The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
NOVEMB R 1976 Com ercial Bank Loans and The Money Supply ... ... .... .... . ... pa e 3 The Concept of Priva e Income ....... . ..... .. ..... page 11 Subscriptions to the Monthly Review are available to the public without charge. Additional copies of any issue may be obtained from the Research Division, Federal Reserve Bank of Kansas City, Kansas City, Missouri 64198. Permission is granted to reproduce any material in this publication provided the source is credited. Commercial Bank Loans and The Money Supply Byl. A. Cacy he lending activity of the nation's commercial banks has been unusually sluggish over the past year or two. Total bank loans in September of 1976 were only moderately higher than a year ea rlier. The growth in the narrowly defined money supply, MI (i.e., publicly held currency plus demand deposits). also has been on the weak side during the past year. Ml increased 4.3 per cent in the year ending September 1976, which is below the rate that would be expected during a period of substantial economic growth accompanied by a sizable, if diminished , inflation rate. Some observers have argued that the weakness in the growth rate of the narrowly defined money supply is closely related to the sluggish behavior of loans at commercial banks. This article examines the relations hip between bank loans and M 1. The first section discusses possible theoretical reasons why the behavior of the two variables may be closely related. Next, the actual behavior of bank loans and M 1 during the last 25 years is examined to determine whether or not the two variables have behaved in similar ways. Finally, the article examines the behavior of bank loans, bank credit, and the broader measures of the money supply to determine if any of these variables show common behavior patterns. T THEORETICAL RELATIONSHIP BETWEEN LOANS AND MONEY Observers who claim that the behavior of the narrowly defined money supply is closely Monthly Review • November 1976 related to that of commercial bank loans hold that M 1 tends to increase when loans are increasing and to decline when loans are declining. These observers offer several reasons for expecting the close relationship to exist. Some of these reasons are potentially val id, while the validity of others is questionable. Certain observers, for example, appear to argue that Ml and loans are necessarily related because the public cannot increase its Ml balances unless banks make loans. It is true, of course, that commercial bank lending is an important source of potential increase in Ml. When banks make loans , the proceeds may be and frequentl y are used by the public to augment its M 1 balances. Bank lending , however, is only one of several sources of potential increase in the narrowly defined money supply. Another source is the investing activity of commercial banks . As in the case of loans, when banks acquire investments, such as U.S. Government securities, the public may use the proceeds to augment its M 1 balances. A third source of potential increase in money balances is the asset-acquiring activities of the Federal Reserve System. When the Federal Reserve buys U.S. Government securities, the proceeds potentially may be used by the public to add to its M 1 balances. Potential increases or decreases in M 1 balances, therefore, result from changes in the assets-investments as well as loans-held by the nation's banking system, i.e. , the Federal 3 Commercial Bank Loans and the Money Supply Reserve and commercia l banks. When the banking system adds to its assets, a potential increase in M 1 occurs; and when t he banking system reduces its assets, a potential decline in MI occurs. Of course, these changes in Ml may not always materialize because the public may not add to M 1 balances when the banking system acquires assets. Instead, the public may add to its time deposits or to other non -M 1 liabilities of the banking system . 1 The potential change in M 1 that results from a change in bank loans during a particular time period can be seen b y examining t he consolidated balance sheet of the banking sys tem . The assets of the b a nking syste m co n s is t of Fed e ral R e erve assets a nd commercial ba nk credit- that is, loa ns plus in vest ments. Liab ilities may be broken d own into M 1 and non -Ml li abi liti es. The poten tia l increase in M 1 during the year ending in June 1976 was $42.5 billion, as the assets of the banking system increased by that amount. (See Table 1.) Of the total increase in assets, only $9.8 billion was accounted for by an increase in bank loans. Th e largest single positive influence on Ml was the $24 .0 billion rise in b ank investments. 2 Because bank lending is only one of several sources of potential increase in the mon ey suppl y, M 1 can increase even though loans do not. Also, loans m ay increase while M 1 remains I Changes in MI balances m ay occ ur in the absence of ch a nges in the loans and investments a nd other assets of th e banking system. For examp le , wh en business firms redeem their large CD's , they may place t h e proceeds in MI balances . In this case. MI in creases even though there h as been no ri se in earning assets. Of course , the reduction in large CD' s a nd the in crease in MI bala nces m ay h ave an indirect impact o n com m ercia l banks. It may. for exam ple , increase th e required rese rves of co mm e rc ia l banks and induce ba nk s to reduce th e ir ea rnin g assets. 2 Tab le I shows that MI did not increase by th e a m ou nt of th e rise in bankin g system assets because a $30.4 billion in crease in non- MI li abi li t ies absorbed most of the rise in assets. Thus. Ml increased o nl y $12. 1 bi llion ($42.5 billion minu s $30.4 billi on) . 4 unchanged. Thus , the behavior of Ml is not necessan·ty related to bank lending activity. Nevertheless, there are several reasons that a close relationship may exist between loans and M 1. One reason is that banks typically require their business loan customers to hold compensating balances. Such balances may be related to the size of the customer's loan so that as loan amounts rise and fall, compensating demand balances rise and fall. Since these compensa ting balances are included in the mo ney supply , their sensitivity to changes in loa n vol ume may res ult in a correlation between the mon ey supply and bank loans. An.ot h e r facto r that would lead to a cor relation between th e m oney supply a nd bank loa ns is that th e d ema nd for mo ney balances may tend to ri se a nd fa ll in line with the d em a nd for loa ns at banks. For exa mple, during a n economic upturn, rising consumer and business spending may be financed in part by loans from commercial banks. At the same time, the expanding economy may lead the public to add to its money balances. Also , during an economic downturn, the public may be inclined to reduce its indebtedness to commercial banks ; and the low economic growth rate may reduce the public's need for money balances. Thu s, the demand for bank loa ns a nd money balances may rise and fall together a nd m ay tend to cause a close correlation between bank loans and the money supply. Finall y, the conduct of monetary policy may tend to produce a correlation between bank loans and the money supply. Monetary authorities may tend to accommodate an expansion of bank loans by increasing reserves during periods in which bank loans are expa nding. Increasing bank reserves , in turn , may give rise to increas ing money balances . By th e sa me token, when bank loans are fa llin g the Federal Reserve may respond by reducing bank reserves , giving rise to a decline in money balances. Therefore, to the extent that Feder.:il Reserve BanK of K:ansas City Commercial Bank Loans and the Money Supply Table 1 CONSOLIDATED BALANCE SHEET U.S. BANKING SYSTEM (FEDERAL RESERVE AND COMMERCIAL BANKS) (In Billions of Dollars) LIABILITIES ASSETS Levels as of June 1976 Federal Reserve Assets : U.S . Govt. securities Oth er assets Total Comm ercia l Bank Credit : L oans I nvest rn ents T ota l Money Supply (M1) 95.6 26 .8 122.4 Large CD's Oth er l iabilit ies Total 511 .7 236 . 1 -- 399 .2 69 .2 99.4 567 .8 747 .8 870 .2 T otal A sse ts Non -M1 Liabiliti es: T ime deposits other than large CD 's 302.4 870.2 Total L iab ilities Ch ang es Dur ing Y ear Ending June 1976 8 .7 Federal R eserve A sse ts : Commercial Bank Credit : Loans Investmen t s Total 12.1 Non-M1 L iabilities : 9 .8 24 .0 -- T otal A sset s 33 .8 42 .5 mon etary policy accommodates alterations in bank loans, a correlation between bank loans and the money su ppl y may result. In brief. compensa ting balances , demand factors. and the conduct of monetary policy may possibly result in a close correlation between bank loans a nd the money supply. Wh ether or not such a correlation has resulted from these or other factors is examined in the foll ow ing section. STORICAL RELATIONSHIP BETWEEN LOANS AND M1 Some broad si mil arities may be observed betwee n the behavior of bank loans and M 1 Monthly Review• November 1976 Money Supply (M1) Time deposits other than large CD 's Large CD 's Other I iabil ities 46 .0 - 13 .1 - 2 .5 To tal 30.4 Total L iabilities 42 .5 during the past 25 yea rs. Both M 1 and tot a l bank loans, as well as business loans. have increased steadi ly throughout most of the period from 1952 through the early part of 1976. (See Chart 1.) The increase in loa ns indicates that commercial bank lending activity has been an important source of potent ial increase in M 1. From 1952 through 1975, for examp le. total loans increased $448 billion and accounted for about 70 per cent of th e rise in the assets of the banking system. Other similarities exist between th e behavior of bank loans and Ml . For example. the upward trends of both loans and M 1 are slowed or in terrupted from time to time. Moreover, m 5 Commercial Bank Loans and the Money Supply Chart 1 OMMERCIAL BANK LOAN5 AND THE MONEY SUPPLY 1952-76 Bil lions of Doi lors p p I 500 400 300 200 100 0 1952 1 54 1 56 1 58 1 60 1 62 1 64 1 66 1 68 1 70 1 72 1 74 1 76 NOTE: The shaded areas in Charts 1 , 2, and 3 pertain to recessions , with the p 's and t's at the top of the charts referring to peaks and t roughs of business cycles. With two exceptions , these reference periods are designated by the National B ureau of Econom ic Research (N BER ) . The fi rst exception is the early-1967 period , which is inc luded here because the behavior of financial variables was quite similar to other recessionary periods. The second exception is that while the NBER has not yet designated a trough for the current recession , the trough is s hown here as the first quarter of 1975 . both cases. these interrupt ions are associated wit h business cycles, as may be seen in Chart 2, which shows the growth rates of total loans and M 1. The growth rates of both of these variables decline fo r several quarters prior to the onset of economic recessions. Also, for both loans and M 1, growth rates tend to accelerate during the first few quarters of economic recoveries. Beyond these broad similarities. however. the behavior of bank loans and M 1 is not closely correlated. For example. the statist ical correlation between simultaneous move ments in the growth rates of these two variables is quite low. Specifically, for a regression involving the current growth rates of M 1 and bank loans 6 from the tirst quarter of 1950 through the second quarter of I 976. the coefficient of determination , R2 , is only .10. (See Table 2.) This means that movements in the growth rate of loans during any quarter explained only about 10 per cent of the movements in the growth rate of M 1 during the same quarter. Regression analysis shows further that M 1 and business loans are not correlated at all. as the R2 between the growth rates of the two va riables is zero. 1 The contrasting behavior of the growth rate s o f loa ns and MI is illustrated by their m ove ments during th e 1973 -75 recession a nd s u bseq u en t period of economic recovery. Federal Reserve Bank of Kansas City Commercial Bank Loans and the Money Supply Chart 2 GROWTH RATES OF BANK LOANS AND THE MONEY SUPPLY Quarterly Averages , 1952-76 Per Cent p I 25 20 15 10 5 0 -5 1952 During the recovery, the money supply grew more rapidly than during the recession. In contrast, bank loa ns grew much more slowly during the recovery than during the recession. In the recovery period from the first quarter of 1975 through the third quarter of 1976. Ml 3 The R2' s in Table 2 are for single variable regressions in which a current money supply va riable was regressed on a current asset variable. For examp le , one regression had Ml as the dependent variable and loans as the independent variable. The origina l data were seaso nally adjusted quarterly data for the period from the first quarter of 1950 t hrough the second quarter of 1976. The observations that entered the regressions were differences between the actual and trend grov.1h rates of the variab le s. Trend gro"'1h rates were determined b y regression analys is . For example, the trend grov.th rate of MI is defined as: T T A ~ In M1t = In M1t - In M1 t - 1 where In MI A is the natural logarithm of the actual level of MI a nd In MIT is the predicted value of the regression equation: ~ In M1 t = a + b In M1 t _ 1 + ut. This procedure allows th e trend growth ra te to change through time . In the case of M l. for examp le. the trend ~onthly Rev ew • l\iovernbe~ 1876 grew at an annual rate of 5.3 per cent, compared with 4.1 per cent in the recessionary period from the fourth quarter of 1973 through the first quarter of 1975. The recovery growth rate of loans was 1. 1 per cent, compared with the recession growth rate of 9. 1 per cent. While the growth rates of bank loans and M 1 are not closely related , there is some correlation between the two variables. An examination of Chart 2 suggests that a time lag may be grow th rate increases from 2.0 per cent in the tirst quarter of 1950 to 6.3 per cent in the second quarter of 1976. The se ries used in the regressions- th e difference between ac tual and tre nd growth rates - ha ve zero meam and have no t re nd . However. the series do contain cycles. as no atte mpt was made to re m ove se rial corre lation. It shou ld be noted . a lso. that the resid uals o f th e regressions in Tabl e 2 a re se riall y corre lated. In add ition to the regre sio ns in Table 2. reg ressions were run using growt h rat es ( ~ in) as observations. In most cases. the growth rate regressions produced hig her R2's than the reg ressions reported in Tab le 2. However. th e growth rates do. in most cases. co ntain trends which contribute to the higher R2· s. 7 Commercial Bank Loans and the Money Supply Table 2 R2 FOR REGRESSIONS OF VARIOUS MONEY SUPPLY MEASURES ON BANK LOANS, INVESTMENTS, AND BANK CREDIT Indep en de nt Varia bl es Dep en dent Varia b les T o t al Loan s Bus iness Loans Inv es tments M1 M2 M4 .10 .00 '. 09 .00 .00 .06 .4 2 .3 0 .10 Loans Pl us Inves tm ents .29 .45 .63 NOTE: M1 consists of demand deposits of commercial banks other than domestic interbank and U .S . Government less cash items in process of collection and Federal Reserve float ; foreign demand balances at Federal Reserve Banks ; and currency outside the Treasury , Federal Reserve Banks , and vaults of commercial banks . M2 c onsists of M1 plus sa vings deposits , time deposits open account , and time certificates of deposit other than negotiable CD 's of $100,000 of large weekly report ing banks . M4 consists of M2 plus large negotiable CD 's . involved in the relationship. Note, in particular, the tendency for a turnaround in the loan growth rate to lag a turnaround in the money growth rate. This tendency is especially evident during periods in which the growth rate of M 1 is accelerating after having been declining. Typically, these turning points in the growth rate of M 1 are near the peaks of business cycles. During the first part of the periods of accelerating Ml growth, the growth rate of bank loans typically either declines or does not change much. The growth rate of loans does not begin accelerating until some time after M 1's growth rate has undergone a reversal of direction. The recent behavior of total loans and M 1 illustrates the tendency for the turnaround in the loan growth rate to lag behind the turnaround in the money growth rate. As usual, the growth rate of M 1 began declining several quarters prior to the onset of the 1973- 75 recession, which began in the last 8 quarter of 1973. M 1's growth rate declined throughout the recession (surprisingly, since M 1's growth rate usually begins accelerating we ll before the end of recessions) and bottomed out in the first quarter of 1975, the recession 's final quarter. In the second quarter of 1975, the growth rate of money accelerated sharply , and grew much more rapidly in both the second quarter and the third quarter of 1975 than in any quarter during the recession. The loan growth rate, however, instead of accelerating during the second quarter of 1975 , continued to decline, with total loans falling during both the second and third quarters of 1975. Statistical ana lys is supports the impression ga in ed from Chart 2 that a time lag may be in vo lved in the relati nship between bank loan and MI . Th e analysi hows that movements in th e grow th rate of MI duri ng any quarter can explain about 29 per cent of th e movements in th e growt h rate of loans two quarters iater. 4 Th is moderately high correlation between current money supply movements and future movements in loans may be du e to a tendency for M 1 movements to lead movements in the general business cycle and for the cycle to coincide with or lead movements in bank loans. RELATIONSHIP BETWEEN BANK CREDIT AND THE MONEY SUPPLY As pointed out ea rlier. increases in the mo ney supply may potentia lly arise when banks add either to their loans or their investments. 4 The 29 per cent is the R2 from a regressi on invol ving time lags. The general form is: t).ln L ~ = t ( t). In M1 ~_ 2 ) where the superscript D denotes differe nces between a ctual and trend growth rat es. Simil a r multi va riate regress ions with up to four lagged MI vari a ble-, in crease the R2 to as high as 35 per cent. Regre ss ion s with loans instead of MI lagged -of th e form t). In M1 t = t (~ In Lt _ 2 ) have R2·s of less than . 10 per cent. rcgarclle~s of the number o f lagged loan variable~ invol ved (up to four lagged va ria b les). Federal Reserve Bank of Kansas City Commercial Bank Loans and the Money Supply Chart 3 GROWTH RATES OF BANK CREDIT AND THE MONEY SUPPLY, M4 Quarterly Averages, 1952-76 Per Cent p t 20 15 10 5 -5 Ll-il..U...1..1.. 1952 1 56 1 58 1 60 1 62 Thus . one would ex pect the correlation between total bank earning assets-i.e., loans plus in vestments-and M 1 to be closer than the correlation betwee n any component of earning assets-such as loans or invest ments-and money. This expectatio n turns out to be correct. The R2 between the growth rates of earning assets during a ny quarter and the growth rate of M 1 in the same quarter is 29 per cent. compared with 10 per cent between loans and M 1. and 10 per cent between investments and M 1. (See Table 2.) The correlation between bank credit and M2 also may be expected to be higher than the correlation between bank credi t and M 1. This is becau se increases in bank ea rning assets may pote ntially be used by the public to a ugment its tim e d eposits as well as its M 1 balances. Wh e th er Ml balances or tim e deposits (except la rge CD 's) a re augme nted , M2 will increase because time deposits ot her than large CD's are includ ed in M2 along with M 1. Also. because Monthly Review • November 1976 1 64 1 66 1 68 1 70 1 72 1 74 1 76 funds arising from increases in earning assets may be placed in large certificates of deposit rather than M2 balances. and because M4 is defined as M2 plus large CD's , the correlation between bank credit and M4 should exceed that between bank credit and M2. These expectations are confirmed by statistic a l ana lysis. Table 2 shows that movements in the growth rate of bank credit explain 45 per cent of the movements in M2, while the R2 between cred it and M4 is 63 per cent. 5 The high correlation between bank credit and M4 may be observed in Chart 3, which plots the growth 5 The R2's reported in Table 2 are for current period regressions. Regressions in volving time lags give somewhat different results. When M2 or M4 is regressed on past values of investments or bank credit, the R2' s are lower than reported in Table 2. Similarly, when investments or bank credit is regressed on past values of M2 or M4 , th e R2· s are lower than in Table 2. However , when loans are regressed on past values of M2 or M4 , the R2 •.,; are somewhat higher than reported in Table 2. This resu lt is similar to that reported for loans and Ml in the text. 9 Commercial Bank Loans and the Money Supply rates of the two variables for the 1952-76 period. Somewhat surprisingly, a fairly close relationship was found between commercial bank investments and M2. (See Table 2.) Movements in the growth rate of investments account for about 40 per cent of the movements in M2. Since the R2 is relatively high between investments and M2, and low between investments and Ml , investments and the time deposit component of M2 must be highly correlated. And. when the time deposit component of M2 is regressed on investments, an R2 of 45 per cent is produced. Evidently , during periods when banks are aggressively adding to their inve tment portfolios, the funds provided to the public are more likely to be used to add to tim e deposits rather than to M J balances. An alternative interpretation of the stat ist ical results is that banks are more likely to acquire investm ents when their time deposits increase than when their demand deposits increase. SUMMA RY Many observers of financial developments argue that the behavior of the narrowly defined money supply, M 1, i closely related to the behavior of commercial bank loans. According to this argument, MI tends to increase when loa ns are increasing and to decline when loans are declining. Some observers appear to argue further that MI and loans are necessarily related because the public cannot increase its MI balances unless banks make loans. This article points out that while increases in bank loans may be and frequently are used by the public to augment Ml balances, bank lending activity is only one of several sources of potential increase in Ml. For example , when banks increase their investments, a potential increase in M 1 occurs because the increase in 10 investments provides the public with funds that may be used to add to money balances. For this reason, Ml may increase even though bank loans do not , and Ml may remain unchanged while loans increase. Thus, the behavior of M 1 is not necessarily related to bank lending activity. Nevertheless, several factors may produce a close correlation between the two variables. These factors include compensating balance requirement , demand factors affecting both loans and MI , and the conduct of monetary policy. An examination of th e historical relationship between loans a nd M 1 ove r the past 25 years uncovered some broad simil a rities between the behavior of the two variables. Both variables have tre nded upward durin g the J952- 76 period and comme rcia l bank lending activity ha s been an important so urce of potential increase in Ml. Furthermore, the upward trends for both MI and loans have been interrupted from time to time and in both cases these interruptions have bee n associated with the business cycle. Beyond these broad similarities, however, the behavior of bank loans and Mt has not been closel y correlated during the 1952-76 period. The statistical correlation is quite low between movements in the growth rate of MI during any period and movements in the growth rate of loan s in that sa me peri od . The correlation is so mewhat high er when th e exis tence of a lagged re lati ons hip is r ecog ni zed. Even so, the correlation is only mod erately high. Furthermore, statistical analysis shows that the correlation is also quite low between growth rates of bank loans and the broader measures of the money supply, such as M2 and M4. As would be expected, on the other hand, a somewhat higher correlation was found between bank credit, which includes both loans and investments , and various measures of the money supply. The correlation between bank credit and M4 is quite high . Federal Reserve Bank of Kansas City The Concept of Private Income By Robert D. Auerbach and Jack L. Rutner T here are a number of frequently used measures of aggregate income or output for the U.S. economy. One of these is gross national product (GNP), which is the one most co mmonly cited. Other less frequently cited aggregates are net national product (NNP) and national income (NI). None of these concepts of income, however, measure the income of the private sector alone since they also include the income of the government sector. There are occasions, though, when a definition of income is needed that excludes the government sector. Such instances arise when one wishes to measure, and perhaps forecast, the demand for goods and services by the private sector , such as money balances, durable goods, vacations, etc. 1 Indeed, the appropriateness of a definition of income that excludes the government sector was originally suggested by Milton Friedman and David Meiselman in connection with the estimation of the demand for money by individuals in the private sector. 2 For estimating such a demand function, it may 1 Of course , for items where a composite demand by the private ector and the government exists, account must also be taken of the government sector. 2 Friedman and Meiselman . "Reply to Ando and Modigliani and to Prano and Mayer," American Economic Review, September 1965 . Set also Robert D. Auerbach , "The In come Effects of the Government Deficit ," unpubli shed Ph.D. dissertat ion , University of Chicago, 1969 . Monthly Review • November 1976 be appropriate to have a measure of income of only the private sector. The purpose of this article is to describe and develop a concept of income of the private sector. The concept of private income outlined here is the total income of individual s including trans fers and net of taxes plus undistributed corporate profits. The first section of the article indicates how this concept can be derived from the GNP accounts. The second section provides an analysis of how private income has behaved in the post-World War II period relative to other commonly used measures of aggregate income. GNP AND ITS RELATION TO PRIVATE INCOME The GNP Accounts To describe the concept of private income, it is useful to begin with an examination of the GNP accounts. For reference, Table 1 presents a summary of the U.S. national income and product accounts for 1975 as published recently by the U.S. Department of Commerce in the Survey of Current Business. The basic purpose of the GNP accounts is to provide a continuing measure of the gross expenditures on final goods and services of the U.S. economy. The underlying rationale of the accounts is embedded in an accounting relationship in which payments equal either receipts or income from the sale of final goods and services. Reflecting this accounting equality, GNP can be computed in one of two 11 The Concept of Pr ivate Income Table 1 SUMMARY NATIONAL INCOME AND PRODUCT ACCOUNTS, 1975 (In Billions of Dollars) Co mp ensa t ion of employees W ages and salaries D isbursem en ts Wage accr u als less d is bu r sem ents Suppl em en ts to wag es an d sa la r ies Emp loy er con tr ibu tions tor soci al i nsurance Oth er labor income 928.8 806 .7 806 .7 .0 122. 1 59 .7 62 .5 Pr op riet ors' income w ith inventory val uat ion and ca pit al cons umpt ion ad j ustm ents 90 .2 Rental inco m e of p ersons with ca pital cons umpt i on adjustm en t 22.4 Personal consumption ex p en ditures 973. 2 Du ra b le goods Nondu ra b le goods Se rvices 131. 7 409 .-1 432.4 183 .7 Gross p riva t e do m estic invest m ent 198.3 147 .1 52 .0 95 .1 51 .2 F ixed in vestment N onre si d enti al St r uctures Pro du ce rs ' durabl e equipment R esidential 14 .6 Change i 11 bu sin ess invento ri es Co rp ora t e profit s w i th in ventory va luation an d capit al consumption adju st m nt s Profi t s before tax Profit s t ax liabili ty Pro f i t s a ft er tax D ivi d en d s Undi str ibut ed p rofits I nventory valuation adjust m ent Capita l cons u m p tion adju st m ent Net interest NATIONAL INCOME 9 1.6 114 .5 49 .2 65 .3 32 .1 33 .2 - 11.4 - 11.5 74 .6 N t export s of good s and servi ces 20 .5 Export s Imp ort s 148.1 127.6 339.0 Gov ern m ent p urc h ases of goods and serv ic es F ed eral Natio nal d e f ense Nondefense State an d local 124.4 84 .3 40 .1 214 .5 1,207 .6 Add endum : Bus ine ss transfer paym ents Ind irect bus iness t ax and non t ax l iab ilit y Less : Subs id ies less cur rent su rp lu s of govern m ent enterpris es Statistical discr ep ancy CHARGES AGAINST NET NATIONAL PRODUCT Cap i t a l cons umpti on allowances w i th capit al cons umpt ion adju stm ent CHARGES AGAINST GROSS NATIONAL PRODUCT 6 .3 138 .7 2. 0 4.4 20.5 11 .9 0.9 7.6 1,355.0 16 1.4 1,516 .3 ways: by summing the values of expenditures of all final products sold , as illustrated on the right side of Table 1; or by summing the receipts or gross income from these sales , as illustrated on the left side of Table 1. As illustrated by Table 1, national income is a narrower measure than GNP. The difference arises because GNP as derived on the right side of the accounts is essentially a consolidated sales figure of all final goods and services sold, 12 Net ex port s of good s an d services N et foreign inv est m ent Perso nal transfers t o for eign ers Interes t and transfer s p ai d b y f oreigners less ca pita l grant s recei ved by U .S. G ROSS NATIONAL PR O DU CT 1,5 16 .3 while NI equals only the value of final goods and services produced in the current period . This difference between the value of goods sold and goods produced , which is shown in Table 1 as adjustments to NI , does not represent income earned by factors of production for producing the current period's output. One such adjustment , for example, is called depreciation and shown in Table 1 as capital consumption allowances and represents the Federal Reserve Bank of Kansas City The Concept of Private Income 3 using up of output from a previous period. In order to avoid double counting output from two different periods, depreciation is subtracted from GNP to obtain NI. When this adjustment and others 4 are made to NI on the left side of the accounts, the total exactly balances with the product side of the acco unts. In other words, total charges against GN P-including factor costs and other costs-exactly equal GNP derived via the right side of the accounts. The concept of net national product is obtained by subtracting one of these items of adjustment, i.e ., depreci ation cha rges, fro m GNP. The Derivation of Private Income As indicated earlier, the concept of private income d sc ribed here is the income of persons incl udin g transfers a nd net of taxes plus undistributed corporate in come. Income acc ru ing to the government sector , both federal and state and local governments, is explicitly excluded in deriving private income. Since the totals of the income and product sides of the GNP accou nts are exactly equal after making certain adjustments, private income can be obtained by subtracting government income from either side of the accou nts. Table 2 shows the relationship of private income to GNP in 1975, together with th e various categories that comprise private income. Many of the categories are identical to those in Table 1. They have simply been rearranged to reflect the se paration of the 3 Output from a previous period which is not consumed in that period is an additio n to the stock of wealth . 4 Other adjustments a re made for indi rect business taxes, business transfer payments, current surpl us of govern ment enterprises less subsidies, a nd fo r an item called statistica l di screpa ncy . Th e prese nce of the statisti cal discrepancy a ri ses beca use of th e different meth od s of accounting used to arrive a t the right and le ft si des of the national income and prod uct acco unt s. It is use d as a balancing item to reconcile both sides. As a technical ;natter , NNP (GNP less deprecia tio n ) is la rger than national income because it includes indirect business taxes (sales and pro perty taxes) . Monthly Review• November 1976 accounts into personal, corporate , and government income. Other categories , however , as well as the values for certain categories. do not appear in the summary accounts of Table 1. Rather , they are contained in supplemental sector accounts, which are published in considerable detail along with the GNP accounts by the U.S. Department of Commerce. Private income derived from the left side of the GNP accounts consists of disposable personal income plus corporate retained income. 5 As shown in Table 2, d is posable personal income amm:~ts to personal income minus personal tax and nontax payments. Persona l income includes most of the item show n previously in Table 1, such as compensation of employees, proprietors' income, rental income , dividends , and net interest. 6 Also included are transfer payments from government and businesses , which are considered income received by persons even though they are not related to the sale of goods and services. Contributions to social insurance are also deducted under the assumption they are government income and not savings to the private sector. Finally, personal taxes and nontax payments are deducted to arrive at disposable income. Tax payments are deducted because they represent income accruing to the government sector. Corporate retained income consists of retained corporate profits whether in the form of cash or invested in inventory or equipment. Corporate dividends, as indicated above , are 5 The statistical discrepancy must be added or subtracted from one side of the accounts in order that both bal a nce , so that private income could also arbitrarily include the statistical discrepancy. Without some additional evidence as to th e so urce of the discrepancy. there is no basis for making a deci sion as to which side of the accounts the discrepancy is to be assigned. 6 The value of net interest received by person s shown in Table 2 is different from the net interest received by all factors of production , which as shown in Table I is corrected for inter-sector transfers. 13 The Concept of Private Income Table 2 RELATION OF PRIVATE INCOME TO GNP, 1975 (In Billions of Dollars) 1,0 58 .1 D isposable p erso na l 111come Personal 111come Compensation of employees Proprietor s· rncome R ental income Dividends Ne t interest received by p ersons Tr ansfers to p ersons B y b usiness 6.3 B y government 168.9 L ess empl oyee and employer con tr ibutions to social insurance L ess wuge accruals plu s d1sbu1sements 1,226 .9 928 .8 90 .2 22.4 32 .1 87.9 185.2 168.8 10.3 33.2 11 .4 - 11 .5 0.9 64.4 PRIVATE I NCOME 339.0 7.6 282 .2 1,072.8 Government n t incom e 282 2 Capital consumption allowances 161.4 1,0 72 .8 282 .2 Government n et i ncome 168 .8 49.2 138.7 109.7 0.0 168 .9 13.3 2.0 1,355 .0 Cap i t al consumption allowances 161.4 G R O SS NA T IONA L PR ODUC T 1,516 .3 shown as part · of personal income. Corporate tax payments are not treated as part of corporate income because they represent income received by the government sector. Thus, adding the value of disposable personal 14 12 0 4.4 PRIVA T E INCOME NET NA TIO NAL PRO DUCT N et foreign investment Government purchases of goods and services Interest and transfers paid by gove1 nment to for eigners less ca µ1t ul giant s rPce1ved by U .S. L es s g vP1nmPnt rw1 income Statistical d iscre pan cy Personal tax and nontax paym ents Corporat e t ax pa yments Indirect business taxes Contributions to social 111surance Wage accruals less disbursem ent L ess governm ent tran sf ers L ess interest p aid L ess subsidie s plus su rplu s of enter p rises 22.3 Deficit of government R ta1n d cor p o1ate 1ncom Undistribut ed corpo ra te profit s Inventory valuation adj ustment Ca pita l consumption adjustment N et private domestic tnvestment Personal transfers to foreigners 109.7 0 .0 Less p e1so nal tdx and nontax p aymen t s 973 .2 Personal consumpt ion expe nd itures GROSS NAT IONA L PRODU CT 1,516 .3 income to corporate retained income (plus a minor statistical discrepancy item) yields the concept of private income. Government net income , which consists of tax receipts less transfers , is shown on the left Federal Reserve Bank of Kdnsas C1•y The Concept of Private Inc me side of Table 2. The components of government income also are shown. 7 When government net income is added to private income , the total equals net national product; and when capital consumption allowances (depreciation) are ad ded to NNP, the total equals GNP. The concept of private income can also be derived from the right side of the GNP accounts . Conceptually, the methodology is the same , i.e., to subtract from the accounts all incom e accruing to the government sector. Accordingly , private income includes personal consumption expenditures, net private domestic and foreign investments, personal transfers to foreigne rs , and the deficit of the government sec tor. The d eficit of the governme nt is ob tained by s ubtrac tin g government net income from gove rnm ent expe nditures. 8 The resu lting va lu e of private income on the expend iture side of the acco unts is exactly equal to the same concept on the income side . By the same token, when government net income and capital consumption allowances (de preciation) 9 are added to private income, the total equals GNP. It shou ld be mentioned that the computation of private income from the right side of Table 2 treats the gove rnm ent sector in a manner 7 Spcci ti ca ll y. a ll perso nal. co rpora te. a nd indirect tax rece ipt~ a re li sted as part o r govern me nt in co me a lo ng with government rece ipt s a ris in g from co n tributi o ns to social in surance . De ducted fr o m these rece ipt s are governm ent domest ic transfe r pay me nts and int e res t p a id b y the governme nt. In terest paid b y th e government e xcludes intere st paid to fo reigners. th e latter being tre a ted ex plicitly o n the expe nditure side of the acco unts. Finally. th e net in come figure for govern m e nt is actually d o mestic net in c:o me since it e xcludes ca pital g rant s received b y the nit ed States . 8 Government ex penditures include expen di t ures for goods and serv ices a nd net govern ment outlays to th e foreig n sectors . The va lu e o r s u ch o utl ays to foreig ners whe n added to net fore ign in ves tmen t a nd perso na l transfers to foreigners equa ls the net e xport o f goods a nd se rvices sh own in Table I. 9 Since capi ta l cons umpti o n all owances are exc luded from pri,·at e in co m e. th e item net pri va te domestic inves tment is used in Table 2 rathe r than gross pri vate domestic in ves tm e nt as sh o wn in Tab le I . Monthly Rev ev, • NovefTlber 1975 analogous to the foreign sector. That is , just as imports are subtracted from exports, government receipts are subtracted from government expenditures. Viewing the government sector this way reveals why deficit spending by government is treated as a stimulant to private income during recessionary periods. In such periods, there is an attempt to stimulate income by increasing exports or by restricting the importation of goods produced abroad. Similarly, there is an attempt to increase sales to the government or reduce government receipts-namely , taxes. 10 A further point is that the concept of private income derived here, because it excl udes the government sector, is not a measure of the tota l ex penditures on final goods and service in th e economy. Indeed , the value of private income in 1975, which is estimated at $1,072.8 billion, wo uld understate total expenditures on final goods and services as measured by NNP and GNP, which are valued at $1,355.0 billion and $1,516.3 billion, respectively. Private income also falls short of national income, valued at $1,207.6 billion in 1975. These aggregate measures, however, often have been used to help measure the demand for goods and services by the private sector. The following section , therefore, examines how these aggregate measures have behaved relative to private income. THE BEHAVIOR OF PRIVATE INCOME To examine the behavior of private income , va lues for the concept were calculated for the post-World War II period 1946- 75. Table 3 contains these values calculated on an annual basis along with values for the major 10 One ca nn o t conc lude. h owever . the precise m ag nitude or sign changes the govern m en t deficit or net ex po rt s will h3\·e o n privat e inco me by looking only at these simpl e acco unting id en titie s. inasmuch as some other compon ents of the accounting identity must also change. and this may ha ve o ffsetting effects. 15 The Concept of Private Income Table 3 PRIVATE INCOME, ANNUALLY 1946-75 (In Billions of Dollars) Sources of Income Year D isposable Personal Income (Ad j usted ) Net Corporate Re taine d Earnings Final Sales Sta tist ical D iscrepanc y Co nsumpt ion Tota l Private Income Net Dom est ic Investments Net Fo re ign Investments Defici t of Government Cu rrent Do llars As a Per Cent of NNP In 1972 Pr ices 1946 1947 1948 1949 1950 157 .2 166.6 185.3 184.8 202 .8 1.9 4.6 9.7 9.6 6.9 0. 7 1.8 - 1.2 1.0 2.0 143 .8 161 .7 174.7 178.1 192.0 16.9 16.8 25 .6 13 .3 29.9 4.6 9.0 2.0 0 .6 - 2.1 - 5.4 - 14.4 - 8.4 3.4 - 8 .0 159 .8 173.0 193 .8 195.4 211.7 81 .7 80 .3 81.2 82 .8 80 .7 334 .8 327 .6 347 .3 351 .0 372.8 1951 1952 1953 1954 1955 221 .9 233 .1 246 .7 251 .4 268 .6 7.0 7.5 6 .4 7.9 12.3 4 .0 2. 7 3.3 3.0 2.5 207 .1 217 .1 229.7 235 .8 253 .7 31 .6 22.5 21 .7 19.6 33 .1 0.3 - 0 .2 - 1.9 - 0.3 0.3 - 6 .1 3 .8 6 .9 7. 1 3.1 232 .9 24 3 .3 256.4 262 .3 283.4 77.0 76 .6 76 .7 78 .7 77 .9 385 .2 393 .3 406 .6 412 .4 441 .0 1956 1957 1958 1959 1960 285 . 7 301 .0 311 .2 329.6 342 .0 9.8 9.2 7.1 12.4 11 .0 - 0 .8 - 0 .2 1. 7 - 0 .2 - 0.7 266 .0 280.4 289.5 310.8 324 .9 32 .1 27 .2 17.8 31 .5 28 .7 1.8 3 .6 0.1 - 2.0 1.7 - 5.2 0 .9 12.6 1.6 - 3.1 294 .7 310.4 320 .0 341.8 352.3 77 .2 77.4 79 .1 77 .6 76 .8 449 .7 458.5 462.9 485 .1 491 .2 1961 1962 1963 1964 1965 355 .2 375 .6 393.4 426 .5 460.5 10.8 16.5 17.9 21 . 7 27.1 1.6 4 .0 3.7 2.2 0.9 335 .0 355.2 374 .6 400.4 430 .2 25.2 34 .7 38 .0 42.0 54 .5 3.0 2.4 3 .2 5.7 4.3 4.3 3.8 0. 7 2.3 0 .5 367.6 396 .1 415.0 450.4 488 .5 77 .5 77.2 76 .5 77 .5 77.5 506 .8 538.1 555 .3 595 .0 633 .2 1966 1967 1968 1969 1970 497 .8 531 .3 574 .0 614 .8 669.4 29.4 26.7 24 .5 19 .2 10 .5 3.2 1. 7 - 0.6 - 3.3 - 2.1 446 .8 490.4 535 .9 579 .7 618 .8 62.8 53.8 57 . 7 63 .7 50 .0 1.6 1.2 - 1.4 - 2.0 0 .5 1.3 14 .2 5.5 - 10.7 8 .5 530.4 559 .7 597.9 630 .7 677 .8 76.7 76.7 75.2 73.9 76 .0 668 .3 688.3 706 .7 712.6 732 .7 1971 1972 1973 1974 1975 725 .5 782.4 880 .2 959 .7 1,057.2 16.9 25 .9 22.6 1. 7 10.3 1.3 1. 7 2.6 6 .6 4.4 668 .2 733 .0 809 .9 887.5 973 .2 61 .2 82.9 102.4 77.3 22.3 - 3 .2 - 9.0 - 0.6 - 3.0 12 .0 17 .6 3 .1 - 6 .3 6 .3 64 .4 743 . 7 8 10.0 905.4 968.0 1,071 .9 77.1 76 .0 76. 1 76 .0 79 .5 769.4 81 0.0 861.0 829.8 846 .8 - - NOTE: Components may not sum to total due to rounding . Also , disposable personal income (adjusted) is adjusted to exc lu de interest paid by consumers and personal transfers to foreigners . components of private income. 11 As is apparent from Table 3, private income has risen 11 The values of private income in Table 3 exclu de personal transfers to foreigners . Hence, the value of private income shown for 1975 ($1 ,071.9 billion) is $0.9 billion less than the va lu e shown in Table 2. This treatment was suggested in Friedman a nd Meiselman , "Reply to Ando and Modiglia ni and to Pran o and Mayer. " 16 throughout the period although it has a somewhat different pattern of growth than other measures of income. Prior to the Korean war private income averaged a bout 81 per cent of NNP. Since that time, however , reflecting the growth of the government sector, private income has averaged only about 77 per cent of NNP, although in 1975 private income jumped Federal Reserve Bank of Kansas City The Concept of Private l'lcome Chart 1 REAL AND NOMINAL PRIVATE INCOME (Quarterly, 1947-75) Bil li ons of Doi lars 1150 950 750 550 350 I50 1--....1....---1.._1...-...1.........i...----1_.1....-_.____,,__L-..J..-....L...---1._..._...1....---L----JL--..J..----'----'-,--..___--'-__.__':---_.__---':-___,__':---........... 1947 1 49 1 51 1 '53 55 1 57 1 59 1 61 1 63 1 65 1 67 1 69 1 71 1 73 1 75 SOURCE: See text . to 79 per cent of NNP. For the entire postwar period, private income has varied from a high of 83 per cent to a low of 73 per cent of NNP. The rising va lu es of several meas ures of aggregate expen ditures on final goods and services in recent yea rs have partly reflected inflationary facto rs. The impact of rising prices on these variables has led increasingly to the use of "real" values, i.e. , values deflated to account for an increase in prices. The last column in Table 3, therefore , also provides a nnu a l values for private income calculated in constant do ll ars. 12 12 The met hodo logy e mpl oye d to ar r ive at the rea l va lu es is described in the appendix . \'1onthly P v ew • Nover1ber 0 16 The distorting effect that changes in the price level have had on private income may be seen in Chart 1. The level of private income is expressed in both current and real terms and is displayed on a quarterly basis for the 1947-75 period. As is clearly shown, the rapid price increases since the late 1960's have led to a sharp rise in private income expressed in current dollars. In contrast, gains in real private income have been relatively more moderate. Since price changes have tended to obscure the underlying behavior of real private income , real values are subsequently employed in comparing private income with other income measures. Chart 2 shows the behavior of real private income relative to three other measures of real 17 The Concept of Private Income Chart 2 FOUR MEASURES OF REAL INCOME (Quarterly, 1947-75) Gross Notional Product \ 1100 900 700 500 300 1947 1 49 1 51 1 53 1 55 1 57 1 59 1 61 1 63 1 65 1 67 1 69 1 71 1 73 1 75 SOURCE: Real GNP, NNP, and NI : Survey of Current Business, U .S . Department of Commerce , Washington , D.C ., January and July 1976 ; Real Private Income : see text. NOTE: The shaded areas in this chart pertain to recessions , with the p 's and t 's at the top of the chart referring to peaks and troughs of business cycles . aggregate income: GNP, NNP , and NI. As is readily apparent, there is a similar upward trend in each of the four measures during the entire 1947-75 period. Closer inspection, however, indicates some dissimilarities between real private income and the other series. During recessions, real private income often tends to remain more stable , or to fall by less, than the other three measures. In the 1953-54 recession , for instance, real GNP, NNP, and NI all declined until the second quarter of 1954, whereas private income reached its low in the last quarter of 1953. Moreover, the decline in 18 private income was much less than in the other three measures. In the 1957-58 recession, private income again declined moderately, while the declines in the other measures were much larger. In the 1960-61 recession and in the 1969-70 recession , private income actually rose somewhat compared to declines in the other measures. Only in the most recent and ve ry severe 1974-75 recession did private income decline sharply along with the other three measures of income. There were two other periods when real private income behaved atypically relative to Federal Reserve Bank of Kansas City The Concept of Private Income Table 4 RATES OF CHANGE AND VARIABILITY OF FOUR MEASURES OF REAL INCOME, 1947-75 Annual Quarterly Standard Standard Inco me Measures Mean Dev iation Mean Dev iation Gross Nati onal Product N et Nati onal Product 3 .38 3 .29 3 .35 3.44 3.20 3.50 3.94 3.16 3 .36 3 .26 3 .27 3.43 4 .58 5 .02 5.46 5.82 Nationa l Income Pr iva t e In come CORRELATION OF RATES OF CHANGE OF FOUR MEASURES OF REAL INCOME, 1947-75 Annu al N et Nati ona l Prod uc t ational Income Pri vate Income GNP NNP .999 .982 .842 .981 .845 Quart er ly NI GNP NNP NI .844 .996 .956 .680 .967 .681 .623 the other measures of real income . The first was during the Korean war (1951-52) and the second was in the midst of the Vietnam conflict (1 968-69). During both of these periods, private income declined while the other measures of income increased. The decline of private income in these periods clearly reflected the inc rease in the government's share of income. In a sense, the increase in the government's sha re of income during these periods of military conflict could be viewed as if the private sector had purc hased more government services in the form of defense. Further evidence on the behavior of real private income relative to the other three income measures is contained in Table 4. As shown in the top portion of the table, the ave rage rates of change of all four measures were very sim ilar during the 1947- 75 period, whether using annual or quarterly data. On the other hand, dissimilarities occur in the Monthly Review • November 1976 variability of the four income series as shown by their respective standard deviations. Using annual data, for example, private income is found to be the least variable; i.e., it only varied within ±3.16 percentage points of its average rate of change. Using quarterly data, though, private income is found to be the most variable of the four income series. The degree of correlation between the rates of changes in each of the four income measures is also shown in Table 4 . On an annual basis, the correlation between private income and the other measures is quite high, although not as high as between each of the other three measures. On a quarterly basis, the differences are quite pronounced. GNP, NNP , and NI are all highly correlated with each other on a quarter ly basis as they are on an annual basis. Private income, though, is much less correlated with GNP, NNP, and NI using quarterly data . These correlations suggest that for quarterly data at least 32 per cent of the deviation in real income is uncorrelated with deviations in the other measures of income; and, hence. these other measures may be inappropriate proxies for private income. SUM M AR Y This article has described a concept of private income and its accounting as well as statistical re lat ionships to other meas ures of income. Essentially, the concept includes income of individuals wit h transfers net of taxes plus corporate retained earnings. Income of the government sector, both federal and state and local governments, is explicitly excluded from private income. Compared to the other measures of income , real private income in most recessions tended to be more stable, or to fall by less. than the other three measures of real income. Moreover. during the entire 1947-75 period, the rates of change of real private income. although similar in average magnitude to the other measures of income. were found to be more variable when using 19 ~.,e Cnnr..m• of Private Income quarterly data than the other income measures . In addition, the correlation between real private income and each of the other three meas ures was also not particularly high. Based on the national income accounts, it a ppears that the concept of private income desc ribed here is a more precise measure of the in come of the private sector than GNP , NNP, or NI . In addition, private income is not as highly correlated with these other measures of income a nd a ppears to follow a somewhat different cyclical pattern than they do . Hence , researchers interested in using a concept of income to analyze the behavior of the private sector may well find the concept of private in co me described here superior to these other meas ures of income . APP NDIX The Computati on of Real Val ues of Private Income The computation of private income in constant purchasing power wa s performed by applying price deflators to the different components of private income . This approach posed som ewhat of a problem because not all of the appropriate price deflators for the components of private income exist . In those instances , the choice of the component deflators was m·ade on a basis that was deemed to be the most appropriate and reasonable . The following list is a summary of the methods used to compute the deflators for the major components of private income . The source of all data used is the Survey of Current Business published by the U.S . Department of Commerce. Components of Real Private Incom e Method s of Computation 1. Real consumption expenditures . . . ...... . From published data 2. Real net domestic investment . ... .. . ... . Real gross domestic investment less real capital consumption allowance 3 . Real net foreign investment . .. . . .. .. . . Nominal net foreign investment/ implicit deflater of imports 20 4. Real deficit .. ..... ... . (1) + (2) + (3) (5) - (6) - (7) 5. Real disposable personal income (adjusted) . . . . . . . ... . . Nominal disposab le personal income (adjusted)/ implicit deflater of disposable persona l income 6 . Real ret ained corporate earnings . . ..... . . . .. . Nominal retained corporate earnings/ implicit deflator of business output 7. Real res id ual . . .. . . . . . Fro m published data The f irst fou r items pertain to the components of real private income ap pearing on t he expend iture side of the accounts and the latter three items pertain to the income s ide . The price deflators for some items are taken d irectly from publ ished data an d for others the methods of computation are rather stra ightforward . In Item 3, for instan ce , no m inal net foreign investment was divided by the impl icit deflator of imports . A deflator for exports was also tried but it gave about the same results as the import deflator. In Item 5, the deflator on disposable personal income appeared most reasonable inasmuch as disposable personal income (adjusted) differs only slightly from total disposable personal income. The computation that presented the most diff iculty was for the government deficit , Item 4. An attempt was orig inally made to deflate the defic it by the price at which the government purchases goods and services . This attempt resulted in some large statistical errors so an alternative method was chosen . The alternative method measures the deficit as the difference between the components taken from the income side , Items 5, 6 , and 7 , and the componentsexcluding the deficit-taken from the expenditure side , Items 1, 2 , and 3. Among the components from the income side is a residual , Item 7, which is analogous to the statistical discrepancy item . It should be pointed out that this ind irect method of computing the real deficit, rather than some alternate direct method , will tend to cause any methodological errors present in other parts of private income to be contained in the real deficit. The deflater for private income , as it turns out , is nearly identical to the deflator on consumption because private income is dominated by the consumpt ion component . Federal Reserve Bank of Kansas City
@FedFRASER