w ill be reflected in the trade account, even if real im port growth is again held to about 3 percent. The outlook for comm odities prices is mixed. Most prices have been falling since the third quarter of 1980, and only grains appear to have much potential for a strong 1981 performance. As a result, the terms of trade for developing countries is pro jected to deteriorate about 2 percent. Moreover, the relatively strong grain prices w ill help only a few and hurt the low income food-im porting countries who may least be able to finance larger deficits. A run-up in oil prices to $40 per barrel would add another $5 billion to the combined developing country deficit in 1981, widening it to around $65 billion. But the $10 billion addition to the oil bill that this price brings would again be concentrated in a few oil-dependent, newly industrialized countries. Moreover, those least developed countries where oil import payments al ready consume a heavy share of export receipts would be forced to cut real imports and th eir eco nomic growth further. Some may not have the option of running a larger deficit. Most of these countries w ill not be helped by the $7 billion increase in receipts that would accrue to the few LDC oil exporters that are not OPEC members as the real oil price rises. The indirect effects of the price hike could add perhaps $1-2 billion to the 1981 deficit, after allowing for a pickup in LDC exports to OPEC members. On the basis of past experience, however, these indirect effects work slowly. Thus, the further slowing of activity in industrial countries brought by higher oil prices would depress developing country exports well into 1982 Digitized FRASER Quarterly R eview/W inter 1980-81 34for FRBNY when even larger LDC deficits would be expected. The above projections for developing country defi cits assume they can be financed. Past and emerging financing trends provide a guide as to how this m ight Chart 5 Financing Non-OPEC Developing Countries Sources of Finance: Major Components * Billions of dollars 50 Private sources 40 Bank lending^ Direct investment 30 20 10 ■ LL 0 40 - O fficial sources 30 — Reserve related 20 - Other bilateral and multilateral credits 10oL lJLi 1973 1974 1975 1976 1977 1978 1979 1980 Estimate Uses of Finance: Major C om ponents* Billions of dollars 60Current account de ficit 20 - Growth of official reserves 1979 1980 Estimate Sources exclude suppliers’ credits and bonds which are more than offset by growth of nonreserve assets on the uses side and by errors and omissions. t Growth of claims of banks in industrial countries as reported to the BIS. f Includes allocations of special drawing rights. Sources: IMF, OECD, and BIS; Federal Reserve Bank of New York estimates and adjustments for country coverage. be accomplished (Chart 5). In the past, bank lending3 has been the major source of finance as well as the source most responsive to changes in LDC deficits. But this bank lending has been concentrated in a few of the more advanced non-OPEC developing countries. Just ten countries4 account for nearly 75 percent of outstand ing international bank credits, and since 1978 four of these countries— Brazil, Mexico, Argentina, and South Korea— have received two thirds of the net new bank lending to the group of more than 100 individual coun tries. The remaining developing countries have relied heavily on official source credits to finance their defi cits. Except for reserve-related lending, mostly from the International Monetary Fund (IMF), these credits grew only slowly during the 1974-75 and 1978-79 peri ods of rising LDC deficits. Bilateral (government-togovernment) lending usually requires legislative ap proval in industrial democracies, and developing country finance often takes low priority in times of economic contraction at home. Multilateral loans and credits (from the World Bank and regional develop ment banks) are linked mostly to project finance and are disbursed only as these projects progress. Financing the $60 billion 1981 deficit anticipated for non-OPEC developing countries, if real oil prices re main constant, does not appear unsurmountable. Prob lems for individual countries doubtlessly would remain, and there would be little room for reserve asset ac cumulation for the group as a whole. However, a rela tively modest growth of official finance and direct investment, along with continued bank lending at its recent rate, would cover the overall deficit. Official source credits should continue to grow, principally 3 Bank lending is defined here to comprise the total increase in claims on non-OPEC developing countries of banks in industrial countries, as reported to the Bank for International Settlements. This is a lending-net-of-repayments concept which includes short-term credits and loans to the private sectors of developing countries which may not carry government guarantees. 4 The ten major non-OPEC developing country debtors to commercial banks are Argentina, Brazil, Chile, Colombia, Korea, Mexico, Peru, Philippines, Taiwan, and Thailand. because of stepped-up IMF and World Bank lending. Official financing is estimated to have grown from $14 billion in 1979 to near $20 billion in 1980. An increase to around $25 billion in 1981 appears reasonable. Recent increases in IMF quotas and guidelines on maximum lending to individual countries, as well as stepped-up disbursements on World Bank project and structural adjustment loans, should make up a good part of this increase. Private source credits would still have to provide nearly $40 billion of the financing under this scenario, mostly in the form of bank lend ing. The growth of bank claims could be somewhat less than the $36 billion reported in 1979 and about in line with the increase now estimated for 1980. This would represent about a 20 percent growth of bank claims on non-OPEC developing countries, somewhat below the average growth rate since 1975. The $40 per barrel oil price scenario calls for only a $5 billion larger combined deficit, but little additional official lending can confidently be expected. An addi tional $5 billion in bank loans might not be out of the question, particularly if the lending spreads were to widen. However, some of the countries that would be hardest hit by the $10 billion rise in the LDC oil bill may already have stretched their borrowing capacity to the limit. Domestic political constraints may make it impossible for them to reduce real imports enough to avoid payments interruptions. Interruptions in trade credit or debt service payments would not entail a broad or permanent default on existing loans. Interruptions, however, would lead to difficult and possibly prolonged periods of negotiation to restructure the debt and re establish credit. During these periods, new credit to the country concerned would be sharply curtailed. Forced import cuts would then reduce the current account deficit to meet available finance. If these interruptions arise in a couple of the countries that account for most bank credits, a drop in the overall rate of bank lending and in the overall deficit is possible. In any event, the increasing incidence of problems in individual de veloping countries could cause a retrenchment of bank lending in general and aggravate the adjustment problems in otherwise sound countries. William J. Gasser FRBNY Quarterly Review/Winter 1980-81 35 Oil Price Decontrol and Beyond Price controls on United States domestically produced crude oil are currently being eliminated,* marking an important step toward resolving our energy problem. Oil price decontrol is one key part of a broader na tional initiative, which includes decontrol of natural gas prices, encouragement of alternative energy sources, and incentives for greater conservation and efficiency. The main purpose of these efforts is to reduce our dependence on increasingly costly and uncertain supplies of foreign oil. Decontrol of domestic oil will have several impor tant effects. First, by October 1981, when decontrol is scheduled to be completed, United States refined petroleum prices will be at least 20 to 30 cents per gallon higher than they would be without decontrol. Second, a price rise of this magnitude should lower United States petroleum usage about 1 million bar rels per day. Third, since higher prices appear to have stimulated United States crude oil production, the total impact of decontrol on United States imports is probably greater than 1 million barrels daily. Fourth, by raising the responsiveness of our oil imports to foreign prices, dropping the controls mechanism raises United States resistance to future foreign price increases. If the completion date for decontrol is moved to earlier in 1981, its full effects would come sooner and more abruptly but in other respects would be basically the same as those outlined here. There is, however, good reason to believe that the mere decontrol of domestic crude oil prices does not * This article was written prior to President Reagan’s recent announce ment immediately ending all price controls on crude oil and petroleum products. As noted in the text, this does not substantially change our conclusions. FRBNY Quarterly Review/Winter 1980-81 Digitized36for FRASER go far enough. Because of the potentially devastating effects of petroleum supply disruptions on the United States economy, the cost of imported oil clearly ex ceeds its dollar price. Further steps beyond decontrol, therefore, are called for to discourage imports. Tax policies which effectively raise the relative price of petroleum in the United States would be a logical ex tension of the decontrol strategy. The price-control mechanism Before examining the implications of decontrol, it is helpful to review the basic elements of the pricecontrol system which is being phased out. United States crude oil price ceilings originated in the gen eral wage-price restraints of the early 1970s, but the basic form of the current controls evolved from the Energy Policy and Conservation Act of 1975.1 Do mestically produced crude oil was divided into two main categories, essentially based on the age and productivity of wells. Oil from older wells, labeled “lower tier” oil, was given a price ceiling below the ceiling for "upper tier” oil, which was produced from newer wells (or stepped-up output from older wells). In 1976, production from small “stripper” wells— wells producing under ten barrels daily— was decontrolled.2 The ceilings kept the average price of domestic oil below the cost of imported oil. Without controls, re1 For a description of how Federal petroleum regulations evolved since the 1930s, see Paul A. MacAvoy, ed„ F e d e ra l E n e rg y A d m in is tra tio n R e g u la tio n , American Enterprise Institute for Policy Research (Washington, D.C., 1977). 2 Oil from the Naval Petroleum Reserve, which has never accounted for more than 1.6 percent of total domestic production, was also exempt from price regulations. finers would be willing to pay a similar price, including transportation costs, for crude oil from both foreign and domestic suppliers. For example, in the fourth quarter of 1978 the average price of foreign oil de livered to United States refiners was $14.77 per barrel. Stripper oil, which accounted for 15 percent of do mestic United States production, received an uncon trolled price of $14.54 per barrel, close to the import price. Due to wellhead ceilings, however, the 35 per cent of United States output classified as lower tier received only $6.14 per barrel. Upper tier oil (exclud ing Alaskan) was priced at $13.00 per barrel and ac counted for 35 percent of domestic oil. Alaskan North Slope oil, which at the time made up 14 percent of United States output, also was technically subject to the upper tier wellhead ceiling but, due to high trans portation costs, actually received a wellhead price less than the ceiling in order to stay competitive with uncontrolled oil from other sources.3 For all domestic oil, the combined average cost to refiners, including transportation, was $10.88 per barrel, well below the average import price. The price ceilings were pegged to the implicit gross national product (GNP) price deflator. Although this has permitted the ceilings to keep pace with the infla tion rate, foreign oil prices since the end of 1978 have risen much more than the general price level, causing the gap between the domestic ceilings and the price of imports to widen. Merely holding down domestic crude oil prices, however, would not guarantee that prices paid by consumers of refined products would be lower. The domestic price level for refined petroleum must be high enough to make it profitable to refine and mar ket not only price-controlled oil but also oil from every other source needed to satisfy total domestic demand, including expensive foreign supplies. Thus, if left alone, refined products prices would reflect the high cost of foreign oil. Refiners of imported oil would cover their costs, and refiners with access to pricecontrolled oil would be in a very profitable situation. To make sure that United States refined petroleum prices were indeed lower, and to remedy the poten tial inequities among refiners, an import subsidy was enacted as part of a system of crude oil “entitle ments” to complement the crude oil price controls.4 3 At the wellhead, Alaskan North Slope oil received an average price of $5.22 per barrel in 1978, compared with $12.15 per barrel for other upper tier oil. At the refinery gate, Alaskan oil generally re ceived at least as much as other upper tier, and often more. 4 In addition, until mid-1976, prices of most refined products were controlled directly. Currently, gasoline is the only major refined product category subject to direct price controls, but these controls apparently are not effectively binding much of the time. Under the entitlements program, refiners of pricecontrolled crude oil pay a uniform per-barrel subsidy to refiners of imported and uncontrolled domestic crude oil. The subsidy lowers the effective cost of refining foreign or uncontrolled domestic oil, thereby lowering the price level of refined products. At the same time, the required payment raises the average effective cost of refining price-controlled oil, making it approximately equal, on balance, to the average effective cost of foreign and uncontrolled domestic oil.5 The average effective cost of all oil to United States refiners, therefore, is below the price of im ported crude. Since the entitlements payments roughly equalize the average effective cost of imported and pricecontrolled crude oil, the size of the subsidy auto matically rises when the import price increases rela tive to domestic price ceilings.6 For example, between December 1978 and May 1980, the average delivered price of imported crude oil to refiners rose from $14.94 per barrel to $34.33 per barrel. Over this period, the lower tier wellhead price ceiling increased only from $5.68 to $6.47, reflecting general price inflation. As a result, the import subsidy, which was $1.27 in Decem ber 1978, jumped to $6.22 per barrel by May 1980. Prices of United States refined products can rise faster than average effective crude oil costs during a tight world market. Since the import subsidy is paid on a uniform per-barrel basis, it does not always fully offset the higher crude oil costs paid by those refiners who are forced to seek supplies from particularly ex pensive foreign sources. Even if most officially posted foreign contract prices remain unchanged during a world shortage, refiners without sufficient contractual supplies may find it unprofitable to turn to more ex pensive sources unless United States refined product prices rise. Under these circumstances, if the extra supplies are needed to meet domestic demand, refined product prices in the United States can rise consider ably, even though the average effective cost of all for eign oil increases much less. The spread between re fined products prices and average effective crude oil costs therefore rises. Conversely, during a glut on the world market, refiners can buy oil for less than the 5 The required payment per barrel of upper tier crude oil is less than the payment per barrel of lower tier oil by just enough to equalize the effective costs of these two categories of crude oil. For a more detailed description of the system, see Kay Sherwood, “Crude Oil Entitlements Program", M o n th ly E n e rg y R e v ie w (January 1977). ‘ More exactly, the entitlements system approximately equalizes the average effective cost of price-controlled oil and the average com bined effective cost of imported and uncontrolled domestic crude oil. Most of the time, however, market forces cause the price of uncontrolled domestic oil to be about the same as the import price. FRBNY Quarterly Review/Winter 1980-81 37 long-term contract prices and, receiving the same perbarrel subsidy as other importers, can force down United States refined products prices relative to the average effective cost of all imported oil. Over the long run, United States refined products prices would be held below the level consistent with average imported oil prices by an amount corresponding to the import subsidy, but in the short run the spread between United States refined products prices and average ef fective crude oil costs can fluctuate in response to shortages or gluts on the world market. For example, suppose the price charged for most imported oil was $30 per barrel but, for domestic de mand to be satisfied, some oil would have to be im ported at $40 per barrel. Unless domestic refined products prices were high enough to make importing the more expensive oil profitable, refiners would not buy it, and the resulting shortage would drive up the price of refined petroleum in the United States until it reflected the $40 per barrel cost less the uniform subsidy. Since the bulk of crude oil was still being bought at controlled domestic prices or at the $30 per barrel import price, the spread between refined prod ucts prices and average effective crude oil costs would widen as well. If, however, supplies of $30 oil subse quently became more abundant, the price of refined products would drop to reflect an effective crude oil cost of only $30 per barrel less the entitlements sub sidy, and the spread between refined products prices and average effective crude oil costs would narrow again. The decontrol process The current process of phasing out all crude oil price ceilings began in June 1979 and is scheduled for com pletion in October 1981. In the month before decontrol started, 83 percent of all domestic production was sub ject to price ceilings— 34 percent lower tier and 49 percent upper tier (including Alaskan). During the phaseout period, lower tier is being gradually reclas sified as upper tier while, simultaneously, upper tier is being gradually freed of price controls entirely. In addition, oil with a high sulfur content, newly dis covered oil, and oil that is difficult and costly to re cover are now free of price ceilings.7 By the middle of 1980, the proportion of domestic output subject to price ceilings was down to 47 percent (15 percent lower tier and 32 percent upper tier). During the first year of decontrol, therefore, the proportion of total domestic 7 For definitions of these new categories of uncontrolled oil, see United States Department of Energy, M o n th ly E n e rg y R e v ie w (September 1980), pages 76, 96-97. 38for FRBNY Digitized FRASER Quarterly Review/Winter 1980-81 output free of price controls rose from 17 percent to 53 percent.8 As crude oil price ceilings are eliminated, the re leased domestic oil receives a price comparable to foreign oil prices. Consequently, the value of the en titlements payments, which equalize average effective foreign and domestic crude oil costs, will fall automat ically to zero as decontrol approaches completion. As the entitlements subsidy on imports disappears, the effective cost of crude oil going into United States re fined products prices will rise to the price of imported oil.9 How much lower would effective crude oil costs and refined petroleum prices have been without de control? This depends on how large the import sub sidy would have been had controls been continued. Suppose, for example, that the delivered price of imported oil, which was $34.48 per barrel in June 1980, reaches just $35 per barrel by October 1981. In this case, under plausible assumptions regarding the path of the continued controls mechanism, by October 1981 the import subsidy would have reached $8 per barrel, or 19 cents per gallon.10 More plausibly perhaps, an October 1981 import price of $39 per barrel would re sult in an import subsidy of $10 per barrel (24 cents per gallon), while a $44 price would imply a $12 per barrel (29 cents per gallon) subsidy. Depending on foreign prices, therefore, by October 1981 the effective cost of crude oil going into United States refined prod ucts would be around 20 to 30 cents per gallon higher than without decontrol. Approximately the same figure 8 Due to high transportation costs, however, the upper tier ceiling on Alaskan North Slope output (15 percent of the domestic total in May 1979) became an effective constraint on wellhead prices only after decontrol had already begun. This reflected the sharp rise in the world market price. ’ The size of the entitlements subsidy can be expressed as the product of (a ) an appropriately weighted sum of lower and upper tier oil as a fraction of all oil refined and (b ) the difference between the average price of all imported and uncontrolled domestic oil eligible for the subsidy and the lower tier price ceiling. As price ceiling coverage is phased out, term (a ) becomes zero, eliminating the subsidy. As noted earlier in this article, however, the rise in world prices during the first part of the decontrol process caused an increase in term (b ) sufficient to produce a temporary rise in the subsidy. Without the phaseout of coverage, of course, this rise would have been larger (and not temporary). 10 Without decontrol, the October 1981 category shares were projected as lower tier, 24 percent; upper tier (excluding Alaskan), 43.5 percent; Alaskan, 16 percent; and uncontrolled, 16.5 percent. Price ceilings and transportation costs were projected to rise at a 10 percent annual rate, and imports were projected to account for 45 percent of the crude oil used in the United States. applies to refined petroleum prices.11 Over the 29month phaseout period (June 1979 through October 1981), therefore, decontrol will have added roughly a penny per month to United States petroleum prices. The actual path of refined products prices since the start of decontrol has differed somewhat from the path of average effective crude oil costs, but this has been mainly due to the successive tightening and loosening of the world market during this period. The Iranian production cutoff at the beginning of 1979 sent spot market prices soaring. Some exporting nations raised prices considerably higher than others, and the price of uncontrolled domestic oil in the United States was bid above the average import price.12 Thus, as crude oil prices from certain key sources rose considerably more than the overall average, the price of United States refined petroleum rose more than the average effective cost of crude oil from all sources together. The spread between refined products prices and aver age effective crude oil costs had already widened by June 1979 when decontrol began, but it continued to increase as the world market remained tight through early 1980 (Chart 1). By summer 1980, spot prices had fallen off and domestic uncontrolled oil had come back into line with average import prices, reflecting a loos ening of the world market.13 As a result, the spread between United States refined products prices and av erage effective crude oil costs narrowed again. During all this period, however, decontrol was making the import subsidy smaller than it otherwise would have been. This in turn raised the effective cost of crude oil from every source, thereby increasing United States refined petroleum prices above what they would have been without decontrol. 11 In the very short run, any reduced usage of petroleum in response to higher prices may lower the profitability of refinery and dis tribution operations, reflecting competition fo ra smaller total amount of business. In the longer run, however, refining and marketing capacity will not be replaced unless the return on such investments justifies the capital costs. Ultimately, therefore, the final products prices will reflect the whole higher cost of crude oil plus the neces sary capital and operating expenses of refining and distributing it. 12 Late in 1978, Libyan and Algerian oils were priced about 10 percent above Saudi Arabian light crude oil, but by the middle of 1979 the price differential had widened to around 30 percent. See Department of Energy, W e e k ly P e tro le u m S ta tu s R e p o rt (August 1, 1980), page 39. In December 1978 the average price, including transportation, of United States stripper oil was $14.57 per barrel, close to the average comparable import price of $14.92. By December 1979, however, stripper oil was selling for $33.43 per barrel, while the average price of imported oil was $28.91 per barrel. 13 By July 1980 the average price, including transportation, of stripper oil was $34.45 per barrel, just under the average import price o f $34.51. The effect on imports Even casual observation confirms that higher prices reduce United States petroleum use. After the first major oil price hike in 1973-74, the rate of growth of United States petroleum consumption slowed dramat ically to 1.7 percent annually during 1973-78, com pared with 4.7 percent over 1949-73. During 1978-80, total consumption declined at a 5 percent annual rate.14 Moreover, the ratio of petroleum use to GNP was 17.2 percent lower in the first three quarters of 1980 than its 1973 level.15 These observations are supported by a statistical analysis of the relationship over time between United States petroleum prices and consumption. The results show that, holding GNP constant, a 10 percent rise in the wholesale price of United States petroleum prod ucts is on average associated with roughly a 2 per cent fall in total usage.16 For example, in the scenario described above with the price of imports reaching $39 per barrel by October 1981, the impact of decon trol on United States products prices (assuming a penny-for-penny pass-through of crude oil costs) is calculated as 24 cents per gallon, which amounts to a 28 percent rise at the wholesale level. This, in turn, should result in a fall of between 5.0 and 8.5 percent 14 Total United States consumption is measured as deliveries of petroleum products from primary storage. The figure for 1978-80 is based on a comparison of the first nine months of 1978 and 1980. Sources: Department of Energy, Energy Information Administration, A n n u a l R e p o rt to C o n g re s s 19 79, Volume Two, page 43, and M o n th ly E n e rg y R e v ie w (June 1980 and December 1980). 15 The ratio of petroleum deliveries (thousands of barrels daily) to real GNP (billions of 1972 dollars) was 13.79 in 1973 and 11.42 over the first three quarters of 1980. 14 Over the period 1975-1 to 1980-11, an ordinary least squares regression was performed, with the following result: C = 1.83 — 0.14P + 0.57Y + 0 .4 1 C (— 1) _ (2.1) (3.2) (2.7) (2.0) R* = 0.86; D.W. = 1.63; SEE = 0.02 C is total petroleum consumption, P is a wholesale price index of refined petroleum products, deflated by the GNP implicit price deflator, and Y is real GNP, all in logarithmic form. The t statistics are in parentheses. The coefficients of the price and income variables are the respective short-run elasticities. The long-run price and income elasticities are — 0.23 and 0.96, respectively, with 87 percent of the effect o f movements in price and GNP on consumption occurring within two quarters and 95 percent within three quarters. The lag structure is admittedly crude. Experimentation with alternatives failed to find a lag structure that was robust with respect to its specification. However, the total effect of price on consumption proved virtually unchanged under the alternative specifications. An autocorrelation correction was performed to check for the possible bias in the D.W. statistic imposed by the presence of the lagged dependent variable, but this caused essentially no change in the coefficients. A statistical appendix, containing sectorally disaggregated estimation results, as well as alternative estimation procedures, is available from the authors. The various methods yield similar results. FRBNY Quarterly Review/Winter 1980-81 39 • barrels daily, equal to 18 percent of the level of imports in August 1980. This, moreover, understates the total effect on imports because United States petroleum output also depends on price. With newly discovered oil now allowed to receive an uncontrolled price, d rill ing activity has stepped up considerably.18 C hart 1 United States Crude Oil Costs and Refined Petroleum Prices Cents per gallon 90 Actual and effectiv e cost of crude oil / S ubsidy A verage im p o rt p ric e X xA verage e ffe c tiv e cost S pread between average effective cost of crude oil and retail petroleum prices C om posite retail price of refined petroleum products 1 0 0 ------------------ ------------ 1978 Source: 1979 I I I I I l I jJ 1980 United S ta te s D epartm ent o f Energy. in total United States consumption, or between 800,000 and 1.35 m illion barrels per day.17 The im pact of decontrol on consumption, therefore, is to reduce im ports of foreign oil by about 1 m illion 17 In August 1980, the approxim ate m idpoint of the decontrol period, the com posite w holesale refined products price was 86 cents per gallon and petroleum consum ption averaged 15.8 m illion barrels per day. A 24 cents per gallon price increase im plies a rise of 28 percent and, using the above elasticity estimate, results in a point estim ate of about 1 m illion barrels per day. The range in the text allow s for one standard deviation around the mean elasticity estimate. 40for FRBNY Digitized FRASER Q uarterly R eview /W inter 1980-81 Effect on the consumer price index The 24 cents per gallon increase in retail prices over the 29-month period of decontrol adds about 6 per centage points to the annualized rate of increase in the consumer fuel and power com ponent of the con sumer price index, using August 1980 as a base level. Since this component accounts fo r about one tenth of the total index, the impact of decontrol on the whole index is to add 0.6 percentage points to its annualized rate of increase between June 1979 and October 1981. Because this does not take into account the pass-through of higher energy costs into the prices of other consumer goods and services, the actual total impact may be somewhat greater. Resistance to future foreign price hikes Under controls, the im port subsidy autom atically rose along with foreign prices, offsetting roughly half of the impact of higher im port prices on the effective cost of crude oil to refiners.19 W ithout the subsidy, any fu ture foreign price hike w ill result in a larger increase in United States refined petroleum prices and, therefore, in a greater reduction of oil imports. This makes it more difficult fo r exporters to raise prices unilaterally, since a given price rise would then require a bigger produc tion cutback. Suppose, for example, that the Organization of Pe troleum Exporting Countries (OPEC) is considering two alternative strategies, one that increases prices by 10 percent and the other that raises prices 12 percent. For the sake of argument, also assume that the sensi tivity of petroleum demand to price changes in the noncom munist world is about the same as it is in the United States. With total noncom munist w orld consumption at about 50 m illion barrels daily, of which 16 m illion is United States consumption, a 10 percent 18 In the first eight months of 1980, 37 percent more oil w ells w e re ' drilled in the United States than in the first eight months of 1979. See Department of Energy, M onthly Energy Review (O ctober 1980), page 50. 19 Under controls, with im ported and uncontrolled dom estic oil accounting for roughly half of refiners’ crude oil inputs, a $2 rise in the im ported (and uncontrolled) price w ould raise the overall average cost by $1. The im port subsidy w ould rise about $1, and the average effective cost of im ported oil would, therefore, be up only $1 on balance. price increase would induce a 2 percent fall in con sumption outside the United States, or 680,000 barrels daily. Due to price controls, however, United States consum ption would fall only 1 percent, or 160,000 barrels daily. Thus, with United States price controls, OPEC would have to cut production by a total of 840,000 barrels daily in order to sustain the 10 percent price increase. Sim ilarly, a 12 percent price rise would require an OPEC production cutback of 1 m illion bar rels daily with United States price controls. Without United States price controls, however, OPEC’s price-raising options would not be so great. W ithout the im port subsidy to mitigate the im pact of price increases on United States petroleum users, cutting current production by 1 m illion barrels per day would sustain only the 10 percent price increase rather than the 12 percent rise possible before. More generally, with world petroleum demand rising because of econom ic growth, OPEC might even be able to sus tain price hikes w ithout cutting current output, but the price rise possible under each alternative production scenario w ill be sm aller w ithout United States controls. Beyond decontrol Crude oil price controls encouraged too high a level of petroleum consumption, discouraged dom estic en ergy production, and increased oil imports. Although the full price of each barrel of imported oil is paid to the exporter, the subsidy makes the refined petroleum appear cheaper to the user. The user may be aware of econom ical ways to reduce consumption through alternatives costing less than the foreign oil. The controls program, however, reduces the incentives to pursue these alternatives, and potential savings go unexploited. If the true cost of foreign oil were no greater than its price, merely removing controls would rectify the problem, for then petroleum users would be motivated to pursue all the alternatives costing less than the unsubsidized price of oil. It is clear, however, that the true cost of foreign oil exceeds its dollar price. Most obviously, our de pendence on imported petroleum leaves the country vulnerable to the threat of econom ic disruption.20 In the 1970s, despite higher petroleum prices, United States dependence on imports rose dram atically as dom estic oil production fell and consum ption was 20 In addition, the more we reduce United States oil consum ption (w hich accounts for nearly 30 percent of world oil output) the more slack this allows in the world market, making it increasingly difficu lt for exporters to maintain or raise their prices. Even if reducing United States oil consum ption initia lly costs more than the dollar price of the oil, the subsequent effect on import prices would make it w orthw hile since the cost of the remaining oil im ports would then be lower than otherwise. Chart 2 United States Petroleum Production and Consumption Millions of barrels per day 2 0 ---------------------------------------- 1949 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 8 0 * * Preliminary for 1980. ^Domestic production includes crude oil, natural gas plant liquids, processing gain, unaccounted-for crude oil, and other hydrocarbons. Source: United States Department of Energy. boosted by the growth of the economy (Chart 2). Do mestic oil price decontrol w ill augment the already ongoing response to higher im ported oil prices in making United States industry, homes, and automo biles more fuel efficient. Nevertheless, the United States has become so dependent on foreign oil that it w ill require a strong, sustained initiative to resolve the long-run problem m eaningfully. Effective new pol icies w ill be needed to make possible both sustained econom ic growth and substantial progress in reduc ing oil imports. A logical and desirable extension of crude oil price decontrol would be a tax to discourage imports. This could take the form of an added tax on gasoline consumption, an oil im port fee, or many other possi bilities. The basic idea is to raise the effective cost (including the tax) of petroleum to a level that more co rrectly reflects the true cost of im porting foreign oil. This would further lower our im ports; the higher the tax, the less foreign oil we would use. Such a tax could then offset other government revenue sources and thus would not require a net rise in overall taxes. In Europe, gasoline is subject to much higher taxes FRBNY Quarterly Review/Winter 1980-81 41 than in the United States. As of July 1980, the tax on a gallon of gasoline was $2.16 in Italy, $1.68 in France, $1.23 in West Germany, and $1.19 in Great Britain, in the United States the average tax in May 1980 was only 14 cents per gallon. Suppose, for example, that an additional one dollar per gallon tax on gasoline in the United States were imposed at the expiration of controls in October 1981. A rough estimate is that this would induce a fall of 12 to 14 percent in United States gasoline consumption.21 This would amount to a reduc tion of demand between 785,000 and 910,000 barrels per day, which is 11 to 13 percent of current United States petroleum imports. An alternative would be to limit petroleum imports directly with an import quota.22 With the petroleum available to the domestic market restricted, the li cense to import petroleum would take on value. The costs associated with securing the import license then would be added to the imported oil price, raising the total effective cost of petroleum on the domestic market, just as a tax would. In this respect, direct limits on imports would be similar to a tax on petro leum. In another important respect, however, direct quotas would be much worse since they would seriously * Price and income elasticities for gasoline demand were estimated as — 0.27 and 0.68, respectively (see the statistical appendix, available from the authors), implying a level of gasoline consump tion in 1981-111 of 6.4 million barrels per day. The range reported in the text allows one standard deviation from the mean in the price elasticity. 22 This analysis of import quotas also generally applies to schemes for directly rationing petroleum among final users, with the cost of rationing coupons analogous to the cost of import licenses. Digitized FRASER Quarterly Review/Winter 1980-81 42for FRBNY undermine our resistance to future foreign price in creases. If exporters raised their price, a petroleum tax would maintain the desired gap between the import price and the effective cost of petroleum on the domestic market, and imports would fall. Under a quota system, however, imports are essentially pre determined. A foreign price increase would simply reduce the value of the import licenses. Unless the quota could be automatically adjusted downward whenever oil prices were raised, the foreign price hike would be, in a sense, completely subsidized, leaving domestic petroleum prices unaffected.23 With United States consumers’ responses eliminated, the sustain able price rise associated with each alternative pro duction scenario of exporting nations would be greater. Conclusion Decontrol is clearly a step in the right direction, but once that is completed new initiatives to reduce oil imports will be required. Replacing the current sub sidy on oil imports with higher taxes on petroleum would help move the United States toward this goal. Unlike quotas, higher petroleum taxes would retain the United States increased resistance to future for eign price hikes. Furthermore, revenues from the pe troleum tax would stay in the country and could re place other sources of funding for government. Only by continuing decontrol’s serious initiative against im ported oil can the United States realistically pursue both economic growth and less dependence on foreign oil. ® If foreign oil prices rose so much that the quotas became irrelevant, then from that point on price increases would no longer be subsidized. Paul Bennett, Harold Cole, Steven Dym Social Security and Savings Behavior Among the most important issues facing the United States economy today is whether existing public policy discourages saving. A central aspect of the problem of insufficient saving and capital formation is the role of the social security system. Many people believe that the United States social security system serves to depress the level of saving in the economy. They point out that a major motivation for saving by individu als is to provide income for retirement. If the need for such saving is reduced because of the existence of Government-sponsored transfers of income to the el derly, then the level of saving may be reduced as well. The proposition is indeed disturbing because it implies that growth of the social security system may result in reduced levels of saving and capital formation and, as a consequence, lower productivity growth and real output growth. Clearly, if these trade-offs exist, the social security system needs to be reexamined. However, it is first necessary to evaluate the logic and evidence underlying the proposition. As it turns out, its veracity is not self-evident on either grounds. The effect of social security on saving involves a diverse and complex set of issues, of which retirement saving is only one. Consequently, the popularity of the propo sition that the social security system depresses saving is not justified. The author is a professor of economics at New York University Graduate School of Business Administration. This article was written while he was a visiting economist at the Federal Reserve Bank New York. The views expressed do not necessarily reflect those of the Federal Reserve Bank of New York. In addition to private retirement saving, the social security system can affect a broad range of household decisions. Thus, its effects on savings behavior remain ambiguous. In particular, social security may affect retirement decisions by inducing earlier retirement, in which case saving during working years may be in creased. Additionally, social security interacts with a whole variety of household investment decisions, such as those involving human capital— schooling, job train ing, health, etc. In this context, social security, which reduces the need for retirement saving, may lead to a shift in the composition of saving toward human capi tal investments. Any apparent negative effect of social security on saving, then, may be because broad areas of capital formation are omitted from measured saving. A related issue, which also implies that social security has a potentially ambiguous effect on saving, is the way in which social security affects the level of intergenerational transfer payments, such as gifts and bequests to children by parents and support to elderly parents by their adult children. Finally, even if the hypothesis that social security reduces savings incentives is true, it is important to consider fully the effects on society of any changes in the social security system. The usually suggested remedy for the savings offset of the existing social security system is to reduce benefits or to increase social security taxes. Both of these could have pro found effects on the level of economic activity and the distribution of income. In the light of these broad consequences, it is not clear that the suggested changes in the system are warranted. FRBNY Quarterly Review/Winter 1980-81 43 The social-security-depresses-saving proposition The argument why social security substitutes for pri vate saving is deceptively simple. It is best explained by examining the lifetime patterns of households’ con sumption and saving. Typically, individuals’ earnings increase with work experience and then decline at retirement. By saving during the most productive years and dissaving during retirement, individuals can main tain a smooth pattern of consumption over their life time.1 If income increases with age until retirement, while consumption is relatively constant, then there are periods of dissaving early in life and after retire ment and a period of saving during mid-life. With such a lifetime allocation, consumption de pends on total wealth or command over resources rather than being constrained by income at any par ticular time. In this context, the concept of wealth is a broad one. In addition to net financial assets and physical assets, wealth includes the present values of future earnings and of benefits to be received from the social security system. These latter items are the value today of earnings or benefits to be received in the future. They are included in wealth because they are part of the individual’s lifetime command over economic resources. When social security benefits are increased, every individual’s overall wealth or life time command over resources also increases. As a result, the typical individual will raise the current level of consumption. Thus, an increase in social security benefits is an increase in wealth which can lead the typical household to reduce the proportion of current income that is saved. The relationship is not, however, quite so simple. It is complicated by the existence of social security taxes, by the effect of social security on retirement de cisions, by the role of intergenerational transfer pay ments, and by the interaction of social security with human capital investment decisions. An examination of these issues reveals that an increase in social security benefits can cause either an increase or a decrease in personal saving. Ultimately, the question of whether social security reduces saving must be settled by em pirical investigation. However, the existing empirical evidence does not address all the issues raised. Social security taxes The role of social security taxes will be examined first. If the social security system were fully funded, which means that the present value liabilities of the 1 In the economics literature, this approach is known as the life-cycle theory of consumption. For a more complete development see, for example, Rudiger Dornbusch and Stanley Fischer, M a c ro e c o n o m ic s (New York: McGraw-Hill, Inc., 1978), pages 146-54. 44 FRBNY Quarterly Review/Winter 1980-81 system are offset exactly by its assets, an increase in future retirement benefits would be matched by an equivalent increase in lifetime tax liabilities. The typi cal individual pays taxes that accumulate in a social security fund. At retirement, this fund is just large enough to pay out retirement benefits over the ex pected remaining lifetime. In the case of such a fully funded system, the individual’s wealth and therefore savings behavior would be unaffected by a benefit change. This is because a benefit increase which adds to wealth would be offset by wealth-reducing increases in taxes.2 Although the social security system as originally envisioned was a fully funded system, this is no longer the case. Generally speaking, social security taxes are set at a level sufficient to pay for current benefits. Since both the size of the population and labor pro ductivity are growing, taxes levied to provide current benefits are less than the present value of future bene fits. Thus, expansion in benefits has increased the net social security wealth held by those currently alive. The opposite effect can occur, if the retired population is large relative to the working population or if benefits accrue to nonearners. It is then possible that an expansion in the benefit structure can require tax increases for current workers, which are more than equivalent to the increase in their expected bene fits. Changes in the age structure of the population after the year 2000 are likely to bring such a situa tion about, since the number of retirees will be ap proaching the size of the working population. Retirement decisions Social security can also affect the decision to work. The current system provides strong inducement to retire at age 65 because retirement benefits are re duced by about 50 cents for every dollar earned over a certain ceiling for those under age 72. Thus, the social security system induces people to retire earlier. To take advantage of the benefit structure, individuals may accumulate additional assets during their working years to provide more retirement income. With a shorter working life, and the prospect of only partial earnings replacement from social security, wage earn ers may increase their pre-retirement saving. Thus, for the typical worker, an increase in the social security benefit structure has a wealth effect which reduces saving and a retirement effect which increases 2 This argument also relies upon some additional rather heroic assump tions, often favored by economists but hardly likely to be true. For example, the rate at which individuals discount future benefits must be equal to the rate of return on saving. In a complex world where taxes and financial market imperfections intervene, individuals may not be indifferent to present taxes as opposed to equivalent future benefits. saving. However, it is unlikely that the additional saving due to induced earlier retirement would be as large as the saving replaced by the social security system. This is because social security benefits are likely to be received for a number of years, while re tirement is likely to be only a few years earlier than it would be in the absence of social security. Thus, the value of benefits will be larger than the additional saving needed for earlier retirement. This comparison assumes that individuals have a clear perception of the magnitude of the increase in wealth due to changes in social security benefits. Such an assumption is un warranted as the benefits to be received by an in dividual are not known with certainty; they depend on his or her earnings and length of life. Thus, the effect of social security on the age of retirement can have important implications for savings behavior. The induced retirement effect of the social security system has an ambiguous effect on aggregate saving for an additional reason. The retirement effect would change the savings behavior of workers and also in crease the relative size of the nonworking population. The total effect on the income and saving of the en tire population has not been explored. Intergenerational transfers The discussion of lifetime planning of consumption patterns did not refer to bequests or to private inter generational transfers of income. These phenomena are widely observed in the real world, and the latter one is of particular concern. Intergenerational transfers of income may well be an important means of provid ing for retirement. Thus, Government provision of re tirement income through the social security system may substitute for private intergenerational income transfers rather than substituting for the intragenerational deferral of consumption (retirement saving).3 To be specific, a situation can be envisioned where, in the absence of social security, elderly persons are pro vided for by income transfers from their working chil dren. With a social security system, the working chil dren make tax payments instead of direct transfers and retirement income for the parents Is provided by the Government. It is conceivable that the two systems are equivalent and the disposable income and saving of both parents and children are the same in each case. 3 This idea has been emphasized by Robert Barro, "Are Government Bonds Net Wealth?”, J o u rn a l o f P o litic a l E c o n o m y (Novem ber/ December 1974), pages 1095-1118. However, aggregate social security benefits are so large that it is difficult to imagine that, in the absence of social security, private transfers would approach the same magnitude. This is an important possibility because it suggests that social security has displaced private transfers rather than private saving and capital formation. The consequences for saving of such an income redistri bution have not been adequately explored but are probably less severe than the wealth effects indicated by the life-cycle approach. Along these lines, it is interesting to note that the social security system may have widespread influ ences on the living patterns of the elderly and the re lationship between the generations. For example, so cial security may encourage the elderly to live alone rather than to share living arrangements with the young. Alternatively, social security may be viewed as a social response to these changes in mores. Social security and human capital The final complication to the basic life-cycle propo sition that social security offsets private saving involves an important element of household savings decisions and lifetime planning that is by and large overlooked in discussions of the social security system. That is, the interaction between social security and capital formation in the form of human capital investments.4 Introducing human capital, particularly investments in education, adds a degree of complexity that has not been explored. This is a serious omission since it is possible that the interaction of human capital invest ments with social security is strong. The strength of the relationship is suggested by the similarities between human capital wealth and social security wealth. Both are nonfungible assets, unlike the financial and physical assets which are viewed as social security substitutes in existing empirical studies. Thus, it is possible that the relationship between these types of assets is as important as their relationship to the standard forms of saving. Additional similarities are that human capital investments, along with re tirement saving, are an important form of life-cycle planning by the family unit and, also, that human capi tal investments are an important form of intergenera tional transfers. It is not evident whether social security wealth and human capital investments should be viewed as substitute or complementary assets. In the first case, social security which provides retirement income could be viewed as an alternative to educating one’s chil dren so that they will have the income to provide re tirement support to their parents. This does not seem to be the appropriate argument because the tendency 4 Sherwin Rosen suggests the possibility of a relationship in “Social Security and the Economy'', T h e C ris is in S o c ia l S e c u rity (San Francisco, California: Institution for Contemporary Studies, 1977). FRBNY Quarterly Review/Winter 1980-81 45 to invest in human capital has increased a great deal since the inception of social security. It is more likely that social security wealth and human capital are com plementary assets. In this case, the advent of the so cial security system, which reduced the burden of saving for retirement, made it possible for the typical individual to devote more resources to saving in the form of human capital.5 Although these hypotheses have not been tested, perhaps an effect of the social security system has been to induce the household sector to channel its resources into human capital investments. Thus, by standard measures, saving did decline, although, due to increases in human capital investments, overall capital formation need not have declined. This argu ment does not obviate the entire issue, if policymakers feel that the induced move from physical to human capital formation has been excessive. There is yet another interrelation between social security and human capital investments. An individual can provide for retirement by accumulating ordinary assets over his working life or by investing in educa tion with the hope that the returns to human capital investment will provide retirement income. As the returns to human capital investments are highly vari able among individuals, there may well be a preference for a less risky means of lifetime planning. Since so cial security reduces the risk of being without income in one’s old age, it may encourage individuals to make investments in human capital. Because of the unavailability of data, there have not been any empirical studies of the relationship of social security to both private intergenerational transfers and investments in human capital. Although there is some evidence indicating that financial support from chil dren to parents is relatively small, it is not clear whether this is a consequence of social security. Data on intergenerational transfers of human capital and the relationship between human capital and other forms of wealth are almost totally lacking. Social security policy If the proposition that social security depresses saving is in fact true, then some changes in social security policy would be appropriate.6 Supporters of the propo sition have suggested changes in the way in which the 5 An elaboration of this argument is found in "Social Security and Investment in Human Capital" by Thomas F. Pogue and L.G. Sgontz, N a tio n a l T a x J o u rn a l (June 1977), pages 157-70. They also present some empirical evidence that the advent of social security has increased human capital investments. 4 For a complete review of all the policy issues, see Bruno Stein, S o c ia l S e c u rity a n d P e n s io n s In T ra n s itio n (New York: Free Press, 1980). Digitized FRASER Quarterly Review/Winter 1980-81 46 forFRBNY system is financed. However, such modifications would have additional undesirable effects on the economy. In general terms, the overall issue is whether the system should be one of intergenerational transfers, essen tially pay as you go, or whether it should be a fully funded annuity system. As the social security system grew, it evolved into a pay-as-you-go system. The trust fund of Government securities, which accumulated in the early years when there were few beneficiaries relative to workers sub ject to the payroll tax, eroded as the Congress in creased benefits. An error made when the 1972 Social Security Act was drafted, compounded the problem by double-indexing the benefit structure.7 Without large increases in payroll taxes, the trust fund was well on its way to bankruptcy. This was rectified by the amendments legislated in 1977 which put social se curity back on a sound pay-as-you-go system.8 Changes in the demographic structure of population over the next fifty years will still put serious financing strains on the system. After the year 2000, there will be a substantial growth of the population above re tirement age relative to the working-age population. The number of persons retired as a percentage of the working population will increase from the present level of about 19 percent to about 30 percent in 2030. The increase will not start until after 2000 and will be even larger if fertility continues at its present low level. Thus, there is a long-term problem of an increasing burden on financing social security pensions, even though the amendments in 1977 reduced the immediate crisis by stopping the growth of the so-called replace ment rate. The replacement rate— the ratio of the median pension benefit at retirement to the median wage prior to retirement— had increased from about 0.3 to over 0.4 in the 1970s because of the indexing procedures. The current legislation will maintain the rate at a constant level of about 0.42. If it had con tinued to increase, much larger increases in the pay roll tax would have been necessary. Proponents of the social-security-retards-privatecapital-formation proposition argue that the system 7 The problem arose from linking both benefits paid and the wage base used to determine initial benefits to changes in consumer prices. 8 The system is still not without its financial problems. There is con siderable pressure in the Congress to roll back the scheduled increases in the payroll tax rate. In addition, continued high inflation could create a cash flow problem for the trust funds by the mid-1980s. In either case, short-run financing from general revenues may be neces sary. For a historical analysis of the social security system, see Martha Dethrick, P o lic y M a k in g fo r S o c ia l S e c u rity (Washington, D.C.: The Brookings Institution, 1979) and Rita Campbell, S o c ia l S e c u rity P ro m is e a n d R e a lity (Stanford, California: Hoover Institution Press, Stanford University, 1977). should pay pensions from an actuarially appropriate trust fund, rather than on a pay-as-you-go basis. In this case, social security wealth would not be fictional but instead would be backed by existing assets. Such a proposal would require substantial tax increases for the fund to accumulate sufficient assets. In essence, there would have to be a larger Government surplus on the consolidated budget as the trust fund accumulated outstanding Government debt. The idea behind this is that it would release funds to the private capital mar kets. However, the effect of such tax increases, in the short run, on aggregate demand and output could be devastating. A basic lesson of Keynesian macroeco nomics is that, although a surplus reduces government demands on the capital market, it can induce a reces sion and lower the overall private-sector demand for capital goods. These latter caveats are understood by the proponents of the trust fund approach who argue that social security should move toward a funded sys tem gradually, as short-term macroeconomic policy permits. The idea that social security should be funded can be criticized on additional grounds best explained by describing the development of the system.9 When the social security system began, the initial generation of beneficiaries received a considerable net transfer since their benefits exceeded their payments. If the argument is that this reduced their saving, then the current generation is producing with a deficiency in the capital stock. By increasing taxes and further de creasing the standard of living of the current genera tion, we may in the long run be able to accumulate a fund and also make up the capital deficiency. This transition may take several generations but, from then on, the system will be funded in the sense that each generation’s benefits are the taxes it accumulated plus interest. Such a proposal imposes the burden of reduc ing consumption to accumulate a fund on the current generation. This was not viewed as desirable forty years ago when the system conferred benefits on the initial generation and does not seem any more ap propriate now. If the current capital stock is considered deficient, there are many other policy approaches to influencing the level of investment, including reduced taxation o f capital income. If there is concern about making the overall tax structure less progressive, it hardly seems appropriate to use payroll tax increases to in ♦The line of argument that follows draws upon the discussion by Mordecai Kurz and Marcy Arvin in "Social Security and Capital Formation: The Funding Controversy” , W o rk in g P a p e rs of the President’s Commission on Pension Policy, 1979. fluence capital formation. There is no specific reason why a society has to make up any capital stock defi ciency that developed when intergenerational trans fers were introduced. It is instead a question of equity and fairness in the design of an overall tax system. Clearly, changes in the distribution of the tax burden promote capital formation, but a society with a concept of distributional equity might not make such choices. Perhaps the most telling blow to the proposal of funding is its impracticality. At current benefit levels and interest rates, the fund would have to approach $1,000 billion, more than the total privately held public debt. Even a gradual fund accumulation would require large changes in the tax structure, with distributional consequences that are not likely to appeal to the public or political decision-making bodies. The cur rent generation is not likely to volunteer to reduce its living standard substantially in order to enlarge the productive capital stock for its heirs. Rather than dwelling upon the relative merits of a pay-as-you-go or funded transfer system, perhaps society should address the issues concerning taxation and capital formation directly. Review of the evidence One of the most problematic aspects of the hypothesis that social security curtails saving and capital forma tion is that casual observation of structural develop ments in the economy since the inception of the social security system provides scant evidence of any such effect. In a sense, the legislation created vast sums of wealth in the economy while the physical assets in ’ the country were unchanged. Over time, one would expect major adjustments in the structure of the econ omy in response to these changes. If there has been an effect on saving, researchers should also be able to detect the effect on capital intensity and on the rates of return to capital. For example, the creation of social security wealth makes physical assets relatively scarce which should lead to larger returns on such assets. Similarly, if social security displaces saving, some downward secular trend in rates of saving and capital formation should have emerged. However, economists have not observed either phenomenon.10 It would be difficult to argue that savings rates have been remark ably steady because increased real returns have offset the depressing effects of social security. Most econo mists have argued that, if anything, real returns to i# There is evidence that the rate of return to schooling, a major com ponent of human capital investments, increased for many years and declined in recent years. This could support the interaction between social security and human capital suggested earlier. FRBNY Quarterly Review/Winter 1980-81 47 capital have declined in the postwar period.11 More formal tests of the proposition that social security depresses saving have been conducted, largely in the context of the life-cycle approach, dis cussed earlier, which showed that wealth is a key de terminant of consumption. Econometricians attempt to measure the impact of social security on saving and consumption by specifying an equation that relates consumption expenditure to social security wealth. So cial security is a savings depressant if the estimated impact of social security wealth on consumption is positive and can be statistically distinguished from a zero effect. A brief description of the results follows. A fuller, but still nontechnical, summary is presented in the accompanying appendix. Current interest in the effect of social security on saving was sparked by Martin Feldstein’s 1974 econo metric study.12 His conclusion that there is a very strong depressing effect has been the basis for all discussion and argument since then. However, an at tempt by Dean R. Leimer and Selig D. Lesnoy of the Social Security Administration to replicate his data uncovered a data error.13 When the—social security wealth variable is corrected, the results are strik ingly different. Feldstein’s conclusion that social secu rity has reduced personal saving by one half and the stock of capital by one third is completely unsubstan tiated with the corrected data. This is important be cause the enormous depressing effect on saving has been widely quoted and supported by many econo mists for six years. Empirical studies have also attempted to measure the effect on labor supply and retirement decisions. Social security may affect saving because it provides an incentive for retirement. The advent of social security makes much of the working public plan for retirement by increasing their saving during working years. Alicia Munnell’s tests of this hypothesis found that the siz able decline in the labor force participation rate for men aged 65 and over (from just under 50 percent when social security was introduced to less than 25 percent by the mid-1970s) had a substantial positive effect on saving. Even if this entire increase were at tributed to social security, the induced increase in saving would offset only about one half of the reduc tion of saving due to social security wealth.14 Clearly, it is difficult to make definite judgments based on aggregate savings data. Since economists do not conduct controlled experiments, it may not be possible to determine what the world would be like without the social security system. The historical com parison of the present economy with the depression era may be inadequate for isolating the effect of the creation of the social security system from all the other changes in the structure of the economy over the past forty years. There are two other types of data which also can be used to investigate the effects of the social security system on saving: data on the savings behavior of dif ferent individuals (cross-section data) and data on the savings behavior in different countries. Cross-section data have been used to investigate the effect of differences among individuals in private pen sion plans and social security benefits and taxes on savings behavior. The evidence concerning the wealth effect of social security on saving is weak.15 Lawrence Kotlikoff suggests that the savings offset predicted by theory is not found in the data because individuals are unable to forecast their social security benefits and their age^of retirement. Others argue that reduced intergenerational transfers and induced retirement ef fects of social security are unlikely to offset the nega tive effect of social security on wealth accumulation. However, even the cross-section results, indicating that individuals with relatively higher social security save less, do not necessarily imply that, after aggregation over the entire population, an increase in the scale of the social security program reduces total saving. Another path of empirical investigation examines differences in both savings behavior and social secu rity systems among countries. Virtually all industrial ized nations have some form of government-sponsored program for transfers to the elderly. Since the cross national differences in savings behavior are large, some analysts have asked whether these differences in savings behavior are to any extent due to differ ences in social security benefits. Most recently, Robert 11 It should also be noted that social security is only one type of fictional wealth. Social security wealth— the present value of future benefits— is fictional because it is not matched either by future con tributions or by the expected earnings from existing assets. The vast unfunded liabilities of private (for some large corporations such liabilities exceed net worth) and government (civil service, military, etc.) pension systems are also forms of fictional wealth. Even more than social security, these wealth components have grown very rapidly in recent years, without any obvious effect on aggregate saving. 14 In the Munnell study, “The Effect of Social Security on Personal Savings" (Cambridge, Massachusetts: Ballinger Publishing Co., 1974), the income coefficient in the consumption relation depended on the labor force participation rate for men aged 65 and over. 12 "Social Security, Induced Retirement, and Aggregate Capital Accumulation” , J o u rn a l o f P o litic a l E c o n o m y (September/October 1974), pages 905-26. w Their results were presented to the annual meeting of the American Economic Association in Denver, September 5-7, 1980. For a report, see “ Economic Diary", B u s in e s s W e e k (September 22, 1980). w For example, see the studies by Lawrence Kotlikoff, “Testing the Theory of Social Security and Life Cycle Accumulation” , A m e ric a n E c o n o m ic R e v ie w (June 1979), pages 396-410, and by Martin Feldstein and Anthony Pellechio, “Social Security and Household Wealth Accumulation, New Microeconomic Evidence", T h e R e v ie w o f E c o n o m ic s a n d S ta tis tic s (August 1979), pages 361-68. 48for FRASER FRBNY Quarterly Review/Winter 1980-81 Digitized Barro and Glen McDonald examined the effect of inter national differences in the ratio of real social security benefits per person over 65 to real income per capita on savings rates.16 They conclude that available cross national data are not rich enough to allow any infer ences about the effect of social security on saving. At this juncture, it is useful to draw some conclu sions concerning the em pirical evidence on the effect of social security on savings behavior. One can only say that there is some highly tentative em pirical support for the hypothesis that social security substitutes for pri vate retirem ent saving. Since private retirem ent saving represents wealth accum ulation which results in capi tal form ation, w hile unfunded social security programs are backed only by the accum ulation of “ fictio n a l” wealth, it is possible that overall capital form ation is depressed. However, there is a complex set of other effects of social security which makes it impossible to give unqualified support to this hypothesis. These ef fects that the em pirical literature has been unable to isolate adequately include retirem ent decisions, the private provision of pensions, other form s of intergenerational transfers, and other types of capital for mation. Conclusions Although this discussion of social security involves a com plex and diverse set of issues, two threads do seem to emerge. Social security should not, at this juncture, be viewed as a substitute fo r private retirem ent saving. The issue is an em pirical one, and the existing evi dence offers only some tentative statistical support for the hypothesis. Furthermore, the evidence is deficient because it omits any serious consideration of the complex relationships between social security and other forms of intergenerational transfers, such as hu man capital investments. The unfunded, or pay-as-you-go, public transfer sys tem should not be viewed as the cu lp rit that has caused a lower than desired capital stock and lag ging productivity growth. Social security is just one part of an overall system of public expenditure and income redistribution that interacts with private sav ings decisions in many ways. The desirability of in ducing more capital form ation is a broad policy issue that should be dealt with in a larger fram ework, par ticu la rly since the extent of any capital form ation effect of social security is, as yet, uncertain. 14 "S ocial Security and Consumer Spending in an International Cross S ection” , J o u rn a l o f P u b lic E c o n o m ic s (A ugust 1979), pages 275-89. Appendix: The Effect of Social Security on Saving T h e re have been s e veral e m p iric a l studies of th e e ffect of so cial s e cu rity on saving w h ich fail to re a c h any consensus. A th o ro u g h te c h n ic a l survey of th e s e stu d ies w as m ade, one by Louis E spo sito, S e cu rity on S avin g: R eview S tates T im e S e rie s D a ta ” , “ E ffect o f S o cial of S tu d ie s U sing U nited Social Security Bulletin (M a y 19 78), p a ges 9-17, and one by N. B u le n t G u ltek in and D en nis Lo gu e, “ S o cial S e c u rity and P erso n al S avin g : S urvey and N ew Private Saving, Social Security versus E v id e n c e ” , G e o rg e M . von F u rs ten b erg , ed . (C a m b rid g e, M a ssac h u setts: B a llin g e r P u b lishing C o., 19 80). A brief, n o n tech n ica l su m m ary of th e m eth o d o lo g y, re sults, and so u rces of th e d is a g re e m e n t is p res en ted here. T h e re is b ro a d a g re e m e n t am o n g ec o n o m ists ab ou t th e g e n eral s p e c ific a tio n of a life -c y c le consu m p tio n fu nction es tim a te d from tim e se rie s d a ta . T y p ic a lly , it ta k e s the fo llow ing fo rm : C t — ao - f on Y D t + a s Y D t_i - f o i W t - f a * S S W t + n t w h ere : C = real p e r c a p ita co n su m p tio n ex p e n d itu re s , YD = real p e r c a p ita d isp o sab le p e rsonal in com e, W = real per c a p ita p e rsonal s e c to r net w orth, SSW = real per c a p ita so cial s e cu rity w e a lth , and Mt = The resid ual o r e rro r te rm . p a ra m e te r es tim a te s e n a b le the to p re d ic t th e e ffe c t on co nsu m p tio n e c o n o m e tric ia n (and hence on saving) of th e v a ria b le s on th e rig h t-h an d sid e of the eq u atio n . For th e qu estio n being c o n s id e re d — th e effect of so cial s e cu rity on sa vin g — th e c o e ffic ie n t on the S S W v a ria b le d e fin e d in th e te x t is o f c ru c ia l in terest. T h e e c o n o m e tric lite ra tu re in clud es m any va ria tio n s on th is eq u atio n , and th e re is so m e co n tro v ersy co n ce rn in g w h ich , if any, a d d itio n a l e x p la n a to ry va ria b le s should be in clu d ed in th e co n su m p tio n re latio n sh ip . T h is is im p o rtan t b e c a u s e th e c o e ffic ie n t o f S S W is sensitive to th e inclusion of o th e r v a ria b le s , such as th e u n em p lo y m ent rate, and to c h a n g e s in th e tim e p erio d of histo ric al d a ta used fo r es tim atio n . Im p o rta n t fo r e v alu a tin g th e m a g n itu d e of an y p a r tic u la r c o e ffic ie n t is th e c o n c e p t of statis tica l sig n ifi cance. W ith o u t providin g a te c h n ic a l e x p la n a tio n , a c o e ffic ie n t is s ta tis tic a lly s ig n ifican t if th e results p ro vid e reaso n a b ly su b stan tial e v id e n c e th a t th e e s tim ated co efficien t differs from ze ro . C h a n g e s in th e s p e c ific a tion of an e q u atio n ca n a ffe c t both th e m a g n itu d e of th e co efficien ts, as sta te d a b o ve , as w ell as th e ir statis tic a l sig n ifican ce. In o u r co ntext, social s e c u rity is a saving s d e p re s s a n t if th e c o e ffic ie n t on S S W is po sitive (i.e., in c re a s e s in S S W raise co n su m p tio n ) and sig n ifi ca n tly d iffe re n t from zero . Paul Wachtel FRBNY Q uarterly R eview /W inter 1980-81 49 August-October 1980 Interim Report (This report was released to the Congress and to the press on December 3, 1980.) Treasury and Federal Reserve Foreign Exchange Operations Coming into the August-October period under review, exchange market participants remained cautious about the outlook for the dollar. Traders were encouraged by the improving trend in the United States current ac count, which had swung from deep deficit to near balance and was expected to move into surplus by late 1980. At the same time, however, they were concerned about the outlook for inflation in the United States. Even though our price indexes were no longer rising as rapidly as before, inflation remained uncomfortably high by historical standards and by comparison with infla tion rates in many other industrial countries. Moreover, it was feared that the improvements in our current ac count and price performance might prove transitory to the extent that they stemmed from the sharp recession which had emerged in the United States earlier in 1980. Meanwhile, discussion of possible tax cuts or of an easing of monetary policy had generated concern in the market that heavy stimulus to the economy might undercut the anti-inflation effort. For its part, the Federal Reserve had phased out the special credit restraints imposed in March, but Chairman Volcker had made it clear that the Federal Reserve would con tinue to adhere to its efforts to slow the growth of money and credit in the United States by placing primary em phasis on bank reserves rather than on interest rates. A report by Scott E. Pardee. Mr. Pardee is Senior Vice President in the Foreign Department of the Federal Reserve Bank of New York and Manager of Foreign Operations for the System Open Market Account. 50 FRBNY Quarterly Review/Winter 1980-81 By August, United States interest rates were rebound ing from their latest lows, and a sudden surge in the growth of the monetary aggregates gave rise to some expectations that United States interest rates might advance even further. Meanwhile, the market’s uncertainties were not lim ited to the outlook for the dollar. Most other major in dustrial countries were afflicted with inflation rates which were too high by their own standards and by substantial current account deficits which had been aggravated by the oil price increases of 1979 and early 1980. The authorities had pursued restrictive policies to deal with these problems. By late summer, eco nomic growth was slackening generally, prompting the authorities in several countries to move cautiously to ward a less restrictive policy stance. But they were reluctant to move too quickly in the direction of ease in view of the need to fight inflation and their efforts to keep interest rates sufficiently high to attract funds from abroad to finance large current account deficits. As a result, interest rates remained high even as mar ket expectations built up that, in view of domestic economic considerations, an easing of monetary pol icy was imminent in several countries. Consequently, an uneasy atmosphere persisted in the exchange markets through August and early Sep tember as traders sought to assess the implications of these economic and financial developments here and abroad. In addition, the sense of unease was heightened from time to time by political events, such as general strikes in Poland and continued tensions in the M iddle East. In this environment, exchange rates fluctuated widely day to day but few clear trends de veloped, with the exception that both sterling and the Japanese yen were bid up by force of heavy capital inflows. Among the currencies participating in the join t float arrangement, the French franc remained near the top of the band and the German mark near the bottom. In the absence of renewed selling pressures on the dollar, the United States authorities took the oppor tunity to acquire currencies to repay debt arising from earlier intervention and to rebuild balances. Operating on days in which the dollar was firm or rising, the United States authorities bought a total of $426.6 m il lion equivalent of German marks in the market, either in New York or in Frankfurt through the agency of the Bundesbank. Over the same period, the Trading Desk purchased an additional $453.6 m illion of marks from correspondents. The Federal Reserve used a portion of these marks, along with previously acquired balances, to repay swap debt to the Bundesbank, which was re duced from $879.7 m illion at end-July to $362.6 m illion on September 15. The remaining acquisitions were added to Treasury balances which increased by $338.1 m illion equivalent. The Federal Reserve also bought small amounts of French francs and Swiss francs in the market and from correspondents. On occasions when the dollar came under selling pressure in Au gust, the United States authorities intervened on five different days, selling a total of $69.6 m illion equiva lent of marks, including $53.9 m illion equivalent from Federal Reserve balances and $15.7 m illion from United States Treasury balances. By mid-September, econom ic indications suggested that the United States was moving out of recession. Although the upturn was welcomed by the markets, it dimmed the prospects for further inflation relief in the near term. Indeed, partly because of rising food prices, the United States inflation rate was expected to accel erate. Moreover, the money and credit aggregates were growing rapidly. In response to this buildup in the demand fo r money, the Federal Reserve was act ing to constrain the growth of bank reserves. Market interest rates clim bed sharply, and on September 26 the Federal Reserve raised the discount rate by 1 per centage point to 11 percent. Strong demand for money and credit persisted through October, putting addi tional upward pressure on money market rates. This advance of United States interest rates was not matched abroad, where, if anything, the authorities were becoming increasingly concerned about slower econom ic growth and the prospect of recession. Con sequently, interest differentials swung increasingly in favor of the dollar against most m ajor currencies, Table 1 Federal Reserve System Drawings and Repayments under Reciprocal Currency Arrangements In m illions of dollars equivalent; draw ings ( + ) or repayments ( — ) System swap com m it ments July 31, 1980 Transactions with Bank of France ............. 166.3 879.7 German Federal Bank .. 1,046.0 August through October 31, 1980 System swap com m it ments O ctober 31, 1980 - 165.2* -0- - 873.0* -0- — 1,038.2* -0- Because of rounding, figures may not add to totals. Data are on a transaction-date basis. * Repayments include revaluation adjustm ents from swap renewals, am ounting to $1.1 m illion for draw ings on the Bank of France and $6.7 m illion for draw ings on the German Federal Bank w hich were renewed during the period. Table 2 United States Treasury Securities, Foreign Currency Denominated In m illions of dollars equivalent; issues ( + ) or redem ptions ( — ) Am ount of com m itments July 31, 1980 Issues August through O ctober 31, 1980 Amount of com m it ments October 31, 1980 Public Series Germany ............................. 5,233.6 -0- 5,233.6 Switzerland ........................ 1,203.0 -0- 1,203.0 T o t a l..................................... 6,436.6 -0- 6,436.6 Data are on a value-date basis. Tabie 3 Net Profits ( + ) and Losses ( —) on United States Treasury and Federal Reserve Current Foreign Exchange Operations In m illions of dollars Federal Reserve Period United States Treasury Exchange S tabilization General Fund account August 1, through O ctober 31, 1980 ............. + 1 4 .0 + 0.1 -0- Valuation profits and losses on outstanding assets and lia b ilitie s as of O ctober 31,1980 . . . + 1 2 .7 — 372.8 + 1 3 8 .8 Data are on a value-date basis. FRBNY Q uarterly R eview /W inter 1980-81 51 prompting flows of funds into dollar-denom inated as sets. Much of this pressure fell on the German mark, in view of Germany’s low nominal interest rates relative to rates abroad and Germany’s sizable current account deficit. Funds were shifted out of marks not only into dollars but into sterling and French francs as well. W ithin the European M onetary System (EMS), the Ger man mark and the French franc were pushed to their respective intervention points, and the Bundesbank and the Bank of France were obliged to absorb sub stantial amounts of marks against francs. At the same time, the EMS currencies as a group declined against the dollar. As a result of the flow of funds into dollar assets, the d o lla r rose in October to end the three-month period up a net 7 percent against the German mark and other currencies in the EMS, 31/2 percent against the Swiss franc, and 1% percent against the Canadian dollar. Over this same period, sterling rose a net 4% percent against the dollar and the yen moved up by 7% per cent. With the dollar in demand, the United States author ities stepped up th eir acquisitions of currencies to re pay debt and rebuild balances. Operations were con ducted in New York, Frankfurt, and on occasion in the Far East. When strong one-way pressures emerged late in October, the Desk intervened, sometimes force fully, in the m arket as a buyer of German marks. Pur chases of marks in the spot market totaled $1,770.7 m illion equivalent between mid-September and end- October. Moreover, as part of the effort to repay debt and rebuild balances, the United States authorities purchased a total of $346.6 m illion of marks from cor respondents, divided about equally between the Fed eral Reserve System and the Treasury, and $132.9 m illion of outright forw ard marks on behalf of the Trea sury. As a result, the Federal Reserve was able to complete liquidation of its swap debt w ith the Bundes bank by the end of the period. In addition to its mark purchases, the United States authorities bought over the three-m onth period $87.5 m illion equivalent of Swiss francs, including $25 m il lion equivalent in the m arket and $62.5 m illion equiva lent from correspondents. Of this amount, $62.6 m illion equivalent was added to System balances and $24.9 m illion equivalent went into Treasury balances. The Federal Reserve also took advantage of opportunities to buy $158.6 m illion of French francs to com plete repayment of its swap debt with the Bank of France. During the August-O ctober period, the Federal Re serve realized $14 m illion in profits on its foreign exchange operations and the Exchange Stabilization Fund (ESF) realized $0.1 m illion. As of the end of the period, the Federal Reserve showed valuation profits of $12.7 m illion on its foreign exchange assets w hile the ESF showed valuation losses of $372.8 m illion on its foreign exchange assets. The Treasury’s general account showed valuation profits, related to the out standing issues of securities denominated in foreign currencies, of $138.8 m illion. SELECTED PAPERS OF ALLAN SPROUL The Federal Reserve Bank of New York has released a representative selection of the published and unpublished w ritings of its third chief executive officer in a 254-page book entitled “ Selected Papers of Allan Sproul” . The book, w hich includes a biographical essay, was edited by Lawrence S. Ritter, Professor of Finance at New York University. A copy is available on request from : Public Information 33 Liberty Street New York, N.Y. 10045 52 FRBNY Q uarterly R eview /W inter 1980-81 Subscriptions to the Quarterly Review are free. M ultiple copies in reasonable quantities are available to selected organizations for educational purposes. Single and m ultiple copies for United States and fo r other Western Hemisphere sub scribers are sent via th ird - and fourth-class mail, respectively. All copies for Eastern Hemisphere subscribers are a irlifted to Amsterdam, from where they are forw arded via surface mail. M ultiple-copy subscriptions are packaged in envelopes containing no more than ten copies each. Quarterly Review subscribers also receive the Bank’s Annual Report. Library of Congress Catalog Card Number: 77-646559