The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Review December 1984 Vol. 66, No. 10 5 The 1981 Personal Incom e Iiix Cuts: A Retrospective Look at Their Effects on the Federal liix Burden 18 Real Interest Rates: What Accounts for Their Recent Rise? 30 Monetizing the Debt T h e Review is p u b lis h e d 10 t im e s p e r y e a r b y th e R e s e a r c h a n d P u b lic In fo r m a tio n D e p a r tm e n t o f th e F e d e r a l R e s e r v e B a n k o f St. L ou is. S in g le-c o p y s u b s c r ip t io n s a r e a v a ila b le to th e p u b lic f r e e o f c h a r g e . M ail r e q u e s t s f o r s u b s c r ip tio n s , b a c k iss u es, o r a d d r e s s c h a n g e s to : R e s e a r c h a n d P u b lic In fo r m a tio n D ep a rtm en t, F e d e r a l R e s e r v e B an k o f St. L ou is, P.O. B o * 442, St. L ou is, M isso u ri 63166. T h e view s e x p r e s s e d a r e t h o s e o f t h e in d iv id u al a u t h o r s a n d d o n o t n e c e s s a r ily r e fle c t o ffic ia l p o s itio n s o f t h e F e d e r a l R e s e r v e B a n k o f St. L ou is o r th e F e d e r a l R e s e r v e S y stem . A r tic le s h e r e in m a y b e r e p r in t e d p r o v id e d th e s o u r c e is c r e d ite d . P le a s e p r o v id e th e B a n k ’s R e s e a r c h a n d P u b lic In fo rm a tio n D ep a rtm en t w ith a c o p y o f r e p r in t e d m a ter ia l. Fed eral R eserve Bank of St. Louis Review December 1984 In This Issue . . . In th e first article in th is R eview , Jo h n A. T atom exam in es w h eth er th e 1981 fed eral tax rate red u ctio n s have red u ced th e fed eral tax b u rd en . C om paring fed eral tax b u rd en s in 1980 an d 1984, T atom co n c lu d e s th a t fed eral tax b u rd en s w ere g reater in 1984 for m ost A m erican fam ilies. B oth th e fed eral in co m e tax "bracket c r e e p ” an d Social Secu rity tax in crea ses from 1980 to 1984 have m ore th an offset th e 23 p e rce n t cu t in average an d m arginal fed eral in c o m e tax rates sin ce 1980. T h e au th o r u ses th is analysis to clarify two m a jo r so u rces o f co n fu sio n about tax ch an g es: First, h e show s th at th e p ercep tio n th at th e 1981 rate red u ctio n s ben efited relatively h ig h er-in co m e groups at th e ex p e n se o f lo w -in com e fam ilies resu lts from bracket creep an d Social Secu rity tax hikes, b o th o f w h ich have raised th e taxes o f lo w -in com e fam ilies d isp roportion ately. T h e actu al rate red u ctio n s, he show s, fell evenly a cro ss in co m e levels. T h e net result w as relatively large in creases in federal tax paid per d ollar o f in co m e for lo w -in com e fam ilies, an d only slight red u ctio n s for relatively h ig h -in co m e fam ilies. Sin ce taxpayers gen erally paid h ig h er average tax rates on larger real in co m es in 1984 th an in 1980, th e exten t to w h ich th e 1981 tax rate re d u ctio n s have co n trib u ted to th e federal d eficit is obviously o p en to q u estio n . P ro p o n en ts o f th e view that th e 1981 tax ch an g es raised th e d eficit ap p aren tly focu s atten tio n on th e rate red u ctio n s alone, acco rd in g to Tatom , and ignore th e o th e r factors that m ore th an offset th e rate red u ctio n s. In th e seco n d article, “Real In terest Rates: W hat A ccou n ts for T h e ir R ecent R ise?”, A. Steven H olland estim a tes b o th sh o rt- an d long-term real in terest rates an d show s th at th ey have b een h ig h er d uring th e 1980s th an in th e previous two d ecad es. M ost o f th is upw ard m ovem ent in real rates o cc u rre d d uring late 1980 an d early 1981. T h e a u th o r th e n e x a m in e s sev era l p o s s ib le fa c to rs th a t a ffe ct re a l in te r e s t ra te s to see if they have played a m ajo r role in th e shift to h ig h er rates. He find s th at an in crea se in th e variability o f m o n ey grow th, w h ich in crea sed e co n o m ic u n c e r tainty an d th e risk p rem iu m on in terest rates, w as m ost clo sely co in cid e n t w ith th e rise in real rates. O th er p oten tial factors, su ch as the m a jo r ch a n g es in cu rren t an d p ro jecte d governm ent deficits and in tax policies, w h ich m any analysts blam e for th e real in terest rate rise, took effect after m ost o f th e upw ard shift in real in terest rates already h ad o ccu rred . In th e third article in this R eview , D aniel L. T h o rn to n reviews th e m ean in g o f th e p h rase "m o n etizin g th e d eb t.” T h e a u th o r p oin ts out th at today, as in th e past, m on etizin g th e d eb t m ean s m o n ey grow th — in d u ced by rapid grow th o f th e federal d ebt — in ex cess o f th at n eed ed to ach ieve so m e m o n eta iy p o licy o b je c tive. C on sequen tly, d eb t m o n etiz a tio n ca n n o t b e an alyzed in d ep en d en tly o f th e objectives o f F ed eral Reserve policy. T h o rn to n p oints out th e in h eren t lim itation s o f using su ch m easu res as grow th o f th e Fed eral R eserve's portfolio o f governm ent debt o r grow th in som e reserve m easu re as evid ence o f d ebt m on etization . He show s how sim p le co rrelatio n s 3 In This Issue Digitized for 4 FRASER betw een su ch m easu res an d d ebt grow th ca n give erro n e o u s “e v id en ce” about d ebt m onetization . Finally, T h o rn to n exam in es w h eth er th e F ed eral Reserve h as m o n etized th e d ebt in re ce n t y ea rs by co n d u ctin g tests o f th e tem p o ral ord erin g o f m oney grow th and d ebt grow th over th e 1 9 6 0 -8 4 period . He find s n o in d ica tio n th at the Fed eral Reserve h a s m o n etized th e d ebt d uring th e p ast d eca d e, w h en th e p re s su re to do so w ould seem to have b een g reater th a n it w as in th e 1960s an d early 1970s. The 1981 Personal Incom e Tax Guts: A Retrospective Look at Their Effects on the Federal Tax Burden Jo h n A. Tatom T -■L HE tax stru ctu re in 1984 is an ex cellen t w ater sh ed from w h ich to asse ss th e effects o f th e 1981 p e r son al in co m e tax ch an g es on th e fed eral tax bu rd en . T h is is th e first y ea r in w h ich th e p h ased red u ctio n o f margined tax rates b ec a m e fully effective; it is th e last y ea r in w h ich th e p erso n al tax stru ctu re w as n o t in dexed. U nd er th e 1981 tax act, th e b rack ets u sed to co m p u te p erso n al in co m e tax liability will be in d exed to inflation beg in n in g in 1985. Sin ce 1981, analysts have exam in ed th e effects o f th e se tax ch an g es u sin g various assu m p tio n s about eco n o m ic p erfo rm an ce. Som e an aly sts fo cu sed only on th e 23 p e rce n t rate red u ctio n s, suggesting th at taxes w ere bein g red u ced . C asual observers q u e s tio n ed th e relev ance o f su ch a view, sin c e it w as dif ficult, esp ecially at th e individual o r fam ily level, to observe any actu al re d u ctio n in tax b u rd en . O th er analysts co m p ared th e rate red u ctio n s to indexing, suggesting th at inflation w ould raise nom in al in co m es a n d add to th e tax b u rd en , roughly offsetting th e effect o f rate re d u ctio n s.1 M ore recently, som e analysts have attem p ted to u se p o st-1981 d ata from in co m e tax re tu rn s to analyze th e im p act o f th e tax rate ch an g es on John A Tatom is an assistant vice president at the Federal Reserve Bank of St. Louis. Thomas A Polimann provided research assistance. 'See Meyer and Rossana (1981), Meyer (1983), McKenzie (1982) and Tatom (1981, 1984) for discussions of the absence of tax reductions due to bracket creep. a ctu al rep o rted tax b u rd en s.2 Ironically, w hile early analyses requ ired a ssu m p tio n s abou t 1 9 8 1 -8 4 e c o n o m ic d evelopm en ts, re ce n t an aly ses often have n e g lected th e effect o f ch an g in g e c o n o m ic co n d itio n s on th e ir co n clu sio n s. T h is article ex am in es th e effects o f th e p erso n al in co m e tax rate re d u ctio n s on th e b u rd en o f federal taxes.3 T h e im p a ct o f assu m p tio n s abou t th e 1 9 8 1 -8 4 eco n o m ic co n d itio n s, p articu larly inflation, is m ini m al sin ce th e se co n d itio n s are n o w largely know n. Alternative assu m p tio n s are em ployed , how ever, to highlight th e im p o rta n ce o f ch a n g es in real in co m e. T h e effects o f th e tax law are stan d ard ized by ex am in ing th e ch a n g e in th e tax b u rd en facing th ree re p re sentative h o u se h o ld s: fam ilies w ith th e 1980 m ed ian fam ily in co m e, an d fam ilies th at earn ed o n e-h a lf or tw ice th e m ed ian level. 2Gwartney and Stroup (1984), Wall Street Journal (April 1984) and the Congressional Budget Office (1984) provide examples of the use of actual data without adjustment for changing economic conditions. The shortcomings of ignoring changing economic conditions in the former two cases are noted in Business Week (1984) and in McCulloch, etal. (1984). XDnly personal income and social security taxes are analyzed here; federal excise and corporate income taxes and state and local government receipts are not. These other taxes have risen substan tially since 1980. From 1980 to the first half of 1984, federal excise tax liabilities rose 41 percent to $55 billion, and corporate income taxes rose 5.7 percent to $74.3 billion. State and local government tax receipts rose from $297.4 to $515.1 billion, a 73.2 percent increase over the same period. 5 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 Table 1 The 1980 Federal Tax Burden at Three Levels of Income 1980 Income Personal Income Tax Average Tax Rate Marginal Tax Rate Employee-Paid Social Security Tax Twice median income One-half median income Median income One wage earner Two wage earners $10,500 $21,000 $42,000 $42,000 $454 $2,505 $9,366 $9,366 11.9% 24.0% 22.3% 43.0% 22.3% 43.0% $1,287 $1,588 $2,575 10.5% 22.1% 18.1% 30.1% 26.1% 43.0% 28.4% 49.1% 16.6% 28.3% 24.2% 36.3% 29.9% 43.0% 34.6% 55.3% 4.3% 16.0% $644 Personal Tax Plus Employee-Paid Social Security Tax Average Tax Rate Marginal Tax Rate Total Tax Burden1 Average Tax Rate Marginal Tax Rate ’Includes personal income tax and employee- and employer-paid social security tax. T h e fed eral p erso n al in co m e tax h as b eco m e in creasingly co m p lex . D ifferen ces in th e eco n o m ic cir cu m sta n ce s an d c h o ic e s m ad e by h o u seh o ld s led to different taxes in 1980 o r 1984 an d to d ifferent tax ch an g es even fo r h o u se h o ld s w ith th e sam e in co m e levels. In tere ste d read ers m ay w ish to pull ou t th e ir ow n 1980 fed eral in co m e tax retu rn an d prelim inary data for 1984 to d eterm in e th e o u tco m e for th eir h o u seh o ld . Are y ou b e tte r off, taxw ise, in 1984 th a n in 1980? Do th e ch an g es in y o u r tax b u rd en sin c e 1980 suggest th at y o u r tax ch an g es are a so u rce o f re cen t an d prosp ective d eficits? THE 1 9 8 0 TAX BURDEN T h e m ed ian fam ily in c o m e in 1980 w as ab o u t $21,000.4 T able 1 show s th e 1980 fed eral p erso n al in- 4ln 1980, the median family money income was $21,023. The median measure indicates the level at which one-half of all families receive more income and one-half receive less. The average size family in 1980 contained 3.27 members and the average number of wage earners per family was 1.63. The range of income in 1980 consid ered here encompasses most families. In 1980, 18.9 percent of families had incomes below $10,000 and 13.5 percent of families had incomes in excess of $40,000. See Statistical Abstract of the United States (1982-83), pp. 432-34. 6 co m e tax an d Social Secu rity tax liabilities for th is level o f in co m e an d for o n e -h a lf a n d tw ice th is m ed ian in co m e. In co m p u tin g p erso n a l taxes, it is a ssu m ed th at th ere are fou r p eo p le (exem ptions) in e a c h h o u s e hold, th at a jo in t retu rn is filed, th a t all in co m e is ad ju sted gross in co m e an d th at th e re are n o o th e r d ed u ctio n s, cred its o r in co m e a d ju stm en ts. In 1980, th e em p lo y ee-p a id So cia l S e cu rity tax eq u aled 6.13 p e rce n t o f w ages u p to a m axim u m of $25,900, w ith an eq u al am o u n t b ein g co lle cte d from th e em ployer. Sin ce th e co st o f em p lo y m en t in clu d es b o th p aym en ts, th e tax b u rd en b o rn e by th e recip ien ts o f th e resp ectiv e in co m e levels are given b o th ways: in clu d in g an d ex clu d in g th e em p lo y er-p aid Social Se cu rity tax. It is th e form er th a t re p rese n ts th e total federal tax b u rd en .5 T h e analysis h ere co n c e rn s w age 5Social security taxes are measured as a percent of “income.” The employer-paid portion, however, is deducted before the income is measured. As a percent of wage earnings up to the maximum tax base, the employer-paid tax is t/1 + t on average and at the margin, where t is the statutory rate on wage “income.” Whether an increase in the employer-paid social security tax is borne from nominal takehome wage reductions or by product price increases is not important here. In either case, the real wage, the purchasing power of wages, is reduced. For discussions of this "incidence" issue, ad well as thorough discussions of the tax system and its effects, see Pechman (1983) and Musgrave and Musgrave (1976). DECEMBER 1984 FEDERAL RESERVE BANK OF ST. LOUIS in co m e; th e overall tax bu rd en , at th e p erso n al level, on su ch cap ital in co m e as dividends, or in tere st is lim ited to th e p erso n al in co m e tax rates. T h e ad d i tio n al taxatio n o f in co m e from cap ital at th e co rp o ra te level, how ever, is generally g reater th a n th e ad d ition al b u rd en o f Social Secu rity taxes sh o w n h ere.6 T h e tax bu rd en is m easu red in tw o w ays: by th e average tax rate an d th e m arginal tax rate. T h e average tax rate is sim ply th e am o u n t o f taxes p aid p e r d ollar o f total in co m e. T h e m arginal tax rate is th e in crea se in fed eral tax liability p e r d ollar o f ad d itional in co m e; it is th e relevant m easu re o f th e im p act o f th e fed eral taxes on in centives to work, save an d invest. B oth m easu res tire sh o w n in table 1. T h e tax calcu latio n s apply to a o n e- o r tw o-w agee a m e r fam ily at th e $10,500 an d $21,000 levels. At $42,000, how ever, th e taxes are ca lcu la ted for bo th o n e-w ag e-eam er an d tw o -w ag e-eam er fam ilies. For th e latter, it is assu m ed th at e a c h w age e a rn e r earn s less th an th e Social Secu rity m axim u m tax b ase o f $25,900 in th at year. If o n e w orker’s earnin gs ex ceed th is b ase in 1980, th en th e relevant m arginal tax rate ap p licab le for th e high w ag e -e am er is th a t in d icated in th e on e-w orker calcu latio n , w hile th e rate ap p licab le for th e low w agee a m e r is th at in d icated for th e tw o-w orker ca lcu la tion. T h e average tax rates for su ch a fam ily are in th e range b o u n d ed by th e average tax rates for th e o n e- or tw o -w ag e-eam er fam ilies. F or exam ple, if o n e w orker earn s $26,000 an d th e o th e r earn s $16,000, th e form er faces an overall m arginal tax rate o f 43 p e rce n t, w hile th e la tter faces a m arginal tax rate o f 55.3 p e rce n t .S u c h a h o u seh o ld h ad tin average tax rate o f 34.5 p ercen t, b ased on th e $9,366 p aid in p erso n al in co m e taxes, th e m axim u m Social Secu rity pay m en t o f $3,175 by th e high w age-eam er, an d $1,962 paid in Social Secu rity for th e low w ag e -e am er for a to tal o f $14,503 on $42,000 o f in co m e. Som e G eneral P roperties o f the Federal Tax Structure T h e data in table 1 provide n o t only a ben ch m ark from w h ich to asse ss 1 9 8 1 -8 4 tax rate ch an g es, but also an illu stration o f so m e im p o rtan t p ro p erties o f th e tax system . M oving from left to right in th e table, o n e observes h o w m arginal and average tax rates rise as in co m e rises, b e c a u se th e m arginal tax rate ex ceed s 6See Joines (1981), for example, for a discussion of the differential taxation of capital and labor income. th e average tax rate. In ad d ition , o n e ca n observe the relative im p o rta n ce o f so cial secu rity taxatio n on both average an d m arginal tax rates. At th e low in co m e, th e em p loy ee-p aid Social S e cu rity tax (on e-h alf th e total) ex ceed s th e p erso n al in co m e tax liability. Even at th e 1980 m ed ian in co m e, th e total Social Secu rity tax liability [(.1226)($21,000) = $2,575) e x c e e d s th e p e rso n a l in c o m e tax liability ($2,505). M oreover, th e Social Secu rity tax is regressive since, at w ag e-in co m e levels above $25,900 in 1980, the m arginal Social Secu rity tax rate is zero. T h u s, th e gap b etw een th e average o r m arginal p erso n al in co m e tax rates an d th e average o r m arginal tax rate m easu res o f th e total b u rd en n arrow s as in co m e m oves above $25,900. F or exam p le, at $42,000 (one w orker), th e dif fere n ce b etw een th e overall tax b u rd en an d p erso n al in co m e tax average rates is on ly 7.6 p e rce n ta g e p o in ts (29.9 - 22.3); for th e m arginal tax rates, th e d ifference is zero. At th e lo w er tw o in co m e levels, th is difference is 12.3 p ercen ta g e poin ts. THE CASE FO R THE PERSONAL INCOME TAX RATE REDUCTIONS Although o n e arg u m en t favoring th e m arginal tax rate cu ts u n d e r th e 1981 tax a ct is essen tially a n o rm a tive case, it ca n b e illu strated u sin g th e d ata in table 1. T h e m arginal tax rates sh o w n ap p e a r to b e “h ig h ,” even at relatively low levels o f in co m e. In th e ca se o f a tw o-w orker co u p le earn in g $42,000, w ith e a c h earn in g less th a n $25,900, e a c h w orker faced a m arginal tax rate o f over 50 p e rce n t (55.3 p ercen t). A stron g er ca s e for th e 1981 ra te-red u ctio n legisla tion ca n b e m ad e b a se d o n w h at w ould have h a p p en ed to tax b u rd en s if th e tax ch a n g es h ad n o t b een m ade. Had n o in c o m e tax rate ch a n g es b ee n a p proved, inflation w ould have p u sh ed all fam ilies into h ig h er tax b rackets. C ou p led w ith existin g provisions for Social Secu rity tax atio n in 1980, th e se in crea ses w ould have raised th e average an d m arginal tax b u r d en substantially, even if th e p u rch a sin g p ow er of fam ily in co m e (real in com e) h ad b ee n u n ch an ged . T h e se effects are sh o w n in table 2.7 In co m e in table 2 eq u als th e 1980 levels a d ju sted for th e 26 p e rce n t in crea se in th e g en eral level o f p rices (co n su m er p rice in d ex for all u rb an co n su m ers) from 7ln 1981, the strongest case for a tax cut was based on the mounting tax burden since 1965. A comparison of the 1980 families tax bur den using 1965 and 1980 rates is given in the appendix. 7 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 Table 2 What the 1984 Federal Tax Burden Would Have Been under the 1980 Personal Income Tax Law: No Change in Real Income 1984 Income Personal Income Tax Average Tax Rate Marginal Tax Rate Employee-Paid Social Security Tax Twice 1980 median income One-half 1980 median income 1980 median income One wage earner Two wage earners $13,230 $26,460 $52,920 $52,920 $923 $3,906 $14,249 $14,249 14.8% 28.0% 26.9% 49.0% 26.9% 49.0% $1,773 $2,533 $3,546 13.7% 24.7% 21.5% 34.7% 31.7% 49.0% 33.6% 55.7% 20.7% 31.7% 28.5% 41.7% 36.7% 49.0% 40.6% 62.7% 7.0% 18.0% $886 Personal Tax Plus Employee-Paid Social Security Tax Average Tax Rate Marginal Tax Rate Total Tax Burden1 Average Tax Rate Marginal Tax Rate 'Includes personal income tax and employee- and employer-paid social security tax. 1980 to 1984; sin ce in co m e rises at th e sam e rate as p rices, n o real in c o m e gain o cc u rs. T h e 1980 tax tab les are u sed to co m p u te th e p erso n al tax liabilities. T h e Social Secu rity tax ca lcu la tio n s in clu d e b o th th e rate in crea se to 13.7 p e rce n t (6.7 p e rce n t for em ployeepaid an d 7.0 p e rce n t fo r em p lo y er-p aid co m p on en ts) an d th e 46 p e rce n t rise in th e tax b ase to $37,800, provided u n d e r th e 1977 an d 1983 Social Secu rity A ct am en d m en ts.8 D esp ite u n ch a n g e d real in co m es, th e fam ilies in table 2 w ould have b e e n su b je ct to su b stan tial ju m p s in th e ir tax bu rd en s from 1980 to 1984 u n d e r th e 1980 tax law. C om p ared w ith 1980, th e to tal tax bu rd en , “Social security taxes have an unusual feature in 1984 only, which does not affect the total burden of taxation, but does affect the calculations of the mix of the tax liability. Under the 1983 amend ments, the Social Security tax rate in 1984 is 14 percent, instead of the 13.4 percent established in 1977 for 1984 or the 13.7 percent used here. The employee-paid portion of 7 percent is actually levied at a 6.7 percent rate, with the remainder (0.3 percent) paid from personal income taxes through a “tax credit” to Social Security funds. For purposes here, the Social Security tax in 1984 is 6.7 percent paid by employees and the employer-paid component is 7.0 percent. 8 m easu red by taxes p e r d o llar o f in co m e, sh o w n at the bo tto m o f tables 1 an d 2, w o u ld have risen by 17.8 p e rce n t for th e m e d ia n -in co m e fam ily (28.5 p e rce n t divided by 24.2 p e rce n t = 1.178), 17.3 p e rce n t for a tw o-w orker, h ig h -in co m e fam ily an d over 22 p e rce n t for th e lo w -in com e an d one-w orker, h ig h -in co m e fam ilies.9 Bracket creep , th e taxatio n o f pu rely in flation-ind u ced ch a n g es in w ages, w ou ld have ra ised th e aver age tax rate for th e p erso n al in co m e tax by over 20 p e rce n t in m o st ca s e s (see in sert o n p ages 10 an d 11). 9These percentage increases in the tax burden measure the rise in taxes as a percent of income, cents paid in taxes per dollar of income, on average. Similar calculations can be made for the marginal tax rate. Besides providing a meaningful measure of changes in the tax burden, percentage changes in the average tax rate provide a convenient approximation to percentage changes in nominal taxes. The latter is roughly the sum of the percentage change in nominal income and the percentage change in the aver age tax rate. Some analysts emphasize percentage-point changes in taxes; for example, a rise in the average or marginal tax rate from 5 to 10 percent is viewed as a 5 percentage-point rise instead of a 100 percent increase in taxes per dollar of income. The data for such calculations are provided in the tables, but the percentage-point calculations are not important here. FEDERAL RESERVE BANK OF ST. LOUIS T h e rise for th e low est in co m e level, from a 4.3 to a 7.0 p e rce n t average tax rate, w ould have b ee n a staggering 63 p e rce n t in crease. Even m arginal tax rates w ould have risen sharply d esp ite th e u n ch an g ed real in co m e. T h e ch an g e from table 1 to table 2 in d icates th at total m arginal tax rates w ould have risen by 12 to 15 p e rce n t u n d e r 1980 tax law s. T h e se relatively large p e rc e n ta g e in c r e a s e s are a s s o c ia te d w ith m u c h sm aller ch an g es in th e m arginal tax rate for th e p e r so n al in co m e tax o f 2 to 6 p ercen tag e p o in ts and a 1.44 p ercen tag e-p o in t in crea se in th e m arginal tax rate for Social Secu rity (12.26 p e rce n t to 13.7 p ercen t). H igher Real In co m e R aises the Federal Tax B urden Of co u rse, average and m arginal tax rates actu ally w ould have in creased m ore th a n th e co m p ariso n of tables 1 and 2 in d icates, b eca u se o f typ ical real in co m e in crea ses and th e progressive p erso n al in co m e tax system . From 1980 to 1984, real GNP p e r cap ita rose abou t 8 p ercen t, o r slightly less th a n 2 p e rce n t p er year. If ea ch o f th e fam ilies in table 2 had ex p erien ced sim ilar grow th in th e ir real in co m es, th e ir in co m es w ould have b ee n 8 p e rce n t h ig h er th a n th o se show n in table 2 an d th e ir tax b u rd en s w ould have b een h ig h er as well, given th e progressive p erso n al in co m e tax. T h e overall average tax rates in table 2 w ould have risen by 2.5 p e rce n t to 4.2 p e rce n t above th o se show n in table 2. F or th e 1980 m ed ian -in co m e fam ily show n in table 2, th e p erso n al in co m e tax average rate, th e co m p o n en t o f th e tax system m o st sensitive to real growth, w ould have risen from 14.8 p e rce n t to 15.7 p ercen t, a 6.1 p e rce n t rise d ue to 8 p e rce n t real g row th.10 At rela tively low in co m es, th e average tax rate is m o st se n si tive to in co m e ch an g es b ec a u se m arginal tax rates e x ceed average tax rates by th e g reatest am o u n t; 8 p e rce n t real in co m e grow th for th e lo w -in com e fam i lies in table 2 w ould raise th e ir p erso n al in co m e taxes m u ch m ore, so th at th e average tax rate w ould rise from 4.3 ce n ts p e r d o llar o f in co m e to 7 ce n ts p er dollar, an 11.4 p e rce n t rise in th e average tax rate. Su ch real in co m e grow th w ould have raised th e average tax rate for th e h ig h -in co m e fam ily in table 2 by abou t th e ,0The rise in average tax rates with unchanged marginal tax rates arises from the fact that additional income is taxed at the marginal tax rate, which exceeds the average tax rate. It is also this discrep ancy that gives rise to bracket creep for purely inflation-induced increases in nominal income. DECEMBER 1984 sam e p e rce n t as th at for th e m ed ia n -in co m e family. None o f th e fam ilies sh o w n in table 2 w ould have m oved in to h ig h er m arginal tax b ra ck ets d ue to typical real in co m e grow th from 1980 to 1984 u n d e r th e old tax law.11 THE 1981 PERSONAL INCOME TAX RATE REDUCTIONS To offset th e escalatin g tax b u rd en d ue to inflation an d th e rise in m arginal tax rates, w h ich red u ced in centives to earn ad d ition al in co m e th rou gh work, sav ing or investm ent, C on gress approved a 23 p e rce n t cu t in till p erso n al in co m e m arginal tax rates to b e p h ased in fully by 1984. F o r o u r p u rp o ses h ere, th e m a jo r co m p o n en ts o f th e 1981 tax a ct w ere a 23 p e rce n t cu t in all m arginal tax rates, p h a sed in as a 5 p e rce n t cu t in O ctob er 1981, 10 p e rce n t in 1983 an d 10 p e rce n t in 1984, an d th e “in d exin g " o f b rack et in co m es an d p e r son al ex em p tio n s b eg in n in g in 1 9 8 5 ,12 Other Provisions o f the E conom ic Recovery Tax Act o f 1981 T h ere w ere o th e r im p o rtan t ch a n g es in th e 1981 tax act, esp ecially th e a d o p tio n o f th e a ccelera ted co st recovery system , ex ten d ed in vestm en t tax cred its and red u ctio n s in tax rates on b u sin ess in co m e. T h ese ch an g es have b ee n highly su cce ssfu l in stim u lating b u sin ess in vestm en t an d productivity grow th, as in ten d ed , an d are n o t ex am in ed h ere.13 Tw o o th e r n o n rate provisions h ad im p o rtan t effects on p erso n al in co m e taxes: th e ex ten sio n o f tax-d eferred in co m e treatm en t th rou gh IRAs an d th e all-savers certificates (July 1981 to N ovem ber 1982), an d an ea rn ed in co m e cred it for tw o -w ag e-eam er fam ilies.14 T h e se are n ot "A $21,023 income increased 26 percent for inflation and 8 percent for real growth in 1980 to yield a 1984 income of $28,608, slightly above the income necessary to move into a new bracket. The conclusion in the text holds for this family due to rounding. This family would have jumped one bracket due to inflation (from a 24 percent marginal income tax rate to a 28 percent rate) and another bracket due to typical real income growth (from a 28 percent rate to 32 percent). '2The 23 percent cut arises because the tax rate was cut to 95 percent of its initial level, then 90 percent of this level, then 90 percent of that rate; the final tax rate is (.9)(.9)(.95) or 77 percent of its original level, a 23 percent cut. Differences due to rounding largely account for the departure from 23 percent for the marginal and average personal income tax rate reductions examined in table 3. 13See Ott (1984), Meyer (1983) and Tatom (1981). Also, see the Economic Recovery Tax Act of 1981 for details of other non-rate provisions affecting the personal income tax. 9 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 How Topical Is B racket Creep? T h e table at right show s th e b rack ets for taxable in co m e for m arried p e rso n s filing jo in t retu rn s u n d er 1980 an d 1984 in co m e tax sch e d u les. T h e in co m e b rack ets w ere u n ch a n g e d from 1980 to 1984, excep t th at th e top tw o w ere p h ased o u t b ec a u se of red u ctio n s in th e in co m e level at w h ich th e m axi m u m 50 p e rce n t m arginal tax rate is achieved . F or a fam ily o f four, th e size o f th e b rack ets sp an s in crea se s in in co m e ranging from 15.6 to 46.7 p ercen t. F ocu sin g on th o se b rack ets up to $109,400 o f tax able in co m e, th e average b racket size is 25.7 p e rce n t o f th e in c o m e at th e b o tto m o f th e b racket. T h is is th e m axim u m ex ten t o f in c o m e gain n ecessary to move from o n e b rack et to th e next. Su ch p e rce n ta g e ch an g es in m o n ey in co m e are qu ite easily o b tain ed over fou r-year p eriod s, w h en inflation p ro ceed s at 6 p e rce n t p e r y e a r o r so. W hen real in co m e rises at 2 to 3 p e rce n t p e r year, bracket ch an g es d ue to real grow th alo n e o c c u r for th e average b rack et size only w ith in 8 to 12 y ears. At th e sm allest b rack et d ifferences taxable in co m es of $16,000 an d $35,200, b rack et m ovem ents p ro ceed m u ch m o re rapidly an d th e m arginal tax rate rises qu ite sharply. U nd er th e 1980 tax law, th e m arginal rate at $16,000 o f taxable in co m e w as 24 p ercen t, and, at $35,200, it w as 43 p e rce n t. W ithout ind ex- form ally analy zed h ere. A n o th er im p o rtan t ch an g e w as to en d th e differential tax treatm en t o f cap ital in co m e for relatively h ig h -in co m e fam ilies. In 1980, m arginal p e rso n al in co m e tax rates o n in co m e from cap ital ro se from 54 p e rce n t to 70 p e rce n t as taxable in co m e ro se from $60,000 to $215,400. T h is d istin ctio n w as d ro p p ed in 1982, so th at all taxable in co m e w as su b je ct to th e sam e m arginal tax rate. 14ln 1984, personal income taxes can be reduced by contributions of up to $2,000 to IRA or deferred income plans that were not allowed for many taxpayers in 1980. As a percent of income, these benefits are, in the limit, equal to the marginal tax rate times $2,000 divided by income. The new deduction for married couples when both work is limited to 10 percent of the lower income up to $30,000. The benefit sub tracts the marginal tax rate times a maximum of one-half of income for a two-wage-earner family. The maximum reduction in the aver age personal income tax rates in table 3 are thus (0.05 x 14 percent) 0.7 percent at the lowest income, (0.05 x 22 percent) 1.1 percent at the median-income level, and (0.05 x 38 percent) 1.9 percent for the high-income family. Digitized for 10FRASER 1980 and 1984 Personal Income Tax Brackets for Persons Married and Filing Joint Returns Taxable income $ $ $ Income1 Percent change in income in bracket 3,400 to $ 5,500 5,500 to $ 7,600 7,600 to $ 11,900 $ 7,400 to $ 9,500 $ 9,500 to $ 11,600 $ 11,600 to $ 15,900 28.3% 22.1 37.1 $ 11,900 to $ 16,000 $ 16,000 to $ 20,200 $ 20,200 to $ 24,600 $ 15,900 to $ 20,000 $ 20,000 to $ 24,200 $ 24,200 to $ 28,600 25.8 21.0 18.2 $ 24,600 to $ 29,900 $ 29,900 to $ 35,200 $ 35,200 to $ 45,800 $ 28,600 to $ 33,900 $ 33,900 to $ 39,200 $ 39,200 to $ 49,800 18.5 15.6 27.0 $ 45,800 to $ 60,000 $ 60,000 to $ 85,000 $ 85,000 to $109,400 $ 49,800 to $ 64,000 $ 64,000 to $ 89,000 $ 89,000 to $113,400 28.5 39.1 27.4 $109,400 to $162,400 $162,400 to $215,4002 $215,400 and over2 $113,400 to $166,400 $166,400 to $219,400 $219,400 and over 46.7 31.9 — ’Includes a $4,000 exemption for four dependents. 2These brackets were phased out under the 1981 tax act. The Effects o f the 1981—84 Rate R eductions W ith th e rate re d u ctio n s in clu d ed in th e 1981 tax act, th e th ree fam ilies sh ow n in table 2 faced th e tax b u rd en sh ow n in table 3.15 C om p ared w ith w hat they 15The marginal personal income tax rate for the low-income family here masks the marginal tax burden at lower incomes. For incomes between $6,000 and $10,000, the earned income credit declines at a 12.5 percent rate on additional income. Thus, for a family of four, the marginal personal income tax rate is 12.5 percent for incomes from $6,000 to $7,400,23.5 percent from $7,400 to $9,600, and 24.5 percent from $9,600 to $10,000. At $10,000 the marginal personal income tax on additional income drops to 12 percent and remains there until income reaches $11,600, where it rises to the 14 percent indicated in table 3. Thus, at the margin, the tax burden on families with incomes from $7,400 to $10,000 exceeded that of 1980 median-income families. The situation is even worse for a head of household with one dependent, where the marginal personal in come tax rate of 23.5 percent begins at an income of $6,000 and rises to 26.5 percent as income approaches $10,000. Bracket creep falls most heavily on persons in these brackets because of both the large difference between marginal and average tax rates at low incomes and the complicated and non-indexed earned income credit. FEDERAL RESERVE BANK OF ST. LOUIS atio n , in flatio n cre a te d a m an ife st p ro b lem o f b racket creep over relatively sh o rt p eriod s o f tim e. B racket creep , how ever, d o es n o t sim ply refer to p erio d ic in flatio n -in d u ced shifts in to h ig h er m ar ginal in co m e tax b rack ets.1 It also in clu d es th e ef fects o f inflation on average tax bu rd en s w ith in a b racket due to in flatio n -in d u ced w age gains. F o r exam ple, co n sid er th e lo w -in com e fam ily in 1980 sh ow n in table 1 in th e text. In 1980, th is fam ily earn ed $10,500, h ad a taxable in co m e o f $6,500 after four p erso n al exem p tio n s an d w as in th e bracket for taxable in co m e th a t rang ed from $5,500 to $7,600. T h e tax in th is b rack et w as $294 plus 16 p e rce n t o f th e ex cess o f taxable in co m e over $5,500. At th e low en d o f th e bracket, th e average tax rate w as 3.1 p ercen t, w hile at th e high en d o f the bracket, th e average tax rate w as 5.4 p e rce n t. T h e low -in com e fam ily at $10,500 paid 4.3 p ercen t. Inflation initially p u sh es u p n o m in al in co m e w ithin th e bracket — in co m e rises from $10,500 to th e top o f th e bracket, $11,600, a 10.5 p e rce n t in co m e in crease. W ith in th e bracket, bracket creep p u sh es th e average tax rate for th e fam ily w ith an u n ch an g ed real in co m e from th e 4.3 p e rce n t aver age tax rate, up to th e 5.4 p e rce n t rate before a bracket rate ch an g e is triggered, fu rth er a cc e le ra t ing th e clim b in th e average tax rate. T h e rise in th e average tax rate w ithin th e bracket arises b ec a u se o f th e fixed n o m in al value o f th e exem p tions, w h ich d eclin e in real value b ec a u se o f inflation and b ec a u se th e m arginal tax rates applied to th e in flatio n -in d u ced in co m e ch an g es exceed 'This point is commonly confused. Bracket creep occurs if mar ginal tax rates exceed average tax rates. Its existence does not depend on rising marginal tax rates. w ould have b e e n (table 2), taxes w ere red u ced su b stantially. F o r th e p erso n al in co m e taxes co n sid ered alone, th e cu ts in average an d m arginal tax rates w ere clo se to th e target. Average tax rates fell by 22.9 to 23.6 p e rce n t for th e th ree fam ily in co m es. Similarly, m ar ginal tax rates fell by 21.4 to 22.4 p ercen t. B ut th e resu lts show n in table 2 never actu ally o c cu rred . A co m p ariso n o f table 3 w ith th e table 1 tax b u rd en s, th e actu al taxes paid in 1980, in d icates th e effect o f th e 1981 rate ch an g es o n actu al tax bu rd en s, w ith n o real in co m e ch an g es. Again, focu sin g o nly on th e p erso n al in co m e tax liability, it ap p ears th at tax b u rd en s w ere red u ced . F o r th e m e d ian -in co m e fam ily, th e average p erso n al in co m e tax rate fell from 11.9 DECEMBER 1984 th e average tax rate. F o r exam ple, for th e 1980 lowin co m e family, th e m arginal rate o f 16 p e rce n t ex ceed ed th e 4.3 p e rce n t average tax rate sh ow n in table 1 in th e text. T h u s, a $1,000 rise in in co m e resulting solely from abou t a 10 p e rce n t in crea se in all p rice s w ould b e taxed at th e m arginal rate o f 16 p ercen t, ad d ing $160 to th e $454 p aid o n th e low er in co m e in stea d o f at th e average rate o f 4.3 p ercen t, o r $43. As a result, taxes o f ($160 + $454) $614 on th e h ig h er in co m e o f $11,500 w ould y ield an average tax rate, o r tax p e r d o llar o f in co m e, o f 5.3 p ercen t. If th e $1,000 gain in in co m e h ad resu lted from real in co m e grow th, n o t from inflation, th e rise in th e tax b u rd en w ould b e co n siste n t w ith th e "verti cal equity" p rin cip le bu ilt in to th e progressive in co m e tax; th is p rin cip le is th at h ig h er real in co m e fam ilies sh ould pay h ig h er average tax rates. W hen th e $1,000 gain reflects in flatio n -in d u ced b racket creep , how ever, fam ilies w ith th e sam e real in co m e w ill pay h ig h er average tax rates after p rices rise th an th ey did before. T h e in tertem p o ral ch a n g e in th e tax b u rd en on a fam ily w ith th e sam e real in co m e violates th e h o rizo n tal eq u ity p rin cip le th at "equals sh ou ld b e taxed equally." T h e sensitivity o f th e average tax rate to ch an g es in in co m e, w h eth er d ue to p rice in crea ses o r real in co m e gains, is in d ica ted by th e ratio o f th e m a r ginal tax rate to th e average tax rate at any level o f in co m e.2 T h is ratio is largest at relatively low in co m e levels. T h u s, a given p ercen ta g e rise in in co m e raises th e average tax rate th e m o st at low in co m e levels; sim ilarly, a given re d u ctio n in real in co m e red u ces th e average tax rate m o re at low in co m e levels th a n at high on es. 2The elasticity of the average tax rate with respect to income is the ratio of the marginal to the average tax rate minus 1. p e rce n t in 1980 to 11.3 p e rce n t in 1984, a 5 p e rce n t red u ctio n ; th e m arginal tax rate fell from 24.0 p e rce n t in 1980 to 22.0 p e rce n t in 1984, an 8.3 p e rce n t cu t. T h e se ch an g es are sh ow n in tab le 4. F o r all th ree groups, th e m arginal tax rates fell, b u t by far less th an th e 22 p e rce n t observed w h en co m p arin g tab les 2 and 3. F o r 1980 m ed ia n -in co m e taxpayers an d high er-in co m e fam ilies, average p erso n a l in co m e taxes d e clin ed , but, again, by m u ch le ss th a n 22 p e rce n t. At the relatively low in co m e level, how ever, th e average tax rate actu ally r o s e from 4.3 to 5.4 p e rce n t, a 25.6 p e rce n t in crease. It sh ou ld be em p h a siz ed th at th e m o d est d eclin es in th e p erso n al in co m e tax rates from 1980 to 1984 11 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 Table 3 The 1984 Federal Tax Burden For Selected 1980 Real Incomes 1984 Income Twice 1980 median income One-half 1980 median income 1980 median income One wage earner Two wage earners $13,230 $26,460 $52,920 $52,920 $711 $2,994 $10,958 $10,958 11.3% 22.0% 20.7% 38.0% 20.7% 38.0% $1,773 $2,533 $3,546 12 . 1% 20.7% 18.0% 28.7% 25.5% 38.0% 27.4% 44.7% 19.1% 27.7% 25.0% 35.7% 30.5% 38.0% 34.4% 51.7% Personal Income Tax 5.4% 14.0% Average Tax Rate Marginal Tax Rate $886 Employee-Paid Social Security Tax Personal Tax Plus One-Half Social Security Tax Average Tax Rate Marginal Tax Rate Total Tax Burden1 Average Tax Rate Marginal Tax Rate 'Includes personal income tax and employee- and employer-paid social security tax. Table 4 Changes in Tax Burdens From 1980 to 1984 for Selected Incomes: No Real Income Growth Twice 1980 median income One-half 1980 median income1 1980 median income1 One wage earner1 Two wage earners1 25.6% -1 2 .5 -5 .0 % - 8 .3 -7 .2 % -1 1 .6 -7 .2 % -1 1 .6 15.2 - 6 .3 - 0 .6 - 4 .7 - 2 .3 -1 1 .6 - 3 .5 - 9 .0 15.1 -2 .1 3.3 -1 .7 2.0 -1 1 .6 -0 .6 - 6 .5 Personal Income Tax Rates Average Marginal Personal Income Tax Plus EmployeePaid Social Security Rate Average Marginal Total Tax Rate Average Marginal ’Percent change; excludes "deduction for a married couple when both work." http://fraser.stlouisfed.org/ 12 Federal Reserve Bank of St. Louis DECEMBER 1984 FEDERAL RESERVE BANK OF ST. LOUIS Table 5 1980-to-1984 Changes in Tax Burdens for Selected Incomes: Real Income Gain of 8 Percent Twice 1980 median income One-half median income 1980 1984' 1980 median income Percent change 1980 Percent 1984’ change One wage earner 1980 1984' Percent change Two wage earners 1980 1984' Percent change Personal Income Tax Rates Average Marginal 4.3% 6.0% 39.5% 16.0 14.0 -1 2 .5 11.9% 12.1% 24.0 22.02 1.7% - 8 .3 22.3% 22.0% -1 .3 % -1 1 .6 43.0 38.0 10.5 22.1 12.7 20.7 21.0 - 6 .3 18.1 30.1 18.8 28.72 3.9 - 4 .7 26.1 43.0 26.4 38.0 1.1 -1 1 .6 28.4 49.1 28.7 44.7 1.1 -9 .0 16.6 28.3 19.7 27.7 18.7 -2 .1 24.2 36.3 25.8 35.72 6.6 - 1 .7 29.9 43.0 31.0 38.0 3.7 -1 1 .6 34.6 55.3 35.7 51.7 3.2 - 6 .5 22.3% 22.0% -1 .3 % 43.0 38.0 -1 1 .6 Personal Income Tax Plus Employee-Paid Social Security Rate Average Marginal Total Tax Rate Average Marginal 'Excludes "deduction for a married couple when both work." 2lncome is $23 below next personal income tax bracket, where the marginal tax rate rises 3 percentage points. sh ow n in table 4 w ere fortu itou s. T h ey o ccu rred p ri m arily b eca u se inflation w as n o t high en o u gh to e n tirely erod e aw ay th e gains from th e p erso n al in co m e tax cu ts for som e fam ilies. T h e 6 p e rce n t average in flation rate over th e fo u r y ears w as w ell b elo w th e 7.8 p e rce n t average rate p ro je c te d by th e ad m in istratio n in 1981. Even th at forecast w as view ed as a rosy s c e n ario at th e tim e; for exam ple, th e C on gression al Budget Office p ro jecte d a 9.8 p e rce n t average ann u al inflation rate for th e fou r y ea rs.16 In stead o f th e 26 p e rce n t rise in p rices an d in co m e th at o ccu rred d ue to inflation sin c e 1980, th e se fo recasts envision ed 35 an d 45.3 p e rce n t in creases, respectively. E ith e r o u tco m e w ould have led to h ig h er average an d m arginal p e r so n al in co m e tax rates for m o st fam ilies in 1984 th an th ey faced in 1980, d esp ite th e 1981 tax cu ts an d u n ch an g ed real in co m es. average tax rates g en erally in crea sed an d th at m ar ginal tax rates fell on ly slightly fo r 1980 m ed ian - and lo w -in com e fam ilies. O nly tw o-w age-eam er, high -in co m e fam ilies a p p ea r to have receiv ed a slight re d u c tio n in th e ir average ta x rate. O n e-w ag e-earn er fam ilies at th e sam e in co m e level fared w orse, on average, b e ca u se th e rise in th e average tax b u rd en d ue to so cial secu rity tax hikes w as larger for fam ilies th at earn ed m ore th an th e m axim u m so cial secu rity tax b a se in 1980. C hanges in the Actual Tax B urden W h en th e so cial secu rity tax b o o sts sin c e 1980 are taken into acco u n t, how ever, even th e m o d est gains cite d above generally d isap pear. At th e bo tto m o f table 4, th e m easu res o f th e to tal tax b u rd en in d icate th at T h e assu m p tio n o f n o real in c o m e grow th u sed to derive th e tax rates in ta b le 3 is a p p rop riate fo r a sse ss ing th e tax cu t effects alo n e. A ctu al ta x ch a n g es from 1980 to 1984, how ever, in clu d e n o t o n ly th e effects of inflation on in co m e an d th e tax law ch an g es, b u t also th e effects o f real in co m e ch a n g es o n in co m e. F am ilies typically earn ed h ig h er real in c o m e in 1984 th a n in 1980 an d paid h ig h er ta x b u rd en s b e c a u s e o f th e p ro gressive in co m e tax. '6See Congressional Budget Office (1981), p. 4. B epresentative a ctu a l tax b u rd en ch a n g es fo r the 1980 m ed ia n -in co m e fam ilies are sh ow n in table 5. T h ere, n om in al in co m e (from tab le 2) h a s b e e n raised 8 13 FEDERAL RESERVE BANK OF ST. LOUIS p e rce n t to reflect th e rise in p e r cap ita real GNP over th e 1 9 8 0 -8 4 p erio d . T h e table provides a co m p ariso n o f 1980 an d 1984 tax b u rd en s assu m in g th is typical grow th. T able 5 show s th at th e average p erso n al in co m e tax rate r o s e from 1980 to 1984 for 1980 m ed ian - and lowin co m e fam ilies. W hen th e h ig h er 1984 Social Secu rity taxes are in clu d ed , th e overall average tax rate r o s e f o r e v e r y g r o u p sh o w n . M arginal tax rates generally d e clin ed slightly over th e p erio d .17 It is cle a r th a t th e rise in th e tax bu rd en from 1980 to 1984, d esp ite th e e n a c te d tax rate red u ctio n s, fell d is p ro p o rtio n ately o n lo w -in com e g ro u p s.18 In tab le 5, th e rise in th e overall average tax rate is sm aller at h ig h er in co m es, raising th e possibility th at som e highin co m e fam ilies actu ally p aid lo w er average tax rates in 1984 th an in 1980. Ind eed , th ere is a “break-even ” 1980 in co m e level o f $55,537 at w h ich th e 1984 average tax rate u n d e r th e assu m p tio n s above eq u als th at paid in 1980. O nly ab o u t 6 p e rce n t o f tax retu rn s h ad an in co m e in e x c e ss o f $50,000 in 1980. M ore im portant, th ese re tu rn s to taled ab o u t 15.9 p e rce n t o f all taxable in co m e. M oreover, th e tax re d u ctio n s from 1980 to 1984 for th e se taxpayers w ere generally qu ite sm all e ith e r as a p e rce n t o f 1980 average tax rates o r in ab solu te p e rce n ta g e-p o in t re d u ctio n s. T h e largest tax red u ctio n s w ere ab o u t 2 p e rce n tag e p o in ts for 1980 in co m es from ab o u t $80,000 to $100,000, w here, u n d er th e assu m p tio n s above, th e average tax w as abou t 40 to 42 p e rce n t in 1980. DECEMBER 1984 first is th at th e tax ra te re d u ctio n s led to low er p e r so n al in co m e taxes for h ig h -in co m e fam ilies bu t little red u ctio n in taxes for lo w -in com e fam ilies. T h e s e c ond m yth is th at p erso n a l fed eral taxes fell from 1980 to 1984 (eith er ab so lu tely o r relative to in com e), th u s con trib u tin g to h ig h er fed eral d eficits. T able 4 clarifies th e so u rce o f th e co n flictin g claim s th a t 1981 tax c h a n g e s e ith e r re su lte d in g rea ter benefits for th o se w ith h ig h er in co m e s o r red u ced m arginal an d average tax rates equally.19 B oth th e p e r son al in co m e an d overall average ta x rate ch a n g es in table 4 in d ica te th a t th e tax in c re a s e s sh o w n th ere fell d isp ro p ortion ately o n lo w er-in co m e fam ilies. T h e dif ferential im p act o f th e tax cu ts sh o w n in table 4, h o w ever, d oes n o t arise from th e tax rate ch a n g es sin ce 1980; in d eed , th e co m p a riso n o f ta b les 2 an d 3 show s th at average an d m arginal tax ra tes w ere low ered by about th e sam e p e rce n ta g e a cro ss in co m e levels by th e tax cu ts en a cte d . T h e d iscrim in ato ry tax ch an g es show n in table 4 aro se from b ra ck et cre e p an d Social Secu rity tax hikes, in crea ses th a t fall d isp ro p o rtio n ately on lo w er-in co m e fam ilies. Fortu n ately, th e g reat est cu lp rit, b ra ck et creep , w as largely elim in ated by th e 1981 tax act, th ou gh n ot u n til 1985.20 T h e se co n d m yth is th at th e tax ch a n g es co n trib u ted to th e su rge in th e d eficit in late 1981 an d 1982, an d to th e m ag n itu d e o f re ce n t an d p ro sp ective defi cits.21 T ab le 5 clearly in d ica tes that, for represen tative fam ilies, th e average tax b u rd en ro se from 1980 to T w o M yths A b o u t th e 1 9 S 1 —S 4 T a x Rate C hanges 19These distributional changes have been noted by Conyers (1984) and Heller (1984), for example. Public d iscu ssio n o f th e 1981 p erso n al in c o m e tax cu ts h as b e e n d o m in ated b y tw o pervasive m yths. T h e “ Proponents of the view that taxes were cut are often leading oppo nents of indexing. See Silk (1984) and Heller (1984), for example. An equally persistent and widespread fallacy concerning the 1981 tax act is that indexing reduces taxes. See Silk, for example. Index ing simply restores “horizontal equity,” the principle that families with equal incomes should be taxed equally. Under indexing, changes in prices from one year to another do not lead to increased average tax rates for families or individuals with unchanged real incomes. Indexing can result in a lower tax burden only if nominal incomes do not keep pace with inflation, that is, if real income falls; a decline in the real tax burden when real income falls, given prices, has been a feature of the U.S. tax system since its inception and is consistent with notions of vertical equity, the tax principle that fami lies with higher incomes should be taxed more than families with low incomes, other things equal. Silk does note, however, the Commit tee for Economic Development’s recognition of the discriminatory impact of bracket creep on low-income families and its removal through indexing. ,7Without rounding the 1980 median income down by $23, the mar ginal personal income tax rate of this group would have risen from 24.0 to 25.0 percent, and the overall marginal rate of this group would have risen from 36.3 percent to 38.4 percent. The maximum marginal tax rate of 50 percent of earned income was achieved at $60,000 of taxable income in 1980 and at $162,400 in 1984. The latter is equivalent to $128,889 in 1980 prices. At earned taxable incomes above this level, the marginal tax rate has been unchanged from 1980 to 1984. 18Business Week (1984) notes that between 1980 and 1984 changes in the distribution of personal disposable real income were such that the top quintile (20 percent of income recipients) gained, while the bottom quintile lost, both by about 8 percent. Families in the second lowest quintile lost close to 2 percent, while those in the third quintile registered a slight gain of about 1 percent. In the fourth quintile, the gain was about 3.5 percent. This pattern reflects the effects of tax changes, spending cuts and the business cycle, with a large share arising from the different increases in the overall average tax rates shown in table 5. http://fraser.stlouisfed.org/ 14 Federal Reserve Bank of St. Louis 2,See Walter W. Heller (1984). He attributes the rise in the deficit to the “huge tax cut” or the "biggest tax cut ever." The alternative cyclical view of recent deficits, which owes much to Heller for its popularization, is developed in Tatom (1984). Hershey (1984) and Harris Bank (1984) echo the frequent claim that personal tax cuts occurred from 1980 to 1984. The former also blames the deficit on such cuts. DECEMBER 1984 FEDERAL RESERVE BANK OF ST. LOUIS 1984. T hu s, p erso n al tax rate cu ts alo n e are not a likely can d id ate as a so u rce o f th e in crea sed federal deficit. W hile p erso n al taxes as a p e rce n t o f in co m e did d e clin e slightly at very high in co m es, th e se red u ctio n s did n o t fully offset th e generally larger in crea ses in tax liabilities o f low er-in co m e groups th at earn th e larger sh are o f in co m e. T h e analysis in d ica tes that, at relatively low in co m es, th e effects o f b racket creep are th e strongest. T h u s, n ot surprisingly, th e 1 9 8 0 -8 4 rise in tax b u rd en s h as b ee n largest at th e lo w est in c o m e levels. T h e se in crea ses w ere rein fo rced b y Social Secu rity tax hikes, w h ich also add d isp ro p ortion ately to th e tax b u rd en o f relatively lo w -in com e h o u se h o ld s an d fam ilies. O f co u rse, federal revenu es w ould have b ee n larger an d th e d eficit co rresp o n d in g ly sm aller in 1984, had th e 1 9 8 1 -8 4 p erso n al in co m e tax rate ch an g es not o ccu rred . A co m p ariso n o f tab les 1 an d 2 show s that 1984 revenues w ould have b e e n abou t 22 p e rce n t larger u n d e r th e old tax sch e d u le. F or fiscal y e a r 1984, actu al p erso n al in co m e taxes a m o u n ted to abou t $300 b illion; this w ould have b ee n ab o u t $85 billio n larger u n d e r th e 1980 tax rates. T h is “lo ss," how ever, w as m ore th an offset by th e effect o f in flation alo n e on federal tax receip ts.22 T h e ap p aren t d eclin e in th e size o f taxes relative to GNP w as largely d u e to th e cy clical d eclin e in th e eco n o m y an d to cu ts in b u sin e ss taxes. Tax reform is high on th e p o litical agenda, b u t som e o f th e im p licatio n s o f th e analysis h ere have n o t b ee n cen tra l to th e d iscu ssio n .23 Su pply-sid e an aly sts co u ld co n clu d e from th e an alysis h e re th at little effective cu ttin g o f m arginal tax ra tes h a s resu lted from th e 1 9 8 1 -8 4 ch an g es. T o th e ex ten t s u ch ch a n g es are d e sirable, a n ew initiative w ould b e in o rd er. At least th ree re cen t reform p ro p o sals in clu d e sh arp re d u c tio n s in m arginal tax rates.24 Against a b ackd ro p o f an in d exed tax system , a n o th e r ro u n d o f s u c h cu ts w ould b e m o re likely to b e effective. SUMMARY AND IMPLICATIONS P ersonal in co m e tax rate red u ctio n s w ere offset by bracket creep and in crea sed Social Secu rity taxes for m o st fam ilies betw een 1980 an d 1984. T yp ical h o u s e hold s, w h o se in co m e m erely kept p a c e w ith inflation an d econ om y-w id e real in co m e gains d urin g th e past fou r y ears, faced h ig h er average tax rates in 1984 th an th ey did in 1980. Although th is m ay seem im plau sible given th e large d eclin es (about 22 p ercen t) in m arginal an d average tax rates provided by th e 1981 tax act, it is easily explained . T h e failure o f tax rates, o n average, to d eclin e is th e resu lt o f b o th th e m assive ex ten t o f bracket creep p ro d u ced by inflation over th e 1 9 8 0 -8 4 period an d th e sharp rise in Social Secu rity taxes sin ce 1980. T h e m ost im p o rtan t u n d e rcu rre n t o f th e analysis h ere is th e role o f in d exatio n in elim in atin g bracket creep . S u ch in d exation, as p rovided in th e 1981 tax act, will begin n ex t y ear. C ontrary to m o st d iscu ssio n s, in d exation will n o t lo w er average tax rates o r taxes p er d o llar o f in co m e, u n less real in co m es d eclin e. Instead , in d exatio n allow s in flatio n -in d u ced in co m e ch a n g es to b e taxed at average tax rates, n o t at h ig h er m arginal tax rates that w ould p u sh u p taxes faster th an in co m es, even if real in co m es are u n ch an g ed . REFERENCES Bureau of the Census. Statistical Abstract of the United States: 1982-83 (103d edition), Washington, D.C., 1982. Bureau of Economic Analysis. 1984), p. 9. Survey of Current Business (August Business Week. "The Reagan Tax Cuts: Were the Supply Siders Right?" (May 28,1984), pp. 68-69. Congressional Budget Office. "An Analysis of President Reagan’s Budget Revisions for Fiscal Year 1982," Staff Working Paper, Congress of the United States (March 1981). _________ “Effects of Major Changes in Individual Income and Excise Taxes Enacted in 1981 and 1982 For Households in Differ ent Income Categories,” Staff Memorandum (March 1984), processed. Conyers, John, Jr. 1984. “What Recovery?” Washington Post, October 3, Economic Diary. “Who Is Better Off Under Reagan — And Who Isn’t,” Business Week (October 22,1984), p. 14. Economic Recovery Tax Act of 1981: Law and Explanation (Com merce Clearing House, August 1981). Gwartney, James, and Richard Stroup. “The Redistributionist Tax Reduction,” Wall Street Journal, June 26,1984. Harris Bank. “ATaxing Development," Barometer of Business (Sep tember/October 1984). Heller, Walter W. “The Unavoidable Issue," Wall Street Journal, October 26,1984. “ See Miller (1984) and Pechman (1984) for discussions of the recent proposals for tax reform. “ For example, see table 2 in Bureau of Economic Analysis (1984) which indicates that cyclically adjusted receipts rose $121.9 billion due to inflation alone in 1981-83. Data for 1984 are not yet available. 24See Wall Street Journal (November 1984). It points out that three major current reform proposals involve reducing the top marginal tax rate for the personal income tax to 25 to 35 percent from the current 50 percent level. At least one of these proposals, however, the Bradley-Gephardt bill, omits indexing. 15 DECEMBER 1984 FEDERAL RESERVE BANK OF ST. LOUIS Hershey, Robert D., Jr. “The Reagan Economic Legacy,” New York Times, October 28,1984. Musgrave, Richard A., and Peggy B. Musgrave. Public Finance in Theory and Practice, 2nd ed. (McGraw-Hill Book Company, 1976). Joines, Douglas H. "Estimates of Effective Marginal Tax Rates on Factor Incomes,” Journal of Business (No. 2,1981), pp. 191-226. Ott, Mack. “Depreciation, Inflation, and Investment Incentives: The Effects of the Tax Acts of 1981 and 1982,” this Review (November 1984), pp. 17-30. McCulloch, J. Huston, James R. Knickman, Terrence E. Kelly, and Robert P. Springer. Letters to the editor of the Wall Street Journal, May 10,1984. McKenzie, Richard B. “Supply-Side Economics and the Vanishing Tax Cut,” Federal Reserve Bank of Atlanta Economic Review (May 1982), pp. 20-24. Meyer, Stephen A. “Tax Cuts: Reality or Illusion?” Federal Reserve Bank of Philadelphia Business Review (July/August 1983), pp. 3 16. Meyer, Stephen A., and Robert J. Rossana. “Did the Tax Cut Really Cut Taxes?” Federal Reserve Bank of Philadelphia Business Re view (November/December 1981), pp. 3-12. Miller, Glen H., Jr. “Alternatives to the Current Income Tax," Fed eral Reserve Bank of Kansas City Economic Review (September/ October 1984), pp. 3-16. Pechman, Joseph A. Institution, 1983). Federal Tax Policy, 4th ed. (The Brookings _________ Options for Tax Reform (The Brookings Institution, 1984). Silk, Leonard. “Federal Deficit and Indexation," New York Times, October 26,1984. Tatom, John A. "We Are All Supply-Siders Now!" this Review (May 1981), pp. 18-30. ________ . “A Perspective on the Federal Deficit Problem,” this Review (June/July 1984), pp. 5-7. Wall Street Journal. "All Supply-Siders Now," November 6,1984. ________ _ 'T ricklenomics," April 11,1984. APPENDIX The 1965 T ax Structure Before 1981, m arginal tax rates u n d e r th e p erson al in co m e tax h ad n o t b ee n altered sin ce 1965.1 T h e in creasingly o n ero u s b u rd en o f th e level o f average and m arginal tax rates in 1980 sh o w n in table 1 in th e text ca n b e see n by co m p ariso n to th e 1965 in co m e tax stru ctu re. Table A .l show s th e th ree representative 1980 fam i lie s’ tax po sitio n s, from table 1 in th e text, b ased on 1965 taxes an d p rice s for o n e-w ag e -e am er fam ilies. In 1965, th e so cial secu rity tax w as only 3.625 p e rce n t on w ages up to $4,800 for b o th th e em p loyee- an d th e em ployer-paid am o u n t. In 1965 p rices, th e 1980 in co m e levels are co n sid erab ly sm aller, bu t p u rch a sin g p o w er h as b e e n h eld co n stan t. At th e sm aller 1965 n om in al earnings, th e 1980 m ed ian real in co m e ex ceed ed th e m axim u m so cial secu rity tax. It sh o u ld b e n o te d that, at th e in c o m e levels given for 1965, th e 1980 fam ilies h ad co n sid erab ly m o re real in co m e th a n sim ilarly p lace d fam ilies in 1965; th e 1965 m edian-fam ily in co m e w as only $6,957. T h e exam p les in table A .l are for fam ilies th at w ere com paratively 'From 1965 to 1981, many changes did occur in the personal income tax. These changes included alterations in standard deductions and personal exemptions, and changes in the incomes associated with brackets. The number of brackets and bracket rates, however, did not change. 16 b ette r off th a n th e ir 1965 co u n te rp a rts; th e ir real in co m es w ere ab o u t 15.6 p e rce n t above th e resp ective m u ltiples o f m ed ian in co m e in 1965. T h u s, th e ir tax treatm en t re p rese n ts h ig h er tax rates fo r in co m e th an th eir 1965 co u n terp a rts. T h e average p erso n al in co m e tax at e a c h in co m e rose su bstan tially from 1965 to 1980. F o r th e 1980 m e d ian in co m e, th e in cre a se is 22.7 p e rce n t o f th e 1965 tax b u rd en o f 9.7 p e rce n t. Even at th e low in co m e, th e average tax b u rd en ro se sh arp ly (19.4 p ercen t). At tw ice th e 1980 m ed ian in co m e, th e average person al in co m e tax rate ro se from 15.1 p e rce n t in 1965 to 22.3 p e rce n t in 1980, a 4 8 p e rce n t in cre a s e in tax es p e r d ollar o f in co m e, d esp ite n o ch a n g e in real in co m e. T h e m arginal p erso n a l in co m e tax rates ro se sharply as well, in crea sin g 6-2/3 p e rce n t at th e low in co m e, 26.3 p e rce n t at th e 1980 m ed ian an d 72 p e rce n t at th e high in co m e. T h e overall tax b u rd en on th e se u n ch a n g ed real in co m es ballo o n ed m u ch m ore. T h e overall m arginal tax rate on th e 1980 m ed ian in c o m e alm o st doubled, rising from 19 p e rce n t to 36.3 p e rce n t. T h e total m ar ginal tax rate at th e low in co m e ro se from 22.3 p ercen t to 28.3 p ercen t, a 2 7 p e rce n t in crea se, w h ile th a t for th e h ig h -in co m e fam ily ro se 72 p e rce n t. T h e overall average tax rates on th e se real in co m e s ro se 53.7 p e r ce n t for th e lo w -in com e fam ily, 72.9 p e rce n t for th e m ed ia n -in co m e fam ily an d 72.8 p e rce n t for th e high- DECEMBER 1984 FEDERAL RESERVE BANK OF ST. LOUIS Table A.1 The Federal Tax Burden on Selected 1980 Real Incomes in 19651 One-half 1980 median income 1980 median income Twice 1980 median income $10,500 $4,021 $21,000 $8,041 $42,000 $16,082 $143 $779 $2,431 1980 Income 1965 Equivalent 1965 Personal Income Tax Average Tax Rate Marginal Tax Rate 1965 Employee-Paid Social Security Tax 3.6% 15.0% $146 9.7% 19.0% $174 15.1% 25.0% $174 Personal Tax Plus One-Half Social Security Tax Average Tax Rate Marginal Tax Rate 7.2% 18.6% 11.9% 19.0% 16.2% 25.0% 10.8% 22.3% 14.0% 19.0% 17.3% 25.0% Total Tax Burden Average Tax Rate Marginal Tax Rate 'Assume one-wage-earner family for Social Security tax calculations. in co m e family. E x cep t at th e high in co m e, th e biggest sh are o f th e in crea se in th e tax bu rd en , o n average or at th e m argin, w as d ue to in crea ses in b o th th e Social Secu rity tax rate an d its tax b ase. At th e relatively highin co m e level, alm ost tw o-third s o f th e overall average an d m arginal tax bu rd en in cre a se o ccu rred d ue to in flatio n -in d u ced bracket creep . Even at th e 1980 m e dian real inco m e, th e ju m p in th e tax b u rd en d ue to b rack et creep w as su bstan tial. In sum m ary, by 1980, m arginal an d average tax rates at all levels o f in co m e h ad risen d ram atically from 1965 levels d ue to rising Social Secu rity tax rates an d its tax base, an d to th e effects o f in flation p u sh in g fam ilies in to h ig h er average an d m arginal p erso n a l in co m e tax brackets. T h e se fo rces co n tin u ed from 1980 to 1984 and, in th e a b se n c e o f th e 1981 tax cu ts, w ould have fu rth er b o o sted th e tax b u rd en . 17 Real Interest Rates: What Accounts for Their Recent Rise? A. Steven H olland IV -L ^1 OMINAL in tere st rates have risen to u n p re c e d en ted levels in th e last five y ears, an d th e co m m o n p erce p tio n is th at ex p e cted real rates o f in terest — rates m inu s ex p e cted inflation — have risen as well. T h e se h ig h er rates are b lam ed fo r a variety o f e c o n o m ic ills in clu d in g re d u ce d cap ital in vestm en t an d slow dow ns in su c h in terest-sen sitiv e secto rs as h o u s ing an d au to m o biles. T h is p a p e r is co n cern e d , first, w ith establish in g th at real in tere st rates have in d eed b e e n h ig h er d uring th e 1980s th a n in th e previous tw o d ecad es and, seco n d , w ith exam in in g p o ssib le ca u se s o f th is m a jo r shift. Potential ca u se s in clu d e ch an g es in th e ex p e cted rate of inflation, m o n etary policy, th e state o f th e econom y, taxes, fed eral b u d g et d eficits an d th e d eclin in g relative p rice o f energy. ESTIMATES OF BEFO R E- AND AFTER-TAX REAL INTEREST RATES T h e real in tere st rate is n o t know n w ith certain ty at th e tim e a secu rity is p u rch ase d , b u t th e p u rch a se r h as an ex p e ctatio n o f it. T h e n o m in al in terest rate, i, is th e su m o f th e ex p e cted real rate o f in terest, r, an d the ex p ected rate o f inflation, p e: (1) i = r + pe.' T h e ex p e cted real rate, th u s, ca n b e estim ated a cc o rd ing to th e form ula: (2) r = i - p', Proxies for th e ex p e cted rate o f in flation frequ ently are b a sed on w eigh ted averages o f p a st inflation rates or th e p re d icted values from regressio n eq u atio n s in w h ich th e in flation rate d ep en d s o n past inflation rates, p ast ra tes o f m o n ey grow th a n d a n u m b er o f o th e r variables.2 B eca u se em p irical resu lts ca n b e se n sitive to a ssu m p tio n s a b o u t th e w ay ex p e cta tio n s are form ed, how ever, a p o ten tially m o re fruitful a p p ro a ch is to u se “o b serv ed ” in flation fo reca sts to estim ate ex p ected in flation.3 In th is a rticle, d ata from surveys o f b o th sh o rt- a n d lo n g -term in flatio n ex p e cta tio n s are u sed to estim ate sh o rt- an d lo n g -term ex p e cted real rates o f in terest. T h is analysis oversim plifies th e problem , sin c e it ap plies only to th e ex p e cted real before-tSQc yield . Sin ce in terest pay m en ts are taxab le as ea rn ed in co m e, th e ex p e cted real after-tax y ield (r*) is: (3) r* = i — ti — p' = ( l - t ) i - p', w h ere t is th e m arginal tax rate. An estim a te o f th e average m arginal ta x rate o n p e rso n a l in c o m e is u sed below to estim ate ex p e cted after-tax real in tere st rates. T h e estim ates p re sen ted in th is article are in ten d ed to rep resen t th e p a ttern o f re ce n t real in terest rate movements, not to provide co m p letely a ccu ra te esti m ates o f real in terest rates at any p o in t in tim e. P o ten tial so u rces o f erro r in th e estim a tes in clu d e (but are n o t lim ited to): (a) m e a su re m en t erro r in calcu latin g th e ex p ected rate o f inflation, (b) th e effects o f different as lon g as an estim ate o f th e ex p e cted in flation rate is available. A. Steven Holland is an economist at the Federal Reserve Bank of St. Louis. Jude L. Naes, Jr., provided research assistance. 2As pointed out by Santoni and Stone (1982), however, the difficulty with this procedure is that any change in economic policy or any structural change or “shock” that affects inflation expectations will not be incorporated in the estimate of expected inflation. 'This equation is a widely used approximation of the "Fisher equa tion.” See Fisher (1965). 3For an example of the sensitivity of empirical results to assumptions about expectations formation, see Holland (1984). Digitized18 for FRASER FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 Chart 1 Nominal and Real 1-Year Interest Rates I960 62 64 66 68 70 72 74 76 78 80 82 1984 NOTE: D a s h e d lines r e p r e s e n t a v e r a g e le vels o f b e f o r e - a n d a f te r -ta x re a l in te r e s t r ates ov er se le cted tim e pe riods. m arginal tax rates a cro ss m arket p articip an ts an d (c) th e difference betw een th e m arginal tax rate ex p ected to h o ld at th e tim e in tere st p ay m en ts are received an d th e cu rre n t rate.4 W h en ever real in tere st rates are re ferred to in th e follow ing d iscu ssio n , it w ill m ean ex p e cted real in terest rates. Estim ates o f Short-Term Real Interest Rates C hart 1 plots n o m in al retu rn s and estim ates o f the b efore- an d after-tax real retu rn s on o n e-y ear Treasu ry secu rities, b a se d o n o n e-y ear inflation fo recasts from "In addition, the return that is relevant for decision-making depends on risk and the tax burden on alternative uses of funds. More will be said about risk later in the article. See Ezrati (1982) and Mehra (1984) for discussions of the implications of taxes on alternative uses of funds. th e Livingston survey from 1960 to th e first h a lf of 1 9 8 4 .5 B etw een 1960 an d 1970, th e n o m in al rate rose from aro u n d 3 p e rce n t to over 7 p e rce n t. E stim ates o f ^Joseph Livingston of The Philadelphia Inquirer conducts a survey of economists each spring and fall, requesting respondents to indicate their predictions of the consumer price index (CPI). Because the survey results published, for example, in June contain predictions for the following December and June, Livingston refers to them as six- and 12-month-ahead forecasts as this article does. Because the respondents to the June survey are thought to know only the April CPI, however, they are actually predicting eight- and 14-month rates of change. For a detailed discussion of the Livingston expectations data, see Carlson (1977). This article uses the data in Carlson’s revised form updated to the present. The nominal interest rates used in the charts and table are the quarterly averages of the rates for the quarter in which the Livingston survey was taken. The same calculations were made for six-month Treasury bills based on sixmonth inflation forecasts. Since the pattern of movements was nearly identical, however, only the one-year rates are reported. The estimate of the average marginal tax rate comes from Chase Econo metrics. 19 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 Chart 2 Nom inal and Real 10-Year Interest Rates I 96 0 62 64 66 68 70 72 th e ex p e cted real rate in d icate th is w as d u e prim arily to h ig h er ex p e cted inflation, sin c e b o th th e beforea n d after-tax real rates a p p ear to have risen only slightly, if at all, over th e period. B etw een 1971 an d 1980, sh o rt-term n om in al in terest rates, on average, w ere m u ch h ig h er th an in th e 1960s; real rates, for th e m o st part, w ere low er. In fact, esti m ated b efo re-tax real rates w ere b elo w 1 p e rce n t from th e se c o n d h a lf o f 1974 to th e first h a lf o f 1978 an d w ere even negative in late 1976 an d early 1977. A fter-tax real rates w ere negative for n early th e en tire p eriod from 1974 to 1980. N om inal rates in creased d ram atically after 1977, w ith in crea ses o f abou t 200 b asis p o in ts o ccu rrin g in late 1978 an d again in late 1979. T h ese in creases, how ever, served o n ly to bring real rates clo ser to th e levels th a t had prevailed before 1974. F rom late 1979 to early 1982, sh o rt-term n om in al in terest rates w ere h ig h er th a n at an y tim e during th e Digitized 20 for FRASER 74 76 78 80 82 1984 1960s o r 1970s. Sh o rt-term real in tere st rates, how ever, did n o t break w ith p re ce d e n t u n til 1981 w h en b efo re tax real rates clim b ed above th e 6 p e rce n t level; th ey co n tin u ed to rise th rou gh early 1982. After-tax real rates beh aved in a sim ilar fash io n and , o n average, have b e e n h ig h er sin c e 1981 th a n in th e previous two d ecad es. T h e d ifferen ce is n o t a s great, how ever, as it is for before-tax real rates. B o th n o m in al an d real rates have d eclin ed sin c e early 1982, b u t th e y rem ain at veiy high levels relative to p ast history. Estim ates o f Long-Term Real Interest Rates W e ex p e ct lo n g -term real in tere st rates to behave in a m a n n er bro ad ly sim ilar to sh o rt-term real rates; if sh o rt-term ra tes rise, lo n g -term ra tes are fo rced u p so th at real y ield s over an y h o ld in g p erio d are co m p a ra DECEMBER 1984 FEDERAL RESERVE BANK OF ST. LOUIS Table 1 Spreads Between Yields on Ten- and One-Year Treasury Securities (2) (1) Date Nominal rate spread (3) After-tax real rate spread i/1978 0.53 0.97 0.82 11/1978 -1 .0 0 -0 .3 8 -0 .0 9 1/1979 -0 .8 3 0.94 1.15 11/1979 -1 .8 2 0.09 0.57 1/1980 0.20 1.81 1.76 11/1980 -1 .4 2 0.50 0.89 1/1981 -1 .3 8 -0 .5 0 -0 .1 1 11/1981 0.54 -0 .0 2 -0 .1 7 1/1982 0.13 -0 .8 8 -0 .9 1 11/1982 1.54 0.07 -0 .3 0 1/1983 1.36 -0 .2 5 -0 .5 8 11/1983 1.74 0.62 0.23 1/1984 1.65 0.46 0.09 ble w h eth er o n e h o ld s sh o rt- o r long-term bo n d s.6 B eca u se o f d ata lim itation s, how ever, it is m u ch m ore d ifficult to g et an a ccu ra te re p rese n tatio n o f th e m ar k et’s ex p e ctatio n o f inflation over th e d istan t future th a n over th e n e a r fu tu re.7 In fact, it is only sin c e 1978 th at a survey o f ex p e cted inflation over p erio d s su b stantially lo n g er th a n a y e a r h as b ee n u nd ertak en . T h e survey, know n as th e D ecision-M akers Poll, provides estim ates o f ex p e cted inflation over th e n ex t five and 10 y ears.8 'This assumes the absence of segmented markets. In other words, there Is a high degree of substitutability between short- and long term securities. This is not meant to imply that the term structure of interest rates does not change over time, only that short- and long term interest rates behave in a broadly similar fashion. 7lt is also more difficult to know the appropriate tax rate to use in calculating the after-tax yield, since interest payments are made much farther in the future. 8Richard Hoey of Drexel Burnham Lambert, Inc., conducts this sur vey of institutional portfolio managers. Each respondent predicts the rate of change of consumer prices over the next five years and over the five subsequent years. The average of the two provides the estimate of expected inflation over the next 10 years. Since 1980, the survey has been conducted at least four times a year. To facilitate comparison with the shorter-term real interest rate estimates, we use data from surveys taken as close as possible to the dates of the Livingston surveys. There is never more than one month’s difference in the dates of the surveys of the short- and long term inflation expectations used in this paper. In 1978 and 1979, there was only one survey in each year (taken near the middle of the year). These two surveys provided data for the estimates of long term inflation expectations for the first halves of 1978 and 1979. Estimates for the second halves of both years were calculated by interpolation. Before-tax real rate spread C hart 2 plots th e n o m in al yield on 10-year Treasu ry secu rities sin c e 1960, as w ell as estim a tes o f th e 10year, before- an d after-tax real rates sin c e 1978 b ased on th e m ea n in flation fo reca sts from th e survey. As exp ected , th e p a ttern o f m ovem en ts in long-term n o m in al rates d urin g th e 1960s an d 1970s is sim ilar to th at in sh o rt-term rates. In particu lar, w h en sh o rt term n o m in al rates sh o t upw ard in th e late 1970s, so did lon g-term n o m in al rates. L ong-term real rates also rea ch ed h eigh ts co m p arab le to th o se o f sh o rt-term real rates in 1981 an d 1982.9 T h u s, it ap p ears th at the in crea se in lo n g -term real rates o ccu rred at roughly th e sam e tim e an d w as o f roughly eq u al size as th e in crea se in sh o rt-term real rates. The Term Structure o f Real Interest Rates N om inal lon g -term rates have b ee n su bstan tially above n om in al sh o rt-term rates sin c e 1982, reversing th e p attern from th e late 1970s an d early 1980s. T h is is illu strated in co lu m n 1 o f tab le 1, w h ic h gives th e difference b etw een th e y ield s on 10-year an d o n e-year Treasu ry secu rities sin c e 1978. C om p arable differ en c e s for before- an d after-tax real rates, respectively, are p resen ted in co lu m n s 2 an d 3 o f th e table. T h e estim ated real term stru ctu re tells an en tirely different story th a n th e n o m in al term stru ctu re. T h ere ’ Five-year rates exhibited a similar pattern. 21 FEDERAL RESERVE BANK OF ST. LOUIS Figure 1 Initia l E q u ilib riu m in the M a r k e t for L o a n a b le Funds DECEMBER 1984 Figure 2 The Effects ol an Increase in the S u pp ly of and Reduction in the Dem and for Loanable Funds o f fu n d s is, for th e m o st part, very little d ifference b etw een sh o rt- an d lo n g -term real rates. In o th e r w ords, th e real "yield cu rv e” — th e relatio n sh ip b etw een th e term to m aturity an d th e real rate o f in terest on secu rities — h as b een m u ch flatter in re ce n t y ears th a n the n o m i nal yield curve. T h e average ab so lu te d ifference b e tw een th e o n e- an d 10-year n o m in al rates from 1978 to 1984 is 109 b asis p o in ts; for b efo re-tax real rates, it is 58 b asis p o in ts, w h ile fo r after-tax real rates it is 59 b asis poin ts. T h e se figures im ply th at long-term real rates have n o t differed su bstan tially from sh o rt-term real rates in re ce n t y e a rs.10 WHY DID REAL INTEREST RATES R ISE? T h e real in tere st rate is d eterm in ed by th e in te ra c tio n o f th e su p p ly o f an d d em an d fo r lo an ab le funds. T h e qu antity o f fu nd s available fo r len d in g (the q u an tity supplied) in c re a se s as th e recil rate o f in tere st in crea ses. T h e qu antity th at p eop le w ish to borrow (the qu an tity d em anded) d e crea se s as th e real rate in crea ses. T h e eq u ilibriu m real rate is th at for w h ich the qu an tity d em an d ed an d q u an tity su p p lied are equal. '“Notice that long-term inflation expectations were substantially lower than short-term inflation expectations from 1978 to early 1981, a period of predominantly rising inflation. This pattern has been re versed for late 1981 through early 1984, a period of generally declin ing inflation. http://fraser.stlouisfed.org/ 22 Federal Reserve Bank of St. Louis In figure 1, th is o cc u rs at th e real ra te r*, w h ere S rep rese n ts th e su p p ly curve an d D re p re se n ts th e d e m an d curve. F a cto rs th a t affect th e p o sitio n s o f th e supply an d d em an d cu rves d eterm in e th e equilibriu m rate. Potentially, th e se facto rs in clu d e th e ex p ected rate o f inflation, m on etary policy, th e sta te o f th e e c o n omy, taxes, fed eral b u d g et d eficits a n d th e d eclin in g relative p rice o f energy. T h e p o te n tia l im p a ct o f ea ch o f th e se factors o n real in tere st ra tes is d iscu sse d below . Expected Inflation W e know th a t e x p e cte d in flatio n affects n o m in al in terest rates. In fact, o u r real rate estim a tes are d e rived by su b tractin g th e ex p e cted in flation rate from th e n om in al in tere st rate. C han ges in ex p e cted in flation, how ever, also have th e p o ten tial to a lter real in terest rates. O ne reason , a sso cia te d w ith M und ell (1963), is th a t h ig h er ex p e cted in flatio n ca u se s p eop le to tran sfer part o f th e ir a sse ts from m o n ey to (higher) in terest-earn in g a ssets, th ereb y in crea sin g th e supply o f lo an ab le fund s an d driving d ow n th e real in terest rate. T h is o cc u rs b e c a u se m o n ey provides a very low o r negative real re tu rn d urin g tim es o f inflation, w h ereas th e retu rn o n in tere st-ea rn in g a sse ts g en er ally k eep s b e tte r p a ce w ith e x p e cte d in flation. A sim i lar n otion , a sso cia te d w ith T o b in (1965), is th a t h ig h er ex p ected inflation ca u ses p eop le to shift part o f th eir m o n ey b a la n ces in to real cap ital. T h is in d u ce s n et FEDERAL RESERVE BANK OF ST. LOUIS in vestm ent in cap ital th at u ltim ately d ep resses the m arginal retu rn on cap ital, red u cin g th e d em an d for lo an able fund s and th e real in tere st rate. An ad d itional argum ent, b ased on th e effect o f ex p e cte d inflation on th e retu rn to cap ital investm ent, is a sso ciate d w ith F eld stein and Su m m ers (1978): H igher inflation drives u p th e re p lacem e n t co st o f cap ital, w hile cu rren t tax law provides for d ep reciatio n allow a n ce s for b u sin esses b ased o n th e h isto rical co st o f cap ital. T herefore, h ig h er ex p e cted inflation resu lts in a low er ex p ected real retu rn o n cap ital investm ent, red u cin g th e d em an d for lo an able fun d s and, c o n s e quently, th e real in tere st rate. T h e se effects are illu strated in figure 2. T h e M undell effect shifts th e su pply curve from S, to S2 (an in crease in supply), resu lting in a d eclin e in th e equilibriu m real rate o f in terest from r* to r*. Sim ilarly, th e Tobin a n d th e F eld stein -Su m m ers effects shift th e d em an d curve from D , to D 2 (a re d u ctio n in d em and ), resu lting in a d eclin e in r* (to r* if b o th shifts o ccu r). T h ere is, how ever, a p o ten tial positive effect o f ex p e cte d inflation on th e real in terest rate th at w orks th rou gh th e p erso n al in co m e tax sy ste m ." U nder th e assu m p tio n th at p eo p le tiy to m ain tain a co n sta n t after-tax real rate, h ig h er ex p e cted in flation lead s to h ig h er before-tax real in terest rates sin c e taxes are a sse ssed on th e n o m in al retu rn .12 T h u s, th e h ig h er th e n o m in al retu rn, th e g reater th e sp read b etw een the b efore- an d after-tax real rates, all o th e r things equal. T h e w id enin g o f th e sp read b etw een b efore- an d after tax real rates as th e n o m in al in terest rate in crea ses ca n b e see n in ch art 1, w h ere th e averages o f th e before- an d after-tax real rates for th e p erio d s 196 0 -7 0 , 1 9 7 1 -8 0 an d 1 9 8 1 -8 4 are given by th e d ash ed lines. Th erefore, w ith th e co m b in atio n o f th e M undellT o b in an d F eld stein -Su m m ers effects an d th e in co m e tax effect, it is n o t p o ssib le to say a p riori w h eth er an in crea se in ex p e cted inflation lead s to h ig h er o r low er before-tax real in tere st rates, alth ou g h w e ex p e ct it to DECEMBER 1984 cau se lo w er after-tax real ra tes.13 From 1960 to 1980, th e co rrelatio n b etw een ex p e cted inflation an d b o th before- an d after-tax real rates o n o n e-y ear T reasury secu rities w as negative an d sta tistically significant: —0.38 for th e b efo re-tax rate an d —0.81 for th e after tax rate. T h is provides su p p o rt for th e M un d ell-Tobin an d F eld stein -Su m m ers effects. From 1981 to 1984, how ever, th e co rrela tio n h as actu ally b ee n positive for th e before-tax rate an d essen tially zero for th e after-tax rate. T h e sam e is tru e for th e co rrela tio n betw een inflation ex p e cta tio n s an d lon g-term real rates over th e 1 9 8 1 -8 4 p erio d .14 F u rth erm o re, d uring th e period o f rapidly rising real rates from 1980 to 1982, long-term inflation ex p e cta tio n s w ere also rising. T h u s, th ou gh th e evid ence on th e effect o f ex p e cted inflation on real in terest rates from sim ple co rrela tio n s is m ixed, it d oes n ot ap p e a r th a t ch a n g es in ex p e cted inflation w ere a m a jo r fa cto r in th e re c e n t rise in real in terest rates. Monetary Policy T h e effect o f m o n etary p o licy o n real rates o f in terest is a s u b je ct o f co n sid era b le controversy. T extbooks typically d escrib e th e im p a ct o f an in cre a se in m o n ey supply on th e real rate as follow s: An in crea se in th e m on ey su pply relative to m o n ey d em an d crea te s an ex cess su pply o f m on ey; in resp o n se, individuals in crea se th e ir p u rch a se s o f secu rities an d goods u ntil th e in terest rate d eclin es by en ou gh to in d u ce th e m to h old th e larger am o u n t o f m oney. T h u s, th e su pply o f loan able fu n d s in crea ses, driving dow n th e real in ter est rate. F u rth erm o re, an ex p an sio n ary m o n etary p o l icy lead s to sh o rt-term in cre a s e s in real in co m e d ue to th e in crea sed d em a n d for goods, w h ich h a s tw o ef fects th a t in flu en ce real rates in o p p o site d irectio n s: (1) th e level o f savings in crea ses, pu ttin g dow nw ard p ressu re on th e real rate, an d (2) th e d em an d for m on ey in crea ses, ca u sin g th e real rate to rise.15 One co n se q u e n ce o f in crea sin g th e grow th rate of th e m o n ey supply, how ever, is a rise in future rates o f inflation an d also in e x p e c t e d future rates o f inflation. "See Darby (1975) and Feldstein (1976). ,3See Makin and Tanzi (1983). 12To see this consider that ’"The correlation coefficients for 1981-84 are: for the one-year before-tax real rate, 0.48; for the one-year after-tax real rate, - 0.06; for the 10-year before tax real rate, 0.38; for the 10-year after-tax real rate, -0 .0 4 . r* = (1 -t)i - p* and that a constant after-tax real return, r*, implies that Ar* = (1 -t) Ai - Ap« = 0. Therefore, Ai = (1/1- t ) Ape. With the tax rate, t, between 0 and 1, this implies that the change in the nominal interest rate, Ai, is greater than the change in the expected inflation rate, Ape. If the tax structure is progressive, then higher expected inflation results in an even wider spread between before- and after-tax real rates. 15For more detail, see Santoni and Stone, and Brown and Santoni (1983). The theory of rational expectations states that a fully antici pated change in the money supply will have no effect on real interest rates. When people forecast money growth and future inflation in an optimal manner — by using all of the information currently available at sufficiently low cost— then the monetary authority is powerless to affect real behavior of any kind unless it is able to fool the public. This implies that only an unanticipated change in money supply affects the real interest rate. See Fischer (1980). 23 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 C hart 3 2 -Q u a rte r M l G ro w th an d 1-Y ear Real Interest Rate Th erefore, th e effect o f in crea sed grow th o f th e m on ey supply on n o m in al rates is likely to be positive in th e long run even if its im m ed iate effect o n real rates is negative. E m pirical evid ence su ggests th at th ere is little, if any, long-term effect o f ch an g es in th e m o n ey supply on real in tere st rates. Hafer an d H ein (1982) fou nd that an initial negative effect o f h ig h er m o n ey grow th on estim ates o f real in tere st rates w as co m p letely offset on e q u arter later. Sim ilarly, San ton i an d Ston e (1982) fou nd n o evid ence to link m o n ey grow th an d real rates over th e long term .'6 C hart 3 p lots th e tw o -qu arter grow th rate o f M l along w ith o u r estim ate o f th e b efo re-tax real in terest ,6Carlson (1982) actually finds a weak positive association between money growth and real interest rates. 24 rate on o n e-y ea r T reasu ry secu rities. T h e first p o in t to m ake is th at th e co rrela tio n s b etw een m o n ey grow th and th e real in tere st rate series are negligible for th e sam ple p erio d u sed in th e ch a rt.17 It is true, how ever, th at real rates o f in tere st beg an to rise in 1980 ju s t after a trem en d o u s re d u ctio n in tw o -q u arter M l grow th. T his red u ctio n w as follow ed by a n equally large in crea se in M l grow th, b u t real in tere st rates co n tin u ed to clim b n o n eth ele ss. T h e d ata illu strated in th e ch a rt suggest a n o th e r possible role for m o n eta iy p o licy in th e d eterm in atio n 17The correlation coefficient for two-quarter M1 growth and the be fore-tax real interest rate on one-year Treasury securities for 196084 is 0.076. The correlation between money growth and the before tax real 10-year rate for 1978-84 is -0.071. Correlations with the after-tax yields on the same securities for the same time periods are -0.157 and -0.004, respectively. DECEMBER 1984 FEDERAL RESERVE BANK OF ST. LOUIS o f real in terest rates: m o re variable m o n ey grow th lead s to h ig h er real rates. T h e exp lan atio n for th is is th at th e instability create d by highly variable m on ey grow th m akes for in creased u n certain ty abou t future retu rn s o n b o th sh o rt- an d lon g-term in terest-earn in g a sse ts an d cap ital an d raises th e d em an d for m on ey relative to th e se a ssets. T h is is, in effect, a re d u ctio n in th e su p p ly o f lo an able funds, w h ich ca u ses an in crea se in real in terest rates.'* real rates o f in terest th at o cc u rre d in 1980 an d 1981. F urth erm ore, th ere is as y e t n o in d ica tio n th at the sh o rt-ru n in stab ility o f m o n ey grow th w as m u ch af fected o n e w ay o r a n o th e r by th e Fed eral R eserve’s shift to a m ore ju d g m en tal o p eratin g p ro ced u re in th e fall o f 1982, an d real in tere st rates have y et to retu rn to th e ir pre-1981 levels.23 A n other w ay to state th is is: len d ers, if they are riskaverse, requ ire th at a g reater “risk p re m iu m ” b e ad d ed to in terest rates in o rd er to offset th e g reater u n c e r tainty asso ciated w ith th e future real re tu rn .'9 T h e ef fect o f m on etary variability on real in terest rates is n ot co m p letely u nam biguous, how ever, sin c e risk-averse borrow ers re d u ce th e ir d em an d fo r lo an able funds as u n certain ty in creases. A re ce n t em p irical stu d y by M ascaro an d M eltzer (1983) su ggests that th e overall effect o f m onetary variability on n o m in al in terest rates is positive. Sin ce th e variability o f m o n ey grow th sh o u ld n o t affect ex p e cted inflation, it follow s th a t th e effect on real in terest rates is positive as w ell.20 The State o f the Econom y A ca su al g lan ce at ch art 3 su ggests th a t m on ey grow th b eca m e su bstan tially m o re variable in 1980, th e sam e y ea r th at real rates o f in terest began to rise. T h e stan d ard deviation o f tw o -qu arter M l grow th is su b stan tially h ig h er for 1 9 8 0 -8 4 th an for 1 9 60-79, 4.1 p e rce n t co m p ared w ith 2.5 p e rce n t. T h e so u rce of g reater m onetary variability is an u n settle d issu e, bu t m an y an alysts attribu te it to th e ch an g e in Fed eral Reserve op erating p ro ced u re th at o cc u rre d in O ctob er 1979.21 O th er events also m ay have co n trib u ted to th e rise in m o n etary variability in clu d in g th e innovation in fin an cial m arkets (such as th e in tro d u ctio n o f NOW, Su p er NOW an d m o n ey m arket d ep o sit acco u n ts) an d th e im p o sitio n and removed o f cred it co n tro ls in 1980.22 T h u s, it ap pears th at an in crea se in th e variability o f m o n ey grow th in 1980 co n trib u ted to th e in crea se in ,8See Friedman and Schwartz (1963). '9The analysis assumes that it is not possible to diversify one’s hold ings in a manner that completely offsets the greater risk associated with monetary variability. “ Mascaro and Meltzer estimate the variability of unanticipated money growth, which turns out to be highly correlated with the variability of actual money growth. 21The Federal Reserve announced on October 6,1979, it would place less emphasis on confining variations in the federal funds rate and more emphasis on reserve aggregates as a sign of its commitment to longer-run restraint on money growth. “ See Hafer (1984) for a discussion of how financial innovations may have affected the accuracy of M1 as a measure of transaction balances. W hen th e eco n o m y en ters a re cessio n , b u sin ess firm s ex p e rie n ce e x c e ss cap acity , an d th e n ee d for ad d ition al cap ital is red u ced . A re d u ctio n in b o th th e d em an d for lo an ab le fu n d s an d th e real rate o f in terest follow s. As th e eco n o m y recovers, so m e firm s begin to p u sh tow ard th e ir ca p a city co n stra in ts, requirin g ad d itional in vestm en t an d in crea sin g th e d em an d for funds. Thu s, h ig h er real in terest rates ten d to a cc o m pany an expan sion . C hart 4 plots a m e a su re o f th e a m o u n t o f “sla ck ” in th e eco n om y, th e GNP gap, along w ith o u r estim ate o f th e before-tax real rate o n o n e-y ea r T reasu ry se c u ri ties. T h e evid ence su ggests th at th e state o f th e e c o n om y h elp s to exp lain m ov em en ts in real in tere st rates b o th before a n d after th e re ce n t upw ard shift in real rates, b u t th e shift itse lf ap p ears to have little to do w ith overall e c o n o m ic co n d itio n s. T h e GNP gap h as a co rrelatio n o f —0.56 w ith th e before-tax real rate for th e period 1 9 6 0 -8 0 , an d —0.44 for 1 9 8 1 -8 4 .24 “ As evidence that the money supply continues to be highly variable, consider the behavior of M1 during 1983 and 1984. M1 grew during the first two quarters of 1983 at a 12.8 percent rate and during the second two quarters of 1983 at a 7.3 percent rate. Similarly, in 1984 the growth rate of M1 was 6.8 percent in the first half of the year, compared with - 0 . 4 percent from June to October. It is generally recognized that the Federal Reserve altered its operating procedure again in late 1982. The post-1982 procedure is not the same as the pre-1979 procedure, however. See Wallich (1984). Another effect of the 1979 change in operating procedure was an increase in the day-to-day variability of nominal interest rates, which adds an additional element of risk in securities markets. This increased variability occurred in late 1979, however, while real interest rates did not begin to rise until late 1980. In addition, the federal override of state usury ceilings effective in March 1980 may have contributed somewhat to higher real interest rates, although there is no reason to think this action would push real rates to levels higher than those during previous periods (such as most of the 1960s and early 1970s) when these ceilings were not binding. 2‘The measure of the GNP gap is the difference between potential and actual GNP as calculated by the Council of Economic Advisers. To get data for 1984, potential GNP was assumed to grow at its average rate for 1960-83,3.44 percent. For the 10-year before-tax real rate, the correlation for 1981-84 is -0.55. For after-tax real rates, the correlations are -0.62 for the 1960-80 period and -0.13 for the 1981-84 period for the one-year rate and -0.37 for the 1981 84 period for the 10-year rate. 25 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 C hart 4 G N P G a p a n d 1-Year Real Interest Rate I960 62 64 66 68 70 B usiness Taxes As n o ted above, th e h ig h er th e retu rn on investm en t in p h y sical cap ital, th e g reater th e d em an d for lo a n able fund s. A tax on b u sin e ss profits re d u ce s th e real retu rn on in v estm en t an d th e d em an d fo r loan ab le funds, th ereb y low ering real in tere st rates. A tax on b u sin ess profits is n o t th e only b u sin ess tax th at af fects in v estm en t an d th e real in tere st rate, how ever. B u sin esses o ften receive tax cred its o r d ed u ctio n s from taxable in c o m e for ce rta in types o f in vestm en t ex p en d itu res. F u rth erm o re, tax d ed u ctio n s to offset th e d ep reciatio n o f cap ital eq u ip m en t an d stru ctu res ca n affect th e investm en t d ecisio n an d real rates, sin ce th e se d ep reciatio n allo w an ces m ay o r m ay n o t reflect 26 72 74 76 78 80 82 1984 th e tru e d ep recia tio n o f th e cap ital. If th e allo w ances overstate th e d ep recia tio n over a p eriod o f tim e, they ten d to sp u r ad d itio n al in vestm en t, driving up th e d em an d fo r lo an ab le fu n d s a n d th e real rate o f in te r est. If allo w an ces provide fo r sm a lle r d ed u ctio n s th an the actu al lo ss from d ep recia tio n , th ey h in d e r invest m ent an d real rates are red u ce d .25 T h e E co n o m ic R ecoveiy T a x Act o f 1981 w as d e signed to sp u r in vestm en t, prim arily by altering th e way in w h ich d ep recia tio n is trea ted for tax p u ip o se s. T h e m ag n itu d e o f th e effect o f th e a ct on in vestm en t is a con troversial issu e, b u t th e re is fairly stron g evid ence “ See Ott (1984). DECEMBER 1984 FEDERAL RESERVE BANK OF ST. LOUIS th at it sp u rred investm ent sp end ing. F o r in sta n ce , th e grow th rate o f real n o n resid en tial fixed investm en t as a p ercen tag e o f real GNP w as 8.7 p e rce n t over th e ex p a n sio n aiy p erio d from th e fou rth q u arter o f 1982 to th e se co n d q u arter o f 1984, up from an average o f only 1.5 p e rce n t over sim ilar p erio d s follow ing th e previous six re cessio n s.26 O ne problem , how ever, w ith co n clu d in g th at th e n ew tax legislatio n is a p rim aiy cau se o f h ig h er in ter est rates is th a t th e legislation w as n o t p assed u ntil August 1981 (although its provisions w ere retroactive to th e beginn ing o f 1981), w hile th e shift in real rates began in 1980 an d w as m ostly co m p lete by August 1981. F o r th is legislation to have b ee n th e p rim aiy factor in th e re ce n t rise in real in terest rates, th e p a ss age o f th e legislation m u st have b ee n p red icted and th e d em a n d fo r lo a n a b le fu n d s in c re a s e d m an y m o n th s in ad vance as th e p red icted future retu rn on cap ital investm ent rose. On th e o th e r h and , th is legis lation co u ld have co n trib u ted b o th to th e rise in real rates th a t o cc u rre d in late 1981 an d early 1982 in th e face o f a severe re ce ssio n an d to th e m ain ten a n ce o f relatively hig h real in tere st rates right up to th e presen t. Federal B udget Deficits G overnm ent borrow ing re p rese n ts an in cre a se in th e to tal d em an d fo r lo an able fun ds. T h is suggests th at real in tere st rates rise as th e size o f th e govern m en t bu dg et d eficit in cre a se s in real term s. O ne rarely sees a positive co rrelatio n betw een th e size o f deficits an d th e levels o f in tere st rates, how ever. T h is is p ro b a bly b ec a u se th ey resp o n d in o p p o site d irectio n s to ch an g es in e co n o m ic co n d itio n s; d eficits ten d to rise d uring b u sin ess re cessio n s an d fall during ex p an sio n s (because tax revenu es and outlays for tran sfer pay m en ts are sensitive to th e state o f th e econom y), w hile in terest rates typically fall d uring re ce ssio n s and rise d uring ex p an sio n s.27 As for th e re ce n t rise in real in terest rates, it is clea r from ch a rt 5 th at th e re ce n t d ram atic in cre a se in th e cy clically a d ju sted bu d g et d eficit did n o t o c c u r u ntil late 1982, by w h ich tim e real an d n o m in al in terest rates h ad begu n to fall. A c lo se r look at th e ch art 26The six previous expansionary periods were IV/1949-11/1951, 11/1954-IV/1955,11/1958—IV/1959,1/1961-111/1962, IV/1970-11/1972 and 1/1975—111/1976. The difference between the growth of the investment-GNP ratio in the current recovery and the average growth in the six previous recoveries is statistically significant. 27See Tatom (1984). in d ica tes th at two m a jo r in crea s es in th e size o f the cy clically a d ju sted d eficit have o cc u rre d in re cen t y ears: o n e in 1975 an d th e o th e r in 1982. N eith er w as a sso cia ted w ith risin g real in tere st rates. T h is d oes n o t n ecessa rily im ply th a t d eficits have no effect on real in tere st rates. Sin ce in tere st ra tes are b a sed on ex p ecta tio n s, ex p e cted fu tu re d eficits co u ld have an im p act o n to d ay ’s real in tere st rates. If on e a ssu m es th e bu dget p ro je c tio n s o f th e C on gression al Budget Office (CBO) are rep resen tativ e o f th e m arket's ex p ectatio n o f fu tu re deficits, how ever, th e n d eficit p ro jectio n s do n ot ap p ear to have b ee n th e m a jo r instigator o f th e re ce n t rise in real in terest rates. T h e CBO rep ort p u b lish ed in Ju ly 1981 p ro je c te d a 1982 d eficit o f less th a n $30 billio n an d s u r p lu s e s in th e next fou r y ea rs grow ing to over $200 b illion by 1986.28 B ecall th at at th e tim e th is rep ort w as w ritten, o u r estim ates o f b o th sh o rt- an d lon g-term b efo re-tax real in terest rates w ere alread y far in e x c e ss o f h isto rica l n orm s an d after-tax real rates h ad risen to n ea r th e ir previous peaks. By February 1982, th e CBO h ad altered its p ro je c tio n s an d w as p red ictin g a d eficit o f nearly $200 billion in 1983, grow ing to n early $300 billion by 1987.29 Yet 1982 w as a y e a r o f g en erally falling real an d n o m i n al in terest rates.30 Like th e ch a n g e in th e tax laws, how ever, ex p e cta tio n s o f fu tu re d eficits m ay b e h elp ing to keep real in tere st rates at levels th at are quite high relative to p a st h isto iy . Declining Relative Price o f Energy Finally, it h as b ee n su ggested th at d ra stic in creases in th e relative p rice o f energy co n trib u te d to th e low real in terest rates o f th e 1970s, w h ich w ould im ply that th e generally falling relative p rice o f energy o f th e 1980s h as co n trib u te d to h ig h er real in terest rates.31 T h e argum en t is th at th e d em a n d for cap ital fell d u r ing th e 1970s b e c a u se o f a re d u ctio n in th e su pply of “ Congressional Budget Office (1981). Carlson (1983) discusses pos sible sources of bias in the CBO’s budget projections. “ Congressional Budget Office (1982). In discussing the reasons for the change in the outlook on the deficits between 1981 and 1983, the Congressional Budget Office (1983, p. 18) says that, “Over the entire five-year period, 60 percent of the change in outlook from budget surpluses to budget deficits can be attributed to the failure of the economy to perform as projected two years ago.” In addition, it says (p. 20) that, “Legislative actions are the second largest reason for differences between the two baselines, accounting for about 30 percent of the change over the five-year period.’’ “ It is possible that higher projected government budget deficits lead to greater expected inflation, in which case higher deficits would cause higher nominal, but not necessarily real, interest rates. 3,See Wilcox (1983). 27 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 C hart 5 Cyclically Adjusted Budget Deficit and 1-Year Real Interest Rate Percent 1960 Billions of dollars S em iann ua l d a t a 62 64 66 68 70 co m p lem en tary energy in p u ts, w h ich resu lted in re d u ced d em an d for lo an able fu nd s an d lo w er real in terest rates.32 O n ce again, how ever, th e tim ing o f th e re ce n t rise in real in terest rates fails to le n d c re d e n c e to th e th eo iy . D uring th e perio d o f m o st rapidly rising real in terest rates in 1980 an d th e first h a lf o f 1981, th e relative p rice o f energy w as still rising rapidly as a resu lt o f the se co n d oil crisis; th e grow th rate o f th e relative p rice o f energy b etw een IV/1979 an d 11/1981 w as 18.3 p e rce n t.33 72 74 76 78 80 82 1984 R ed u ctio n s in th e relative p rice o f energy did n o t b e gin u n til late 1981, after m o st o f th e in cre a se in real in terest rates alread y h ad o ccu rred . CONCLUSIONS “ See Tatom (1979) for a discussion of the impact of energy shocks on investment. T h e 1980s have se e n u n p re c e d e n te d b eh avior in several key e co n o m ic variables, th e m o st n o tab le being in terest rates. A ccord ing to estim a tes o f real in terest rates b a sed on surveys o f ex p e cted inflation, b o th sh o rt- an d lo n g -term recil rates ro se to reco rd levels in 1981 an d 1982 and , alth ou g h th e y have d eclin ed som ew h at sin c e th en , have n o t re tu rn ed to th e levels o f th e 1960s an d 1970s. “ The measure of the relative price of energy is the producer price index for "fuels and related products and power” divided by the business sector deflator. A co m p a riso n o f estim a tes o f b efore- an d after-tax real in terest ra tes in d ica tes th a t th e overall p a ttern o f http://fraser.stlouisfed.org/ 28 Federal Reserve Bank of St. Louis FEDERAL RESERVE BANK OF ST. LOUIS th e ir m o v e m e n ts h as b e e n s im ila r. T h e s p re a d b e tw e e n th e b e fo re - a n d a fte r-ta x re a l ra te s in c re a s e d o ver m u c h o f th e s a m p le , h o w e v e r, as n o m in a l in te re s t ra tes (a n d e x p e c te d in fla tio n ) in c re a s e d . T h e re fo re , a fte r-ta x re a l in te re s t ra te s h a v e n o t b e e n n e a r ly as h ig h re la tiv e to p re v io u s e x p e rie n c e as b e fo re -ta x re a l ra tes . N o n e th e le s s , th e y h ave b e e n h ig h e r o n averag e th a n th e y w e r e in th e 1960s a n d m u c h h ig h e r th a n in th e 1970s. T h e p h e n o m e n o n m o s t c lo s e ly c o in c id e n t w it h th e rise in re a l ra te s w a s a n in c re a s e in th e v a ria b ility o f m o n e y g ro w th , w h ic h in c re a s e d e c o n o m ic u n c e r ta in ty a n d th e ris k p r e m iu m o n in te re s t ra tes . M a jo r ch a n g e s in c u r r e n t a n d p ro je c te d g o v e rn m e n t b u d g e t d e fic its a n d in ta x p o lic ie s h a p p e n e d a f t e r m u c h o f th e u p w a r d s h ift in re a l in te re s t ra te s a lre a d y h a d o c c u rre d , b u t m a y h a v e c o n tr ib u te d to s o m e a d d itio n a l u p w a r d m o v e m e n t. C h a n g e s i n e c o n o m ic c o n d itio n s h ave b e e n a m a jo r in flu e n c e o n th e m o v e m e n t o f re a l in te re s t ra te s s in c e 1981; p e rio d s o f s lo w g r o w th o r re c e s s io n h a v e p r o d u c e d fa llin g re a l ra tes , w h ile e x p a n s io n s h a v e p u s h e d re a l ra te s u p w a r d . DECEMBER 1984 Ezrati, Milton J. “Inflationary Expectations, Economic Activity, Taxes, and Interest Rates: Comment," American Economic Review (September 1982), pp. 854-57. Feldstein, Martin. “Inflation, Income Taxes, and the Rate of Interest: A Theoretical Analysis,” American Economic Review (December 1976), pp. 809-20. Feldstein, Martin, and Lawrence Summers. “Inflation, Tax Rules, and the Long-Term Interest Rate," Brookings Papers on Economic Activity (1:1978), pp. 61-99. Fischer, Stanley. “On Activist Monetary Policy with Rational Expec tations,” in Stanley Fischer, ed., Rational Expectations and Eco nomic Policy (The University of Chicago Press, 1980), pp. 211-47. Fisher, Irving. Appreciation and Interest (1896), Reprints of Eco nomic Classics (Augustus M. Kelley, Bookseller, 1965). Friedman, Milton, and Anna Jacobson Schwartz. A Monetary His tory of the United States 1867-1960 (Princeton University Press, 1963). Hafer, R. W. “The Money-GNP Link: Assessing Alternative Trans action Measures,” this Review (March 1984), pp. 19-27. Hafer, R. W., and Scott E. Hein. "Monetary Policy and Short-Term Real Rates of Interest,” this Review (March 1982), pp. 13-19. Holland, A. Steven. “Does Higher Inflation Lead to More Uncertain Inflation?” this Review (February 1984), pp. 15-26. Makin, John, and Vito Tanzi. “The Level and Volatility of Interest Rates in the United States: The Role of Expected Inflation, Real Rates, and Taxes," Working Paper No. 1167 (National Bureau of Economic Research, July 1983). Mascaro, Angelo, and Allan H. Meltzer. “Long- and Short-Term Interest Rates in a Risky World,” Journal of Monetary Economics (November 1983), pp. 485-518. REFERENCES Brown, W. W., and G. J. Santoni. “Monetary Growth and the Timing of Interest Rate Movements," this Review (August/September 1983), pp. 16-25. Carlson, John A. “A Study of Price Forecasts,” Annals of Economic and Social Measurement (Winter 1977), pp. 27-56. Carlson, Keith M. “The Mix of Monetary and Fiscal Policies: Con ventional Wisdom Vs. Empirical Reality,” this Review (October 1982), pp. 7-21. _________ “The Critical Role of Economic Assumptions in the Evaluation of Federal Budget Programs,” this Review (October 1983), pp. 5-14. Congressional Budget Office. Baseline Budget Projections: Fiscal Years 1982-1986 (U.S. Government Printing Office, July 1981). _________ Baseline Budget Projections for Fiscal Years 1983-1987 (GPO, February 1982). Mehra, Yash. “The Tax Effect, and the Recent Behaviour of the After-Tax Real Rate: Is It Too High?” Federal Reserve Bank of Richmond Economic Review (July/August 1984), pp. 8-20. Mundell, Robert. “Inflation and Real Interest,” Journal of Political Economy (June 1963), pp. 280-83. Ott, Mack. "Depreciation, Inflation and Investment Incentives: The Effects of the Tax Acts of 1981 and 1982, "this Review (November 1984), pp. 17-30. Santoni, G. J., and Courtenay C. Stone. "The Fed and the Real Rate of Interest," this Review (December 1982), pp. 8-18. Tatom, John A. “Energy Prices and Capital Formation: 19721977,” this Review (May 1979), pp. 2-11. _________ “A Perspective on the Federal Deficit Problem," this Review (June/July 1984), pp. 5-17. Tobin, James. “Money and Economic Growth,” Econometrica (Oc tober 1965), pp. 671-84. _________ Baseline Budget Projections for Fiscal Years 1984-1988 (GPO, February 1983). Wallich, Henry C. “Recent Techniques of Monetary Policy," Fed eral Reserve Bank of Kansas City Economic Review (May 1984), pp. 21-30. Darby, Michael R. “The Financial and Tax Effects of Monetary Pol icy on Interest Rates,” Economic Inquiry (June 1975), pp. 266-76. Wilcox, James A. “Why Real Interest Rates Were So Low in the 1970’s," American Economic Review (March 1983), pp. 44-53. 29 Monetizing the Debt Daniel L. Thornton M 1 * -B-ONETIZING th e d eb t" co n ju re s up tearsom e im ages o f excessive m oney stock grow th resulting from Fed eral R eserve p u rch a se s o f T re asu iy debt. M any an alysts fear th at d ebt m o n etizatio n m ay pro d u ce u nd esirab le ec o n o m ic co n se q u e n ce s, su ch as m ore rapid inflation and, thu s, h ig h er n om inal in ter est rates. T h e re ap p ears to b e so m e con fu sio n , h ow ever, over w hat d ebt m o n etizatio n m eans, w h eth er or to w hat ex ten t th e F ed eral Reserve has p u rsu ed a policy o f d ebt m o n etizatio n in th e p ast an d w hat the b est in d ica to r o f d ebt m o n etizatio n is. T h e se q u es tio n s are o f in ten se in terest, w ith poten tially large d ef icits loom ing on th e h o rizo n th a t co u ld put in crea sed p ressu re on th e F ed eral Reserve to m o n etize th e debt in th e future. T h e p u rp o se o f this article is to clarify th e m eanin g o f th e p h rase "m o n etizin g th e d ebt " an d to d eterm in e w h eth er th e Fed eral Reserve h as m o n etized th e d ebt sin ce 1960. As w e w ill see, th e p o licy objectives o f th e m o n etary au th ority play an im p o rtan t role in d eter m ining w h eth er th e Fed eral Reserve will m o n etize th e debt. W e also will sh ow th at care m u st be taken not to co n fu se d ebt m o n etizatio n w ith grow th in th e Federal R eserve’s portfolio o f governm ent debt. MONETIZING THE DEBT: WHAT DOES IT MEAN? In large m easu re, th e p h rase "m o n etizin g th e d e b t” grew ou t o f th e e x p e rie n ce o f th e Fed eral Reserve im m ediately after W orld W ar II. At th e tim e, th e Federal Reserve h ad a tacit co m m itm e n t to th e U.S. T re a su iy to Daniel L. Thornton is a senior economist at the Federal Reserve Bank of St. Louis. John G. Schulte provided research assistance. Digitized for30 FRASER stabilize the T reasu ry 's co st o f fin an cin g th e w ar debt. After th e war, individuals began liqu id atin g th eir h o ld ings o f Liberty B on ds. B eca u se o f its agreem en t w ith the Treasury, th e Fed eral Reserve p u rch a se d su b sta n tial am o u n ts o f govern m en t d e b t.1 T h ese p u rch a se s in crea sed th e reserves o f th e ban k ing system and, co n seq u en tly , th e m o n ey stock; the Fed eral Reserve w as said to have m o n etiz ed th e debt. In M arch 1951, th e F ed eral Reserve an d th e T re a su iy rea ch ed an a cc o rd w h ereb y th e Fed eral Reserve e sta b lish ed its in d e p e n d e n c e .2 Sin ce th en , th e Fed eral Re serve has b een free to p u rsu e its policy ob jectiv es in d e p e n d e n t o f th e d eb t fin a n cin g n ee d s o f th e T reasury.3 W ith th e n et fed eral d eb t (NFD) — total d ebt m inu s holdings o f govern m en t a g en cies an d tru sts — at nearly $1.3 trillion an d w ith h isto rically high deficits, in b o th n om in al an d real term s, th ere is co n c e rn that the rapidly rising d ebt will put upw ard p ressu re on in terest rates, in d u cin g th e F ed eral Reserve to in 'The Federal Reserve's holdings of government debt more than tripled from 1943 to 1946. See Historical Statistics of the United States (1975), p. 1116. 2See Ahearn (1963), pp. 16-21. 3Actually, at a more abstract level, the question of the independence of monetary and fiscal policies is open to debate. Sargent and Wallace (1981) use the government budget constraint to argue that the monetary authority must ultimately monetize deficits. This argu ment has been challenged recently by Darby (1984), and some evidence has been supplied recently by Barth, Iden and Russek (1984). Furthermore, the budget constraint can be used to argue that the seignorage associated with Federal Reserve open market operations requires a compensatory change in government expend itures or taxes. This latter point is discussed in Horrigan (1983). The seignorage associated with open market operations is easily illus trated from the budget constraint suggested by Thornton (1984). FEDERAL RESERVE BANK OF ST. LOUIS crea se th e m oney supply m ore rapidly th an it o th e r w ise w ould or, perh ap s, shou ld .4 Today, as in th e im m ed iate post-W orld W ar II p e riod, th e p h rase “m onetizing th e d e b t” m ean s m oney grow th in d u ced by attem p ts to m o d erate th e effects of rapidly grow ing governm ent d ebt on in terest rates. By definition, o p en m arket o p eratio n s (buying an d selling governm ent secu rities in th e m on ey an d cap ital m ar kets) rep resen t d ebt m onetization, that is, th e re p la ce m ent o f governm ent d ebt w ith m oney. O pen m arket p u rch a ses and debt m onetization, therefore, are often taken to be synonymous/' T h is view is en h a n c e d by th e fact that o p en m arket o p eratio n s are u sually c o n sid ered th e p rin cip al tool throu gh w h ich th e Fed eral Reserve in flu en ces th e m oney supply, so th at ch an g es in Fed eral Reserve policy are likely to be reflected initially in its portfolio o f governm en t debt. F or th ese reason s, analysts so m etim es look at th e grow th o f the Fed eral R eserve’s portfolio o f governm ent debt, the ratio o f th e Fed eral R eserve’s hold ings o f d ebt (FHD) to NFD, o r sim ilar m easu res as in d icato rs o f d ebt m o n eti zation. T h e se m easu res, how ever, give too little a tte n tion to th e goals o f policy and th e n atu re o f th e m oney stock m ech anism / It is cle a r from o u r definition th at d ebt m o n etizatio n ca n n o t be analyzed sep arately from th e o bjectiv es of Federal Reserve policy. A ssum e, for exam ple, th at th e Fed eral Reserve is targeting m o n ey grow th to achieve p rice level stability. Fu rtherm ore, assu m e th at real in co m e is grow ing at a faster rate th an velocity so that m oney grow th m u st be positive. If th is m o n ey grow th is achieved throu gh o p en m arket p u rch a se s o f govern m ent d ebt w hile th e debt is sim u ltan eo u sly in c re a s ing, th e co rrelatio n b etw een th e grow th in th e Fed eral R eserv e’s p o rtfo lio an d g o v ern m en t d eb t grow th w ould give the false ap p e aran ce o f d ebt m o n etizatio n .7 In th is exam ple, debt m o n etizatio n actu ally o ccu rs “See Tatom (1984) for a historical survey of the deficit. 5ln principle, any debt can be purchased. In practice, however, the Federal Reserve primarily purchases marketable debt of the U.S. Treasury. 6These measures are singled out here because they are most fre quently used in the popular press. Other measures, such as growth of total reserves or the monetary base, suffer from this same defi ciency, as well as some of the deficiencies noted in the following discussion. See Blinder (1983), Dwyer (1984) and Barth, Sickles and Wiest (1982) for examples of the various measures that have been employed in empirical studies of this question. T h e astute reader will recognize that this implies that open market purchases of debt are not strictly required for debt monetization to occur. This can be argued in a number of ways. At a rudimentary level, assume that the Treasury has the power to print money, so that deficits can be financed either by issuing debt or, as a substi tute, printing money. Printing money directly is as much debt mone- DECEMBER 1984 only if th e Fed eral Reserve m odifies its p rim aiy o b je c tive o f p rice stability b ec a u se it fears th at d ebt grow th will bo o st in terest rates. Fu rth erm ore, in o rd er to claim that th e F ed eral Re serve has m o n etized th e debt, o n e w ould have to ar gue both th at th ere is a positive relatio n sh ip betw een actu al o r an ticip ated in terest rates an d d eb t grow th and that th e Fed eral Reserve h ad m odified its p rim aiy m on ey stock grow th ob jectiv e in re sp o n se to actu al or perceived upw ard p ressu re on in terest rates. In this in stan ce, th e a sso cia tio n b etw een th e difference in the actu al an d targeted grow th rates o f m on ey an d the grow th o f NFD w ould provide ev id en ce o f d ebt m o n e tization/ T h u s, u sin g th e grow th o f FHD o r th e ratio of FHD to NFD alo n e as in d ica to rs o f d ebt m o n etizatio n cou ld b e m isleading. If th e Fed eral Reserve ach ieves its d esired m o n ey grow th objective, it is n o t m on etizin g th e debt, even if m on ey grow th is achieved solely throu gh o p en m arket p u rch a se s o f governm ent debt. Alternatively, su p p o se th e Fed eral R eserve’s in ter m ediate p o licy o b jectiv e is to peg in terest rates at som e d esired level/ T h e n th e Fed eral Reserve m o n e tizes th e d ebt o n ly w h en ch a n g es in th e debt, ceteris paribus, p ro d u ce ch a n g es in in terest rates in th e sam e d irection . T h at is, if in crea s es in th e d ebt put upw ard pressu re on in tere st rates, th e Fed eral Reserve will m o n etize th e d ebt u n d e r an in terest rate target. OBSTACLES TO IDENTIFYING DEBT MONETIZATION In ad d ition to th e n ee d to a cco u n t for th e explicit or im plicit targets o f m o n etary policy, th ere is a n o th er co n sid eratio n th at m akes th e grow th o f FHD, the ratio tization as if debt were first issued to finance the deficit, then repur chased (later) through note issue. Of course, the Treasury cannot issue notes directly. Indeed, the Federal Reserve cannot even purchase government debt directly from the Treasury. Consequently, all deficits must initially be fi nanced through debt issue. This initial debt issue increases the demand for credit. If this drives interest rates upward, the Federal Reserve can lessen the effect by increasing the supply of credit, using any of its policy tools. The long-run effects of Federal Reserve activities, however, depend on the tool used due to possible wealth effects and the seignorage associated with open market operations. See Thornton. 8This does not imply that the Federal Reserve has the ability to hit its money target exactly. It requires only that there be a systematic relationship between these errors and debt growth. The identifica tion of this process could be complicated, however, if the uninten tional errors are associated directly or indirectly with debt growth. 9lt is difficult to conceive of a situation in which the Federal Reserve could control interest rates in anything but the short run; neverthe less, this is a common conception of the transmission mechanism of monetary policy. 31 FEDERAL RESERVE BANK OF ST. LOUIS o f FHD to NFD an d sim ilar m easu res even less reliable as in d icato rs o f d ebt m o n etizatio n : m oney grow th d oes n o t n e c essa rily requ ire grow th in th e Fed eral R eserve’s p ortfolio o f govern m ent secu rities. T h u s, th e link betw een d ebt m o n etizatio n an d the grow th o f FHD m ay b e m u ch w eak er th an co m m o n ly im agined. C on sid er th e sim p le m o d el o f m o n ey growth, M = rh + B, w h ere th e grow th o f m oney, M, is th e su m o f the grow th o f th e m o n ey m u ltiplier, m, and th e grow th o f th e a d ju sted m o n etary base, B. If ad ju sted b ase grow th w ere achieved en tirely th rou gh o p en m arket op era tio n s an d if th e m u ltip lier w ere co n sta n t (i.e., rh = 0), m oney grow th w ould be equal to th e grow th in the F ed eral R eserve’s portfolio o f governm en t debt. If the m u ltip lier w ere rising, how ever, m o n ey grow th w ould exceed portfolio grow th: if th e m u ltip lier w ere falling, portfolio grow th w ould e x ceed m o n ey grow th. T h e re fore, th e ex ten t to w h ich th e Fed eral Reserve is m o n e tizing th e d ebt ca n n o t b e d eterm in ed sim ply by o b serving th e g row th rate o f th e F ed eral R eserve’s portfolio o f governm ent secu rities. M ultiplier m ove m ents m u st be co n sid ere d b ec a u se su ch m ovem ents w eaken th e link b etw een p ortfolio grow th an d d ebt m on etizatio n . Base Growth an d D ebt M onetization O th er facto rs affect th e link b etw een a d ju sted base an d portfolio grow th and, thereby, m ake th e c o n n e c tion betw een d ebt m o n etizatio n and th e grow th o f the Fed eral Reserve's portfolio even m ore ten u o u s. Even if th e m u ltip lier is co n stan t, th e grow th rate o f m oney n eed n o t co rresp o n d clo sely w ith grow th in th e F ed eral R eserve’s h old ing s o f govern m ent debt. O ne o f th e se factors is ch an g es in reserve requ ire m ents, su ch as th o se m an d ated by th e M onetary C on trol Act o f 1980. T h e se reserve req u irem en t ch an g es are reflected in th e reserve ad ju stm en t m agnitu d e (RAM).10 In crea se s in RAM in crea se the base, w hile red u ctio n s re d u ce it. C onsequently, ch an g es in RAM m ay cau se th e Fed eral Reserve to bu y m ore o r less governm ent d ebt th a n it oth erw ise w ould to achieve its m onetary grow th o bjectiv e u n d e r a m o n etary tar geting p roced u re. DECEMBER 1984 th e Fed eral Reserve an d F ed eral Reserve flo at." S e cu lar m ovem ents in th e se facto rs ca n resu lt in ad ju sted b ase grow th th at is fa ster o r slow er th a n th e grow th o f th e Fed eral R eserve’s portfolio. C on sequ en tly, d ata on th e grow th rate o f th e F ed eral R eserve’s portfolio is n ot n ecessarily a good in d ica to r o f th e ex ten t to w h ich the F ed eral Reserve is m o n etizin g th e debt. An Illustration o f T hese Relationships T h e im p o rta n ce o f th e se facto rs is illu strated in ch a rts 1 -4 . C hart 1 sh ow s th e ratio o f FHD to NFD an n u ally from 1960 to 1983. T h is ratio in crea sed from 1960 to 1974 an d d eclin ed th ereafter. T h u s, generally speaking, th e Fed eral Reserve p u rch a se d governm ent secu rities at a m o re rap id p a c e th a n th e grow th o f NFD up to 1974, b u t at a m u ch slow er p a c e afterw ards. If th is ratio w ere u sed as th e so le in d ica to r o f d ebt m o n etization, o n e w ould likely co n c lu d e th at th e Fed eral R eseive m o n etized th e d ebt from 1960 to 1974, th en reversed this policy. T h e sam e co n c lu sio n w ould em erg e if only th e grow th rate o f th e Fed eral R eserve’s portfolio w ere co n sid ere d .12 Yet M l grow th w as abou t 4.8 p e rce n t during th e form er period an d ab o u t 7.2 p e rce n t d uring th e la tter p eriod . T h u s, th e grow th o f m o n ey w as slow er in th e first perio d th a n in th e seco n d , d esp ite th e fact th at grow th o f FHD w as faster in th e first period th an in th e seco n d , b o th in ab so lu te term s an d relative to th e grow th o f NFD. T h is inverse relation ca n b e exp lain ed , in large part, by m ovem ents in th e m o n ey m u ltiplier, RAM, d ep o si tory in stitu tio n s' borrow ings an d float. T h e se series are p resen ted in ch a rts 2 -4 . T h e m u ltip lie r d e c lin e d m o re o r le s s stea d ily throu gh 1974. Over th e sam e period, RAM w as fairly stable, first rising th e n d rop p in g slightly. Borrow ing w as fairly stab le th rou gh 1972, th e n in crea sed d ram at ically in 1 9 7 3 -7 4 . F loat in c re a s e d m o d estly th rou gh 1972, th en d eclin ed by about $1 b illion d uring 1973 and 1974. On net, a m o d est am o u n t o f m o n eta iy base w as su p p lied to th e ban kin g sy stem from 1960 to 1974 throu gh borrow ings an d float, w h ile RAM d rain ed a m o d est am o u n t o f m o n etary b a se from th e system over th is period. W ith th e ex cep tio n o f borrow ings d uring 19 7 3 -7 4 , how ever, n o n e o f th e se facto rs w as A d justed b ase grow th also is affected by o th e r fa c tors, su ch as d ep ository in stitu tio n borrow ing from "There are other factors that affect base growth; however, quantita tively they are typically less important than those noted. ’“See Tatom (1980) for a discussion of the adjusted monetary base and RAM. Digitized for32 FRASER l2The compounded annual rate of growth of FHD was 6.89 percent from 1974-83, compared with 8.23 percent from 1960-74. FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 Chart 1 Ratio of Federal Reserve Holdings of Federal Debt to Total Debt A nnual data p a rtic u la rly large relativ e to th e d e c lin e in th e m u ltiplier.'3 C onsequently, even if th e policy objective had b een zero m on ey growth, th e Fed eral Reserve still w ould have h ad to m ake su b stan tial n et o p en m arket p u r ch a se s o f governm ent secu rities to offset th e d eclin e in the m ultiplier. Thu s, th e in crease in th e ratio o f FHD to NFD m ight reflect n o th in g m ore th an the Fed eral Re serve's n eed to m ake p u rch ase s o f governm ent debt (to offset m u ltip lier m ovem ents) in e x c e ss o f the grow th o f NFD over th is period. After 1974, a n u m b er o f factors re d u ce d th e Fed eral R eserve's n eed to engage in o p en m arket p u rch ases. T h ere w as a significant in crea se in RAM th rou gh 1 9 7 4 78, follow ed bv an even m ore d ram atic rise after the 13For example, the multiplier declined from 3.13 in 1960 to 2.78 in 1974. Given the average level of the adjusted monetary base of $64.82 billion over this period, the decline in the multiplier had an impact equivalent to a $7.64 billion drain on the adjusted monetary base on average over this period. M onetary C ontrol Act o f 1980. At th e sam e tim e, both borrow ings an d float in crea sed d ram atically through 1975-79, th en d eclin ed th rou gh 1983. W hile th e m u lti p lier co n tin u ed to d eclin e th rou gh 1980, th e rate of d eclin e w as m ore m o d erate th an before. Sin ce 1980, th e m u ltip lier h as rem ain ed relatively u n ch an ged . It is easy to see h ow m o n ey grow th cou ld have a ccelera ted sin ce 1974 even th ou gh th e ratio o f FHD to NFD has fallen .14 T h u s, th e grow th o f FHD co u ld b e u sed as an in d ica to r o f d ebt m o n etizatio n only if th ese o th e r fac tors affecting m o n ey rem ain u n ch an g ed . ,4No attempt is made here to explain why the various changes in the multiplier, RAM, the float or borrowings occurred. Nevertheless, some of these changes can be readily explained. For example, the increase in borrowings in 1974 was due, in part, to the Federal Reserve's efforts to shore up the banking system after the collapse of the Franklin National Bank (see the Board of Governors of the Federal Reserve System [1974], pp. 740-41). Likewise, the signifi cant jump in RAM after 1980 can be attributed directly to the Mone tary Control Act of 1980, while the significant decline in the float after 1979 is associated with Federal Reserve efforts to improve the check-clearing process. 33 Chart 2 Money M ultiplier A nnual data Actually, it w ould be legitim ate to u se th e grow th of th e FHD as an in d ica to r o f d ebt m o n etizatio n if the Fed eral Reserve estab lish ed it as an in term ed iate p o l icy target. Sin ce th is h as never b ee n done, the p o ssib il ity is n o t co n sid ered h e re .1' A Note on the Interest Rate and Liquidity Effects T h e read er is cau tio n ed th at this analysis is an illu s tration o f d ebt m o n etizatio n u n d e r th e u sual textbook d escrip tio n o f co u n te rcy lica l m o n e ta iy policy. As su ch , it an d th e em p irical analysis that follow s are heavily d ep en d en t on th e e x iste n ce o f two effects that find little su p p o rt in em p irical stu d ies."4T h e first is the effect o f ch an g es in th e governm ent d ebt on in terest ,5During the period from October 1979 to October 1982, the Federal Reserve used nonborrowed reserves as an operating target. Money growth, however, was its intermediate policy target. ,6For evidence on the liquidity effect, see Brown and Santoni (1983), Melvin (1983) and the references cited in these articles. For evi dence on the relationship between debt growth and interest rates, see Evans (1984), Blinder and the references cited in Blinder. http://fraser.stlouisfed.org/ 34 Federal Reserve Bank of St. Louis rates. T h e story o f d ebt m o n etiz a tio n told above rests on th e idea th at in c re a s e s in d ebt issu e by th e T re a sury raise th e d em an d for cred it relative to th e supply; co n sequ en tly, in terest rates rise. W hile it is beyon d th e sco p e o f th is article to delve in to th e se argum ents, som e e co n o m ists believe, an d th e bulk o f em p irical work suggests, th at in cre a s e s in d ebt have n o effect on in terest rates. If th is is true, th en in crea ses in debt w ould n o t pu t p ressu re on th e F ed eral Reserve to m o n etiz e u n d e r an y p o licy regim e — u n less, o f co u rse, th e Fed eral Reserve believes that d ebt in crea se s ca u se in terest rates to rise. T h e seco n d effect im p licit in this analysis is th e socalled liquidity effect, an initial d eclin e in in terest rates a sso cia te d w ith an u n e x p ecte d a ccelera tio n in th e grow th rate o f m oney. W hile th e liquidity effect has b een isolated em pirically, estim a tes suggest th at it is w eak an d short-lived. T h e evid en ce fu rth er suggests th a t th e lo n g e r-ru n e ffect o f a c c e le r a te d m o n ey grow th is higher, n ot low er, n o m in al rates o f in terest. If th e Fed eral Reserve believed this, it w ould b e c o n siderably less an xio u s to m o n etize d ebt in creases, re- C h a rt 3 Reserve Adjustment M a g n itu d e (RAM) and Float at Depository Institutions A n n u a l d a ta C h art 4 Total Borrowings at Federal Reserve Banks 1 9 6 0 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 1983 ' 35 FEDERAL RESERVE BANK OF ST. LOUIS alizing th at th e lo n g er-ru n c o n se q u e n ce s o f its actio n s w ould be h ig h er in flation an d h ig h er n o m in al in terest rates. WHAT IS THE EVIDENCE ON DEBT MONETIZATION? Ideally, to d eterm in e w h e th e r th e Fed eral Reserve h a s m o n etize d th e debt, deviations b etw een actu al an d d esired m o n etary targets (say m o n ey growth) sh ou ld b e co m p ared w ith d ebt grow th. P ractical c o n sid eratio n s m ake su ch a co m p ariso n difficult. If the Fed eral Reserve esta b lish ed target rates for th e grow th of m o n ey th at w ere ch an g ed infrequently, the growth rate o f m o n ey m ight rep resen t a useful proxy for the deviations b etw een actu al an d d esired m o n ey grow th. U nfortunately, d uring m o st o f th e p o st-acco rd ex p eri en ce w ith co u n te rcy clica l m o n etary policy, th e evi d en ce su ggests th a t th e F ed eral Reserve seld o m fo cu sed exclu sively on a m o n etary aggregate target. In stead , th e F ed eral Reserve h as given co n sid era b le em p h asis to in terest rates as th e prim ary interm ed iate p o licy target. O nly during th e p ast 15 y ears h as the Fed eral Reserve given m o n etary aggregates m u ch ex p licit atten tio n . In Ja n u a iy 1970, th e F ed eral O pen M arket C om m ittee (FOMC) ex p ressed a d esire to p lace in creased em p h asis o n th e grow th o f ce rta in m o n e tary aggregates, but ex p licit targets for th e aggregates w ere not estab lish ed until 1975. From O cto b er 1979 to O cto b er 1982, th e FOMC em p h asized th e grow th of th e m o n etary aggregates even m ore; how ever, the w eight given to th e various aggregates ch an g ed over th is period . C on sequen tly, it is difficult to find an ex ten d ed period over w h ich m o n etary p o licy objectiv es are su fficiently stable to draw stron g in feren ce s about w h eth er th e F ed eral Reserve h as m o n etized th e debt. D espite th e se difficulties, w e provide so m e evidence, w h ich sh o u ld be regard ed as descriptive, o f th e rela tio n sh ip b etw een m o n ey grow th an d d ebt grow th during th e past tw o an d o n e-h a lf d ecad es. T h e em p irical investigation u n d ertak en h ere differs from th e usual p ro ced u re, w h ich is to estim ate a "Fed eral Reserve reactio n fu n ctio n .” T h e re actio n fu n ctio n is an eq u ation th at p resum ably re p rese n ts th e Fed eral R eserve’s re sp o n se to variables affecting its p o licy d e cisio n s. Stud ies th at have u sed th is type o f eq u ation have p ro d u ced in co n clu siv e resu lts an d suffer from tw o p ro b lem s.17 17For a more precise definition and a review of some of the empirical literature, see Dwyer. For other reviews of empirical studies, see Blinder and Barth, Sickles and Wiest. http://fraser.stlouisfed.org/ 36 Federal Reserve Bank of St. Louis DECEMBER 1984 First, m any o f th e se stu d ies have u sed reserve or b ase grow th as th e m o n etary p o licy variable. Sin ce th e evid ence su ggests th at th e F ed eral Reserve h as never targeted explicitly o n th e se variables, ch a n g es in th e grow th rates o f th e se variables sh o u ld n o t b e relied on to provide evid en ce o f d eb t m o n etizatio n . S econ d , th e se stu d ies in clu d e only co n te m p o ra n e ous values o f b o th th e m o n etary policy variable and th e m easu re o f fed eral d eb t grow th. S in ce th e se varia b les are co n sid ere d only sim u ltan eou sly , it ca n n o t be d eterm in ed w h eth er m o n etary grow th c a u s e s d ebt grow th o r vice versa. T h e ca u sa tio n ru n n in g from m on ey to d eb t is likely b ec a u se m o n ey grow th affects p rices, ou tp u t and n om in al in terest rates w h ich , in turn, feed back to d ebt grow th."1O f co u rse, no sta tisti cal p ro ced u re ca n esta b lish cau sality. T h e re is an ea s ily im p lem en ted p ro ced u re, how ever, w h ich ca n be u sed to test for th e tem p o ra l ord erin g o f tw o o r m ore variables. T h e p ro ced u re is called a test o f G ranger cau sality .'3 The Test P ro cedu re T h is test p ro ced u re ca n b e illu strated by u sin g the grow th rate o f M an d NFD. Let M, an d NFD, d en o te th e grow th rates o f m o n ey an d th e n et fed eral debt, re spectively, in th e cu rren t p eriod an d le t M,_,, M , ..., NFD, ,, N F D , ..., d en o te values o f th e se variables in previous period s. T h e test for G ranger cau sality ru n ning from M to NFD a m o u n ts to regressin g th e cu rren t value o f NFD on past values o f itself an d M, an d testin g th e h yp oth esis th at all o f th e co efficien ts on th e past values o f M are zero. T o te st th a t G ranger cau sality ru n s from NFD to M, th e cu rre n t value o f M is re g ressed on previous values o f itself a n d NFD, an d th e h yp oth esis th at th e co efficien ts on th e past values of NFD are zero is tested . If th e la tte r h y p oth esis is re jected , w hile th e form er is not, th e n it is said that grow th o f th e NFD G ran ger-cau ses (tem porally p re ced es) m o n ey grow th. If th e fo rm er is re je cted , w hile th e la tte r is n ot, th e n m o n ey grow th is said to Grangerca u se (tem porally p reced e) grow th o f th e NFD. If b o th are rejected , no tem p oral ord erin g ca n b e estab lish ed (i.e., th ere is feed b ack b etw ee n M an d NFD). If n eith e r ca n be re je cted , th e series are n ot tem p orally related (i.e., th ey are said to b e in d ep en d en t). 18See Dewald (1984), Carlson (1984) and footnote 3. 19Friedman and Schwartz (1963) were among the first to try to estab lish the temporal ordering between macroeconomic variables. De spite its name, it is now recognized that this procedure is not literally a test of causality, nor is it a test of statistical exogeneity. See Zellner (1979) for a discussion of causality, and see Jacobs, Learner and Ward (1979) and Wu (1983) for a discussion of the relationship between Granger causality and statistical exogeneity. FEDERAL RESERVE BANK OF ST. LOUIS T ests o f G ranger cau sality w ere perform ed over th e I/1960-IV/1983 period using tw o m easu res o f debt grow th th at have b ee n u sed in th e reactio n fu n ctio n literatu re. T h e first is th e grow th o f NFD, d iscu ssed earlier. T h e se co n d is th e high -em p loym ent budget d eficit (HEBD).2" T h e d eficit and ch an g es in th e NFD differ by th e so -called off-budget item s. T h e se item s are o m itted from th e official rep o rts o f th e deficit, d esp ite th e fact th ey requ ire d ebt issue. In ad dition, th e ch an g es in th e NFD and th e HEBD differ in th at th e latter is a d ju sted for cy clical factors, w hile th e form er is not. C onsequently, ch an g es in NFD m ay m isrep resen t th e p ressu re to m o n etize th e debt b eca u se th ey are n o t cy clically ad ju sted . In o th e r w ords, a given ch an g e in NFD is likely to be asso ciated w ith a m u ch sm aller effect on in terest rates if it o ccu rs in th e co n tra ctio n rath er th an th e ex p an sio n p h ase o f the cycle. F urtherm ore, sin c e a relatively larger portion of d eficits are cy clica lly in d u ce d , th e s e cy clic a l in flu en ces m ay be dom inant.-' If th e se cyclically in d u ced ch an g es in d ebt are not a sso ciate d w ith rising in terest rates, th ere is no p ressu re to m o n etize the debt. T hu s, b ec a u se th e cy clical effects have not b een co n tro lled for, th ere m ay be no tem poral ordering ru nning from NFD to m oney. C hanges in m on ey growth, m oreover, have b ee n show n to in d u ce cy clical sw ings in ec o n o m ic activity, so w e shou ld n ot be su r p rised to find a stron g effect ru n n in g from m o n ey to in co m e to NFD. To a cc o u n t for th e effects o f cy clical factors, lags o f ou tp u t grow th and lags o f th e inflation rate are in clu d ed in so m e o f th e tests o f G ranger ca u sality.-2 DECEMBER 1984 issue.23 F u rth erm o re, sin c e th e HEBD is cyclically a d ju sted , ch a n g es in past o u tp u t sh o u ld n o t affect tests o f G ranger cau sality ru n n in g from m o n ey to HEBD; past ch a n g es in p rices, how ever, m ay affect th e se tests. Finally, b ec a u se th e q u estio n o f d ebt m on etization is tied clo sely to th e policy ob jectiv es o f th e Federal Reserve, it is im p o rtan t to take a cc o u n t o f th e se policy objectives. T h u s, th e te sts o f G ranger cau sality w ere co n d u cte d over th e en tire period I/1960-III/1983 and over th e su b p erio d III/1972-III/1983, d uring w h ich at least som e co n sid era tio n w as given to m on ey stock objectives.24 B eca u se o f th e sh o rtn e ss o f th is period, it w as n ecessa ry to re strict th e sea rch to six lags on ea ch variable an d to in clu d e only th ree lags o f ou tp u t grow th and inflation. Em pirical R esults T h e G ranger cau sality tests w ere perform ed on quarterly grow th rates o f M l an d NFD an d on the quarterly grow th rate o f M l an d HEBD, follow ing a p ro ced u re o u tlin ed in T h o rn to n and B atten (1985).21 T h e sign ifican ce levels co rresp o n d in g to th e c a lc u lated F -statistics o f th e G ranger tests are rep o rted in tab les 1 -6 .26 T h e sig n ifican ce levels are p resen ted b e ca u se th e sig n ifican ce o f th e F -statistics vary w ith th e “ For example, the change in NFD in fiscal 1983 of $202.8 billion was made up of a $188.8 billion on-budget deficit and a $14.0 billion offbudget deficit. See Economic Report of the President (1983). Also, see Allen and Smith (1983). 24The FOMC stated its desire to place increased emphasis on certain monetary aggregates at its January 1970 meeting; however, the estimation period begins in 111/1972 to be conservative and to allow for the six lags of both variables. T h e advantage o f u sing th e HEBD is th at it is ad ju sted d irectly for cy clical factors. It too m ay m isrep re sen t th e p ressu re to m o n etize th e debt, how ever, b e cau se th e off-budget item s are o m itted . C onsequently, it m ay b e significantly sm aller th an th e am ou n t o f debt 25The fact that these tests ignore the question of whether changes in the debt affect market interest rates is particularly important in inter preting the results. If changes in debt have no effect on interest rates, we should not expect to find a temporal ordering running from debt growth to money growth. If changes in debt have an effect, we may or may not find such a temporal ordering. Thus, the lack of a temporal ordering running from debt to money growth could result either from a lack of an interest rate effect or from a refusal on the part of the Federal Reserve to monetize the debt. 20The data for HEBD ends in 111/1983, so the tests of Granger causality involving this variable were performed over this shorter period. Al though there are other ways to carry out these tests, work by Geweke, Meese and Dent (1983) and Guilkey and Salemi (1982) indicate that the procedure used here is preferred. Furthermore, in a rational expectations view, the Federal Reserve might anticipate the deficit and increase money growth in advance of the actual increase in the debt. In this case, money growth might precede debt growth, but we find no evidence of this temporal ordering. 21See Tatom (1984). 22Lags of past inflation are included based on the finding reported by Blinder and on the work of Horrigan and Protopapadakis (1982) and others who find that much of the measured deficits are related directly to inflation. It could also be argued that the lag from money growth to inflation is long. Therefore, the lags of past inflation may simply be a proxy for even longer lags of money growth. In addition to NFD, a relatively new measure, the cyclically ad justed federal debt calculated by deLeeuw and Holloway (1983), was used. The qualitative results with this variable were unchanged from those using NFD, so they are not reported here. ^Tests of Granger causality should be conducted with time series that are covariance stationary. When the autocorrelation functions of M1, NFD and HEBD were investigated, the series appeared station ary. When the Granger causality tests were undertaken including a time trend, however, the trend variable was always significant at the 5 percent level, suggesting that the series are not stationary. When the tests were performed on first differences of M1, NFD and HEBD, the time trends were uniformly insignificant. With one exception noted below, however, these results were not qualitatively different from those using the growth rates of M1 and NFD and the level of HEBD. The latter results are reported because they are easier to interpret. 37 DECEMBER 1984 FEDERAL RESERVE BANK OF ST. LOUIS Table 1 Significance Levels for Granger Causality Tests of M1 on NFD for the 1/1960—IV/1983 Period M1 on NFD Lags of NFD Lags of M1 1 2 3 4 5 6 7 8 g 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 .074 .080 .082 .221 .061 .038* .276 .318 .191 .203 .208 .203 .035* .041’ .041* .156 .068 .059 .204 .214 .277 .268 .255 .247 .046* .048* .047* .168 .110 .095 .327 .348 .367 .383 .401 .398 .089 .094 .091 .269 .179 .165 .478 .503 .493 .523 .562 .564 .124 .132 .121 .342 .235 .224 .609 .631 .636 .662 .680 .667 .190 .202 .187 .465 .343 .329 .726 .747 .757 .779 .785 .765 .269 .283 .263 .566 .396 .392 .756 .778 .804 .826 .842 .836 .264 .275 .257 .538 .355 .369 .745 .768 .814 .836 .840 .828 .229 .235 .228 .479 .306 .322 .758 .779 .830 .853 .851 .846 .260 .271 .260 .511 .349 .366 .743 .766 .823 .843 .851 .843 .310 .325 .308 .566 .407 .425 .774 .795 .862 .880 .889 .881 .332 .338 .319 .589 .422 .440 .779 .797 .873 .891 .900 .893 NFD on M1 Lags of M1 Lags on NFD 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 .248 .244 .195 .241 .245 .251 .246 .253 .300 .337 .462 .430 .514 .510 .420 .502 .509 .517 .509 .519 .578 .616 .757 .734 .000* .000* .000* .002* .001* .001* .001* .002* .002* .003* .002* .002* .000* .000* .000’ .000* .000* .000* .001* .001* .001* .001* .001* .001* .000* .000* .000* .001* .001* .001* .001* .001* .002* .002* .001* .001* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .000* .001* .001* .000* .000* .000* .000* .000* .000* .000* .001* .001* .001* .001* .002* .001* .001* .000* .000* .000* .001* .001* .001* .001* .002* .003* .004* .002* .001* .000* .001* .001* .002* .002* .002* .003* .003* .005* .006* .003* .002* .001* .001* .001* .003* .002* .003* .003* .004* .006* .008* .003* .002* 'indicates significance at the 5 percent level degrees o f freedom . T h e o u tco m e s that are significant at th e 5 p e rce n t level are d en o ted by an asterisk. For exam ple, in th e regressio n o f M l on NFD in table 1 the e n tiy for th ree lags on ea ch variable is .047. T h is in d i cates th at th e h y p o th esis that NFD d oes not Grangercau se M l can b e re je cte d at th e 5 p e rce n t sign ifican ce level for this lag sp ecificatio n . W hen th e lag length is in creased to four, how ever, th is h y p oth esis can n o t be re je cted b eca u se th e en tiy , .269, is g reater th an .05. T h e sign ifican ce levels b ased on sim ple b id irec tional tests o f G ranger cau sality b etw een M l and NFD an d M l and HEBD are p resen ted in tab les 1 an d 2, respectively. T h e resu lts in table 1 in d icate a strong u n id irectio n al effect ru n n in g from M l to NFD. Only seven o f th e 144 F -tests for th e in flu en ce o f NFD on M l rep o rted w ere significant at th e 5 p e rce n t level. None 38 o f th ese seven lag stru ctu res, how ever, w as c h o se n by a co m m o n ly u sed lag-length sp ecifica tio n criterio n .27 B ecau se NFD is n ot cy clically a d ju sted an d is likely to b e affected by ch a n g es in real o u tp u t an d p rices in d u ced by ch a n g es in th e m o n ey supply, it is not too surprising that th e tem p o ral ord erin g ru n s from M l to NFD/" 27The lag-length selection criterion used here is the final prediction error. See Thornton and Batten, and Batten and Thornton (1984). 2Slt is somewhat surprising, however, that the same qualitative result is obtained for the cyclically adjusted debt measure. This suggests the possibility that this cyclically adjusted measure does not capture all the effects of past output and price level growth. This conjecture is supported by the fact that the significance levels are greatly increased when three lags of output growth and inflation were in cluded in these specifications. In any event, there is no evidence of a temporal ordering running from cyclically adjusted debt to money. DECEMBER 1984 FEDERAL RESERVE BANK OF ST. LOUIS Table 2 Significance Levels for Granger Causality Tests of M1 on HEBD for the 1/1960-111/1983 Period_________________________________________ M1 on HEBD Lags of HEBD Lags of M1 1 2 3 4 5 6 7 8 9 10 11 12 2 3 4 5 6 7 8 9 10 11 12 .000* .000* .000* .000* .000* .000* .002* .003* .004* .004* .005* .006* .001* .001* .001* .002* .001* .001* .010* .011* .015* .015* .021* .022* .003* .004* .004* .004* .002* .002* .021* .025* .028* .030* .041* .041* .003* .003* .004* .002* .001* .001* .009* .010* .016* .018* .028* .028* .007* .008* .008* .005* .003* .003* .018* .021* .032* .037* .054 .053 .015* .016* .017* .010* .007* .006* .032* .036* .052 .060 .085 .082 .018* .019* .019* .010* .008* .008* .023* .027* .035* .040* .057 .049* .028* .030* .030* .018* .013* .013* .041’ .046* .058 .066 .092 .080 .015* .018* .018* .009* .005* .004* .021* .022* .018* .022* .031’ .027* .021* .024* .023* .012* .005* .005* .018* .020* .020* .022’ .031* .028* .035* .039* .037* .019* .009* .009* .026* .030* .030* .033* .044* .039* .045* .050 .051 .029* .016* .016* .040* .044* .041* .044* .058 .057 HEBD on M1 Lags of M1 of HEBD 1 2 3 4 5 6 7 8 9 10 11 12 _____________ 1 1 2 3 4 5 6 7 8 9 10 11 12 .171 .208 .195 .209 .358 .358 .360 .243 .200 .138 .159 .169 .379 .446 .362 .380 .604 .614 .615 .482 .440 .328 .342 .366 .298 .339 .444 .435 .599 .597 .582 .460 .437 .276 .307 .418 .270 .297 .365 .390 .692 .696 .689 .533 .503 .349 .415 .514 .277 .307 .401 .429 .805 .815 .810 .661 .642 .474 .547 .661 .150 .166 .183 .195 .408 .432 .451 .306 .324 .267 .284 .364 .169 .188 .219 .235 .426 .450 .463 .247 .285 .263 .292 .325 .225 .250 .292 .311 .504 .528 .541 .340 .382 .355 .380 .409 .292 .324 .356 .372 .586 .610 .621 .421 .451 .406 .424 .406 .181 .204 .286 .282 .528 .552 .562 .418 .451 .450 .483 .465 .195 .216 .294 .308 .592 .616 .624 .484 .521 .513 .509 .476 .197 .222 .320 .335 .666 .689 .699 .558 .595 .591 .581 .483 'indicates significance at the 5 percent level T h e results in table 2 for th e HEBD, how ever, in d i ca te u n id irectio n al cau sality ru n n in g from HEBD to M l. T h e h yp othesis th at HEBD h as n o im p act on M l w as re je cte d for nearly eveiy lag sp ecificatio n co n sid ered , w hile th e h y p oth esis th at M l h as no effect on HEBD w as never re je cted . T hu s, th is m easu re suggests th at m on ey grow th resp o n d s to cyclically ad ju sted ch a n g es in th e debt. The G ranger Tests Extended f o r Cyclical Influences In th e sim p le te sts o f bivariate G ranger cau sality p re se n te d above, th e o bserv ed feed b ack b etw een m on ey grow th and NFD o r th e cau sality ru n n in g from HEBD to M l co u ld be th e result o f th e clo se a sso cia tio n b etw een th e se variables an d fa cto rs not a c co u n te d for by th e eq u atio n . In o rd er to guard against th is possibility, th e tests w ere rep ea ted adding th ree and th e n six lags o f th e grow th rates o f p rices an d real output as ad d ition al variables."1 T h e results for th e eq u atio n s w ith six lags are p re sen ted in tables 3 an d 4. T h e resu lts in table 3 in d icate “ The possibility that money growth responds to either past output growth or inflation can be argued two ways. First, such variables could represent a Federal Reserve reaction function response, e.g., high past rates of inflation or output growth induce the Fed to slow the rate of M1 growth. Second, the money supply could be endoge nous (at least over short periods of time like a quarter), i.e., related to other variables in the system like interest rates. Since interest rates are positively related to both inflation and output growth, the money stock should move with these variables. If the second case were correct, there should be a positive relationship between past infla tion and money growth; however, Blinder reports a negative rela tionship. We find the same result, although it is not reported here. 39 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 Table 3 Significance Levels for Granger Causality Tests of M1 on NFD for the 1/1960—IV/1983 Period, with six lags of P and X M1 on N fr) Lags of NFD Lags of M1 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 8 7 8 9 10 11 12 .463 .505 .486 .665 .464 .401 .577 .529 .355 .356 .410 .407 .255 .279 .188 .409 .340 .340 .358 .363 .472 .477 .520 .529 .248 .255 .202 .464 .474 .484 .509 .509 .561 .561 .612 .616 .376 .358 .297 .562 .629 .621 .642 .643 .722 .723 .758 .764 .478 .473 .369 .634 .707 .729 .757 .760 .825 .827 .843 .850 .611 .606 .498 .729 .805 .826 .841 .842 .892 .894 .900 .907 .708 .697 .584 .811 .875 .892 .904 .905 .942 .944 .948 .952 .673 .611 .470 .653 .606 .690 .738 .741 .809 .812 .813 .824 .680 .637 .546 .734 .677 .758 .810 .814 .875 .877 .877 .886 .754 .723 .644 .814 .763 .832 .868 .871 .922 .924 .924 .930 .734 .687 .564 .712 .710 .780 .821 .817 .903 .906 .913 .920 .706 .601 .477 .678 .626 .690 .751 .736 .852 .852 .867 .875 n £d on M1 Lags Of M1 Lags of NFD 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 .477 .466 .397 .394 .395 .419 .469 .474 .589 .604 .708 .728 .415 .419 .368 .388 .401 .401 .413 .410 .370 .400 .487 .519 .060 .069 .055 .059 .063 .060 .067 .071 .125 .128 .094 .102 .061 .065 .043* .046* .049* .063 .065 .069 .137 .136 .093 .102 .091 .097 .072 .076 .082 .102 .112 .117 .220 .218 .147 .158 .104 .113 .089 .096 .099 .123 .134 .138 .204 .202 .138 .147 .046* .053 .040* .044* .047* .063 .070 .074 .143 .154 .106 .110 .077 .087 .065 .071 .076 .098 .109 .115 .198 .210 .160 .164 .113 .127 .094 .101 .108 .136 .147 .155 .237 .254 .164 .173 .125 .141 .088 .092 .098 .124 .135 .144 .233 .254 .166 .171 .178 .198 .130 .136 .142 .177 .192 .202 .310 .335 .231 .238 .130 .144 .085 .085 .085 .109 .127 .135 .224 .245 .145 .155 •indicates significance at the 5 percent level that M l and NFD are in d ep en d en t series. T h ere are ju st seven in sta n ce s w h ere th e F -tests o f NFD on M l are significant, an d n o n e o f th ese w ere selecte d by the lag-length sp ecificatio n criterion:'" T h u s, o n ce output grow th an d inflation are a cc o u n te d for, th e re is virtu ally no evid ence o f a sep arate effect o f m o n ey grow th on d ebt grow th an d n o evid ence o f cau sality from NFD to M l. the tests of the effect ru n n in g from HEBD to M l are su bstantially larger w h en grow th rates o f output and p rices are a cco u n ted for. N evertheless, th e HEBD p ro vides som e evid en ce o f d ebt m o n etiz a tio n n o t evident w h en NFD is used. Results f o r III/1972-IV /1983 T h e resu lts in table 4 are sim ilar to th o se in table 2, in that M l has no effect on HEBD, w hile HEBD c o n tin u es to G ran ger-cau se M l. A co m p ariso n o f tab les 2 an d 4, how ever, show s th at th e sig n ifican ce levels for T h e resu lts for th e III/1972-III/1983 period, in w h ich m ore em p h asis w as p la ced on th e m o n etary aggre gates, are rep o rted in tab les 5 an d 6. (Only th e resu lts w ith lags o f in flatio n an d o u tp u t g row th are r e p o rted .)1' T h e se resu lts in d ica te that, over this period, ^When first differences of growth rates are used, there is no area of the lag space where the hypothesis can be rejected. 31When no lags of inflation and output growth are used, the results indicate unidirectional causality running from M1 to NFD and inde pendence between M1 and HEBD. 40 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 Table 4 Significance Levels for Granger Causality Tests of M1 on HEBD for the 1/1960-111/1983 Period, with six lags of P and X M1 on HEBD Lags of HEBD Lags of M1 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 .014* .016* .021* .012* .023* .026* .033* .033* .044* .045* .056 .057 .046* .054 .071 .042* .064 .069 .086 .088 .126 .129 .159 .165 .105 .117 .140 .087 .104 .120 .146 .151 .184 .190 .227 .242 .018* .024* .026* .010* .009* .009* .011* .012* .024* .024* .035* .038* .038* .048* .053 .022* .018* .018* .022* .024* .046* .044* .062 .068 .068 .084 .090 .041* .035* .035* .042* .045* .083 .079 .107 .117 .035* .046* .037* .015* .011* .013* .012* .013* .025* .023* .032* .035* .055 .072 .059 .027* .019* .022* .022* .023* .041* .038* .053 .057 .058 .081 .070 .027* .016* .017* .017* .018* .028* .026* .036* .038* .090 .123 .104 .041* .020* .024’ .021* .022* .034* .035* .046* .050 .133 .173 .147 .063 .034* .038* .034* .035* .053 .055 .070 .075 .090 .117 .122 .052 .045* .051 .049* .051 .078 .079 .100 .103 HEBD on Ml Lags of M1 Lags of HEBD 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 .059 .079 .096 .097 .253 .257 .258 .119 .108 .068 .075 .086 .112 .157 .151 .166 .370 .358 .382 .204 .207 .128 .124 .144 .076 .099 .135 .155 .281 .279 .346 .196 .194 .083 .087 .118 .143 .182 .237 .267 .432 .429 .509 .323 .318 .156 .163 .212 .223 .271 .321 .353 .446 .450 .521 .375 .365 .216 .221 .259 .268 .322 .331 .379 .393 .382 .416 .343 .344 .239 .234 .277 .361 .425 .440 .489 .484 .480 .487 .372 .375 .309 .311 .348 .453 .526 .550 .601 .592 .586 .586 .475 .481 .410 .415 .457 .538 .617 .617 .687 .682 .678 .676 .567 .565 .485 .496 .520 .413 .483 .564 .653 .737 .733 .737 .655 .653 .587 .597 .621 .466 .526 .605 .671 .791 .794 .798 .707 .709 .641 .629 .652 .329 .390 .517 .574 .790 .787 .814 .712 .721 .671 .655 .633 'indicates significance at the 5 percent level M l is in d ep en d en t o f bo th NFD and th e HEBD. T here w as n o portion o f th e lag sp ace co n sid ere d in w hich the h yp o th eses co n stitu tin g th e G ranger te sts cou ld b e re je cted . H ence, th ere is no evid ence o f d ebt m o n e tization over this period for eith e r m easu re o f debt grow th. T hu s, th e HEBD result, w h ich in d icates that th e Fed eral Reserve h ad m o n etized th e d ebt over the 1 9 6 0 -8 3 period, ap p ears to resu lt from th e Fed eral R eserve’s in terest rate target p ro ced u res over nearly the first h alf o f th e period — a period w hen debt m o n etizatio n is m ore likely to be an in h e ren t resu lt of attem p ts to in flu en ce in terest rates. An investigation of a period for w h ich it is m ore relevant to co n sid er the qu estio n o f debt m o n etizatio n y ield s no evid en ce that th e Fed eral Reserve h as m o n etized th e debt. CONCLUSIONS T h e p u rp o se o f th is article w as to cle a r up co n fu sio n that often ch a ra cte riz e s d iscu ssio n s o f d ebt m o n etiza tion an d to provide so m e evid en ce on th e q u estio n o f w h eth er th e F ed eral Reserve has m o n etized the debt. Specifically, it w as p o in ted ou t th at th e p h rase "m one tizing th e d eb t" m e a n s m on ey grow th in ex cess o f that required to ach ieve so m e policy ob jectiv e th at is in d u ced by rapid grow th in th e federal debt. It w as n o ted th at th e ratio o f Fed eral Reserve debt hold ings to n et fed eral debt, o r o th e r su ch m easures, ca n n o t b e u sed alo n e as evid en ce o f d eb t m o n etiz a tion. C hanges in th e m o n ey m u ltip lier an d factors that affect co m p o n e n ts o f th e m o n etary b a se w ill in flu en ce 41 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 Table 5 Significance Levels for Granger Causality Tests for M1 and NFD for the 111/1972—IV/1983 Period, with three lags of P and X M1 on NFD Lags of NFD Lags of M1 1 1 2 3 4 5 6 .310 .377 .395 .524 .315 .229 2 3 .236 .254 .180 .242 .223 .221 .415 .435 .325 .366 .360 .361 1 2 3 4 5 6 .574 .611 .489 .532 .531 .532 5 6 .719 .749 .628 .661 .667 .676 .830 .852 .752 .775 .789 .796 5 6 NFD on M1 Lags of M1 V Lags of N f D 4 1 2 .884 .782 .776 .804 .816 .813 4 3 .296 .276 .342 .498 .512 .530 .329 .347 .526 .676 .682 .695 .296 .301 .562 .673 .675 .688 .328 .347 .605 .680 .681 .676 .212 .244 .441 .542 .516 .526 Table 6 Significance Levels for Granger Causality Tests for M1 and HEBD for the III/1972-III/1983 Period, with three lags of P and X M1 on HEBD Lags of HEBD Lags of M1 1 2 3 4 5 6 1 2 3 4 5 6 .433 .473 .495 .317 .367 .336 .737 .768 .788 .597 .659 .629 .741 .799 .842 .732 .833 .813 .782 .843 .879 .681 .779 .751 .826 .875 .914 .808 .885 .865 .857 .913 .941 .854 .933 .910 HEBD on M1 Lags of M1 Lags of HEBD 1 2 3 4 5 6 1 2 3 4 5 6 .183 .234 .299 .316 .764 .765 .411 .494 .558 .586 .911 .915 .088 .109 .156 .168 .307 .323 .167 .202 .273 .289 .425 .444 .272 .320 .400 .421 .530 .548 .108 .125 .164 .186 .178 .174 http://fraser.stlouisfed.org/ 42 Federal Reserve Bank of St. Louis FEDERAL RESERVE BANK OF ST. LOUIS the grow th o f th e Fed eral R eserve’s portfolio o f govern m ent secu rities for any given policy o bjective in ways that co n fo u n d attem p ts to d eterm in e th e ex ten t of d ebt m o n etizatio n taking place. Tw o co m m o n ly u sed m easu res o f debt grow th, the grow th o f n e t fed eral d ebt and th e h igh-em ploym ent budget deficit, w ere u sed to test w h eth er m oney grow th p re ce d e s d eb t grow th, o r vice versa. T h e resu lts for th e III/1972-IV/1983 period, during w hich th e Fed eral Reserve p laced m o re em p h asis on th e monetary^ aggregates th an it h ad in previous years, show s n o evid en ce o f d ebt m o n etizatio n by th e F ed eral Reserve u sin g eith e r d ebt m easu re. F o r th e en tire 1 9 6 0 -8 3 period, th ere is evid en ce o f d ebt m o n etizatio n for th e h ig h -em p loy m ent d eficit m easu re, bu t n ot for grow th o f th e n e t fed eral debt. T h u s, th e only evid ence of d ebt m o n etizatio n o cc u rs d uring th e period o f in terest rate targeting, w h en d ebt m o n etizatio n is to be ex p ected if in crea ses in th e federal d ebt put upw ard pressu re on in terest rates. T h e read er is cau tio n ed , how ever, in th at actu al m on ey grow th, rath er th an deviations o f actu al from desired m o n ey grow th, w as u sed in th e se tests. Sin ce th e d ebt m o n etizatio n h as to do w ith m ovem ents away from p o licy objectives in d u ced by actu al o r p e r ceived p ressu re o f rapid d ebt grow th on in tere st rates, th e critica l im p licit assu m p tio n here, an d in m ost p re vious stu d ies o f d ebt m o n etizatio n , is that actu al ch an g es in m o n ey grow th proxy su ch m ovem ents. REFERENCES Ahearn, Daniel S. Federal Reserve Policy Reappraised: 1951-1959 (Columbia University Press, 1963). Allen, Stuart D., and Michael D. Smith. “Government Borrowing and Monetary Accommodation,” Journal of Monetary Economics (No vember 1983), pp. 605-16. Barth, James R., George Iden and Frank S. Russek. "The Eco nomic Consequences of Federal Deficits: An Examination of the Net Wealth and Instability Issues,” processed (1984). Barth, James R., Robin Sickles and Philip Wiest. "Assessing the Impact of Varying Economic Conditions on Federal Reserve Be havior,” Journal of Macroeconomics (Winter 1982), pp. 47-70. Batten, Dallas S., and Daniel L. Thornton. “How Robust Are the Policy Conclusions of the St. Louis Equation?: Some Further Evi dence,” this Review (June/July 1984), pp. 26-32. Blinder, Alan S. “On the Monetization of Deficits," in Laurence H. Meyer, ed., The Economic Consequences of Government Deficits (Kluwer-Nijhoff, 1983), pp. 39-73. Board of Governors of the Federal Reserve System. "Announce ments,” Federal Resen/e Bulletin (October 1974), pp. 740-41. Brown, W. W., and G. J. Santoni. "Monetary Growth and the Timing of Interest Rate Movements,” this Review (August/September 1983), pp. 16-25. Carlson, Keith M. "Money Growth and the Size of the Federal Debt," this Review (November 1984), pp. 5-16. DECEMBER 1984 Darby, Michael R. “Some Pleasant Monetarist Arithmetic," Federal Reserve Bank of Minneapolis Quarterly Review (Spring 1984), pp. 15-20. deLeeuw, Frank, and Thomas M. Holloway. "Cyclical Adjustment of the Federal Budget and Federal Debt," Survey o f Current Business (December 1983), pp. 25-40. Dewald, William G. “Deficits and Monetary Growth,” Federal Re serve Bank of Atlanta Economic Review (January 1984), pp. 1120 . Dwyer, Gerald P. cessed (1984). "Money Creation and Federal Deficits," pro Economic Report of the President, February 1983. ment Printing Office, 1983). (U.S. Govern Evans, Paul. “Do Large Deficits Produce High Interest Rates?" processed (1984). Friedman, Milton, and Anna J. Schwartz. A Monetary History of the United States 1867-1960 (Princeton University Press, 1963). Froyen, Richard T. "A Test of the Endogeneity of Monetary Policy,” Journal of Econometrics (July 1974), pp. 175-88. Geweke, John, Richard Meese and Warren Dent. “Comparing Al ternative Tests of Causality in Temporal Systems: Analytic Results and Experimental Evidence,” Journal of Econometrics (February 1983), pp. 161-94. Guilkey, David K., and Michael K. Salemi. “Small Sample Proper ties of Three Tests for Granger-Causal Ordering in a Bivariate Stochastic System,” The Review of Economics and Statistics (No vember 1982), pp. 668-80. Historical Statistics of the United States: Colonial Times to 1970, part 2. (GPO, 1975). Horrigan, Brian R. “Pitfalls in Analyzing Deficits and Inflation," Fed eral Resen/e Bank of Philadelphia Working Paper No. 83-4 (1983). Horrigan, Brian R., and A. Protopapadakis. “Federal Deficits: A Faulty Gauge of Government's Impact on Financial Markets," Fed eral Reserve Bank of Philadelphia Business Review (March/April 1982), pp. 3-16. Jacobs, Rodney L., Edward E. Learner and Michael P. Ward. "Diffi culties with Testing for Causation," Economic Inquiry (July 1979), pp. 401-13. Melvin, Michael. "The Vanishing Liquidity Effect of Money on Inter est: Analysis and Implications for Policy," Economic Inquiry (April 1983), pp. 188-202. Sargent, Thomas J., and Neil Wallace.. “Some Unpleasant Mone tarist Arithmetic," Federal Reserve Bank of Minneapolis Quarterly Review (Fall 1981), pp. 1-17. Tatom, John A. "A Perspective on the Federal Deficit Problem," this Review (June/July 1984), pp. 5-17. _________ "Issues in Measuring An Adjusted Monetary Base,” this Review (December 1980), pp. 11-29. Thornton, Daniel L. “The Government Budget Constraint with En dogenous Money," Journal o f Macroeconomics (Winter 1984), pp. 57-67. Thornton, Daniel L., and Dallas S. Batten. “Lag-Length Selection and Granger Causality Between Money and Income," Journal of Money, Credit and Banking (forthcoming, May 1985). Wu, De-Min. “Tests of Causality, Predeterminedness and Exoge neity,” International Economic Review (October 1983), pp. 547-58. Zellner, Arnold. "Causality and Econometrics,” in Karl Brunner and Allan H. Meltzer, eds., Carnegie-Rochester Conference Series on Public Policy, Volume 10 (Amsterdam: North-Holland 1979), pp. 9-54. 43 FEDERAL RESERVE BANK OF ST. LOUIS DECEMBER 1984 FEDERAL RESERVE BANK OF ST. LOUIS REVIEW INDEX 1984 JANUARY Michael T. Belongia and Thomas H. Gregory, “Are Op tions on Treasury Bond Futures Priced Efficiently?” G. J. Santoni, “ Employment Trends in St. Louis: 195482” Dallas S. Batten and Daniel L. Thornton, “ How Robust Are the Policy Conclusions of the St. Louis Equation?: Some Further Evidence” AUGUST/SEPTEMBER FEBRUARY Dallas S. Batten and R. W. Hafer, “Currency Substitu tion: A Test of Its Importance” Michael T. Belongia, “The Dairy Price Support Program: A Study of Misdirected Economic Incentives” G. J. Santoni, “ Interest Rate Risk and the Stock Prices of Financial Institutions” A. Steven Holland, “ Does Higher Inflation Lead to More Uncertain Inflation?” A. Steven Holland, “The Impact of Inflation Uncertainty on the Labor Market” R. Alton Gilbert, “Calculating the Adjusted Monetary Base under Contemporaneous Reserve Requirements” R. W. Hafer, “ Examining the Recent Behavior of Infla tion” MARCH OCTOBER K. Alec Chrystal, “A Guide to Foreign Exchange Mar kets” Dallas S. Batten and Michael T. Belongia, “The Recent Decline in Agricultural Exports: Is the Exchange Rate the Culprit?” R. W. Hafer, “The Money-GNP Link: Assessing Alterna tive Transaction Measures” APRIL K. Alec Chrystal, “ International Banking Facilities” G. J. Santoni, “A Private Central Bank: Some Olde En glish Lessons” Michael T. Belongia, “ Money Growth Variability and GNP” MAY R. Alton Gilbert and A. Steven Holland, “ Has the Deregu lation of Deposit Interest Rates Raised Mortgage Rates?” Dallas S. Batten and Mack Ott, “What Can Central Banks Do About the Value of the Dollar?” K. Alec Chrystal, “ Dutch Disease or Monetarist Medi cine?: The British Economy under Mrs. Thatcher” Michael T. Belongia and G. J. Santoni, “ Hedging Interest Rate Risk with Financial Futures: Some Basic Princi ples” Daniel L. Thornton, “An Early Look at the Volatility of Money and Interest Rates under CRR” NOVEMBER Keith M. Carlson, “ Money Growth and the Size of the Federal Debt” Mack Ott, “ Depreciation, Inflation and Investment Incen tives: The Effects of the Tax Acts of 1981 and 1982” John A. Tatom, “ Interest Rate Variability: Its Link to the Variability of Monetary Growth and Economic Perfor mance” DECEMBER JUNE/JULY John A. Tatom, “The 1981 Personal Income Tax Cuts: A Retrospective Look at Their Effects on the Federal Tax Burden” John A. Tatom, “A Perspective on the Federal Deficit Problem” A. Steven Holland, “ Real Interest Rates: What Accounts for Their Recent Rise?” R. W. Hafer, “ Money, Debt and Economic Activity” Daniel L. Thornton, “ Monetizing the Debt”