The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Vol. 73, No. 5 September/October 1991 3 T h e U .S . B a la n c e S h e e t: W h a t Is It a n d W h a t D o e s It T e l l Us? 19 D i v i s i a M o n e t a r y S e r v ic e s I n d e x e s f o r S w itz e r la n d : A r e T h e y U s e fu l f o r M o n e t a r y T a rg e tin g ? 3 4 T h e U .S . C u r r e n c y S y s te m : A H is to ric a l P e rs p e c tiv e THE FEDERAL A RESERVE J Z \ l \NKof M TSTA O VIS 1 F e d e r a l R e s e r v e B a n k o f St. L o u is R e v ie w September/October 1991 In T h is Issu e . . . Business accounting relies greatly on the balance sheet as a tool for analyzing a firm ’s financial health. Similar practices do not prevail in na tional econom ic accounting. Econom ic analysts instead rely mainly on the GNP accounts as a m easure of national econom ic perform ance. In the first article in this Review, "T h e U.S. Balance Sheet: W hat Is It and W hat Does It Tell Us?” Keith M. Carlson provides an overview of the U.S. balance sheet. He explains its basic stru ctu re and gives some examples of its usefulness in examining econom ic questions. One of the most im portant uses of the balance sheet, says Carlson, is to analyze the role of financial stru ctu re in the process o f econom ic grow th. The ch ief benefit of the U.S. balance sheet, as currently prepared, is that it forces the u ser to take a long-term perspective to detect changing econom ic trends. * * * In the second article in this Review, “Divisia M onetary Services Indexes for Switzerland: Are They Useful for M onetary Targeting?” Piyu Yue and Robert Fluri derive alternative m onetary aggregates fo r Switzerland that are based on econom ic aggregation theory, Divisia M l and Divisia M2. Noting that the historical relationship betw een the Swiss National Bank’s principal policy instrum ent, the m onetary base, and inflation ap pears to have broken down in the 1980s, they com pare the p erfo r m ance of their two new aggregates and simple-sum M l and M2 in ex plaining Swiss inflation. They find that Divisia and simple-sum M2 p er form somewhat b etter than the M l aggregates; however, th eir evidence suggests that simple-sum M2 cannot be controlled. They conclude that simple-sum M2 is unlikely to b e an adequate substitute for the m onetary base for policy purposes, but that the oth er aggregates are potentially w orthy of fu rth er investigation. * * * One of the principal roles of the Federal Reserve System is to provide the United States w ith currency. This role places the Fed at the cen ter of the U.S. cu rren cy system, a system that provides the foundation for our m onetary and financial systems. In the third article in this issue, “The U.S. Currency System: A Historical Perspective,” Steven Russell uses historical examples to define cu rren cy and cu rren cy systems, describe the various form s that curren cy can take and identify the distinctive features of the U.S. cu rren cy system. He also provides a short history of the U.S. curren cy system from its origins through the end of the Civil W ar. In recen t years, Russell notes, there has alternative curren cy systems. Many of the proposed are patterned after systems that of the m odern U.S. curren cy system, says been considerable interest in alternatives that have been existed in the past. D efenders the author, frequently SEPTEMBER/OCTOBER 1991 FEDERAL RESERVE 2 p ortray it as the outgrow th of a process of natural selection: as the U.S. econom y evolved, less efficient systems w ere rejected by the public in favor of m ore efficient ones. This portrayal interprets the very fact of the m odern system ’s existence as evidence of its superior efficiency. Russell’s account suggests that econom ic Darwinism has not played a dominant role in the development of the m odern U.S. cu rren cy system. Historically, U.S. cu rren cy systems seem to have risen and fallen for reasons th at had m ore to do w ith political crises—particularly, m ajor w ars—than with efficiency considerations. * * * BANK OF ST. LOUIS 3 Keith M. Carlson Keith M. Carlson is an assistant vice president at the Federal Reserve Bank of St. Louis. Thomas A. Pollmann provided research assistance. The U.S. Balance Sheet: What Is It and What Does It Tell Us? 1 5 USINESS ANNUAL REPORTS provide two basic accounting statem ents—a balance sheet, which is also term ed a statem ent of condition, and an incom e statem ent. A firm 's balance sheet lists the dollar value of its assets and liabilities as of a specific date. A firm ’s incom e statem ent lists its revenues and expenses (the difference being profit) for a year. Similar statem ents are prepared on a national level in the United States. Analogous to a firm ’s incom e statem ent, a na tion's production of goods and services for a year (as well as its spending and saving deci sions) are sum m arized in its gross national pro duct (GNP) accounts. Analogous to a firm ’s bal ance sheet, the U.S. balance sheet lists the dollar value of assets and liabilities fo r U.S. residents. The flows that are identified in the GNP ac counts and elsew here are linked to changes in the levels of assets and liabilities reported in this balance sheet. The GNP accounts receive the m ost attention simply because they focus on cu rren t produc tion and incom e, w hich in turn, affects and is affected by, the level of employment. These ac counts provide vital inform ation on the shortrun perform ance of the economy. On the other hand, the U.S. balance sheet generally receives little attention. This might be because it is in complete, including only nonhum an wealth (see shaded insert on page 5), and seem s to be m ore appropriate fo r long-term analysis. The purpose of this article is to provide an overview of the U.S. balance sheet. Its stru ctu re is explained and its usefulness is illustrated by examining trends in some individual balance sheet items. Fu rth er examples of its usefulness are given by examining balance sheet ratios such as the financial interrelations ratio, the net foreign balance ratio, the ratio of business capital to household capital and the relation of net w orth to inflation. TH E STANDARD U.S. BALANCE SHEET A balance sheet shows the position that a busi ness or household, or the econom y as a whole, has reached as a result of its past activity. It reflects flows o f real and financial activity plus any revaluations of stocks because of price changes. Table 1 summarizes the U.S. balance sheet fo r 1990 as currently prepared by the Board of Governors of the Federal Reserve System .1 G en era l D efin itio n s A balance sheet usually shows all assets and all liabilities, with the difference called net worth. 1Board of Governors (1991). This is called the C.9 release. SEPTEMBER/OCTOBER 1991 4 Table 1 Standard U.S. Balance Sheet: 1990 (billions of dollars)1 Beginning-ofyear value End-ofyear value Tangible assets Reproducible assets Residential structures Nonresidential structures Inventories Consumer durables Land at market value Net foreign assets Foreign assets owned by U.S. residents Minus: U.S. assets owned by foreigners U.S. monetary gold and SDRs $16,017.2 12,163.6 4,546.0 4,633.0 1,050.8 1,933.8 3,853.6 -670.1 852.5 1,522.6 21.0 $16,241.7 12,501.3 4,603.9 4,811.0 1,062.7 2,023.7 3,740.4 - 694.5 941.0 1,635.5 22.0 1.4% 2.8 1.3 3.8 1.1 4.6 -2 .9 NATIONAL NET ASSETS (consolidated) $15,368.1 $15,569.2 1.3% Private net worth (consolidated) Household net assets Nonfarm noncorporate business Farm business Nonfinancial corporate business Private financial institutions net assets U.S. holdings of foreign corporate stock2 Minus: Foreign holdings of U.S. corporate stock Public sector net assets State and local governments U.S. government3 Unallocated financial assets $18,104.5 9,899.3 1,949.2 720.4 3,778.4 1,757.2 95.8 257.4 -2,323.4 -99.9 -2,223.5 -251.3 $18,573.0 10,019.6 1,919.0 758.2 3,870.8 2,005.4 90.5 218.4 -2,616.5 -137.8 -2,478.7 -259.4 2.6% 1.2 -1 .5 5.2 2.4 14.1 -5 .5 -15.2 NATIONAL NET WORTH $15,368.2 $15,569.2 1.3% National Net Assets Percent change — 10.4 7.4 4.8 National Net Worth, by sector — — — — 'Source: Board of Governors of the Federal Reserve System (1991). 2AII corporate equities were subtracted in calculating net assets of households and financial institutions, includes sponsored credit agencies and monetary authority. T he U.S. balance sheet is unusual, however, because the largest category of assets, human wealth, is not included. National net w orth, as currently estimated, is all nonhum an wealth. For the national balance sheet, assets are divided into two types—tangible and financial. Tangible assets are econom ic goods that yield a stream of services in kind. Plant and equipment, housing, consum er durables and land are all ex amples of tangible assets. Financial assets are claims or rights to amounts of money now or in the future. For every financial asset owned by FEDERAL RESERVE BANK OF ST. LOUIS an econom ic unit there is a corresponding lia bility owed by another econom ic unit. Financial assets can be categorized as fixed claims or vari able claims; variable claims are called equities. C urrency and deposits, notes, bonds and m ort gages are all fixed claims. Corporate stock and the net w orth of noncorporate business are equities. For a closed economy, financial assets equal financial liabilities. For an open economy, national net assets, or national net w orth, is the total of all tangible assets, U.S. m onetary gold and SDRs (special drawing rights created and distributed by the International M onetary Fund) 5 National W ealth and the U.S. B alance Sheet A nation’s w ealth is defined as the value of all resources that contribute to the produc tion of goods and services. The broadest defi nition would include the goods and services resulting from both m arket and nonm arket activities. To b etter understand the process of w ealth accum ulation, econom ists have classified w ealth as hum an vs. nonhum an and tangible vs. intangible. The following classifi cation is typical:1 Tangible wealth Human Nonhuman Intangible wealth Human Nonhuman Tangible w ealth is m aterial in form , that is, it is "touchable.” Its quality, or productivity, depends greatly on intangible factors. Human tangible w ealth is essentially a population count, w hich might also take into account the average age of the population. Nonhuman tangible wealth is most fam iliar and consists of structures, durable goods, inventories and natural resources. This type of w ealth is some tim es classified as reproducible (man-made) and nonreproducible (natural). For all tangible form s of wealth, it is simply not possible to estim ate value w ithout con sidering the effect of intangible factors. The most fundam ental intangible embodied in human wealth is a person’s stock of knowl edge. Likewise, health conditions or the state of medical knowledge, have an im portant bearing on the value of society’s human wealth. Still another example of intangible human w ealth is the mobility of the popula tion which reflects socio-political factors af fecting the job m arket. T h ere are also other form s of intangible wealth that are difficult to classify as human or nonhum an. These in clude personal and national security, freedom , equity, privacy and a system o f property rights.2 Nonhuman tangible wealth is easy to under stand, although estimating its value is difficult 'Kendrick (1976). 2Juster (19/3), pp. 40-48. because of interaction with the state of tech nology w hich is an intangible factor. The value o f stru ctu res and equipm ent reflects the state o f productive know-how. Similarly, the value of physical resources like land, sub soil assets, forests and w ater depend on the state of technology. Because so many form s of w ealth are im m easurable (or at least appear to be), wealth estimation has generally been limited to its nonhum an com ponent. Furtherm ore, such estim ates have been generally restricted to m arket-oriented activities. The im portance of human w ealth to national w ealth has been established clearly, but th ere is little agree m ent on how to m easure it. Human w ealth is not traded in a m arket in the same sense as property, so its value must be determ ined in directly. Tw o fundam ental approaches have been used: (1) accum ulate past investm ents in education and training and allow fo r deprecia tion, or (2) estim ate the discounted value of future income. Tw o of the best-known studies of human wealth in the United States are by Kendrick (1976) and Jorgenson and Fraum eni (1989).3 T he table below com pares their estim ates of the distribution of wealth for 1969 (the latest y ear for w hich they both provide estimates). Distribution of Private National Wealth: 1969 JorgensonKendrick Human wealth Tangible Intangible Nonhuman wealth Tangible Intangible Total wealth per capita (1990 dollars) 69.2% 21.5 47.7 30.8 29.0 1.8 $8,572 Fraumeni 93.9% N.A. N.A. 6.1 N.A. N.A. $92,540 Kendrick estim ates that U.S. human wealth (including both tangible and intangible) is 3Also see Eisner (1989), whose estimates of human wealth as a proportion of the total lie between those of Kendrick and Jorgenson and Fraumeni. SEPTEMBER/OCTOBER 1991 6 about 2.3 tim es that of total nonhum an wealth. Jorgenson and Fraum eni, on the other hand, estim ate that hum an w ealth is 15.4 tim es that of nonhum an wealth. The methodologies un derlying the estim ates are very different with the Jorgenson-Fraum eni study including esti mates of the value of nonm arket activities such as household production and leisure time. Despite considerable advances in the under standing of w ealth, m easurem ent and estim a and net foreign assets—a balancing item (the difference betw een foreign assets owned by U.S. residents and U.S. assets owned by foreigners). T h e S ta n d a rd U.S. B a la n ce S h e e t in 1990 Table 1 shows 1990 beginning-of-year and end-of-year values in cu rren t dollars. The dif feren ce reflects saving and investm ent flows (net of depreciation) during the year plus revalu ations of existing assets.2 For the econom y as a whole, the bottom line is national net assets, or national net w orth. It is clear that national net assets are dominated by tangible assets.3 In fact, w ith the recen t rapid grow th in foreign ownership of U.S. assets, the value of tangible assets on U.S. soil has exceeded the value of the nation’s net assets (owned by U.S. residents) since 1983. On the net w orth side of the standard balance sheet, no liabilities are shown. National net w orth is defined as the sum of private net w orth, public sector net assets and unallocated financial assets (plus a balancing item for foreign and U.S. holdings of each oth er’s corporate stock). Private n et w orth is b rok en down fu rth er by sector. Household “n et assets” and private finan 2The Board's C.9 release also contains sector balance sheets for households, farm, nonfarm, noncorporate and corporate business, private financial institutions and the rest of the world. For definitions of sectors, see Board of Governors (1980). 3Tangible assets for federal, state and local governments are not included in the C.9 release. For further discussion of the government balance sheet, see Boskin, Robinson and Huber (1989). FEDERAL RESERVE BANK OF ST. LOUIS tion of many form s o f wealth are still in an early stage of development. As a result, a com plete accounting of U.S. wealth is not yet available on a regular basis; published na tional balance sheets are generally limited to the m arket value of nonhum an wealth. De pending on the estim ates, this portion rep re sents betw een 6 p ercen t and 31 percent of total national wealth. In this sense, the U.S. balance sheet is far from complete. cial institutions sector "n et assets” are presented along w ith the net w orth of each type of busi ness.4 This is useful fo r analysis, but one must keep in mind that households are the ultimate ow ners of businesses. Our focus below is on long-term trends in the United States, but one development in 1990 is w orthy of m ention. As shown in table 1, na tional net w orth grew only 1.3 percent or, given the increase in GNP prices (GNP implicit price deflator) of 3.9 percent, it declined 2.6 percent in real term s. This w eak perform ance is attribu t able, in part to, declines in real estate prices. T r e n d s in N ational N et W orth a n d GNP One possible use of balance sheet inform ation is to view national n et w orth as a m easure of m acroeconom ic perform ance over time. Does it yield inform ation that differs from that o f GNP? Many years ago, Raymond Goldsmith and Robert Lipsey concluded that "National balance sheets are not intended as a device to m easure econom ic grow th over time, but they are essential to study the relations betw een the financial super stru ctu re and the real in frastru ctu re, w hich constitute an im portant aspect o f econom ic grow th.’’5 W ith 30 years of new data, does this conclusion still hold? 4Equities and pension fund reserves are subtracted from household net worth to obtain net assets. Private financial institutions’ net assets are obtained by subtracting cor porate equities and adding pension fund reserves to their net worth. 5Goldsmith and Lipsey (1963), p. 25. Also, see Goldsmith (1985, 1982, 1969, 1966). 7 Figure 1 GNP and National Net Worth Ratio scale Billions of 1982 dollars Ratio scale Billions of 1982 dollars 2500 2000 1500 1000 1950 1955 1960 1965 Figure 1 shows real national net w orth (NNW) and real GNP from 1948 to 1990.6 Over the full period, the grow th rates are the same—3.2 p er cent. Real NNW is a relatively sm ooth series, while real GNP displays considerable volatility. Real NNW shows little cyclical movement, in dicating that tangible assets are valued over a long horizon, even if they are utilized m ore or less intensively during the business cycle. Table 2 sum m arizes and com pares GNP and NNW during the 1948-90 period. The referen ce periods w ere chosen to conform with different inflation experiences. The 1948-64 period was one of relatively low inflation, averaging 2.2 percent per year. From 1964 to 1981, inflation accelerated; it was 3.1 percent in 1965 and 9.4 percent in 1980. The 1981-90 period is more difficult to define—disinflation from 1981 to 1985 and then m oderately accelerating inflation over the past five years. 1970 1975 1980 1985 1990 lar rates in both nom inal and real term s. Periods of accelerating or decelerating inflation, how ever, produced differing grow th rates fo r the two m easures. From 1964 to 1981, both nom inal and real NNW grew faster than GNP. During the 1980s, the opposite occurred; NNW grow th slowed relative to that of GNP in both nominal and real term s. Based on the post W orld W ar II experience, NNW shows the same long-term trends as GNP. W hile the grow th rates of NNW and GNP can deviate fo r several years, fo r purposes of long term analysis they appear to give the same answ ers. Goldsmith's and Lipsey’s conclusion from almost 30 years ago appears to be confirm ed. AN ALTERNATIVE VERSION OF TH E U.S. BALANCE SHEET During the period of relatively low inflation from 1948 to 1964, GNP and NNW grew at simi Since GNP accounts are limited to data on the cu rren t production of goods and services and generally omit financial transactions, they are of 6The C.9 release also includes data for 1945-47, but these years are omitted because of distortions caused by the transition to peace from World War II. Real NNW was calculated by using the Department of Commerce constant-cost net stock of fixed private capital, assuming that the real value of land changed at the same rate as real GNP and deflating all other components with the GNP deflator. SEPTEMBER/OCTOBER 1991 8 Table 2 A Comparison of GNP and NNW as Measures of Macroeconomic Performance (compounded annual rates of change) Measure 1948-64 1964-81 1981-90 1948-90 GNP GNP deflator Real GNP Population Real GNP per capita NNW NNW deflator Real NNW Population Real NNW per capita 6.1% 2.2 3.9 1.7 2.1 5.8 2.1 3.6 1.7 1.9 little use in studying financial superstructure and the relation betw een real and financial assets.7 To achieve the purpose of understanding relations betw een financial and real assets, the national balance sheet has to be viewed m ore broadly than NNW. Financial assets play a key role in the econom ic development of m arket econom ies, enabling the tran sfer of lenders’ excess purchasing pow er to finance spending for borrow ers. If an econom ic unit consum es less than its income, it accum u lates assets or retires debt, thereby adding to its net w orth. If an econom ic unit chooses to ac cum ulate tangible assets less than its saving, it must accum ulate financial assets or reduce lia bilities. Similarly, if an econom ic unit chooses to accum ulate tangible assets in excess of saving, it m ust borrow or sell o ff financial assets. Because of this fundam ental role fo r financial m arkets, it is useful to focus on national balance sheets that include the financial asset-liability stru ctu re.8 A full understanding of the forces that are driving NNW requires a supplem entary analysis of financial assets and liabilities. A G o ld sm ith -T y p e U.S. B a la n ce S h ee t in 1 9 9 0 Table 3 sum m arizes the U.S. balance sheet in a Goldsmith-type form at. It uses the same data that are in the Board of Governors report, but H'he term “ financial superstructure” is attributable mainly to Goldsmith. It refers to all aspects of the system of finan cial markets that channels the funds of savers into invest ment. For more detailed discussion, see Board of Gover nors (1980). Digitized forFEDERAL FRASER RESERVE BANK OF ST. LOUIS 9.4% 6.5 2.7 1.1 1.6 10.9 7.2 3.4 1.1 2.3 6.7% 3.6 2.9 1.0 1.9 4.2 2.0 2.2 1.0 1.2 7.6% 4.2 3.2 1.3 1.9 7.5 4.1 3.2 1.3 1.9 sums the assets and liabilities for the separate sectors—households, businesses and private fi nancial institutions. Even though there is double counting, this procedure preserves detailed in form ation about financial assets and liabilities. Although the approach used h ere follows Goldsmith in principle, it is incom plete because the Board of Governors rep ort does not provide estim ates of tangible assets held by governm ents. By adding privately held financial assets to tangi ble assets, the value of tangible assets held by business and the equity claims on business held by households are both included. The assets and liabilities are fo r the private sector, so the balancing item is really a m ean ingless residual that includes governm ent liabilities, rest-of-w orld liabilities, as well as equities and net w orth of the private sector. Consequently, this balance sheet is not offered as a substitute for the standard Board of Gover nors version, but as a supplementary summary of the Board’s rep ort with a focus on private sec tor assets and liabilities. T he 1990 values in table 3 indicate that private sector holdings of financial assets w ere 1.68 tim es the value of tangible assets at the beginning of the year and 1.70 times at the end of the year. The sector breakdow n of this ratio sum marizes the stru ctu re of the U.S. economy. The household sector’s holdings of financial assets are about tw ice th eir holdings of tangible assets. 8For a discussion of the importance of keeping financial assets and liabilities on the national balance sheet, see Goldsmith and Lipsey (1963). 9 Table 3 Goldsmith-Type U.S. Balance Sheet: 1990 (billions of dollars)1 Beginning-ofyear value End-ofyear value Percent change Tangible assets Reproducible assets Households Nonfinancial business Private financial institutions Land Households Nonfinancial business Financial institutions Financial assets Households Nonfinancial business Private financial institutions $16,000.8 12,147.2 5,598.7 6,271.3 277.2 3,853.6 1,458.6 2,376.5 18.5 26,813.6 13,854.5 2,434.5 10,524.6 $16,223.2 12,482.9 5,760.2 6,417.9 304.8 3,740.3 1,315.8 2,404.4 20.1 27,571.8 14,091.4 2,566.1 10,914.3 1.4% 2.8 2.9 2.3 10.0 -2 .9 -9 .8 1.2 8.6 2.8 1.7 5.4 3.7 TOTAL ASSETS $42,814.4 $43,795.0 2.3% Liabilities, Equities and Net Worth Liabilities (private sector) Households Nonfinancial business Private financial institutions Other liabilities, equities and net worth2 TOTAL LIABILITIES, EQUITIES AND NET WORTH $18,510.0 3,620.7 4,634.3 10,255.0 24,304.4 $42,814.4 $19,390.3 3,961.3 4,840.4 10,588.6 24,404.7 $43,795.0 4.8% 9.4 4.4 3.3 0.4 2.3% Assets 'Source: Board of Governors of the Federal Reserve System (1991). Follows Raymond Goldsmith’s procedure of combin ing (adding) sector balance sheet items to derive a ‘total’ balance sheet. This balance sheet is incomplete, however, because it omits assets held by government and foreigners. Calculated as a residual. Reflects considerable double counting and also includes liabilities of government and foreigners. The nonfinancial business sector, on the other hand, holds financial assets equal to about 30 percent of its tangible assets. Financial institu tions, almost by definition, hold very few tangi ble assets. Financial assets can be com pared with liabilities to show the net m onetary cred itor status of the different sectors. Households hold financial assets about 3.7 times as large as their liabilities, w ith fixed claims almost 1.5 tim es as large as liabilities. Nonfinancial businesses have liabilities almost tw ice as large as their financial assets, w ith the ratio about the same w hen variable claims and liabilities are subtracted. Private fi 9“ Credit market instruments is a core group of debt claims that is the principal medium used by nonfinancial sectors in raising funds through formal credit channels.” [Board of Governors (1980), pp. 42-43.] It includes all government nancial institutions are net m onetary creditors, although not to the extent that households are. Reproducible assets are divided into the same categories as in table 1 and in the GNP ac counts—residential structures, nonresidential plant and equipment, inventories and consum er durables. Financial assets can be divided into many types, but h ere they are grouped into five categories—cu rren cy and deposits, credit m arket instrum ents, equities (both corporate equities and the n et w orth of noncorporate business), reserves (pension fund and life insurance) and other (which includes security and trade credit).9 securities, corporate and foreign bonds, mortgages, con sumer credit, bank (not elsewhere classified) and other loans. SEPTEMBER/OCTOBER 1991 10 Figure 2 Distribution of Tangible Assets Percent T r e n d s in T otal A ssets Postw ar trends in total assets are summarized by charting the com ponents of both tangible assets and financial assets relative to their respec tive totals. T a n g i b l e a s s e t s — Figure 2 sum m arizes the m ajor com ponents of tangible assets in 1982 dollars relative to the total of tangible assets, but broken down by type of asset as in table 1 (the same as in the GNP accounts). Vertical lines (1964 and 1981) correspond to the inflation episodes in table 2. For inventories and land, the trend generally has been downward throughout the postw ar period. In the case of consum er durables, the trend is upward throughout the period. For fixed residential in vestm ent, the trend was upward b efore 1964 but has been downward since then. Fixed nonresidential investm ent has trended upward, although there appears to be a flattening in the 1980s. 10Land was not included because it is fixed and nonreproducible. Its real value can change but not its quantity. "Higher rates of inflation increase the uncertainty associated with rates of return on financial assets, making tangible assets more attractive to investors. See Cagan and Lipsey (1978). FEDERAL RESERVE BANK OF ST. LOUIS Percent The relationship betw een the grow th of real reproducible tangible assets and inflation was exam ined.10 The conventional wisdom is that in vestors view tangible assets as a good hedge against inflation.11 Table 4 shows the correlation coefficients for the year-to-year percent change in reproducible tangible assets and inflation. None of the coefficients is significant and posi tive fo r each of the subperiods. Explaining trend movements in the com ponents of tangible assets is apparently m uch m ore complex than indi cated by a simple inflation model. F i n a n c i a l a s s e t s — Privately held financial assets w ere collected into categories as shown in figure 3. C urrency and deposits (broadly defined) drifted downward from the end of World W ar II until the early 1960s, stabilized until 1972, shifted to a higher level and then fell from 1984 to 1990. This category reflects a num ber of financial innovations throughout the period, notably certificates of deposit in the 11 Table 4 Correlation Coefficients: Percent Change in Reproducible Tangible Assets (1982 dollars) and Inflation1____________ 1948-64 1948-90 1965-81 1982-90 Consumer durables Residential structures Nonresidential plant and equipment Inventories -.1 6 -.4 2 * .08 .23 .30 .60* .52* .66* -.6 4 * -.2 8 -.5 0 ’ .04 -.8 6 * -.8 8 * -.5 7 -.3 5 'Significant at the 5 percent level 'Inflation is year-to-year percent change in GNP deflator. Figure 3 Distribution of Financial Assets Percent Percent 45 1960 1965 1960s, m oney m arket mutual funds in the 1970s and the paym ent of interest on checkable deposits in the 1980s.12 The trend of credit m arket instrum ents was quite flat until the mid-1970s, shifted to a higher level until the early 1980s and rose sharply from 1982 to 1990. Because federal securities are an 1975 1980 im portant com ponent of credit m arket instru ments, about half of the increase in the 1980s can be traced to the rapid grow th in the federal debt. The equity portion of financial assets shows a pattern generally the opposite o f that for credit m arket instrum ents. The downward trend in 12For a brief financial history of the United States, see Council of Economic Advisers (1991), chapter 5. SEPTEMBER/OCTOBER 1991 12 Table 5 Correlation Coefficients: Percent Change in Financial Assets (1982 dollars) and Inflation1_______________________________ 1948-90 1948-64 1965-81 1982-90 Currency and deposits Credit market instruments Life insurance and pension fund reserves Equities Other financial assets -.1 3 -.4 2 * -.41 -.6 9 * -.2 6 -.31 -.1 9 -.8 4 -.3 8 * -.2 0 -.1 5 -.5 6 * .15 .57* -.1 8 -.1 5 -.3 7 -.0 3 -.2 8 -.8 3 ‘ Significant at the 5 percent level ’ Inflation is year-to-year percent change in GNP deflator. equities began in the high-inflation 1970s, but continued through the disinflation of the 1980s.13 These trends suggest a com plem entary relation betw een equities and credit m arket instrum ents (including governm ent securities). The total of these two categories has varied betw een 60 and 70 percent of all financial assets since W orld W ar II. Life insurance and pension fund reserves rose gently until 1980, and then accelerated in the 1980s. This recen t acceleration is consistent with a num ber of explanations. One would be that it represented a favorable long-term plan ning response to the deceleration of inflation. Another would be the demographics of the de cade which included a rise in the average age of the population.14 The residual com ponent of financial assets, called “oth er,” reflects mainly trade and security credit. This category moved upward slowly but steadily until the mid-1970s and then stabilized. As with reproducible tangible assets, coeffi cients w ere calculated for the correlation b e tw een the percent change in financial assets in 1982 dollars and inflation. These results are summarized in table 5. Most of the coefficients are negative, although most are insignificant. Even though nominal financial assets tend to in crease with inflation, their grow th is generally outpaced by inflation so that in real term s there is an inverse relationship. 13lt was formerly believed that corporate stocks were a hedge against inflation. Fischer and Modigliani (1978) suggest that this changed when investors realized that higher inflation carries with it a higher real tax burden. Digitized for FEDERAL FRASER RESERVE BANK OF ST. LOUIS SOME USES OF THE BALANCE SHEET The U.S. balance sheet covers a relatively small portion of the nation’s wealth. However, it can yield insights into particular relationships that cannot be fully analyzed using inform ation only from GNP accounts. T h e accumulation of flows into stocks provides a built-in long-term perspec tive that is generally missing w ith GNP accounts. By lengthening the tim e perspective, balance sheet inform ation can shed new light on some commonly held perceptions about economic trends. F in a n cia l In terrela tio n s Ratio One of the most im portant applications of balance sheet inform ation is the calculation and analysis of the financial interrelations ratio.15 This ratio m easures the size of the financial superstructure relative to the real in frastru c ture. Specifically, it is the ratio of the value of financial assets to the value of tangible assets. The financial interrelations ratio provides a fram ew ork for the analysis of the relationship betw een financial development and economic growth. However, as Goldsmith points out, “Eco nomic grow th is so complex a phenomenon, ob viously determ ined or influenced by basic fac tors of a physical, technological, and masspsychological nature, that an attempt to isolate the effects of apparently secondary forces such ' “Carlson (1990). 'sGoldsmith (1966) 13 Figure 4 Financial Interrelations Ratio1 fin a n c ia l assets relative to tangible assets (both in 1982 dollars) as the ch aracter of financial institutions and the nature of credit practices does not promise suc cess.”16 Generally, the argum ent is that a rise in the interrelations ratio indicates a broadening of the range of financial assets and institutions. This prom otes the flow of saving into its most productive uses w hich stimulates econom ic grow th and increases productivity.17 Figure 4 shows the interrelations ratio for the 1948-90 period. The factors influencing its move m ent are num erous and complicated, although inflation appears to have played a role. Prices of tangible assets, the denom inator in the ratio, tend to increase m ore than other prices during periods of accelerating inflation, and by a lesser amount w hen it decelerates. The ratio fell to its postw ar low during the high-inflation period of the 1970s before rising during the disinflation of the 1980s. Such an explanation is simplistic because a full analysis of the interrelations ratio would consider all other factors entering into its 16Goldsmith (1969), p. xi. 17For more detailed discussion of this theory, see Goldsmith (1969), pp. 390-401. Also see Shaw (1973). 18Goldsmith (1969), p. 97. 19Recently an argument has been offered challenging the determ ination. Nonetheless, inflation is a factor influencing the ratio.18 On the other hand, real GNP grow th does not appear to be related systematically to the ratio, especially since the mid-1970s. Thus, even though the financial in terrelations ratio shows interesting movements in the postw ar period, it is only a starting point in the analysis of financial stru ctu re and eco nomic grow th.19 One facet of the interrelations ratio that has produced con cern in the 1980s is the rapid grow th of credit m arket debt in the private sec tor. Expansion of debt permits m ore spending than otherw ise, but adds to the severity of a recession w hen the pace of econom ic activity slows. To maintain debt payments, households and businesses have to restrain their spending or default on their loans. W idespread loan de faults could endanger the econom ic health of the financial system. notion that growth in financial structure is always beneficial. See Fingleton (1991). SEPTEMBER/OCTOBER 1991 14 Figure 5 Credit Market Debt Relative to Tangible Assets Percent Figure 5 puts the 1980s expansion of debt into perspective. All credit m arket debt of households, private nonfinancial business and private finan cial institutions is included and m easured against total privately held tangible assets. Debt expanded at an 11 percent annual rate from 1983 to 1990, pushing the ratio of debt to tangi ble assets to a historical high of 0.52. The ex tent of the increase is dampened somewhat w hen the ratio is calculated w ith 1982 dollars, but 1990 is still at a historical high; it appears to be leveling off, however. W hether this debt burden is “too high” will probably not be answ ered until the strength and duration of the recovery from the recen t recession is clear. In fla tio n a n d th e D istrib u tio n o f N et W orth An additional use of balance sheet inform ation is to analyze the effect of inflation on the net w orth of various sectors. The standard theory of such effects is outlined in the shaded insert 20No attempt is made here to measure anticipated inflation However, based on current procedures, the variances of change in inflation and the unanticipated change have been found to be similar. See Ball and Cecchetti (1990), p. 242. FEDERAL RESERVE BANK OF ST. LOUIS Percent at right.20 Figure 6 shows the distribution of private net w orth among sectors. The proportion of private net w orth held by households gradually increased from 1948 until inflation accelerated sharply in the mid-1970s; it then declined until 1981. Since then, households' share of net w orth has risen as the disinflation continued through most of the decade. These responses are typical of a sector that is a net m onetary creditor. The nonfinancial business sector has shown a variety of long-term trends, but the response to the acceleration and deceleration of inflation is similar for the three subsectors because they are all m onetary debtors. Farm business has been in a long-term decline throughout the post w ar period, interrupted by a slight increase from 1971 to 1980 w hich prim arily reflected a rise in farm real estate values. Nonfarm non cor porate business declined as a share of private net w orth until 1976, increased until 1980, and 15 Inflation and the B alance Sheet units are net m onetary creditors. Net m one tary creditors are harm ed by unanticipated inflation because the purchasing pow er of their m onetary assets declines m ore than the real value of their m onetary liabilities. Simi larly, net m onetary debtors benefit from un anticipated inflation. These effects take place without any action on the part of the eco nomic unit and rep resen t a passive redistribu tion of wealth. Inflation has many effects, but most dis cussed are its effects on the distribution of income and w ealth.1 If inflation w ere fully and correctly anticipated, nominal interest rates on fixed claim assets would adjust to com pensate for the declining purchasing pow er of the principal. Generally, however, in terest rates have not com pletely com pensated for inflation. As a result, periods of inflation have been accom panied by arbitrary tran s fers of real net w orth from net m onetary creditors to net m onetary debtors w hen infla tion is accelerating, o r vice versa w hen it is decelerating. Even if inflation is anticipated and reflected in nominal interest rates, added uncertainty about future prices can affect econom ic deci sions.2 As a result, econom ic units will at tempt to redistribute their assets for protec tion from inflation. This will elevate prices of tangible assets relative to financial assets. The key factors in determ ining w hether econom ic units will benefit from , or be harm ed by, inflation are (1) w hether the inflation is anticipated and (2) w hether the econom ic 2See Cagan and Lipsey (1978). 1See Bach and Stephenson (1974), Cagan and Lipsey (1978), Conard (1964), Fischer and Modigliani (1978) and Kessel and Alchian (1962). Figure 6 Distribution of Private Net Worth Percent Percent 6 0 1--------- ---------60 Households1 Nonfinancial corporate business Nonfarm noncorporate business Farm business Private financial institutions' 1950 1955 1960 1965 1970 1975 1980 1985 1990 ’ Net assets SEPTEMBER/OCTOBER 1991 16 Figure 7 Net Foreign Assets Relative to Total Assets Held by the Private Sector since then has moved back to its 1976 level. Nonfinancial corporate business is difficult to characterize for the full period. Its response to the acceleration and deceleration of inflation seems quite clear, with its net w orth proportion increasing during the acceleration and then fall ing back during the disinflation. The net w orth of private financial institutions does not seem to be affected m uch by swings of inflation, con trary to the well-known problem s of the savings and loan industry. Private finan cial institutions are net m onetary creditors, but the difference betw een their financial assets and their liabilities is very small.21 F o r e ig n O w n ersh ip o f U.S. A ssets Another application of the U.S. balance sheet is to exam ine con cern about the accumulation 21It is difficult to relate to the private financial sector because of its heterogeneity. It consists of commercial banking, savings institutions, insurance (including private pension funds and state and local government retirement funds) and other (including finance companies and mutual funds). 22This concern and several others are examined in Ott (1989). 23These trends in foreign ownership are unlikely to continue. For example, assets of U.S. business acquired or FEDERAL RESERVE BANK OF ST. LOUIS of U.S. assets by foreigners during the 1980s. A common perception is that foreigners could eventually own m ore than 50 percent of business capital leading to the potential for foreign con trol of the U.S. economy. This would threaten U.S. econom ic sovereignty and national securi ty.23 Balance sheet data can b e used to examine this concern. Figure 7 shows n et foreign assets as a percen tage of total assets in the United States. It is clear that the proportion of U.S. assets held by foreigners has increased sharply since the early 1980s. Foreign direct investment, how ever, has increased from only 0.4 percent in 1980 to about 1 percent of U.S. total assets in 1990. W ith con tinued grow th at this pace, foreign ow nership would not exceed 50 p ercen t fo r 800 years.23 Rather than "signaling an econom y in decline, established by foreign investors dropped by 25 percent from 1989 to 1990. See Fahim-Nader (1991). 17 Figure 8 Ratio of Business Capital to Household C apital 1 1950 1955 1960 1965 1970 1975 1980 1985 1990 1Nonresidential plant and equipment divided by residential structures plus consumer durables (all in 1982 dollars) such investm ent by foreigners is a m easure of the econom y’s vigor.”24 On a worldwide basis, foreign direct investm ent increases econom ic w elfare by moving resources from less to m ore productive uses. R atio o f B u s in e ss Capital to H o u s e h o ld Capital Another concern that developed in the 1980s was that the United States was not channeling its saving into the "right” kind of investm ent.25 Mainly because of a com bination of inflation and tax shelters, savings w ere directed tow ard household capital rath er than productive busi ness capital.26 Business investm ent in plant and equipm ent is a m ajor vehicle fo r increasing eco nom ic growth. It boosts productivity by provid ing m ore capital per w orker and also embodies technological improvements. Household capital, on the oth er hand, provides services to con sum ers but does not add directly to productive capacity. Figure 8 shows the ratio of nonresi- ^O tt (1989), p. 63. 25This view is developed in Rutledge and Allen (1989). 26For a discussion of the effects of the Tax Reform Act of 1986, see Fazzari (1987) and Slemrod (1990). dential plant and equipment to household capi tal—the sum of consum er durables and residen tial structures. Following W orld W ar II, households enlarged their stock of capital until 1964. From 1964 to 1970, grow th of business capital stock exceeded that of household capital stock. For the next 12 years, business and household capital grew at roughly the same rate. Since 1982, how ever, the grow th of household capital has exceeded that of business capital. Boosting the overall level of saving is the prim ary vehicle for stimulating econom ic growth. T h ere is also potential for faster grow th by designing policies that direct the flow o f saving away from household capital into business capital.27 SUMMARY Economic analysts rely mainly on the nation's GNP accounts as a source of inform ation on econom ic perform ance. For purposes of under- 27For a more complete discussion of saving and its role in economic growth, see Cullison (1990) and Harris and Steindel (1991). SEPTEMBER/OCTOBER 1991 18 standing the forces at w ork in the determ ina tion of cu rren t production, GNP accounts are indispensable. Generally overlooked, how ever, is another source of inform ation—the nation’s balance sheet. Business accounting relies greatly on the balance sheet as a tool for analyzing a firm ’s financial health. Similar practices do not prevail in national econom ic accounting. W hat would appear to be one of the m ost im portant items in the U.S. balance sheet is the m easure of national n et w orth. W hen com pared with GNP as a m easure of long-term econom ic perform ance, how ever, it does not seem to of fer m uch added inform ation. Probably the most im portant use of balance sheet data is to analyze the role of financial stru ctu re in the process of econom ic grow th. A variety of other questions, how ever, can also be exam ined by developing ratios of particular balance sheet items. The chief benefit of the U.S. balance sheet, as it is currently prepared, seems to be that it forces the u ser to take a long-term perspective to detect changing trends. W hat appeared to be m ajor concerns during the 1980s sometimes took on a different in terp reta tion w hen viewed from the perspective of the U.S. balance sheet over the entire post-W orld W ar II period. REFEREN CES Bach, G. L., and James B. Stephenson. “ Inflation and the Redistribution of Wealth,” Review of Economics and Statistics (February 1974), pp. 1-13. Ball, Laurence, and Stephen G. Cecchetti. “ Inflation and Uncertainty at Short and Long Horizons," Brookings Papers on Economic Activity, 1 (1990), pp. 215-45. Board of Governors of the Federal Reserve System. Balance Sheets for the U.S. Economy 1945-90 (March 1991). _______ . Introduction to Flow of Funds (June 1980). Boskin, Michael J., Marc S. Robinson, and Alan M. Huber. “Government Saving, Capital Formation, and Wealth in the United States, 1947-85,” in Robert E. Lipsey and Helen Stone Tice, eds., The Measurement of Saving, Investment, and Wealth (University of Chicago Press, 1989), pp. 287-356. Cagan, Phillip, and Robert E. Lipsey. The Financial Effects of Inflation (Ballinger Publishing Company, 1978). Carlson, Keith M. “ On Maintaining a Rising U.S. Standard of Living Into the Mid-21st Century,” this Review (March/April 1990), pp. 3-16. Conard, Joseph W. “The Causes and Consequences of Infla tion,” in Inflation, Growth and Employment, A Series of Research Studies prepared for the Commission on Money and Credit (Prentice-Hall, 1964), pp. 1-144. Council of Economic Advisers. 1991 Annual Report (GPO, February 1991). FEDERAL RESERVE BANK OF ST. LOUIS _______ . 1987 Annual Report (GPO, January 1987). Cullison, William E. “ Is Saving Too Low in the United States?” Federal Reserve Bank of Richmond Economic Review (May/June 1990), pp. 20-35. Eisner, Robert. The Total Incomes System of Accounts (University of Chicago Press, 1989). Fahim-Nader, Mahnaz. “ U.S. Business Enterprises Acquired or Established by Foreign Direct Investors in 1990,” Survey of Current Business (May 1991), pp. 30-39. Fazzari, Steven M. “Tax Reform and Investment: How Big an Impact?” this Review (January 1987), pp. 15-27. Fingleton, Eamonn. “ Highly Speculative,” Atlantic (June 1991), pp. 22-25 . Fischer, Stanley, and Franco Modigliani. “Towards an Understanding of the Real Effects and Costs of Inflation,” Weltwirtschaftliches Archiv (1978), pp. 810-33. Goldsmith, Raymond W. Comparative National Balance Sheets: A Study of Twenty Countries, 1688-1978 (University of Chicago Press, 1985). _______ . The National Balance Sheet of the United States, 1953-1980 (University of Chicago Press, 1982). _______ . Financial Structure and Development (Yale Univer sity Press, 1969). _______ .“The Uses of National Balance Sheets,” Review of Income and Wealth (June 1966), pp. 95-133. Goldsmith, Raymond W., and Robert E. Lipsey. Studies in the National Balance Sheet of the United States, Volume I (Princeton University Press, 1963). Harris, Ethan S., and Charles Steindel. “ The Decline in U.S. Saving and Its Implications for Economic Growth,” Federal Reserve Bank of New York Quarterly Review (Winter 1991), pp. 1-19. Jorgenson, Dale W., and Barbara M. Fraumeni. “ The Ac cumulation of Human and Nonhuman Capital, 1948-84,” in Robert E. Lipsey and Helen Stone Tice, eds., The Measurement of Saving, Investment, and Wealth (University of Chicago Press, 1989), pp. 227-86. Juster, F. Thomas. “A Framework for the Measurement of Economic and Social Performance,” in Milton Moss, ed., The Measurement of Economic and Social Performance (Na tional Bureau of Economic Research, 1973), pp. 25-109. Kendrick, John W. The Formation and Stocks of Total Capital (National Bureau of Economic Research, 1976). Kessel, Reuben A., and Armen A. Alchian. “ Effects of Infla tion,” Journal of Political Economy (December 1962), pp. 521-37. Ott, Mack. “ Is America Being Sold Out?” this Review (March/April 1989), pp. 47-64. Rutledge, John, and Deborah Allen. Rust to Riches (Harper and Row, 1989). Shaw, Edward S. Financial Deepening in Economic Develop ment (Oxford University Press, 1973). Slemrod, Joel, ed. Do Taxes Matter? The Impact of the Tax Reform Act of 1986 (MIT Press, 1990). 19 Piyu Yue and R obert Fluri Piyu Yue is a research associate scholar at the 1C2 Institute, University of Texas at Austin. Robert Fluri is an economist at the Swiss National Bank. This article was written while both authors were visiting scholars at the Federal Reserve Bank of St. Louis. Lynn Dietrich provided research assistance. Divisia M onetary Services In dexes for Sw itzerland: Are They Useful fo r M onetary Targeting? F r o m 1973 TO 1989, INFLATION in Switzerland was roughly one-half that in the United States. For example, consum er prices in Sw itzer land rose 3.5 percent per year during this 16-year period com pared with the 6.6 percent average annual rise in U.S. consum er prices. Similarly, Swiss wholesale prices rose at an an nual rate of 2 p ercen t during this period in con trast to the nearly 6 percent annual average increase in the U.S. producer price index. In deed, the Swiss inflation experience, along with that of W est Germany, is often cited as an ex cellent example of the gains that accrue to a nation whose central bank conducts m onetary policy by announcing—and achieving—a target fo r the grow th of a m onetary aggregate. The Swiss central bank has used m onetary base grow th rate targets since 1980. Because the historical relationships betw een m onetary base grow th and econom ic activity have changed m arkedly since the end of the 1980s, the Swiss have begun to reconsider the use of annual m onetary aggregate targets, and are considering the potential usefulness of broad er m onetary aggregates as indicators of m onetary policy. This paper develops two alternative broader monetary aggregate m easures, Divisia M l and M2, for Switzerland and com pares their poten tial usefulness as m onetary indicators w ith the Swiss M l and M2 aggregates as usually defined. First, however, we show why questions have been raised about the continued usefulness of the annual m onetary base grow th rate target in Switzerland. W e then discuss the methodology underlying the Divisia approach to constructing m onetary aggregates and use this methodology to derive Swiss Divisia M l and M2 m easures. Next, we examine the relationship betw een Swiss inflation and the grow th rates of Swiss M l and M2 and the Swiss Divisia M l and M2 aggregates to determ ine their relative usefulness as m onetary policy indicators. Finally, we exa mine the relationships betw een these various m onetary aggregates and the m onetary base to assess the extent to which the Swiss central bank could control their grow th. SOME BACKGROUND ON THE SWISS MONETARY BASE Prior to 1986, the relationship betw een the m onetary base and econom ic activity in Sw itzer land was quite close; this link, how ever, has broken down since then. T he sudden change SEPTEMBER/OCTOBER 1991 20 ■I Figure 1 Inflation and Swiss Monetary Base Growth Percent Percent can be illustrated, in part, by looking at the relationship betw een inflation and the grow th rate of the m onetary base as depicted in fig ure 1. Until 1986, Swiss inflation movements lagged about th ree years behind corresponding variations in the base m oney grow th rate. Since 1986, how ever, this pattern no longer holds. For example, the sharp drop in Swiss inflation in 1986 was attributable to substantial reductions in the prices of im ported goods. Consequently, it appears that m onetary base grow th neither contributed to this decline nor provided any w arning that it would occur; indeed, the grow th of the Swiss m onetary base was virtually cons tant from 1984-1987. Similarly, movem ent in the Swiss m onetary base from 1987-1989 (in particular, the sharp drop in 1988) yielded neither w arning nor ex planation of the sharp rise in Swiss inflation in 1989. This change followed two significant insti FEDERAL RESERVE BANK OF ST. LOUIS tutional innovations in the Swiss banking sys tem. First, a new electronic interbank payment system (Swiss Interbank Clearing System, SIC) was introduced in the sum m er of 1987. Then, on Jan uary 1, 1988, reduced reserve req u ire m ents on Swiss bank deposits w ent into effect. In response to these changes, Swiss banks have sharply reduced their reserve balances at the Swiss National Bank (SNB). Swiss bank reserves dropped from m ore than SF8 billion (about $5.5 billion) at the end of 1987 to SF3 bil lion (about $2.1 billion) by the end of 1989. As a result of changes in the relationship b e tw een the m onetary base and inflation, the con tinued usefulness of the monetary base as a monetary policy indicator has been questioned. One suggestion is to rely more on broader mone tary aggregates as monetary policy indicators. 21 MONETARY AGGREGATION Generally, central banks worldwide use essen tially identical procedures to con stru ct th eir na tions’ m onetary aggregates. They first define the specific aggregate—that is, they determ ine which financial com ponents it will include—and then they simply “add” its selected com ponents to gether. Not too surprisingly, these m onetary ag gregates are called "simple-sum” aggregates. Simple-sum aggregation has been criticized for failing to distinguish betw een the differing de grees of m onetary (transaction) services and store-of-value services provided by the com ponents in the m onetary aggregate. Presumably, only the form er (that is, m onetary or tran sac tion) services should be included w hen a m one tary aggregate is considered. Friedm an and Schw artz (1970) have described this problem: This [summation] procedure is a very special case of the more general approach discussed earlier. In brief, the general approach consists of regarding each asset as a joint product having different degrees of "moneyness,” and defining the quantity of money as the weighted sum of the aggregate value of all assets, the weights for individual assets varying from zero to unity with a weight of unity assigned to that asset or assets regarded as having the largest quantity of "moneyness” per dollar of aggregate value. The procedure we have followed implies that all weights are either zero or unity. The more general approach has been suggested frequently but experimented with only occasional ly. We conjecture that this approach deserves and will get much more attention than it has so far received, (pp. 151-52) As Friedman and Schwartz surmised, economic aggregation theory and statistical index num ber theory have been used to provide both theoreti cal and em pirical solutions to the problem of m onetary aggregation.1 This research has led to the development of alternative m onetary aggre gates, in particular, the Divisia m onetary service m easure.2 T h e Divisia In d e x N u m b e r Index num bers are widely used to provide a single broad m easure for a disparate collection of items. Well-known examples of index numbers are the industrial production index, the consum er price index and the producer price index. These index num bers depend upon both the prices and quantities of items included in the in dex because the values of commodities involved are determ ined by their physical quantities and corresponding prices. Because quantities of financial assets are m easured in term s of "dollars,” simply adding the balances of various m onetary com ponents would appear to be a natural approach to m ea suring m onetary aggregates. Consequently, it is not surprising that simple-sum m onetary aggre gates have been used extensively throughout the world. However, econom ic theory suggests that various m onetary aggregate com ponents differ in term s of their “liquidity” and, thus, may have substantially different effects on eco nom ic activity. If this is so, the simple-sum procedure may actually be inappropriate for m easuring “m onetary service flow s” in the na tion. Instead, an alternative approach that in volves calculation of Divisia indices may provide superior alternatives to m easuring m onetary ag gregates w hen com pared with the traditional simple-sum m onetary m easures. Theoretically, the Divisia index num ber is derived from the econom ic aggregation theory and first-order conditions for utility optimiza tion. An expanded discussion of the Divisia ap proach appears in appendix 1 of this paper. Em pirically, the Divisia index num ber is estim ated from a nonlinear function of the quantities and the corresponding p rices of individual com ponents that create the aggregate. M oreover, its grow th rate is a linear combination of the grow th rates of its com ponents, w here the weights (or coefficients) on the com ponents are their average ex p en d itu re shares. In com parison, the simple-sum index is the linear sum of the quantities of the components in which the weight (coefficient) given to each com ponent is unity and their prices have no effect on the index. The grow th rate of the simple-sum index is also a linear com bination of the grow th rates of its ’ See Barnett (1980). 2Use of Divisia monetary aggregates appears in Barnett and Spindt (1982), Donovan (1978), Farr and Johnson (1985), Belongia and Chalfant (1989), among others. SEPTEMBER/OCTOBER 1991 22 com ponents, w here the coefficients are equal to the quantity shares.3 D ifferences betw een the behavior of Divisia and simple-sum aggregates stem from the dif feren t weights assigned to the grow th rates of the com ponents, w hich m easure their contribu tions to the m onetary aggregates. Coefficients called “shares” are expressed by the notations S * a n d Sit for various com ponents used to d eter mine the Divisia and simple-sum indexes, re spectively, in the discussion that follows. In calculating the Divisia index, the share of each com ponent is the ratio of the expenditures on the m onetary service flows it provides to the total expenditure on m onetary service from all com ponents in the aggregate; as such, it represents an "expenditure share." In contrast, shares fo r com ponents in the simple-sum index are equal to the quantities of the balances held in each com ponent divided by the total balances of all com ponents in the aggregate. In general, these tw o types of shares yield d ifferent values and move diversely over time. For example, the expenditure share of tim e deposits in Swiss M2 in January 1980 was 0.0856; its quantity share, in contrast, was 0.3681. In January 1988, how ever, the expenditure share of time deposits in Swiss M2 w as 0.2823, while its quantity share was 0.4527. For the com ponents of simple-sum m onetary aggregates, the only data required to compute their respective shares are quantities of the com ponents themselves. In contrast, for Divisia aggregates, th e p rices of the com p onen ts— w hich involve their interest rates—must also be obtained in ord er to com pute their expen diture shares. T h e U ser Costs o f th e M o n eta ry A ssets As noted above, one m ajor problem involved in computing Divisia index num bers for m one tary aggregates is in determ ining the relevant prices of the individual m onetary assets that 3By definition, the simple-sum aggregate is denoted such that SIMit = I mcit dSIM/SIM, = I (mcit/SIMt)(dmcit/mcit) where me; is the quantity of the i- th component. Thus, approximately ln(SIM,)-ln(SIM,_,) = IS it(in(mcit)-ln (m cit_1)), FEDERAL RESERVE BANK OF ST. LOUIS make up the aggregate. In econom ic aggregation theory, m onetary assets are treated as com mod ities and their prices are defined similarly to rental prices of durable goods. In this approach, it is assumed that people receive m onetary ser vices from holding m oney to finance their con sumption. In doing so, they forego higher yields typically available on other financial assets. W hile m onetary services are considered con sumed during some given period, money stock (like any durable good) is not generally con sumed during this period. Because m onetary services are flow variables—n ot stock variables— they should be evaluated by their rental prices or u ser costs. T h erefore, Divisia index num bers can be used to m easure the m onetary service flows provided by various m onetary assets in the econom y only if the u ser costs of these assets can be correctly defined and accu rate ly measured. The appropriate u ser costs of m onetary assets are based on m icroeconom ic theory and are derived by examining the representative con sum er’s optimal intertem poral consum ption pat tern and m onetary asset portfolio allocation.4 These user costs are m easured as the oppor tunity costs of foregone in terest associated with holding funds in different types of m onetary assets. The opportunity cost is obtained by com paring each asset’s rate of retu rn to that on a b en c h m a rk asset with the highest rate of return. Under the relevant consum er theory, the b en chm ark asset is assumed to provide no liquidity or other m onetary services. Because it is held only for accum ulating and tran sferrin g wealth across time, its in terest rate is the highest in the economy. Consumer theory, how ever, does not specify other characteristics of the benchm ark asset that would enable research ers to identify the actual benchm ark asset to be used in em pirical studies. Barnett and Spindt (1982) have suggested that, while hum an capital might best fit the theoreti cal concept of the benchm ark asset, no satisfac tory em pirical data exists on its rate of retu rn. In their research , they found that where Sit = mcit/SIMt is the quantity share of the i- th component. This equation shows that the growth rate of the simple-sum aggregate is a linear combination of the growth rates of its components with weights equal to the fractions of the quantities of the components to the simplesum aggregate. 4See Barnett (1978). 23 Figure 2 Selected Swiss Interest Rates Percent Percent Rt = m ax[rbaa, r jt, i = 1,2, .. n] many research ers have used the approach adopted by Barnett and Spindt. provided the best available proxy for the theo retical benchm ark rate, w here r baa is Moody’s series of seasoned Baa corporate bond rates and rit is the own rate of retu rn on each of the com ponents of L (the broadest U.S. m onetary ag gregate defined by the Federal Reserve Board).5 Although Donovan (1978) used the nominal rate of retu rn on "bonds” to com pute the rental price of interest-bearing money fo r Canada, In this paper, we use the Barnett-Spindt ap proach to generate a proxy for the Swiss ben chm ark asset rate (see appendix 2 for fu rth er details). The benchm ark asset is either the long term Swiss bond or short-term Euro-Swiss deposits, depending on w hich yield is higher. Thus, as shown in figure 2, the benchm ark asset was long-term Swiss bonds b efore 1980, Euro-Swiss deposits during 1980-1981 (due to an 5ln the user cost formula, R, is the maximum available yield in the economy on any monetary asset which is a uniquely defined theoretical maximum available yield. Empirically, the proxy variable is defined by a long-term bond yield relative to the rates of return on all monetary components. The need for a long-term bond yield in measuring short term holding-period yields is demonstrated by R. Shiller (1979). The Baa bond rate is used as a representative yield for long-term debts of average risk. See Barnett and Spindt (1982). SEPTEMBER/OCTOBER 1991 24 inverted term stru ctu re of Swiss interest rates during this period), long-term Swiss bonds from 1982 to 1987 and Euro-Swiss deposits again in 1988-1989.6 T he form ula fo r the real user costs of m one tary assets in period t is expressed as uit = (Rt - r it)/(l + Rt), w here uit is the u ser cost of the i-th m onetary asset, Rt is the nominal interest rate on the ben chm ark asset and r itis the nominal interest rate on the i-th m onetary asset in period t.7 COMPARISON O F THE BEHAVIOR OF SWISS DIVISIA AND SIMPLESUM MONETARY AGGREGATES Divisia m onetary aggregates and simple-sum aggregates fo r M l and M2 w ere calculated us ing the Swiss m onetary data described in appen dix 2 .8 The monthly simple-sum and Divisia m onetary aggregates are indexed to equal 100 in June 19 7 5 .9 Figures 3 and 4 show the 12m onth grow th rates of these aggregates and the Swiss consum er price index from Ju n e 1976 to D ecem ber 1989. W hile Divisia and simple-sum M l indices display virtually identical grow th rates in this period, the grow th rates of Divisia and simple-sum M2 indices differ substantially beginning in 1979 (see figure 4). From 1980-1981, for example, the grow th of simple-sum M2 rose rapidly while that of Divisia M2 slowed markedly. From 1982 to 1983, in contrast, the opposite pattern can be observed in their respective grow th rates. In 1989, how ever, the divergent pattern observed from 198081 occurs once again. These widely divergent grow th rates over ex tended periods fo r the simple-sum and Divisia M2 suggest that discussions about the appropri ateness of alternative procedures used to con 6Needless to say, such shifts in the benchmark asset raise questions about the validity of this approach and have resulted in criticisms of Divisia indices by a number of economists. We do not address this issue in this paper. H'he nominal user costs of monetary assets usually are ex pressed as Ui, = P, (R, —r„)(1 - t,) 1 + R ,(1 - t,) where uitis the user cost of the i- th monetary asset in period t, R, is the benchmark rate in period t, rit is the nominal interest rate on the i- t h monetary asset; xt is the Digitized forFEDERAL FRASER RESERVE BANK OF ST. LOUIS struct m onetary aggregates are not m erely "academ ic.” In general, the direction and mag nitude of grow th in m onetary aggregates are presum ed to provide useful inform ation about the cu rren t stance of m onetary policy and the future course of econom ic conditions. Such ex trem e d ifferences in the grow th of alternative M2 m easures (as shown in figure 4), however, may produce considerable difficulty in assessing that information. S im p le S u m Vs. Divisia M o n eta ry A g g re g a te s : W hat’s th e D if f e r e n c e ? In Switzerland, M l consists of cu rren cy (C), demand deposits with banks (DB) and demand deposits with the postal giro system (DP). To com pare simple-sum and Divisia M l, we calcu lated the shares (Sit, S*t) fo r each com ponent of M l. For m ost of the period, the respective shares of each m onetary com ponent fo r simplesum and Divisia M l moved so uniform ly that their respective contributions to these aggregates are roughly equal. T h erefore, the grow th in simple-sum and Divisia M l was essentially the same over the sample period (as already noted in figure 3). In May 1989, how ever, the explicit interest rate on demand deposits with the postal giro system (DP) rose from zero to tw o percent, reducing the user cost of DP, U3 (as shown in figure 5). Since expenditure shares depend both on quantities of the com ponents and their user costs, DP’s share (S *) fell, reducing its weight in calculating Divisia M l. T h erefore, since the in troduction of explicit in terest paym ents on DP had no effect on its weight in calculating simple-sum M l, the values S * and S3 diverged after May 1989. Since the values of S *a n d S3 are quite small, how ever, the difference b e tw een the grow th of simple-sum and Divisia M l after May 1989 is trivial. Figure 5 shows that the u ser costs of the th ree Divisia M l com ponents (U1-U3) follow marginal income tax rate, and p , is the true cost-of-living index used to deflate all nominal quantities to real quan tities. Since taxes are not considered here, we use the simplified formula. W eak separability conditions should be satisfied first to calculate a meaningful aggregate. However, we did not conduct weak separability tests; instead, we used the ac tual Swiss definitions of M1 and M2 and simply assumed that they are admissible aggregates. For details, see Belongia and Chalfant (1989). 9See appendix 2. 25 Figure 3 Inflation and M1 and Divisia M1 Growth Percent Figure 4 Inflation and M2 and Divisia M2 Growth SEPTEMBER/OCTOBER 1991 26 Figure 5 User Costs of Currency, Demand Deposits, PGS, and Time Deposits wmmmm similar m ovements from 1975 through 1989, especially fo r the u ser costs of curren cy and de mand deposits with banks. Thus, the relative user costs for the Divisia M l com ponents are nearly constant, making simple-sum M l as use ful as Divisia M l over this period.10 Although the expenditure and quantity shares w ere similar fo r the M l com ponents, low er u ser costs fo r tim e deposits in M2 explain the divergent patterns shown earlier betw een simple-sum and Divisia M2. W e can illustrate the im portance of changes in the econom ic en vironm ent on the weights used by examining the d ifferent behavior of Divisia and simple-sum M2 over these tim e periods: Jan uary 1979 - 10This result is consistent with Hicks’ (Hicks, 1946) conclu sion that “ when the relative prices of a group of com modities can be assumed to remain unchanged, they can be treated as a single commodity.” FEDERAL RESERVE BANK OF ST. LOUIS D ecem ber 1981; Jan uary 1982 - November 1987; D ecem ber 1987 - D ecem ber 1989. At the begin ning of each period, th ere was a significant change in Swiss m onetary policy as m easured by sharp movements in the Swiss m onetary base. During these periods, changes in the econom ic environm ent w ere reflected in the levels of short-term and long-term in terest rates. As noted earlier (in figure 2), we used the three-m onth Euro-Swiss Franc rate as the short term rate, the Swiss governm ent bond yield as the long-term rate and the benchm ark rate was equal or close to the higher of these two rates in any specific period. The grow th rate of the Swiss m onetary base over these periods was previously shown in figure 1. 27 Figure 6 Simple-Sum Share S4 and Divisia Share SJ of Time Deposits and User Cost U4 Percent T h e F irst P e rio d : J a n u a r y 1 9 7 9 to D e c e m b e r 1981 During this period, the SNB’s response to ris ing Swiss inflation was sharply slow er grow th in the Swiss m onetary base (resulting in lower inflation in 1983-84). The abrupt rise in short term Swiss interest rates produced an inverted yield curve for the next two years. The dramatic increase in interest rates on tim e deposits re duced their user costs (U4) to nearly zero as their in terest rates approached the benchm ark rate (see figures 2 and 5). How did this interest rate movem ent affect the m onetary aggregates? Asset holders shifted from low er yielding securities into time deposits (TD) causing the quantity of tim e deposits to in crease dram atically and the quantity of demand deposits w ith banks (DB) to decrease substantial ly. Furtherm ore, asset holders also shifted funds into time deposits from other financial assets not included in M2. This sharp rise in the quan tity of time deposits is indicated by the surge in their share in simple-sum M2, S4 (shown in figure 6). These changes produced quite different r e sults in the Divisia M2 m easure, how ever. As in terest rates on time deposits increased relative to other interest rates, time deposits had lower opportunity costs and the m onetary service flows from a given quantity of time deposits naturally fell. Thus, despite the large increase in time deposits, the expenditure share of the m onetary service flows from tim e deposits ac tually declined during this period (see S^in figure 6), as did the grow th of Divisia M2 (see figure 4). T h e S e c o n d P e r io d : J a n u a r y 1 9 8 2 to N o v e m b e r 1 9 8 7 During this period, the rate of inflation de clined to low er levels and an extended period of SEPTEMBER/OCTOBER 1991 28 expansionary grow th in the m onetary base began in Jan uary 1982. The term stru ctu re of Swiss interest rates resum ed its norm al shape, w ith short-term in terest rates below long-term interest rates. Figure 5 shows that the user costs of currency, demand deposits and demand deposits with the giro system (U„ U2, U3) began to fall in 1982. This reflects the fact that the difference betw een their interest rates and the benchm ark rate was declining, while the user cost of time deposits (U4) increased as its in terest rate fell relative to the benchm ark rate. As noted previously, the contributions o f each com ponent to the Divisia and simple-sum M2 ag gregates are determ ined by their shares. W e on ly display the quantity and expenditure shares fo r time deposits (S * and S4) in figure 6 fo r illus tration. In 1982, the quantity share of time de posits (S4) fell substantially, while its expenditure share (S*) rose sharply. This resulted in the pos itive grow th of Divisia M2 and negative grow th of simple-sum M2 shown in figure 4 for 1982. Divergent movem ents in Divisia and simple-sum M2 occu rred again in 1985 w hen the simple-sum M2 growth was positive, while Divisia M2 growth w as almost zero. During the rest of this period, Divisia and simple-sum M2 moved similarly. These th ree episodes suggest that, at least for a m onetary aggregate as broad as Swiss M2, dif feren t aggregation procedures produce m one tary m easures that can move quite differently and generate very distinct interpretations for the stance of policy and the likely cou rse of econom ic conditions. T h erefore, it is im portant to know which o f the potential broader m one tary aggregate m easures are m ore closely re lated to key econom ic conditions. COMPARISON O F TH E P E R F O R MANCE OF TH E DIVISIA AND SIMPLE-SUM AGGREGATES To determ ine w hether a m onetary aggregate can be used as a m onetary policy target, two questions m ust be answered. First, is th ere a satisfactory relationship betw een the m onetary aggregate and some key econom ic variable, such as inflation or nominal GDP or GNP? Second, is the m onetary aggregate strongly related to something that is directly controllable by the m onetary authority? W e exam ine both these questions in this section. In fla tio n a n d M o n eta ry A g g re g a te s T h e T h ird P e rio d : D e c e m b e r 1 9 8 7 to D e c e m b e r 1 9 8 9 Swiss inflation rose from 2 percent throughout most of 1988 to nearly 5 percent by the end of 1989. W hile both short- and long-term interest ra te s ro se over this period, sh o rt-term rates rose relative to long-term rates. M oreover, the m ajor institutional changes that took place re duced demand fo r the m onetary b a se.11 In r e sponse to these events, u ser costs of the first two com ponents (U, and U2) rose, while the user cost of tim e deposits (U4) fell (figure 5); these movements w ere similar to those in the first period. However, as m entioned earlier, the user cost of demand deposits in the postal giro sys tem (U3) fell in May 1989 w hen the interest rate jumped from zero to two percent. Consequent ly, the Divisia and simple-sum M2 m easures moved in opposite directions; simple-sum M2 rose sharply, while Divisia M2 fell substantially (figure 4). "S ee the data description in appendix 2. ,2See Batten and Thornton (1983); Thornton and Batten (1985). Digitized for FEDERAL FRASER RESERVE BANK OF ST. LOUIS To evaluate the first question, the relationship betw een selected Swiss m onetary aggregates and inflation are compared. Specifically, quarterly Swiss inflation rates w ere regressed on distri buted lags of selected Divisia and non-Divisia m onetary aggregates. Because the sample is rel atively small and because we would like to in clude enough lags to capture the significant ef fect of money grow th on inflation, the Polynom ial Distributed Lag (PDL) estimation technique was used.12 Ideally, it is desirable to use one of the com monly used lag-length selection m ethods for choosing both the lag length and the degree of the polynominal. However, for two of the m one tary aggregates, simple-sum M l and Divisia M l, the equations exhibited significant serial correla tion. This com plicates the application o f these procedures fo r these aggregates. Because of this, w hen these aggregates w ere used, several speci fications of both lag length and polynominal 29 Table 1 PPL Regressions of Inflation on Money Growth Growth Rate of Growth Rate of Simple-sum Divisia M1 M2 M1 M2 PDL order 1 1 Lag 1 1 Coefficients of Distributed Lags 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 0.00898 0.01232 0.01566 0.01900 0.02233 0.02567 0.02901 0.03235 0.03569 0.03903 0.04237 0.04571 0.04905 0.05238 0.05572 0.05906 0.06240 0.06574 0.06908 0.02740 0.02721 0.02702 0.02683 0.02664 0.02646 0.02627 0.02608 0.02589 0.02570 0.02552 0.02533 0.02514 0.02495 0.02476 0.02458 0.02439 0.02420 0.02401 0.00917 0.01257 0.01597 0.01938 0.02278 0.02618 0.02958 0.03298 0.03639 0.03979 0.04319 0.04659 0.04999 0.05339 0.05680 0.06020 0.06360 0.06700 0.07040 -0.01316 -0.00653 0.00010 0.00672 0.01335 0.01998 0.02661 0.03323 0.03986 0.04649 0.05311 0.05974 0.06637 0.07299 0.07962 0.08625 0.09287 0.09950 0.10613 Sum t-Stat1 0.74154 (3.1168) 0.48838 (8.1203) 0.75596 (3.1877) 0.88322 (6.0924) AR(1) S.E. Adj-R2 D-W 0.4673 0.00697 0.5751 1.9811 0.0763 0.00611 0.6729 1.9409 0.4596 0.00695 0.5776 1.9773 0.2253 0.00625 0.6579 1.9249 S.E. is the standard error of the regression. 1 The critical t-statistic value at the 5 percent significance level is 1.645. degree w ere estimated. Specifications with rela tively long lags and relatively low polynominal degrees produced the highest adjusted R-square. To keep the specifications comparable with those o f simple-sum and Divisia M2 (which w ere cho sen using the FPE criteria), results with lag lengths of 18 and first-degree polynominals are presented.13 To com pare the long-run relationship of the m onetary aggregates on inflation, we estimated the selected PDL models and computed the sum of coefficients of the distributed lags to test w hether the sum of the lagged money growth coefficients was significantly different from z ero . The estim ated coefficients and respective statistics are shown in table 1. The results in table 1 show that m onetary ag gregates influenced Swiss inflation over periods o f up to fou r or m ore years. W ith the exception o f simple-sum M2, the m onetary aggregates have roughly similar values for the sum of the coefficients on th eir distributed lags. T he hypo thesis that the sum of the lag coefficients is 13The identical polynominal degree was obtained for M2 and Divisia M2 by the Pagano-Hartley technique if a T-statistic of 2.0 is used for the critical value. SEPTEMBER/OCTOBER 1991 30 Table 2 F-Statistics for PDL Models1 Growth Divisia M2 Divisia M2 and rates and M1 Divisia M1 M2 and M1 M2 and Divisia M1 F-Stat. H, h2 1.8349 7.3115 1.8349 7.1715 0.0605 4.0645 0.0726 3.9585 'At the 5 percent significance level, the critical F value is 3.29. zero was rejected in all cases at the 5 percent significance level.14 Because M l and Divisia M l are included in th eir respective M2 counterparts, we investi gated w h eth er the non-M l com ponents of M2 them selves added significant explanatory pow er in the inflation equation. Thus, we defined the u n restricted model as regressions of inflation on both the M2 and M l PDLs; the restricted models w ere those with regressions of inflation on the M2 or M l PDLs, respectively. The actual F-statistics and their 5 percent significance level critical values are shown in table 2. H, is the hypothesis that inflation can be explained by Divisia M2 or M2 alone; H2 is the hypothesis that inflation can be explained by M l aggregates alone. These hypotheses are tested against the corresponding u nrestricted PDL model that in flation is explained jointly by simple-sum and Divisia M2 PDL an d the M l aggregates PDL.15 The results in table 2 show that the data failed to reject H„ but did reject H2. This result sug gests that the broad er aggregates M2 and Divisia M2 b etter explain Swiss inflation than does M l or Divisia M l alone. C ontrollability As noted earlier, the practical use of a m one tary aggregate as an intermediate target depends on its controllability. Even if some m onetary ag gregate shares a close relationship with inflation 14The hypothesis that the sum of the lag coefficients was unity failed to be rejected in all cases except for simplesum M2. Thus, except for simple-sum M2, the results did not reject a one-to-one long-run relationship between the growth of the monetary aggregate and the rate of inflation. The results for the monetary base are similar to those of the Swiss monetary aggregates; its adjusted R2 (0.4880) was lower than those displayed in table 1. As shown in the first part of this paper, the relationship between inflation FEDERAL RESERVE BANK OF ST. LOUIS or nominal GDP, it would be of little use as a m onetary target if its grow th can not be con trolled by m onetary authorities. Since the cen tral bank controls the m onetary base, the rela tionships between it and the broader Swiss mone tary aggregates are examined. Because the cross correlations betw een the grow th rates of m onetary aggregates and the grow th rate of the m onetary base show a longlag pattern, we used the PDL models to estimate their relationships. Again, because of significant serial correlation, various specifications of lag length and polynominal degree w ere estimated. However, they share the same qualitative pro perties. Table 3 displays the estim ates of the 16-lag and first-degree PDL models. Results show that the grow th rates of M l, Divisia M l and Divisia M2 are statistically signif icant in relation to the grow th rate of the m onetary base. However, a significant long-term relationship betw een the grow th rate of the m onetary base and simple-sum M2 is rejected at the 5 percent level of significance. In addition, the contem poraneous grow th rate of the m onetary base is positively and signifi cantly correlated to the grow th rates of both M l aggregates and the Divisia M2 aggregate. For simple-sum M2 grow th, how ever, the con tem poraneous and initial lagged grow th rates of the m onetary base have negative effects on its and growth of the monetary base slipped considerably dur ing 1986-89 (the end of the period examined); earlier, however, there had been a close link between Swiss infla tion and long-run growth in the monetary base. Because the estimation period covers 1975-89, the recent "breakdown” in the monetary base growth-inflation rela tionship does not dominate the results. 15For both M2 and M1 aggregates, we take the same number of lags, 18, in the PDL models. 31 Table 3 PDL Regressions of Growth Rates for Monetary Aggregates on Growth Rate of Monetary Base____________________________ Growth Rate of Growth Rate of Simple-sum Divisia M1 M2 M1 M2 PDL order 1 1 1 1 Lag Coefficients of Distributed Lag for the Growth Rate of the Monetary Base 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 0.39216 0.36648 0.34079 0.31511 0.28942 0.26374 0.23805 0.21237 0.18668 0.16100 0.13531 0.10963 0.08394 0.05826 0.03257 0.00689 -0.01880 -.07127 -.05610 - .04092 - .02574 -0.01057 0.00461 0.01979 0.03496 0.05014 0.06532 0.08049 0.09567 0.11085 0.12602 0.14120 0.15637 0.17155 0.38461 0.35937 0.33414 0.30890 0.28367 0.25844 0.23320 0.20797 0.18273 0.15750 0.13227 0.10703 0.08180 0.05656 0.03133 0.00610 -0.01914 0.20584 0.19467 0.18351 0.17235 0.16118 0.15002 0.13886 0.12769 0.11653 0.10537 0.09420 0.08304 0.07188 0.06071 0.04955 0.03839 0.02722 Sum t-Stat 3.17360 (2.3518) 0.85237 (0.5770) 3.10648 (2.3538) 1.98099 (3.3068) AR(1) S.E. Adj-R2 D-W 0.57811 0.01852 0.42888 1.97733 0.55397 0.02317 0.48930 1.98227 0.57459 0.01821 0.42632 1.98631 0.40863 0.01137 0.40400 2.12645 S.E. is the standard error of the regression. The critical t-statistic for the 5 percent significance level is 1.645. growth. Indeed, significant positive correlation shows up only th ree years after the changes in the m onetary base. broader m onetary aggregates as m onetary policy targets. CONCLUSION This paper examined the potential usefulness of Swiss M l and M2 m onetary aggregates com pared with Swiss Divisia M l and M2 aggregates derived from econom ic aggregation theory. We showed that M l and Divisia M l generally dis played similar movements over time and w ere related similarly both to Swiss inflation (which justifies their potential usefulness as a target) and to the m onetary base (which means that their grow th was potentially controllable by the Swiss National Bank). The relationship betw een the Swiss m onetary base and inflation in Sw itzerland has becom e more uncertain in recen t years. This phenom e non has generated considerable interest in using M2 and Divisia M2, however, displayed sub stantially different behavior over time and, at certain key times, yielded substantially different signals about the stance of m onetary policy. These results suggest that the Swiss National Bank can significantly influence the grow th of M l, Divisia M l and Divisia M2 through changes in the grow th of the m onetary base. Long-term simple-sum M2 grow th, how ever, does not ap pear to be influenced by grow th of the m one tary base. SEPTEMBER/OCTOBER 1991 32 More im portantly, M2 grow th was statistically unrelated to the grow th of the m onetary base. The results suggest that M l, Divisia M l and Divisia M2 would be suitable fo r fu rth er study if the Swiss National Bank is interested in the possibility of using broader m onetary aggre gates to replace m onetary base targeting. How ever, these results indicate that M2 is unlikely to provide an adequate substitute fo r m one tary policy purposes. REFEREN CES Barnett, William A. “ The User Cost of Money,” Economics Letters, no. 2, 1978, pp. 145-49. _______ . “ Economic Monetary Aggregates: An Application of Index Number and Aggregation Theory,” Journal of Econometrics: Annals of Applied Econometrics 1980-3, supplement to the Journal of Econometrics (September 1980), pp. 11-48. Barnett, William A., and Paul A. Spindt. “ Divisia Monetary Aggregates: Compilation, Data, and Historical Behavior,” Staff Study no. 116 (Board of Governors of the Federal Reserve System, May 1982). Batten, Dallas S., and Daniel L. Thornton. “ Polynomial Distributed Lags and the Estimation of the St. Louis Equation,” this Review (April 1983), pp. 13-25. Belongia, Michael T., and James A. Chalfant. “ The Changing Empirical Definition of Money: Some Estimates from a Model of the Demand for Money Substitutes,” Journal of Political Economy (April 1989), pp. 387-97. Diewert, W. E. “ Exact and Superlative Index Numbers,” Journal of Econometrics (May 1976), pp. 115-45. Donovan, Donal J. “Modeling the Demand for Liquid Assets: An Application to Canada,” International Monetary Fund Staff Papers, Vol. 25, no. 4, (December 1978), pp. 676-704. Farr, Helen T., and Deborah Johnson. “ Revisions in the Monetary Services (Divisia) Indexes of the Monetary Ag gregates,” Staff Study no. 147 (Board of Governors of the Federal Reserve System, December 1985). Friedman, Milton, and Anna J. Schwartz. “Monetary Statistics of the United States: Estimates, Sources, Methods,” New York, Columbia University Press, 1970. Hicks, John Richard. Value and Capital, 2nd ed. (Oxford, Eng.: Clarendon Press, 1946). Rich, Georg. “Swiss and United States Monetary Policy: Has Monetarism Failed?” Federal Reserve Bank of Richmond Economic Review (May/June 1987), pp. 3-16. ________“Geldmengenziele und schweizerische Geldpolitik: eine Standortbestimmung. In: Schweizerische Nationalbank: Geld, Wahrung und Konjunktur,” Quartalsheft No. 4 (Dezember 1989), pp. 345-60. Shiller, Robert J. “The Volatility of Long-Term Interest Rates and Expectations Models of the Term Structure,” Journal of Political Economy, (December 1979), pp. 1190-219. Schweizerische Nationalbank. Revision der Geldmengenstatistik. In: Geld, Wahrung und Konjunktur. Quartalsheft No. 1, Marz 1985. Thornton, Daniel L. and Dallas S. Batten. “ Lag-Length Selection and Tests of Granger Causality between Money and Income,” Journal of Money, Credit and Banking, (May 1985), pp. 164-78. Appendix 1 Divisia Indexes T h ere are two types of Divisia index numbers: the continuous-time version and the discrete time version. Continuous-time Divisia index numbers are derived from microeconomic theory; discrete-time Divisia index numbers are approxi mations of the continuous-time version. To understand how continuous-time Divisia in dex numbers are derived, consider the case where economic agents want a measure that aggregates a group of n commodities in the economy. The quantities of the goods are expressed by the vec tor q=(q„ q,, ... , qn); their corresponding prices are denoted by the vector p = (p„ p2,..., pn). Economic aggregation theory states that the aggregator function is a utility function g(q) to be maximized subject to the budget constraint, first-order necessary condition for utility maximi zation is (2) dg(q)/dq = A pi( where A is the Lagrange multiplier. Because the aggregator function is linear homogeneous, Eul er’s equation is satisfied such that (3) Z (dg/dq)q = g(q). i~l Substituting equation 2 into equation 3 yields Ai=11 q.Pi = g<q) and n A E = g(q). Hence, A = g(q)/E and (i) 1£=1 qiPi = g(q)flp) = E, (4) dg(q)/dq = Pig(q)/E. where f(p) is the price aggregator function and E is the total expenditure on the specific goods. The Taking the total differential of the aggregator function g(q) yields FEDERAL RESERVE BANK OF ST. LOUIS 33 (5) dg(q) = I (dg/dq,)dq, i=l Thus, substituting equation 4 into 5 yields time, the continuous-time Divisia index must be transform ed into a discrete-time version to make it useful. The discrete time approximation of equa tion 7 is n (6) dg(q) = 1 pdqg(q)/E and i=l n dg(q)/g(q) = I (9) ln(g(t))-In(g(t-l)) = I S'flnlqlt)) - ln(q(t-l))] i=l pqdq/Eq, 1=1 If we set Sj = pq/E and call it the i-th good’s value share in the total expenditure, equation 6 can be transform ed into (10) g(t) = g(t-l)EXp( I S^n(q(t)) - ln(q(t-l))]l ' i=l ' (7) d(ln(g(q))) = S Sjddnfq,)). i=l where S*t = (Sit + Sit.,)/2, Solving equation 7 for g(t) yields n Sit = pq/E = p.q/ ( I 1= 1 (8) g(t) = EXP/ ( £ S,(t)clfln(qi(t))l) dt. xi=l ' Equation 8 is the continuous-time Divisia index. Because economic variables are observed and measured in discrete time rather than continuous p ,q ), and S*t is the average expenditure share in the two adjacent time periods. Equations 9 and 10 are the discrete-time Divisia index equations used in calculating the Swiss Divisia M l and M2 monetary aggregates. Appendix 2 Data To calculate the Swiss Divisia M l and M2 mone tary services indexes, we used the seasonally adjusted monthly Swiss M l and M2 series and their components consistent with the definitions established in 1975 and incorporating the revision that occurred in 1985 (for m ore details, see Schw eizerische Nationalbank, 1985). The monetary aggregates consist of the fol lowing assets held by individuals and non-bank institutions: M l: Currency in circulation (C) Demand deposits with banks (DB) Demand deposits with the postal giro system (DP) M2: M l plus time-deposits (TD) MB: Seasonally adjusted monetary base, defined as the sum of banks reserves and banks notes in circulation. In terest R ates To compute the user costs of monetary assets, we need the assets’ own rates of return, the benchmark rate of return and the cost-of-living index. Own R ates c Zero DB DP 0.25 percent 1975:06 - 1989:04: Zero 1989:05 - 1989:12: 2 percent TD three-month rate on time-deposits with large banks (monthly average) Cost-of-living In d ex Monthly Swiss Consumer Price Index (CPI) calculated by the Federal Statistical Office (Bundesamt fuer Statistik). B e n c h m a rk R ate The highest rate in each period from the fol lowing interest rates: the secondary market yield on cantonal bonds, interest rates on cash cer tificates with the cantonal or large banks and short-term Euromarket-Swiss franc interest rates. Short-term rates became the benchm ark rates during 1979:12 - 1982:04 and 1988:12 - 1989:12, when the Swiss yield curve was inverted. SEPTEMBER/OCTOBER 1991 34 Steven Russell Steven Russell is an economist at the Federal Reserve Bank of St. Louis. Lynn Dietrich provided research assistance. The U. S. Currency System: A Historical Perspective i HE USE OF CURRENCY in transactions is a regular part of our daily lives and a basic feature o f o u r eco n o m ic system . T h e im p o rta n ce of cu rren cy derives both from its obvious role in daily transactions and from the somewhat m ore subtle role of the cu rren cy sy stem as the basis for our m onetary and financial systems. The curren cy system is so fundam ental to econom ic activity that we tend to give it little thought. Few of us would have an easy tim e imagining w hat alternative system s m igh t be like o r w hy th e y m ight be desirable. Indeed, it seems likely that most of us, if pressed, would offer the opinion that the present cu rren cy system is the only one that is feasible—or at least, the only one that is desirable. This article has th ree purposes. T he first is to define the term “cu rren cy ” and explain the special im portance of cu rren cy and the cu rren cy system to our economy. The second is to describe the U.S. cu rren cy system —the system that governs the form s, uses and roles of curren cy in the m odern United States. This description will be preceded by a catalog of the form s curren cy has taken at vari ous points in the past, so that the m odern U.S. system em erges as a set of selections from a menu of choices provided by history. This procedure is intended to suggest that alternative m enu selec tions w ere possible—that the curren cy system which actually evolved in the United States is not the only one that could have evolved. The article's third and most ambitious purpose is to present a b rief but com prehensive account of the historical 1U.S. monetary history from the end of the Civil War through modern times has been chronicled quite extensively, notably by Friedman and Schwartz (1963). In addition, most of the key FEDERAL RESERVE BANK OF ST. LOUIS development of the U.S. cu rren cy system. This account focuses on the period b efore and during the Civil W ar.1 Its prim ary goal is to provide the reader w ith historical context that may improve his understanding of the m odern cu rren cy system. The historical account has a second purpose, however. The development of the U.S. curren cy system is often characterized as a process of slow but steady advancement: older institutions and practices, having failed to m eet the demands of th e ir tim es, w e re re p la ced by m o re e ffic ie n t successors. This "gradual progress” ch aracteriza tion implies that the m odern cu rren cy system m eets the needs of our econom y m ore efficiently than could any of the alternatives suggested by history. The historical account is intended to help determ ine w h eth er this ch aracterization is valid, and w h e th e r rela tiv e e ffic ie n c y co n clu sio n s should be based on it. WHAT IS CURRENCY? One approach to defining cu rren cy is to contrast it w ith so m eth ing w h ose d efin itio n is clo sely related, but m ore fam iliar: money. Most people have been exposed at some point to an econom ist’s definition of money; it usually reads something like “things that serve as media of exchange” or "things that function as means of paym ent.” While all cu rren cy is money, all money is not currency. Currency can be defined as m oney which circu- decisions that determined the basic form of the U.S. currency system were arguably made before 1865. 35 lates, or passes from hand to hand. ("Circulation” w as on ce com m only used as a synonym fo r currency.) Formally, a type of money can be said to circu late if it usually p asses in e x ch a n g e fro m one person to another w ithout third-party verification. One easy way to illustrate the d ifference betw een circu la tin g and n o n -circu la tin g m on ey is to contrast dollar bills, w hich circulate, w ith checks, which do not.2 A dollar bill may pass from one person to another many tim es in d ifferent tran sac tions. The only people involved in each transaction are th e b u y e r and se ller. T ra n sa c tio n s using checks require m ore complex arrangem ents. It is unusual for a check, w ritten by one person in paym ent to another, to be offered in paym ent to a third person. Instead, the second person usually deposits the check in a bank account. His bank and the first person’s bank then conduct a “clearing” transaction w hich, if successfully completed, vali dates the paym ent.3 example, commodities such as wampum (colored beads), tobacco, w heat and rice w ere used as curren cy at different places and tim es.4 Gold and silver, th e “p re cio u s m e ta ls ,” had a ttra ctiv e p ro p e rtie s — p o rta b ility , m alleab ility and d u r ability—w hich ultimately made them the cu rren cies of choice in most early economies. Coin C u r re n c y As the volume of transactions involving gold and silver increased, people began to divide these metals into pieces of readily recognizable size and shape, called coins. T he earliest coin-producing facilities (mints) seem to have been privately oper ated.5 In m ost countries, how ever, the govern m ent eventually took over coin production. Commodity C urrency The rationale behind the governm ent takeover may well have included the belief that govern ment-issued coins would be m ore uniform , and m ore reliable, than their privately issued cou n ter parts. Early governm ents, however, could have reso lv ed p ro b lem s o f d iv ersity and fra u d by regulating private mints and inspecting private coins, in essentially the same way that govern ments have long regulated and inspected other industries.6 A m ore compelling reason fo r govern m ent coin monopolies, how ever, was the desire to earn revenue from seign iorage—from periodically shortweighting or debasing the cu rren cy .7 Unless a government had a coinage monopoly, its attempts to earn substantial revenues from seigniorage would have been frustrated as the public aban doned its coins in favor of those minted by its private competitors. The earliest form s of cu rren cy w ere c o m m o d i ties (widely traded goods). In colonial America, for T h e p re v a len ce o f g ov ern m en t c u rre n c y monopolies gave rise to the twin concepts of a In the m odern United States, only dollar bills and coin s, issu ed by a g en cies o f th e fe d e ra l governm ent, fit the definition of currency. Earlier in our history (and that of many other nations) the n u m b er of a lte rn a tiv e typ es o f c u r re n c y w as la rg e r, and in clu d ed item s issu ed by p riv ate organizations. The next section presents a brief catalog of some of the varieties of cu rren cy that have existed in the past. WHAT FORMS CAN CURRENCY TAKE? 2Strictly speaking, economists think of the accounts against which checks are drawn (the demand deposits) as money, rather than the checks themselves. 3Typically, a person who is offered a newly written check in payment (the second party) will ask the check-writer (the first party) to present identification and will record information from the identification presented. The second party will deposit the check in his bank account. His bank will “ clear” the check by sending it to the bank against which it is drawn, and demanding payment in cash. The two banks are the “ third parties” which are actually involved in most transac tions using checks. The clearing transaction is necessary to verify that the check is drawn on an account that contains sufficient funds. If the check “ bounces,” it has failed the verification test. The amount of the check will not be credited to the second party’s account, and he will use the information from the check-writer’s ID to pursue him for some alternative form of payment. [The reason the second party will rarely try to pass the check along to a third party is that the third party is unlikely to accept it. (If you doubt this, try passing such a “ third-party check” at your local grocery store.) A third party typically has no easy way of obtaining reliable identification from the (absent) first party.] 4For an extended discussion of the role of commodity money in the colonies, see Nettels(1934), chapter VIII. 5Feavearyear (1963) describes early English currency as follows: “ At the beginning of the eighth century the currency consisted of small silver coins varying in design according to the fancy of the individual moneyer.” (p. 7) 6Adam Smith (1776/1937) points out that before coins evolved, governments often stamped ingots of precious metal to certify their purity (pp. 24-25). 7A coin is said to have been “ shortweighted” if it is minted with less than its official metallic weight, but represented as having exactly that weight. A coin is said to have been “ debased” if it is minted as a mixture of genuine monetary metal and common scrap metal, but represented as pure monetary metal. These fraudulent practices were sometimes practiced by private mints as well. For a discussion of govern ment seigniorage motives, see Timberlake (1991), pp. 3-5, 50-51. SEPTEMBER/OCTOBER 1991 36 national cu rren cy and a national m onetary unit. Typically, a governm ent would define a basic m onetary unit as a fixed quantity of gold or silver. It would then mint coins in d en om in ation s that w ere multiples or fractions of this unit and w ere scaled appropriately in size and weight. Most nations had an extended period during w hich governm ent-issued coins w ere the only form of currency. One problem with these purecoin cu rren cy systems was that they had difficulty handling transactions of widely differing scales. If, for example, coins w ere denominated so that a single coin of m oderate w eight could be used to purchase an inexpensive item (say, an apple), then the coins necessary to purchase an expensive item (say, a carriage) w ere necessarily quite heavy. One com m on way in w hich governm ents tried to solve this problem was by establishing bim etallic coinage system s. In th ese sy stem s coin s o f low value contained a relatively inexpensive m etal (typically silver), while larger-value coins w ere com posed of a m ore expensive m etal (typically gold). The two types of coins w ere referred to collectively as sp ecie. The U.S. experience w ith specie cu rren cy illus trates most of the concepts just described. The U.S. Constitution gave Congress the pow er to “coin money, and regulate the value th ereo f”—a provision w hich has been universally interpreted as prohibiting the states eith er from minting coins directly or from authorizing private parties to do so.8 Shortly after the Constitution was ratified, Congress enacted legislation that defined the basic m onetary unit, the dollar, as either a fixed weight of gold or a (different) fixed weight of silver. The fed era l g o v ern m en t th e n opened a m int th at produced dollar coins in accordance w ith these definitions. The mint also produced silver “quar te rs” containing one-fourth the amount of silver in a silver dollar, five-dollar gold pieces containing five times the amount of gold in a gold dollar, and so on.9 The U.S. Mint continued to produce full-bodied gold coins until the early 1930s, and full-bodied silver coins until the mid-1960s. (A full-bodied coin con tains a quantity of m etal whose m arket value is equal to the face value o f the coin.) 8U.S. Constitution, Article I, Section 8. 9See Huntington and Mawhinney (1910), pp. 474-79. Gold dollar coins were not actually minted until 1849. See Carothers (1930), pp. 105,109, and Huntington and Mawhinney (1910), pp. 508-09. FEDERAL RESERVE BANK OF ST. LOUIS As the magnitude o f econom ic activity increased, the weight of the gold coins necessary for a m ajor purchase, or even the quantity that a relatively wealthy person might desire to have on hand, becam e unm anageably large. Coins also tended to w ear away or have their edges clipped. A fter a few years, coins of the same denomination could be significantly different in size.10 These problem s made coins increasingly unsatisfactory, even for relatively small-scale transactions. Bills o f E x c h a n g e An obvious solution to the “weight problem s” of the coin cu rren cy system was to find or create lightweight objects that, while not made of coins themselves, had know n values in term s of coins. Objects like this already existed: they w ere p ro m is so ry n otes—contracts betw een b orrow ers and lenders calling for the repaym ent of fixed sums (in coin) at fixed future dates. One special type of prom issory note, the bill o f exchange, was readily adapted fo r use as currency. Bills of exchange grew out of com m ercial tran sac tio n s in w h ich m e rc h a n ts w ou ld a rra n g e to purchase goods from other m erchants for delivery at fixed future dates (for example, in 90 days). Often the seller could not afford to produce and/or deliver the goods unless he received immediate payment, while the buyer was reluctant to pay for the goods b efore receiving delivery. One solution to this problem was an exchange of contracts. The seller would con tract to deliver the goods at the date in question, while the buyer would con tract to pay the purchase price at the delivery date. The latter con tract took the form of a conventional prom issory note. This exchange of contracts may not seem to have addressed the seller’s immediate problem : to obtain the cu rren cy needed to finance the produc tio n and/or tra n s p o rt o f his goods. Suppose, however, that the seller, armed w ith his prom is sory note, sought to purchase m aterials from a supplier. He could then w rite out another credit instrum ent—a bill of exchange—calling on the m erchant who had issued the prom issory note to pay th e su p p lier th e p u rch a se p rice o f th e m a te ria ls, plus an allo w an ce fo r in te re s t, in 90 10For a description of the clipping problem in pre-eighteenth century England, see Feavearyear (1963), pp. 5-6, and Macaulay (1877), volume V, pp. 85-93. 37 days. This process was called draw ing a bill; the original goods seller was called the d ra w er and the issuer of the prom issory note the d raw ee. The drawee would accep t (agree to cover) the bill as long as its value was less than that of the prom is sory note. He would indicate his acceptance by endorsing the bill.11 By accepting the bill, the supplier was, in effect, lending the seller the value of the m aterials the latter had “purchased.” The supplier, how ever, usually did not expect to hold the bill until it came due. Instead, he planned to pass it along to some one from whom he wished to purchase goods; this person might pass it along to someone else, and so on, until the bill m atured. The last person in the chain would demand payment from the drawee. In betw een, the bill served as paper cu rren cy .12 Notice that drawing a bill was analogous to w riting a check, with the draw ee of the bill playing the same role as that of a bank on which a check is drawn. It seems to follow that bills of exchange should not have circulated, for precisely the same reasons that m odern checks do not circulate. T h ere was a basic d ifference betw een a bill o f exchange and a m odern personal check, however. Because no one would accept a bill unless it was endorsed by the relevant drawee, the question of "bad ch ecks”—checks w ritten by individuals with insufficient funds—did not arise. Stated differently, an accepted bill was purely a liability of the drawee; a person offered a bill in paym ent did not need to be concerned about the creditw orthiness of the d raw er.13 It is true that bills w ere occasionally d ish o n o red by their drawees, just as m odern banks occasion ally fail.14 As long as the bills w ere drawn against well-known m erchants with established reputa tions in com m erce, how ever, failures w ere un common. Consequently a person who accepted a bill in paym ent could be reasonably confident 11See Clough and Cole (1941), pp. 77-78. 12ln England, bills of exchange played a prominent role as means of payment during the seventeenth and eighteenth centuries; see Feavearyear(1963), pp. 160-62. In several English districts, they retained this role well into the nineteenth century; see Clapham (1944), II, pp. 90-91, 97-98, Viner(1937), p. 123, and Feavearyear(1963), p. 165. 13lt is worth noting that while it is usually difficult to negotiate a third-party check drawn on an individual's bank account, a check drawn on the account of a government agency or prominent local corporation may be easy to negotiate. that, if he did not have occasion to pass it along, he could redeem it w hen it came due.15 Although bills of exchange becam e an im portant adjunct to coin currency, a num ber of problems limited their usefulness. Since they w ere typically drawn in fairly large denominations (of the sort appropriate fo r trade betw een m erchants), they w ere not well suited fo r small-scale transactions. And, as the volume of trade in a given region increased, it becam e less and less likely that a person proffered a bill would be familiar, either personally or by reputation, with the m erchant against whom it was drawn. Consequently, disho nored bills becam e a m ore serious problem , and people becam e hesitant to accept them in payment. A less fundamental, but still annoying, problem was that w henever a bill changed hands, interest had to be calculated and deducted from its face value. This fairly involved calculation required consideration of both the rem aining term on the bill and the m arket rate of interest. B a n k N otes The transactions problem s w ith bills of exchange created opportunities for private entrepreneurs to profit by providing paper cu rren cy in m ore con venient form s. Suppose an enterprising m erchant with a good reputation sold small bills of exchange in retu rn fo r specie and used the proceeds to buy large bills with the same m aturity dates. The proceeds of the large bills would then provide a fund out of w hich the small bills could be redeemed. Because small bills w ere m uch m ore convenient for exchange purposes than large bills, they w ere slightly m ore valuable, per dollar of face value, to their holders.16 As a result, small bills could be sold at smaller percentage discounts (lower interest rates) than large bills. It followed that the total purchase price of the large bills necessary to cover a given face value of small bills was smaller than the total sale price of the small bills. This d ifference in total prices represented the m erch an t’s profits. 14The existence of federal deposit insurance prevents smallscale personal depositors from being endangered by bank failures. Before 1935, however, this was not true. 1Contemporary criminals sometimes forged merchants’ acceptances, just as modern criminals sometimes forge checks. The severity of penalties for forgery limited the scale of this problem, however. 16For a careful description of the logic behind this statement, see Wallace (1983). SEPTEMBER/OCTOBER 1991 38 The m erchant had now becom e a ban ker, and the institution he operated a b a n k o f issu e—a financial interm ediary whose liabilities consisted prim arily of paper cu rren cy .17 The small bills cam e to be know n as b a n k notes. The m erchant could increase his profits from note issue by reducing the risk that he would default on his notes. This would reduce the “risk prem ium ” that small billholders demanded, and enable him to sell the bills at smaller discounts. One strategy fo r accom plishing this was to diver sify h is la rg e b ill p o rtfo lio as ex ten siv ely as possible. A nother was to provide, or to obtain from investors, some capital to act as a cushion against defaults on the large bills. A basic problem w ith the schem e just described was that the tim e and effort necessary to compute th e a p p ro p ria te d iscou n t on a b an k n o te w as usually large relative to the face value of the note. This reduced the usefulness of notes in tran sac tions and discouraged people from purchasing them .18 One way to solve this problem was to issue notes with characteristics so appealing that their h o ld ers w ould b e w illing to fo rg o in te re s t on them . How could this be accomplished? Since we have assumed (perhaps too quickly) that proper diversification and capitalization made the risk on bank notes negligible, the need for interest on them arose purely out of their holders' tim e p r e fe r e n c e —th eir desire to be com pensated for giving up their money (in this case, their specie currency) for fixed periods. Suppose, however, that a m erchant prom ised to redeem his notes on d em a n d (at any time) instead of at a fixed future date. Since the pu rchaser of such notes could reclaim his specie w henever he chose, he would not be giving it up for any fixed period, and would have no reason to demand interest. The bills could then be sold at p a r (undiscounted). How could a m erchant make such a convertibility com m itm ent credible? Clearly, he would need to hold back some of the (specie) proceeds of his note sales fo r use as reserv es. These reserves would not have to be large, how ever, because as long as note holders w ere confident that they cou ld redeem 17For a description of banks that dealt in bills of exchange, see Feavearyear(1963), pp. 162-65. ,8White (1987) provides an analysis of the transactions problems associated with interest-bearing currency. 19Calomiris and Kahn (1991) construct a formal model in which note- or deposit-holders can use redemption demands as a device for preventing bank frauding by forcing a preemptive FEDERAL RESERVE BANK OF ST. LOUIS their notes, there was no particular reason why they w ould do so. A fter all, the holders had bought the notes because they w ere m ore convenient for exchange purposes than specie. Notice that th ere is some circularity in the argu ment just presented. Convertibility, it asserts, was necessary to prevent noteholders from demanding com pensation for giving up their specie, w hich they had been holding fo r use as money. But these people had exch an g ed th e ir sp ecie fo r n otes precisely because the notes w ere a m ore conven ient form of money! This paradox m akes it seem possible that convertibility is not really necessary; indeed, th e re are both historical and theoretical reasons for suspecting that it may not be. In p rac tice, how ever, the vast m ajority of private banks of issue have attached convertibility com mitments to their n otes.19 T he Bank of England, fo r example, began circulating convertible notes shortly after it received a royal ch arter in 1695. These notes b e cam e the principal paper cu rren cy of the relatively developed reg io n su rro u n d in g L ondon (the “M etropolis”).” 20 G o v ern m en t P a p e r C u r re n c y Governments eventually acquired a role in the paper cu rren cy system by regulating the issuance of private paper currency, and/or by issuing paper cu rren cy directly. The motives for this decision w ere essentially the same as those w hich drove governm ents to acquire a monopoly over coinage; some com bination of a desire to improve effi ciency by facilitating the development of uniform and reliable paper money, and a desire to earn revenue by regulating or replacing the private banking system. This revenue has been earned in a variety of ways. In some cases, governm ents have earned substantial sums by granting private institutions the right to issue paper cu rren cy in retu rn for some kind of financial consideration. (See, for example, the discussion of the establish m en t o f th e Bank o f England w h ich ap p ears below.) In oth er cases, revenue has been earned through direct cu rren cy seigniorage, in w hich the governm ent issues paper cu rren cy to purchase goods and se rv ice s, o r th ro u g h in d ire c t seig- liquidation. Incomplete information problems make it impos sible for these agents to detect fraud without a liquidation. While this model is not reasonable in every historical context, it represents a first step toward explaining the prevalence of convertibility. 20For the early history of the Bank of England, see Clapham (1944), volume I. 39 niorage, in w hich the governm ent issues paper cu rren cy to purchase and retire its own bonds. Government paper curren cy can take a variety of form s. The earliest form o f governm ent paper c u r r e n c y —and u ntil quite re c e n tly , th e m ost common form —was rep resen tativ e currency. A governm ent curren cy is said to be "rep resen ta tive” if it is issued under a convertibility com m it ment; that is, a governm ent prom ise to redeem the curren cy in specie, at par and on demand. Rep resentative curren cies are the government-issued an alog u es to p riv ate, c o n v e rtib le b an k n otes. W h ile they have u sually b een issued by g o v e rn m en t-o rg an ized " c e n tr a l b a n k s ,” th e y have sometimes been issued directly by the govern ment. The United States, for instance, had directly issued representative cu rren cy during 1879-1913 (the U.S. n otes, o r " g r e e n b a c k s ,” w h ich w ere issued by th e T re a su ry ) and re p re se n ta tiv e currency issued by a central bank during 1914-1933 (the Federal Reserve notes, w hich w ere issued by the Federal Reserve Banks).21 Governm ents have also issued curren cy that is not convertible into specie, or anything else. This type of cu rren cy is often referred to as fia t cu r rency.22 During the Civil W ar, both the Union and the Confederacy issued fiat cu rren cy to finance p a rt o f th e ir m ilitary p u rch a ses. T h e U nion cu rren cy was the greenback m entioned above. M odern U.S. c u r re n c y is also fia t in n a tu re. Federal Reserve notes (our dollar bills) have not been convertible for dom estic holders since 1933; since 1971, they have not been convertible fo r any holders w hatsoever. The Federal Reserve Banks 21The greenbacks were first issued in 1863, but were not convertible until 1879. Federal Reserve notes were convert ible for domestic holders from the establishment of the Federal Reserve System in 1914 until March 1933. They remained convertible for certain foreign holders until 1971. 22A distinction is sometimes made between inconvertible government currencies that are issued in purchase of assets (and so form the liabilities side of a “ balanced” balance sheet), and currencies which are issued in purchase of goods and/or services. Currencies of the former type are referred to as fiduciary. Many economists believe that currencies derive much of their value from the assets which back them. Descriptions of this view appear in Smith (1985b) and Russell (1989a). It suggests that fiduciary currencies may be less likely to decline in value (that is, to depreciate) than fiat currencies. 23Strictly speaking, the Federal Reserve System pays for the Treasury securities it purchases by issuing claims on the Federal Reserve Banks. These claims can be redeemed in currency—Federal Reserve notes—which can be held by the general public as cash balances, or by commercial banks as reserves. Alternatively, the claims can be converted into demand deposits at the Federal Reserve Banks, which can issue most of these notes in purchase of U.S. Treasury securities.23 WHAT IS A CURRENCY SYSTEM? A nation’s cu rren cy sy stem can be defined as the set of laws, conventions and practices that deter mine the form and role of cu rren cy in the nation’s economy. A com plete description of a nation’s cu rren cy system would provide answ ers to ques tions like: "W hat things does the econom y of this nation use as cu rren cy?”, "W hat sorts of institu tions (private and/or government) are perm itted to issue cu rren cy under the nation’s laws?”, "W hat role (if any) does the nation’s governm ent play in defining the econom y’s curren cy unit, or in p re serving its value?", and ’’W hat is the nature of the relationship betw een the nation’s cu rren cy system and its m onetary and financial system s?” HOW DOES TH E U.S. CURRENCY SYSTEM W O RK ? This section will provide a b rief summary of the history and legal fram ew ork of the U.S. cu rren cy system. It will focus on a pair of legal restrictions that play a critical role in shaping the system. These restrictions would be prim e candidates for revision or repeal if the system w ere to be refo rm ed or deregulated. As previously noted, the U.S. Constitution gave Congress exclusive pow er to define a national m onetary unit and produce coined currency. In addition, the states w ere explicitly prohibited from issuing paper cu rren cy directly.24 The also be used by commercial banks as reserves. The decision concerning how the claims are divided between these competing uses is made by the private sector. At present, currency held by the public, or as reserves, accounts for about 85 percent of total claims on the Federal Reserve Banks, while U.S. Treasury securities account for about 75 percent of their total assets. In addition, the economic implications of the scheme for paying for these securities just described are identical to those of an alterna tive scheme under which the System paid for Treasury securi ties with newly issued Federal Reserve notes, and the private sector decided how much of this currency to retain and how much to deposit with the Reserve Banks. The Federal Reserve Act prohibits the System from purchasing newly issued Treasury securities—an action that would amount to issuing currency (and/or Reserve Bank deposits) to finance government purchases. [See the defini tion of “ indirect currency seigniorage” presented earlier in this section.] 24Article I, Section 10 of the U .S. Constitution denies the states the power to “ emit Bills of Credit;” this was almost universally understood to prohibit them from issuing their own currency. SEPTEMBER/OCTOBER 1991 40 Constitution was silent, how ever, on two ques tions that ultimately becam e controversial: Does the federal governm ent have the right to issue paper currency? Do eith er the federal govern ment or the states have the right to authorize private institutions to issue paper cu rren cy—do they have the right, th at is, to grant ch arters to private banks? Shortly after the Constitution was ratified, the states began to ch a rte r private banks o f issue.25 In 1791, and again in 1816, the federal governm ent chartered a single private bank—the Bank of the United States. For the next three-quarters of a century, the bulk of the paper curren cy that circu lated in the U.S. was issued by state banks; virtu ally all of the rem ainder was issued by the United States Bank. The rights of the federal governm ent and the states to ch a rter private banks w ere even tually affirm ed (in separate decisions) by the U.S. Suprem e Court.26 In 1865, Congress imposed a tax on note issue by state banks that was high enough to make the activity unprofitable. This action, w hich cam e one year after Congress had established a system of fed era lly c h a rte r e d b a n k s o f issu e called the National Banking System, was evidently intended to put an end to state banking.27 A nother wartim e innovation was the issuance, beginning in 1862, of "green backs.” For the next 50 years, the U.S. stock of paper curren cy consisted almost entirely of national bank notes and greenbacks.28 The Civil W ar produced a dram atic expansion of the federal governm ent’s role in, and pow ers over, the U.S. m onetary system. In the years immedi ately following the w ar, the right of the federal governm ent to play this role, and to exercise these exp an d ed p o w ers, w as a ffirm ed in a se rie s of 25For an exhaustive list of banks chartered by the states prior to 1837, see Fenstermaker(1965). 26Key decisions upholding the right of the states to charter banks are Craig v. Missouri (1830) and Briscoe v. the Bank of the Commonwealth of Kentucky (1837). Key decisions regarding the right of the federal government to charter banks are McCulloch v. Maryland (1819) and Farmers and Mechanics Bank v. Dearing (1875). 27State banking survived because it proved possible for many state banks to convert to pure deposit banking, which was not taxed prohibitively. 28They were eventually supplemented by substantial quantities of silver certificates, and by minor quantities of federal govern Digitized forFEDERAL FRASER RESERVE BANK OF ST. LOUIS Supreme Court decisions. The w ar converted a political system in w hich the m onetary pow ers of federal governm ent w ere sharply circum scribed into one in w hich they w ere virtually unlimited. Almost half a century later, the federal govern m ent’s m onetary pow ers w ere wielded in dram atic fashion w hen Congress passed the Federal Beserve Act o f 1913. This legislation established 12 "Federal Beserve Banks” that collectively constitu ted a "cen tral bank” fo r the United States.29 The Beserve Banks issued a new form of representative paper cu rren cy called "Federal Beserve notes.” T h e se n o tes b ecam e th e b asis fo r th e U.S. c u r re n c y system . During the first two decades following the passage of the Federal Beserve Act, the national banks retained the right to issue limited quantities of notes. In 1935, however, the national banks’ issue rights expired; Congress declined to renew them, and made provisions fo r the gradual retire ment of all national bank notes still outstanding.30 Since 1935, the Federal Beserve Banks have been the only U.S. organizations authorized to issue paper cu rren cy on a regular basis.31 T h e re s u lt of th is h isto ric a l p ro ce ss can be sum marized as the first of two basic legal restric tions w hich govern the U.S. cu rren cy system: the federal governm ent has a legal m o n o p o ly over the issuance of currency, w hether in coin or paper form. The second basic legal restriction involves the relationship betw een currency, which is now exclusively federally issued, and "m oney” of other so rts, w h ich co n tin u e s to b e provided by th e p riv ate se cto r. P riv ately issu ed m on ey is r e quired to be convertible (redeem able at par and on ment currency of other sorts. See Friedman and Schwartz (1963), pp. 124-34, and Timberlake (1978), chapter 10. 29The text of the act appears in the Annual Report of the Federal Reserve System for 1914. 30See Friedman and Schwartz (1963), p. 442. 31ln principle, Congress retains the right to authorize the Treasury to issue paper currency directly. It has declined to do so, however, since the establishment of the Federal Reserve System. 41 demand) in governm ent cu rren cy .32 In practice, privately issued money consists of deposits at com m ercial banks and thrift institutions that are potentially convenient as media of exchange—i.e., that are readily transferable (checkable) and avail able in small denominations. The governm ent requires that these deposits be convertible, and they are referred to as "demand deposits.” If demand deposits are to be convertible into governm ent curren cy, they m ust be denominated in the same units, and have the same m arket value per unit, as governm ent currency. Consequently, the convertibility restriction, com bined with the governm ent’s cu rren cy monopoly, imposes a com m on denominational and value standard on all U.S. money. The denom inational standard is of course the "dollar,” the basic unit of governm ent curren cy; the value standard is the purchasing pow er of a dollar (or any fixed num ber of dollars) of this currency. It is w orth noting that governm ents usually attem p t to e n h a n ce th e a cce p ta b ility o f th e ir cu rren cy by making it legal tender. Legal tender laws either require or strongly encourage people to accept governm ent cu rren cy in paym ent of nom inal debts—debts denom inated in national cu rren cy units. In the United States, both coins minted by the Treasury and Federal Reserve notes are legal tender. HOW DID THE MODERN U.S. CURRENCY SYSTEM D EVELO P? Scien ce h a s b e e n slo w to adm it th e d iffe r e n t explan atory w orld o f h istory into its d om ain —an d ou r in te r p r e ta tio n s h a v e b e e n im p o v e r is h e d b y this o m is s io n . S c ie n c e h a s a ls o t e n d e d to d e n ig r a te history, w hen f o r c e d to a con fron tation , by reg ard ing an y invocation o f contingen cy a s less elegant o r less m ean in gful than explanation s b a s e d directly on tim eless “law s o f n atu re.’’ —Stephen Jay Gould (1989). 32During the earliest decades of U.S. banking history, the convertibility requirement was largely implicit. Early in the nineteenth century, however, states began to pass legislation which explicitly imposed convertibility on the banks—or, alternatively, to include convertibility requirements in bank charters. The federally chartered United States Banks had convertibility requirements in their charters. Convertibility requirements were standard features of state charters issued under the “ Free Banking” laws of 1836-63. The federal E n glish O rigins S y n o p s is : The origins of the m odern U.S. cu r rency system can be traced in large part to Eng land. Many im portant features of the U.S. cu r rency system w ere based on English models. The early history of paper curren cy in England was dominated by the governm ent’s need for specie revenues to finance its foreign wars. This need caused the governm ent to establish two princi ples—Bank of England monopoly, and strict specie convertibility—as the basis for England’s system of paper currency. These principles had a profound effect on the evolution of paper curren cy and banking in the United Kingdom, and later in the United States. In England, the notion o f organized note issue seems to have arisen during the latter part of the seventeenth century. At the time, England had had a gov ern m en t-m on op oly co in c u r re n c y system for several centuries, and had begun to develop a paper cu rren cy system based on bills of exchange.33 During the last decade of the seven teenth century several groups of entrepreneurs recognized an opportunity to profit by providing a m ore convenient paper currency. Each of these groups sought royal ch arters for banks of issue. Horsefield (1960) singles out fou r groups for special study. One of them, led by William Paterson, proposed a bank w hich would lend convertible n otes on co m m ercia l se cu rity . T h e new b an k w as called th e Bank o f England. It re ce iv e d a c h a r te r in 1 6 9 5 , and has o p e rate d co n tin u o u sly sin ce; it is now th e c e n tra l b a n k o f th e U nited K ingdom .34 T h ree oth er groups, led by Hugh Chamberlen, Joh n Briscoe, and Joh n Asgill and Nicholas Barbon, respectively, proposed “land banks” which would lend inconvertible notes on the security of land and other real property. The land banks of Briscoe and Asgill-Barbon actually operated fo r a short time during 1695-96. In the latter year, they w ere consolidated pursuant to a schem e to secure a royal ch arter by raising £2,000,000 in specie to be len t to th e B ritish g o v ern m en t, w h ich w as des- charters issued during the late nineteenth century, under the National Banking Act, also included convertibility requirements. 33For a description of early English monetary history, see Feavearyear (1963), chapters l-IV. 34The definitive history of the Bank of England has been written by Sir John Clapham (1944, two volumes). SEPTEMBER/OCTOBER 1991 42 perate fo r funds. T he ch arter of this “National Land Bank’’ required that it raise half of the specie loan prior to beginning operation. W hen it proved unable to do this, the ch arter lapsed and the schem e fell apart.35 The fear that a public accustom ed to coin cu r ren cy would not accept inconvertible notes c e r tainly played a role in the collapse of the Na tional Land Bank and other land bank schemes. Nevertheless, even Horsefield (1960), who is generally unsym pathetic to the concept of land banking, points out that “the m ajor cause of these events was an accident o f tim e.”36 The Bank of England, w hich had obtained its ch arter by m eans of a similar com m itm ent to provide specie, had drained the capital m arket of funds. This problem was exacerbated by the onset of a com m ercial crisis, w hich forced the Bank of England to suspend specie convertibility of its notes.37 In 1697 the Bank of England obtained, in retu rn for a fu rth er extension of credit to the governm ent, a form al com m itm ent that Parlia m ent would authorize no other banks so long as the Bank existed. Its urgent desire for this co m m itm en t su ggests th a t it co n tin u ed to regard land banking as a viable competitive threat. Clapham (1944) w rites that “the General Court [the directors of the Bank of England] wanted no m ore Land Banks.”38 In 1708, in retu rn fo r fu rth er loan com m itm ents, the m o nopoly grant was “reenacted and made more precise.” Parliam ent explicitly prohibited any firm consisting of m ore than six partners from issuing notes in England.39 Thereafter English note issue was dom inated by the Bank of En gland. T h e small “cou n try b an k s” operated in 35See Horsefield (1960), chapters 14-16, and Clapham (1944), volume I, pp. 33-34. “ Horsefield (1960), p. 246. 37lbid., pp. 246-47. 38Clapham (1944), vol. 1, p. 47. 39See Clapham (1944), p. 65, and Feavearyear (1963), pp. 167-68. 40During 1797-1821 (the era of the Napoleonic Wars), the Bank of England suspended specie payments. Although its notes were not officially legal tender, they became so operationally. Specie virtually disappeared from circulation, most payments were made in Bank of England notes, and other English banks redeemed their notes in Bank of England notes. See Feavearyear (1963), pp. 182-85, and Viner(1937), p. 154. 41Macaulay (1877) provides a colorful and illuminating para graph describing this relationship (IV, pp. 551-2). 42Clapham (1944), describing the Bank’s first summer, observes that “ what the government—like the Bank—most Digitized for FEDERAL FRASER RESERVE BANK OF ST. LOUIS its shadow, and evolved along strictly convertible lines.40 The development o f the British cu rren cy sys tem can be properly understood only in the context of the symbiotic relationship betw een the British governm ent and the Bank o f En gland.41 The Bank regarded its paper curren cy monopoly as critical to its profitability, and was willing to make large financial concessions to the British governm ent in order to protect and extend it. The British governm ent, on the other hand, was willing to grant the Bank a monopoly because it needed the Bank’s financial assist ance—in particular, to help it obtain specie to finance foreign w ars.42 Under the circu m stances, it was profoundly in the in terest of both parties fo r governm ent liabilities to be identified as closely as possible w ith Bank liabili ties, and fo r Bank liabilities to be identified as closely as possible with specie.43 The simplest and most certain way to achieve this was for the Bank to lend extensively to the governm ent and make its notes strictly convertible.44 In Scotland the situation was quite different. The convertible, com m ercial Bank of Scotland was chartered by the Scottish Parliam ent in 1695. This bank, unlike the Bank of England, w as sta tu to rily u ninvolved in g o v ern m en t finance. It was granted a 21-year note issue mo nopoly (which was not renewed). In 1705, both Hugh C h am b erlen , w ho had now m oved to Scotland , and Jo h n Law , w ho w as la te r to achieve m onetary infam y in France, proposed land banks fo r Scotland. Both proposals w ere rejected because they involved notes that w ere legal ten d er—a status Parliament was unwilling wanted in the summer of 1696 was not a circulation of notes but cash, hard cash for the Army in Flanders” (vol. 1, p. 39). 43Clapham (1944) points out that under the Act of 1697, which formalized the Bank’s monopoly, “ forgery of the Bank’s notes was to be punished with death, the penalty for clipping or coining the King's money. Bank notes were not yet the King’s money, but they were getting near to it.” (vol. 1, p. 50). In October of 1698, he writes, the English Treasury agreed to “ 'receive such bills of the Bank of England commonly called Bank Bills ... provided the said bills are not at any discount. ’ The time was getting nearer when the Bank would circulate the Exchequer Bills for the Treasury, cash them on demand, accept them as deposits, make generous advances on their security, and even pay a dividend on them.” (vol. 1, p. 56; my emphasis). And in 1710, just two years after it had acted to further strengthen the Bank’s monopoly status, the British Parliament passed an Act “ ‘for engaging and obliging the Bank of England ... to exchange all Exchequer bills for ready money on demand.’ ” (vol. 1, p. 67). 44Santoni (1984) asserts that “ the Bank’s contract with its customers to redeem its notes at a fixed price in terms of gold 43 to grant private liabilities.45 Apparently, the failure of the English land bank schem es had created a belief that inconvertible notes would be accepted only if they w ere legal tender. As we shall see, this belief was also widespread in the American colonies—w here experience ulti mately refuted it. During the eighteenth and early nineteenth centuries, Scottish banking was considerably m ore competitive than English banking. The m ajor banks fought bitter "note duels,” present ing their com petitors’ notes for payment in an effort to drain their specie reserves and force them to retren ch . One defensive response to these duels was the issuance b y Scottish banks of notes w hich contained an “option clause”—a clause that granted them the right to defer specie payments for a fixed period in retu rn for legal in terest.46 Experim ents like the option clause might well have led to fu rth er departures from convertibili ty in Scotland, w here both the public and the governm ent w ere m ore com fortable with banks and paper curren cy than their counterparts in England. During the financially troubled years of the early 1760s, how ever, the option clause and other "irregu lar” Scottish banking practices a ttra c te d u n fa v o ra b le a tte n tio n in England (whose Parliament had absorbed that of Scot- was a voluntary arrangement” (p. 15). He justifies this asser tion by noting that the gold standard was not imposed by British law until 1821, more than a century after the Bank was founded. While this is certainly true, it is also true that until the early nineteenth century much of British bank regulation was implicit rather than explicit. One illustrative example involves the monopoly status of the Bank of England. Most historians describe the Bank as having possessed a monopoly over joint-stock (corporate) banking in England. However, the Act of 1708 gave it a monopoly only over jointstock note issue. Nevertheless, none of the deposit banks in England attempted to organize in corporate form. Feavearyear (1963) explains this by noting that, despite the lack of an explicit legal prohibition against joint-stock deposit banking, “ there can be no question whatever that the inten tion was to give the Bank of England a monopoly of joint-stock banking, and that had any other institution of more than six partners attempted to carry on banking in England in any manner whatever at any time during the first half of the century it would have been suppressed” (pp. 167-68). Simi larly, despite the lack of an explicit legal requirement that bank notes be convertible, there can be little question but that the intention was that they should be convertible, and that any attempts by the Bank of England, or any of the English country banks, to issue inconvertible or semiconvertible notes would have been suppressed—just as attempts of this sort by Scottish banks were suppressed (see below). Santoni also asserts that the establishment of an official gold standard occurred as a result of “ the Bank’s continuous prods to an unwilling government.” He justifies this assertion by noting (1) that the deadlines set by Parliament for resump land in 1707). In 1765 the British Parliament stepped in with an act prohibiting notes con taining an option clause, or any other depar tures from strict convertibility.47 C olonial O rigins S y n o p s is : The Am erican colonies experim ented with a variety of cu rren cy systems based on inconvertible notes issued by colonial govern ments. During the early eighteenth century the British governm ent began to regulate these sys tems. British regulation forced some of the col onies to back their notes m ore carefully, and eventually prevented all the colonies from m ak ing their notes legal tender. By the end of the colonial period many of the colonies had deve loped successful and popular cu rren cy systems. These systems w ere based on inconvertible notes w hich w ere carefully backed, and w ere not legal tender. Conditions in early colonial America dif fe re d fro m th o se in England ev en m ore profoundly than did conditions in Scotland. In the colonies, the most pressing m onetary problem was a specie shortage: the quantity o f sp ecie th e co lo n ists w e re able to re ta in seems to have been insufficient to m eet their tion were repeatedly postponed, and (2) that the Bank resumed paying specie for small notes in 1817, and Parlia ment intervened in 1819 to prohibit it from doing so. However, Viner (1937), Clapham (1944), and Feavearyear all report that the government’s primary goal in postponing resumption was to avoid embarrassing the Bank. [Viner writes that “the govern ment continued to refuse to obligate the Bank of England to resume cash payments, and both government and Bank were obviously waiting for the course of events to disclose the auspi cious occasion for resumption” (p. 172, my emphasis); Feav earyear notes that “the Government refused to allow the Bank's hand to be forced, and repeatedly extended the term of the Restriction Act” (p. 214).] These authors also describe the resumption attempt cited by Santoni as a failure, and the government prohibition as an attempt to protect the Bank. Finally, both Clapham (1944, vol.2, p. 70) and Feavearyear (pp. 221-22) report that the Bank opposed the legislation which actu ally compelled it to resume specie payments. 45For a history of the early years of Scottish banking, see Checkland (1975) or White (1984). For a description of the Scottish land bank proposals and their fate, see Horsefield (1960), pp. 175-78, 215-16. 46The option clause has attracted a good deal of academic attention in recent years. See White (1984), pp. 25-30, 141-42, Rockoff (1986) and Dowd (1988), for example. 47See Checkland (1975), pp. 118-21; White (1984), pp. 29-30. The legislation originally proposed ruled out the option clause only; it was later amended to rule out any notes not redee mable on demand. SEPTEMBER/OCTOBER 1991 44 needs for a medium of exchange.48 During the 1650s, the M assachusetts Bay Colony attem pted to allay the shortage by operating its own mint (which produced the renow ned "Pine T ree Shill ings”). The British governm ent viewed this ac tion as usurping a royal prerogative, how ever, and forced M assachusetts to close the m int.49 The colonies also experim ented with commodity currencies of d ifferent types; these included wampum (Indian beads), rice and tobacco.50 Despite the public's need for m ore convenient means of paym ent, the introduction of a new form of governm ent paper cu rren cy was m oti vated in the first instance by the fiscal exigen cies of a colonial governm ent. In 1696, the M assachusetts legislature experienced great difficulty financing an expedition against the French in Canada. It decided to issue "bills of credit" in the form of paper curren cy to use to purchase supplies. These bills w ere neith er con vertible nor ultimately redeem able in specie; they could, however, be used to defray future tax liabilities. This financing expedient proved quite successful, and the colony used it rep eat edly during the ensuing 50 years. W ithin a very few years, other colonies began to adopt the practice—first in New England, and later else where. By 1730 or so, bills of credit had become the principal currency of the American colonies.51 The earliest colonial bills w ere issued, like these M assachusetts bills, in anticipation of fu tu re taxes. A fter a few years, how ever, certain colonies began experim enting with bills that w ere issued on loan. Typically the issuing colo ny would pass laws providing that relatively small sums in new bills could be lent to individ uals w ho w ere able to provide land or other sorts of property as collateral. (Often these loans w ere m ortgage loans and w ere intended in part to encourage the colonists to settle and improve land.) These "loan office” or "land bank” 48For discussions of the specie shortage, see Nettels (1934), pp. 202-207, and Brock (1975), pp. 1-9. 49See Felt (1839), Bullock (1900), Chapter III, Breckinridge (1903), pp. 55-56 and Nettels (1934), p. 276. 50None of these experiments proved particularly satisfactory. For a description of colonial experiments with commodity currency see Nettels (1934), chapter VIII, and Brock (1975), pp. 9-16. 51For an encyclopedic account of colonial currency history prior to 1764, see Brock (1975). 52For discussions of colonial land banking, see Davis (1900), Kemmerer (1939), Thayer (1953), Billias (1959), Brock (1975) and Smith (1984), among others. Digitized forFEDERAL FRASER RESERVE BANK OF ST. LOUIS issues becam e increasingly popular during the first half of the eighteenth cen tu ry.52 As previously indicated, the legislation auth orizing the emission of tax anticipation or loan office bills was typically accom panied by legisla tion providing fo r th eir eventual retirem en t— either by imposing future taxes which the bills could be used to pay, or setting out the term s according to w hich the loans would be secured and repaid. The legislators clearly believed that it was these retirem ent com m itm ents that con veyed value to the bills.53 Unfortunately, th ere w ere often great political and financial incen tives for the colonies to violate these com m it ments by declining to levy or collect the future taxes, by declining to collect the loan payments, or by stretching out the period over w hich loans could be repaid. W hen these things happened, the bills would often depreciate in value relative to specie and goods.54 The extent of the depreci ation was typically m easured by the discount on paper cu rren cy relative to specie currency; that is, by the difference, in percent, betw een unity (one) and a fraction equal to a given quantity of specie cu rren cy divided by the quantity of paper cu rren cy it could be sold for in the open m arket. If it took 50 shillings in paper curren cy to purchase 40 shillings in specie currency, for instance, then the discount on paper curren cy was 20 percent. During the early decades of the eighteenth century, many colonial curren cies ex perienced significant depreciation. In some cases, the depreciation was quite severe.55 Currency depreciation became particularly con troversial because most colonies gave their bills of credit the status of legal tender. Legal tender laws compelled creditors to accept bills at face value in payment of debts. If, for example, a loan agreement called for a repayment of 500 shillings at the end of five years, the lender could be forced to accept 500 shillings in bills of credit, 53Nettels (1934) writes that “ In the opinion of the colonists, the principal factor affecting the specie value of their paper was the provision made for redeeming it from tax revenues” (pp. 257-58). 54For analyses of the link between backing and depreciation, see Smith (1984,1985a, 1985b) and Russell (1988). 55Data on the specie prices of the currencies of various different colonies are presented by Brock (1975) and Smith (1984,1985a, 1985b), among others. 45 even though he might prefer specie. The pen alties for refusing to accept bills were relatively harsh: the creditor might forfeit the entire amount of the debt or, in some cases, a multiple thereof.56 The original idea behind legal tend er laws was to protect borrow ers, and to reduce the fre quency of lawsuits, by providing a method of repaym ent w hich was beyond legal challenge.57 W hen unexpected depreciation occured, how ever, legal tend er laws tended to benefit debtors (by reducing the real value of their obligations) at cred itors’ expense. This made them popular with farm ers and oth er debtors, and unpopular with creditor interests. The creditor interests included a number of British merchants who did business with the colonies. Many of these m erchants were wellconn ected in Great Britain; th e ir com plaints, which were seconded by those of indigenous merchants and creditors, received sympathetic attention from the British colonial administration, and eventually from the British Parliament. After 1730, the colonial administration began to issue regulations eliminating or restricting the right of particular colonies to issue new bills or (more fre quently) to make them legal tender. As the prob lem of depreciation worsened, however, Parlia ment considered comprehensive legislation. The Currency Act of 1751 deprived the New England colonies of the right to issue legal tender bills and greatly restricted their powers to issue paper cur rency of any description. In 1764, a second Cur rency Act extended the legal tender prohibition to all the colonies.58 56See Bullock (1900), p. 131, Nettels, (1934), p.265, and Russell (1988), pp. 47-48. 57See Breckinridge (1903), p. 52, and Hurst (1973), p. 40. West (1 978) stresses the role of colonial paper currency in providing a “ means of settlem ent,” but does not mention legal tender laws explicitly. 58Even before the blanket legal tender prohibition, the British government had intervened to prevent particular colonies from making their currencies legal tender. It also intervened to force some of the colonies to back their legal tender curren cies more carefully with future tax receipts, and to prevent others from issuing currency on loan. See Davis (1900, vol. I and II), Ferguson (1953), Ernst (1973), Brock (1975), Smith (1984,1985b) and Russell (1988). For the history of the Currency Acts, see Davis (1900), Greene and Jellison (1961), Ernst (1973), Brock (1975), Smith (1985) and Russell (1988). 59ln 1767, Ben Franklin wrote that “ On the whole, no method has hitherto been formed to establish a medium of trade, in lieu of money, equal, in all its advantages, to bills of credit, funded on sufficient taxes for discharging it, or on land secu rity ... and in the mean time made a GENERAL LEGAL TENDER.” [Franklin (1971), p. 354; his emphasis.] Ferguson Many colonies responded to the legal tender prohibitions by issuing non-legal tender bills of credit. Although many contem porary analysts be lieved that giving the bills legal tender status was essential to preserve their value, this does not seem to have been the case in practice.59 The non-tender bills remained quite stable in v a lu e far more stable, in many cases, than their legal tender predecessors. This was particularly strik ing because many non-tender issues took the form of land banks—a mode of issue the British regarded as particularly prone to depreciation.60 During the decade prior to the Bevolution, the colonies appeared to be moving toward a system of non-legal tender land bank currency.61 R ev o lu tio n a n d R eo rga n iza tio n S y n o p sis : The Revolution completely disrupted the evolution of the American currency system. The Continental Congress was forced to finance wartime expenditures by money creation—a policy which led to a virtual hyperinflation. The war, and the depression that followed it, produced financial problems for both state governments and the general public. One symptom of these problems was large public and private debts— many of which were held by the domestic prop ertied classes. This situation, combined with mem ories of the recent inflation, created fears among the m em bers of th ese classes th at popularly elected state governments would adopt monetary policies designed to partially repudiate these debts. Representatives of the propertied classes domi nated the Constitutional convention. They moved (1953) writes that “ The restraining act of 1764 ... prohibited legal tender laws and required that existing legal tender currencies be sunk at their expiration dates. Many colonies protested, in the belief that the legal tender feature was an essential prop to their currency. Experience was to show, however, that the restriction did not materially impair the workings of the currency system” (p. 177). 60For discussions of the strong performance of nontender paper currency, see Ernst (1973), Smith (1984b) and Russell (1988). 61See Ferguson (1953), pp. 177-180. Inconvertible government currency which was not legal tender is of special historical interest because it was issued under circumstances that approximated relatively closely the circumstances under which inconvertible private currency might have been issued. This was especially true when, as was frequently the case, the currency was issued on loan rather than in anticipation of taxes. Russell (1988) argues that the success of government, non-legal-tender, inconvertible land banking before the Revo lution provides indirect evidence that private inconvertible banking might have been feasible after the Revolution, had it been legally permitted. SEPTEMBER/OCTOBER 1991 46 to prevent repudiation by prohibiting the states from issuing their own currency, or from making privately issued cu rren cy legal tender. During the Revolutionary W ar, the Continental Congress was the Am erican central governm ent, and bore prim ary responsibility for conducting and coordinating the w ar effort. It also faced a critical financing problem : under the Articles of Confederation, it lacked the pow er to levy taxes. (Colonial opposition to British taxation had been one of the most im portant causes of the rebellion.) During the early stages of the Revolutionary W ar, the Congress attem pted to subsist on voluntary contributions from the colonies. W hen this source of revenue proved insufficient, it began to issue bills of cred it—the renow ned "continentals”— which w ere backed by little m ore than the pious hope that the states would eventually provide funds, or authorize tax levies, to retire them. The likelihood of such retirem ents becam e ever m ore distant as the quantity of continentals increased and the states supplemented them w ith their own cu rren cy issues. Both form s of paper curren cy began to depreciate—at first gradually, and later very rapidly. By the end of the w ar, they w ere virtually w orthless.62 During the b rief “critical period” betw een the end of the w ar (in 1783) and the ratification of the U.S. Constitution (in 1789), the newly independent states began to reorganize th eir finances and con sider the problem of providing a paper currency. A num ber of states issued or seriously considered issuing bills of credit in anticipation of taxes or on loan—m uch in the m anner of the prerevolutionary colonies.63 Unfortunately, the continental hyperinflation had fractu red the prerevolutionary consensus reg ard in g th e u sefu ln ess o f p ap er c u rre n c y . People who had accepted continentals or contin ental-denom inated securities from governm ent or private parties w ere outraged that the states ap 62For a good account of the history of the Continentals, and indeed of Revolutionary War finance, see Ferguson (1961). See also Calomiris (1988). 63For general discussions of currency issues by (incipient) states during the critical period, see Nevins(1924), Ferguson (1961), Nettels (1962), Russell (1988) and Schweitzer (1989). 64For descriptions of public attitudes toward paper currency during the critical period, see Libby (1894), Hammond (1957) and Ferguson (1961). 65Most of the other colonies that issued legal tender bills also experienced serious depreciation—though not on the scale of Digitized forFEDERAL FRASER RESERVE BANK OF ST. LOUIS peared to have no intention of redeem ing them at anything close to the values at w hich they had traded during the early years of the w ar. Indeed, the propertied classes cam e to view paper cu rren cy as a device by which popularly elected govern ments sought to perm it the common people to escape the burden of their public and private debts. (The lengthy trade depression that followed the end of the w ar had increased private debt burdens.) W ealthy Am ericans becam e terrified that the state legislatures, which w ere now free from British restraint, would rep rise the Revolu tionary experience by issuing large volumes of inadequately backed legal tender bills—bills which would rapidly depreciate, and w hich could be used to retire debts at a fraction of their real value. As a result, proposals to issue paper cu rren cy that would have received consensus support before the Revolution now becam e the subjects of intense political controversy.64 In Rhode Island, radical populists gained control of the legislature. They confirm ed the w orst fears of the anti-paper m oney conservatives by issuing a legal tender currency, and then engineering a rapid inflation that seemed clearly designed to en ab le b o rro w e rs to escap e th e ir d e b ts .65 In M assachusetts, an agrarian insu rrection (Shay's Rebellion) erupted as a result of the refusal of the legislature to issue legal tender paper cu rren cy .66 As it happened, the Constitution was w ritten and ra tifie d d urin g a p eriod o f co n serv a tiv e ascendancy— a reaction against excesses of the sort epitomized by events in M assachusetts and Rhode Island.67 The conservatives desired a “h ard ” c u rre n c y im m une from depreciation. As a result, the fram ers of the Constitution w ere not content merely to deprive the states of the right to issue legal tender bills; instead, they w ere prohibited from issuing cu rren cy of any kind. Specie was established as the new nation’s sole legal tender cu rren cy—and, in the minds of many, as the nation's only legitimate currency. Rhode Island. On the other hand, colonies that issued nontender bills experienced little or no depreciation. For a discussion of this question, see Russell (1988). For descrip tions of Rhode Island’s post-revolutionary currency policy, see Phillips (1865), Bates (1898), Nevins (1924) and Ferguson (1961). 66The classic study of Shay’s Rebellion is Taylor (1954). 67This point is made by Ferguson (1961), pp. 249-250, Nevins (1924), p. 537, and Schweitzer (1989), pp. 319-320. 47 Thus the peculiar historical circum stances of the post-revolutionary critical period had a pro found and lasting im pact on the nature of the U.S. m onetary system. The traum a of the Revolution made the curren cy system controversial, and ulti mately produced a system very d ifferent from the relatively uncontroversial system of late colonial times. Indeed, it seems likely that had the colonies been able to escape British domination without fighting an expensive w ar, or had the principal casu s bellu m not been one w hich required that the w ar be financed by m eans w hich sowed the seeds of a divisive struggle betw een classes, the United States might have begun its existence with a decen tralized cu rren cy system based on (non-legal tend er?) bills of credit issued by state governm ents.68 The Constitution was silent on the question of privately issued currency. Indeed, during the years immediately following its ratification, issu ance of small-denomination liabilities w hich might circulate as cu rren cy (which might "pass cu rren t,” to use the contem porary phrase) was regarded as a right of all free persons. By the second decade of the nineteenth century, how ever, the legislatures of m ost of the states had acted to eliminate or greatly restrict th atrig h t.69 T h ere w ere at least two reasons for this. One was the problem s caused by irresponsible, or downright fraudulent, private issues. Another, w hich was perhaps m ore com pel ling, was the desire of the state legislatures to re serve the right of note issue to state-chartered banks.70 The Constitution had also been silent regarding the right of the states (or the federal government) to ch arter banks of issue—perhaps because private b an k in g had little h isto ry in th e co lo n ies. In M assachusetts, the most econom ically sophisti cated colony, efforts to organize a private land 68Many historians believe that resentment over British efforts to regulate colonial currency practices played a major role in stimulating the Revolution. [See Bullock (1900), pp. 56-59, Davis (1900), vol. 1, chapter XXI, vol. 2, pp. 256-61, Brock (1975), pp. 561-63, Billias (1959), p.42, and Ernst (1973), pp. 359-60, for example.] This makes it seem very ironic that the currency restrictions the U.S. Constitution imposed on the states were generally more restrictive than any the British had ever imposed. [One exception is that the states could charter private banks of issue, something the British had prohibited the colonies from doing. It is not completely clear that the framers actually intended to authorize state-chartered banking, however (see below).] 69See Hammond (1957), pp. 27-29,184-85 and Fenstermaker (1965a), pp. 21-22. 70Both these motives are mentioned by Hammond (1957), pp. 27-29,159-60,184-85. See also Fenstermaker (1965a), pp. bank along the lines of public land banks began late in the seventeenth century and persisted episodically for the next five decades. The m ercan tile community was somewhat skeptical of land banking, how ever, since it threatened to compete with their own lending activities. [M erchants in the coastal cities provided a good deal of trade credit to the m erchants and farm ers of the interior.] In addition, the colonial authorities (both adminis trative and legislative) w ere reluctant to give up their monopoly over paper cu rren cy—partly out o f fear that a form of m oney issued outside of offi cial control might be subject to manipulation, and partly out of con cern that it might reduce poten tial revenues from seigniorage. Although various private land bank projects received considerable popular support, they w ere unable to surm ount this political opposition.71 By 1740, how ever, the restrictions on colonial issues w hich had been imposed by the British governm ent had becom e so onerous (land bank issues, in particular, had been entirely prohibited) that the M assachusetts legislature was willing to ch arter a private land bank. W hile the land bank project received broad support from the public, it was vehem ently opposed by the colony's governor (a creatu re of the British), who viewed it as w eak ening the mother country’s control over the colony’s economy. The land bank was also opposed by British m erchants, who saw it as a th reat to their n ear monopoly over trade credit. Both groups ap pealed for relief to Parliament, which responded by enacting legislation prohibiting the establishment of banking corporations anywhere in the colonies.72 The legislatures of the newly independent states saw chartered banking as a means by which they could provide their citizens with paper currency while at the same time (in many cases) providing 15-16. The importance of the latter motive is indicated by Hammond’s comment that restrictions on unincorporated note issue were enacted “ on the complaint of chartered banks” (p. 184). 71In 1714 there was a well-organized and determined attempt to organize a private land bank—an attempt which was supported by some influential British merchants, and received the endorsement of the Board of Trade. The govern ment of Massachusetts responded by establishing a public land bank. See Billias (1959), pp. 3-5, Nettels (1934), pp. 271-275, Davis (1900), volume I, pp. 56-61, volume II, pp. 82-91, Ernst (1973), pp. 27-28, and Metz (1945), chapters 3,4. 72See Billias(1959), Davis(1900), volume II, pp. 130-261, Hammond (1957), pp. 24-25, Brock (1975), pp. 123-27, Ernst (1973), pp. 34-35, and Metz (1945), chapter 10. SEPTEMBER/OCTOBER 1991 48 themselves with revenue. Of course bank charters would be m ore valuable to their holders, and thus issuing charters would be more lucrative for the states, if the charters conveyed an exclusive right to issue paper currency. This accounts for the prohi bitions against private issues from other sources.73 But what sort of banks should the states charter? While it might seem that the colonies' extensive experience with public land banks should have led them, as states, to charter private land banks, they did not in fact do so. Two factors may help explain this situation. The first is that the right of the states to charter banks of any sort was not altogether clear; indeed, the view that they did not have this right was widely held.74 Doubts about state charter rights seem to have existed on two levels. Many, and perhaps most, informed Americans believed that the Constitution established specie as the only legitimate form of "money." The question then becam e "W hat is ‘m oney’? ” and, in particular, "Are bank notes ‘m on ey?” A conservative view was that the category "m oney” did include bank notes, and indeed paper bills of all descriptions, so that paper curren cy in any form was proscribed. A m ore m oderate view was that bank notes con v ertible in sp ecie w ere not money, but m erely its “rep resen ta tive,” and thus w ere not constitutionally prohibi ted.75 This view implied that private land banking conducted along colonial lines, w hich is to say through the issuance of inconvertible notes, in volved direct creation of money, and was th e re fore unconstitutional—even if the bills w ere not legal ten d er.76 A second possible reason why the states did not opt for some form of land banking was that, given 73The motives (and actions) of the states in this regard were similar to those of any license-granting monopolist. The strategy was evidently based on the British example. During the eighteenth century, the British government had repeat edly extracted large payments, or loans on favorable terms, from the Bank of England in return for extending or strength ening its monopoly on note issue. A distinctively American variant of this strategy was for a state government to require a bank’s organizers to cede the state an equity interest in return for granting the bank a charter. Sometimes the interest was ceded gratis, and sometimes merely on favorable terms. This practice was particularly common in the southeastern states. In several of these states bank dividends accounted for a substantial portion of state government revenues. [See Fenstermaker (1965a), pp. 17-20. For a thorough analysis of the importance of dividends from bank stock in state govern ment finance during this period, see Sylla, Legler and Wallis (1987).] 74For a general description of the nature and source of doubts about the constitutionality of state banks see Hammond FEDERAL RESERVE BANK OF ST. LOUIS their British-imposed lack of experience with p rivate b an k in g , th e ir only m odels fo r b an k c h a rte rs w e re th o se o f B ritish b a n k s—w h ich meant, for all intents and purposes, the ch arter of the Bank of England. Indeed, the ch arters of many of the earliest state banks w ere virtual carbon copies o f th e Bank of England 's c h a r te r . T h e attractiveness of the British model may have been enhanced when, shortly after the Constitution was ratified, the Federal governm ent decided to seek a ch arter for a single “National Bank” along the lines of the Bank o f England. This institution was called the "Bank of the United States,” and was established in 1791 with a 20-year charter. Though the U.S. Bank was basically a private organization, the Federal governm ent held a minority interest, and the Bank was expected to provid e a v a riety o f fin a n cia l se rv ices fo r th e g o v ern m en t in addition to its p riv ate lending activities.77 The early state banks seem to have been intended as state versions of the U.S. Bank. This is reflected in the fact that until the ch arter of the U.S. Bank expired in 1811, most states ch ar tered just one or, at most, a handful of banks.78 The failure of the first U.S. Bank to secure a new ch arter was due to a com bination of doubts about its co n stitu tio n a lity , su sp icion of its p ow er, and d isco m fo rt w ith th e fa c t th a t m uch o f its stock w as fo re ig n -o w n e d .79 T h e dem ise o f th e b an k coin cid ed w ith a p eriod o f n atio n al eco n o m ic exp an sion asso ciated p artly w ith th e im p act of the N apoleonic W ars on com m od ity p rices, and p artly w ith th e se ttle m e n t of th e w e ste rn (trans-A ppalachian) reg ion . A cro ss th e U nited S tates, and p a rticu la rly in th e new states and (1957), pp. 103-13, 564-71, and Hurst (1973), pp. 11-12, 141-45. 75See Hammond (1957), pp. 61,105-06, Gallatin (1879), pp. 254-55, (1879), p. 379, Bancroft (1831), p. 40. 76The relatively radical view that private or even public land banking was consistent with the Constitution, so long as the states did not try to make the notes of such institutions legal tender (or even, perhaps, ifthey did), did not become popular until the economically troubled period following the Panic of 1819. 77For the history of the first U.S. Bank, see Holdsworth (1910). 78For information concerning state banks chartered before 1819, see Fenstermaker (1965b). 79See Hammond (1957), pp. 209-26, and Holdsworth (1910). 49 te r r ito r ie s of the W est and South , b an k s w ere c h a rte r e d in larg e n u m b e rs.80 T h ese n ew b a n k s d iffered fro m the esta b lish ed b a n k s in one c r it ical re sp e c t: th e y w e re o rg an ized p rim arily to provid e c re d it to fa r m e r s .81 E c o n o m i c D e p r e s s i o n a n d Its C onsequences S y n o p s is : The lengthy and severe depression that followed the Panic of 1819 placed great strains on the U.S. banking and cu rren cy system. The crisis exp o sed ab asicin con sisten cv betw een two goals of the developing banking system: specie convertibility of bank cu rren cy , on one hand, and liberal extension of farm cred it, on the other. Many of the southern and w estern states, in w hich farm erspred om inated , responded by exp eri m enting w ith system s in w hich banks issued in convertible notes. W hile some of these systems w ere clearly not viable, others appear to have had prom ise. The fed erally ch a rtered Bank of the United States intervened to put an end to all of them , how ever. At this point in U.S. history, a conflict arose be tween the needs of economic development and the devotion to “hard money” which grew out of the Revolution. Economists since the time of Adam Smith had understood that banks which issued convertible notes, and thus were vulnerable to runs, could not safely lend to farmers: farm loans were typically long term, illiquid and relatively risky.82 The need for farm credit was sufficiently great, and public under standing of banking sufficiently slight, however, that the legislatures of the southern and western states 80At the time, the western region included western Pennsyl vania, Ohio, Kentucky, Tennessee, Indiana, Illinois and Missouri. (The following account applies most closely to the experience of the last five states listed.) The southern region included Virginia, North Carolina, South Carolina, Georgia and Alabama. (The following account applies most closely to the experience of North Carolina, South Carolina and Georgia.) For information concerning state banks chartered before 1837 see Fenstermaker (1965a). 81For a discussion of the development of agricultural (and other types of long-term) banking see Hammond (1934), Hammond (1957), pp. 676-80, and Redlich (1951), vol. 1, pp. 11-13, 44-45. 82Many farm loans were seasonal loans to finance planting or harvesting the crop. Variation in weather conditions and crop prices could make such loans quite risky, and by contemporary commercial standards they were relatively long term. However, a good deal of farm credit involved much longer-term loans to finance the purchase and/or clearing of land, the purchase of equipment (and in the South, slaves), etc. Most loans for these purposes were “ accommodation loans.” An accommodation loan did not permitted (and indeed encouraged) their banks to finance large quantities of farm credit by issuing convertible notes.83 The collapse in agricultural prices which occurred during 1818-19 (and led to the Panic of 1819) made the two mandates of the southern and western banks—prompt specie redemption of notes, and liberal extension of farm credit—impossible to recon cile.84 The price collapse produced widespread loan defaults and runs on banks. Since agricultural loans were impossible either to collect or to sell in a short time, the runs could be stopped only by suspending specie paym ents—by refusing to redeem bank notes in specie on demand. Even after the runs subsided, however, the defaults represented tremendous losses for the stockholders of banks that were willing, or could be compelled, to honor their convertibility commitments. Indeed, a large number of western banks had become insolvent.85 The managers of the southern and western banks responded by declining to resume payments, and their notes continued to trade at substantial discounts in the open market. The holders of these notes were forced, in effect, to bear some of the financial losses associated with the Panic. The governments of the western and southern states responded to this situation in very different ways. Most western states had banking systems composed of large numbers of relatively small banks. Since the losses associated with the price collapse and Panic were particularly heavy in the West, most western banks were insolvent, or nearly so; their notes were trading far below par. Many westerners viewed the events of 1818-19 as a conspiracy on the part of the "monied interests” to ruin them and seize have a fixed term; the borrower was expected to pay an “ installment” equal to a fixed fraction of the principal (typi cally, 10 percent) every 90 days. In practice, installments could be deferred and/or reduced, so that the lifespan of an accommodation loan could greatly exceed the term im plied by these conditions. Crop loans might also be ex tended as accommodation loans, with the presumption that they would be fully repaid at the end of the season. Very often adverse circumstances made this impossible, however. See Fenstermaker (1965a), pp. 47-49, Redlich (1951), volume I, p. 11, Holder (1937), pp. 119-22, and Russell (1989a), pp. 69-73. 83See Hammond (1957), pp. 178-83, Redlich (1951), pp. 9-12. 84For descriptions of and data on the price collapse, see Cole (1938), Berry (1943), Smith (1953) and Russell (1989c). 85A list of the banks which failed during or shortly after the Panic can be compiled from information presented by Fen stermaker (1965a). Berry (1943) provides information con cerning bank failures in Ohio during this period. SEPTEMBER/OCTOBER 1991 50 their property. The banks, they believed, had been agents of this conspiracy. Consequently, the w estern legislatures moved to revoke their ch arters and/or to fo rce them to liquidate. Pri vate banks w ere replaced by monolithic, statemanaged organizations called "Banks of the State” or “relief banks.” They w ere supposed to lend inconvertible notes w hich would be given quasilegal tend er status by their states.86 The history of the relief banks was brief, con troversial, and generally undistinguished. Since their m andate was to lend liberally to financially distressed farm ers, the m arket value of their loan portfolios was low relative to the nominal value of th eir outstanding notes. This circu m stance was reflected in the deep discounts on the notes.87 The lending standards of the Illinois relief bank w ere so lax, and its efforts at collec tion so ineffectual, that its notes soon becam e virtually w orthless.88 Both the legislation that created the relief banks and the "stay law s” w hich made their notes quasi-legal ten d er w ere of doubtful con stitutionality. These laws w ere challenged vig orously in state and federal courts. In Kentucky, Tennessee and Missouri, these challenges ulti mately led to the demise of the relief banks. Some of the federal cou rt challenges w ere o r chestrated by the second Bank of the United States. In 1816, Congress had responded to fi nancial problem s created by the W ar of 1812 by chartering a second U.S. Bank. The new bank was (again) a private institution, though the federal governm ent held a sizable m inority interest. The term of its ch a rter was 20 years.89 The U.S. Bank's ch a rter required its notes to be strictly convertible. The Bank’s m anagem ent believed that it could circulate convertible notes in the w est and south only if the notes of its lo cal com petitors w ere also convertible. The m an agem ent also believed that acquiring a large local circulation was essential if the Bank w ere 86The best available account of the Panic and its aftermath is Rothbard (1962). The legislatures of these states enacted “ stay laws” which provided that foreclosures and other legal actions for the collection of delinquent debts (“ executions,” to use the contemporary description) could be delayed for long periods—typically a year or more—unless the creditors in question were willing to accept Bank of the State notes at par. See Rothbard (1962), chapters II, III. 87Data on the discounts on the notes of the Kentucky relief bank are available from Berry (1943) and Sumner (1896). Fenstermaker (1965a) presents somewhat less complete in formation on the discounts on other relief bank currencies. Digitized for FEDERAL FRASER RESERVE BANK OF ST. LOUIS to earn a profit on its southern and w estern operations, or effectively perform its duties as the payments agent of the federal governm ent.90 The Bank consequently made strenuous efforts to force the w estern banks to retu rn to the specie standard. This policy allowed it to extend the scope of its activities in the W est, while striking a pose as the defender of sound cu rren cy. In the meantim e, bitter political controversy over the redistributive implications of the activi ties of the Banks of the State greatly reduced their effectiveness.91 In many ways, the relief banks w ere direct descendants of the public land banks that had been quite successful, and had enjoyed consen sus public support, during the years preceding the Revolution. Unfortunately, the extraordinari ly adverse econom ic circum stances w hich led the w estern states to circum vent the Constitu tion by creating these institutions also served to ensure th eir ultimate demise. The relief banks w ere created by the state legislatures for the purpose of relieving the financial plight of their constituents, rath er than to provide a sound currency, or to earn revenue (which is to say profits) fo r the states.92 This m andate, com bined with the depth of the agrarian distress, made it impossible for the banks’ m anagers to resist making too many loans to troubled farm ers— who w ere already burdened with debt and whose ability to repay was very doubtful. The same circum stances led the states to supplement the relief bank legislation w ith stay laws and related provisions. These provisions made the banks extrem ely controversial, earning them the enmity of creditor interests in general and the pow erful United States Bank in particular. The failure of the relief bank experim ent had an effect on public attitudes tow ard m onetary issues that was rem iniscent of public reaction to the Revolutionary hyperinflation. People becam e increasingly suspicious of banks and paper cur- 88See Rothbard (1962), pp. 41-42, 80-83; see also Dowrie (1913). 89See Catterall (1902), chapter I. 90See Catterall (1902), pp. 96-99, Redlich (1951), pp. 109-10, 124, 128-29, 440-44, Temin (1969), p. 49, and Fraas (1974). 91For accounts of the demise of the relief systems, see Rothbard (1962), Chapter III, Hammond (1957), pp. 282-285, and Sumner (1896). 92See Fenstermaker (1965a), pp. 26-27, Rothbard (1962), pp. 81-83, 85-86, 102-03, and Sumner (1896). 51 rency, and increasingly enam ored with "hard m oney”—specie, or bank notes rigidly convert ible into specie. The demise of the relief banks also gave the Bank of the United States a virtual monopoly over banking in m uch of the W est. notes.94 These notes dropped to variable discounts (against specie) in the open market. Variation in these discounts seem to have reflected changing market conditions—much like the variation in modern national currency exchange rates.95 The southern states tried a d ifferen t—and ini tially, at least, m ore promising—experim ent with inconvertible banking. The southern banks, unlike their w estern counterparts, w ere large but few in num ber; frequently, they operated b ranch es across their respective states. The southern state governm ents held large minority interests in these banks. Since the dividends on the shares provided an im portant source of state revenue, the state legislatures had no desire to see the banks close down.93 In addi tion, since the financial distress that accom pa nied the Panic was less severe in the South than in the W est, the popular outcry against the banks was somewhat less strident there. Finally, because of their large size and branch systems, the banks of the region w ere m ore effectively diversified than their w estern counterparts, and requirem ents imposed by their ch arters had kept them relatively highly capitalized. Economic historians have usually viewed the suspensions as irregular events completely in consistent with the m aintenance of m onetary and financial stability.96 T h ere are good practi cal and theoretical argum ents against this view, however. On a practical level, the banks stayed in business, and continued to supply badly needed currency and credit, despite the depressed condi tions created by the collapse in prices and the financial panic. This situation stands in marked contrast to that of the W est. There, the banking system collapsed, and a scarcity of cu rren cy and credit threatened to bring econom ic activity to a standstill. W hen the southern banks suspended the con vertibility of their notes, the governm ents of the southern states did not force them to liquidate, or even to close down. Instead, the banks w ere permitted, and often encouraged, to continue to do business—lending, collecting on loans, and conducting other financial transactions, all through the medium of their now-inconvertible 93See note 73 above. 94For a description of banking in North Carolina after the suspension, see Russell (1989a). 95Note discount data for North and South Carolina state banks during 1817-1829 are provided by Russell (1989a). Fenster- maker (1965a) provides less complete data for all the southern states. 96For expressions of this view see Gouge (1833), Sumner (1896) and Klein (1974). 97Before the suspensions, the notes of southern state banks had traded at or near par with specie; afterwards, they traded at variable discounts. Thus it seems clear that the suspen sions exposed noteholders to risks they had not previously borne. The author believes that after the suspensions, the market priced bank notes in much the same manner as modern mutual fund shares. This sort of pricing scheme would have linked the value of a bank’s notes to the value of its assets—a link which was absent under convertibility. [See Russell (1989b,c).] Other theories of inconvertible note pricing have broadly similar implications, however. The most popular alternative theory is that the notes of a suspended bank were priced as risky titles to future specie, payable if and when the issuing bank resumed payments. Since the state of a bank’s portfolio was probably the biggest single factor influencing the prospect that it would be able to re On a theoretical level, the suspensions shifted some of the burden of the banks' portfolio risk from shareholders to noteholders.97 The price collapse and panic had revealed the true extent of the risks the banks faced, and had ex acer bated the already acute scarcity of financial capital—particularly concentrated financial capi tal—in the relatively undeveloped southern states.98 Under these circum stances, it seems doubtful that current or future bank shareholders would have been willing to continue to b ear all of the banks’ portfolio risk—particularly in light of the heavy losses that a prompt retu rn to specie payments would have im posed.99 sume, this theory also implies that the holders of inconvertible bank notes were bearing portfolio risk. And because all of the portfolio risk had to be borne by the holders of the banks’ liabilities, any risk borne by noteholders reduced the risk born by shareholders (and vice-versa). Another alternative theory is that bank notes traded at a discount during suspension because holders’ inability to convert them into specie temporarily reduced their useful ness, relative to specie, as media of exchange. Advocates of this theory typically view suspensions as a sort of “ safety valve” protecting fractional reserve banking systems oper ating under laissez-faire—operating without, that is, govern ment deposit insurance and/or a government lender of last resort. For expressions of this view see Friedman and Schwartz (1963), pp. 163-68, 324-32, and Dwyer and Gilbert (1989). 98The charters of contemporary banks specified minimum denominations for shares which were usually well out of reach of the common people. The rationale behind these minima is not entirely clear. See Russell (1989a), pp. 35-36. "T he North Carolina banks, in particular, decided to close down when specie payments were finally imposed on them in 1828-1829. See Russell (1989a), pp. 25-32, 78-80, Holder (1937), pp. 250-51, and Flanagan (1934). SEPTEMBER/OCTOBER 1991 52 A case can be made that the post-Panic suspen sions began a process w hich, had it been allowed to proceed unhindered, might have enabled the South to develop an alternative banking system which was peculiarly suited to its distinctive needs. Because the southern econom y was dominated by agriculture, banks could be useful to southern econom ic development only if they w ere able to make farm loans in large volumes. Farm loans, however, w ere relatively risky—and under con vertibility, these risks w ere born e almost exclu sively by bank stockholders. These stockholders had both the opportunity and the means to lend outside the South, and could be induced to take these large risks only in retu rn fo r high average rates of retu rn. Such rates would have made bank credit too expensive fo r many farm ers—and usury laws might have prevented the banks from charging them in any case. One solution to this problem was the development of an alternative banking system w hich could bring the diffuse financial capital of the com m on people into the risk-bearing process. Inconvertible priv ate banking seem s to have had the potential to provide such a system .100 Unfortunately for the southern banks, both the federal governm ent and its financial agent, the Bank of the United States, regarded the suspen sions with almost unalloyed hostility. The U.S. Treasury D epartm ent was deeply (and somewhat irrationally, under the circum stances) committed to fiscal arrangem ents under w hich payments to the federal governm ent (for taxes, land purchases, etc.) w ere made exclusively in par cu rren cy .101 Southerners, how ever, w ere used to making governm ent paym ents in local (bank) currency, w hich was no longer trading at par. The Bank of the United States, w hich was charged with the 100This argument is presented in detail in Russell (1989a,b). 101It seems likely that the exchange problems described here could have been avoided if the Treasury had been willing to accept and disburse state bank notes at their market rates of discount. A similar situation had arisen after the general suspension, of specie payments which occurred in August of 1814, near the end of the War of 1812. During the months after the suspension, the Treasury needed funds to service debt held by residents of New England. New England was the only region of the United States in which the banks had not suspended specie payments, so New England bank notes were trading at par with specie. Unfortunately, most of the federal government’s revenue was received in the mid-Atlantic region. The mid-Atlantic banks had suspended, and their notes were trading well below par. The Treasury insisted on accepting local cur rency at par in payments to the federal government, and on disbursing such currency only where it would be ac cepted at par. This forced it into temporary default on its debt service payments, despite its large balances of midAtlantic bank notes. See Catterall (1902),pp. 4-7, and Ban croft (1831), pp. 47-49. FEDERAL RESERVE BANK OF ST. LOUIS responsibility of receiving and clearing the pay ments, found itself wedged very uncom fortably in betw een. If the Bank accepted discounted state bank notes at par for governm ent payments, the Treasury would insist that it clear them at par, and the Bank would take large exchange losses. If it did not accept state bank notes at all, it would at the very least offend the people of the South—a region w h ere it greatly desired to extend its busi ness—and might well materially increase their econom ic troubles. The Bank would also offend the Treasury, its patron, w hich wished to ensure that federal payments could be made in currency readily available to the public.102 Finally, if the Bank accepted state bank notes at th eir m arket rates of discount, it would be accepting a situation which, it believed, prevented it from operating profitably and effectively in the region.103 The Bank’s problem s would have been solved if the southern banks had resum ed specie payments promptly after the Panic. W hen they did not, it launched a campaign to force them to do so. It continued to accept discounted state bank notes in payment of federal debts and, w hen it had accu mulated them in large quantities, presented them at the counters of the state banks fo r payment. W hen the southern banks refused to pay, the U.S. Bank filed suit against them in federal cou rt.104 While the suspensions had no formal legal validi ty—in principle, each bank note was redeemable at par and on demand, and a bank which declined to redeem its notes had defaulted on its debts—they were implicitly (and sometimes explicitly) tolerated by state legislatures and courts. The U.S. Bank, how ever, was in a position to sue in federal courts, which provided the state banks no such protection.103 102For a general description of the U.S. Bank’s problems as a federal collection and clearing agent, see Catterall (1902). For accounts of its disputes with the banks of Ge orgia and North Carolina, see Govan (1937), Heath (1954), Holder (1937) and Russell (1989a). These accounts are based on correspondence between the U.S. Bank, the Treasury Department, and the state banks that is recorded in the American State Papers, Finance, Volume 4. 103See note 90 above. 104Govan (1937) and Russell (1989a) describe federal court suits filed by the U.S. Bank against state banks in Georgia and North Carolina, respectively. 105See Hammond (1957), pp. 283-84, and Russell (1989a), pp. 42-43. The efforts of various southern and western states (notably Georgia) to evade adverse federal court de cisions concerning the U.S. Bank were ultimately rejected by the U.S. Supreme Court. See Catterall (1902), pp. 88-91, Govan (1937), and Hammond (1957), pp. 263-68, 272-73. 53 The efforts of the U.S. Bank slowly forced the banks of the various southern states to resume payments: Virginia and South Carolina in 1823, Georgia in 1825, Alabama in 1827 and North Caro lina in 1828. In view of the conventional wisdom regarding suspensions, it should come as no sur prise that most economic historians have regarded the Bank’s resumption campaign as virtuous and constructive; the Bank is lauded, in particular, for having created a “uniform national currency.’’106 Southern farmers, and other southerners whose livelihood was based, directly or indirectly, on farming—groups which collectively comprised the bulk of the region’s population—had less reason to sing the Bank's praises. The cost of resumption was that state banking systems (or those portions of them which survived) became reluctant to lend to farmers. Since there were few alternative sources of credit available, farm loans becam e substan tially m ore difficult to obtain.107 It should thus be equally unsurprising that, just a few years later, the southern states w ere in the forefron t o f the opposition to the Bank’s efforts to secure a new ch a rte r.108 A B a n k W ar a n d th e B is e o f F r e e B a n k in g S y n o p s is : The political controversy that led to the demise of the Bank of the United States had a profound effect on public attitudes tow ard the banking and cu rren cy system. The American public becam e suspicious of any hint of monopoly pow er in banking and of any link betw een the federal governm ent and private banks. It also becam e increasingly devoted to the concept of hard money. One outgrow th of these attitudes was that many states adopted laws providing for “Free Banking.” Free banking laws encouraged entry into banking, and resulted in the establish ment of large num bers of banks. The free banks w ere heavily regulated, how ever; their notes w ere to be carefully secured, and strictly convertible. O ther results of changed public attitudes w ere the "Specie Circular” and the "Independent T reasu ry.” The federal government became reluctant to accept 106See in particular Catterall (1902), Redlich (1951), Ham mond (1957) and Temin (1969). 107For the case of North Carolina, see Russell (1989a), pp. 71-74. In 1828, Georgia established the Central Bank, a state-owned institution designed to extend long-term loans to farmers and planters. Heath (1954) ascribes this deci sion to a shortage of long-term commercial bank credit that developed during the 1820s. b an k c u r re n c y in p aym en t, and attem p ted to conduct its financial operations w ithout the aid of banks. The election of 1828 tran sferred control over the federal adm inistration from the Whigs, led by defeated President Joh n Quincy Adams of Mas sachusetts, to the Dem ocrats, led by President elect Andrew Jackson of Tennessee. The Demo crats regarded them selves as the party o f the com mon (and thus largely agricultural) people; they had long been advocates of competitive, decentral ized state banking and opponents of the Banks of the United States. T he party also contained a hard money faction which was deeply suspicious of banking of any kind. Jackson him self seems to have had somewhat ambiguous feelings tow ard banking. On one hand, many of his principal advisors w ere m en who had defended the relief banks and bitterly resented the damage the U.S. Bank had done to the banks and people of the w estern states. Some of these men w ere now con nected with the state banks, and thus tended to form ulate policies w hich favored their interests vis-a-vis those of the U.S. Bank. On the oth er hand, Jackson is said to have been personally opposed to banking and paper curren cy of any sort; late in his administration he took actions that greatly in creased the problem s of the state banking system s.109 Jackson ’s State o f the Union message in 1829 cam e out against the rech a rter of the United States Bank. (The Bank's ch arter did not expire until 1836, but its friends in Congress had begun to agitate fo r an early rech arter act.) Jackson argued that the Bank’s constitutionality was doubtful, and that its concentrated financial pow er was inconsistent with the tenets of representative dem ocracy. Jackson ’s message m arks the beginning of the "Bank W a r,” a period of five years or so during w hich the W higs (who controlled Congress) attem pted to defend the Bank against the increasingly vituperative attacks of the Dem ocratic adm inistration. During this period, the "Bank question” becam e the single biggest issue in national politics. The Dem ocrats 108Catterall (1902), pp. 164-65, 235, Hoyt (1960) and Wilburn (1967), pp. 7-11, 17-19, describe southern support for the Bank War. 109For descriptions of Jackson’s attitude toward banking in general, and the United States Bank in particular, see Cat terall (1902), pp. 182-85, Schlesinger (1953), pp. 76-78, Redlich (1951), pp. 162-71, Hammond (1957), pp. 346-50, and Temin (1969), chapter 2. SEPTEMBER/OCTOBER 1991 54 simultaneously exploited and encouraged public feeling against the banks so as to build up their political pow er base at the Whigs' expense. After Jackson was re-elected in 1832 the issue cam e to a head; Congress passed legislation rech artering the Bank; Jackson vetoed the legislation, and Congress failed (narrowly) to override his veto. The second U.S. Bank effectively ceased to exist as a national institution w hen its federal ch a rter expired in 1836; its relationship w ith the federal governm ent had been severed, and its pow er thus greatly reduced, two years earlier.110 The demise o f the U.S. Bank raised two im por tant questions: how should the federal govern m ent adm inister its financial affairs, and how should the states regulate their banking systems? T he first question arose because the U.S. Bank had acted as the financial agent fo r the U.S. T reasury— in other words, as the federal governm ent’s bank. The second arose because the U.S. Bank no longer existed to regulate the state banks and, in particu lar, to en force specie payments. By the second half of the 1830s, Jacksonian hard-money notions had becom e so pervasive that few people w ere p re p ared to a c c e p t a re tu r n to rela tiv ely la iss e z f a i r e b an k in g , o r to a c u r re n c y w h ich m ight com e to co n sist larg ely of in co n v ertib le b an k n o tes. T h e e ffe c t of th e Ja c k so n A d m in istra tio n ’s an ti-U .S. B an k cam p aig n on a fin a n cia lly u n so p h istica ted p u blic had b een to e x a c e rb a te its dou bts ab ou t and su sp icion of all b a n k s and all p a p er c u r r e n c y .111 The Jackson and Van Buren Administrations [Martin Van Buren, who was elected president in 1836, had been Jackson’s vice president] responded to the first problem w ith tw o policy initiatives. The first was an executive order called the Specie Circular, issued in 1836, shortly b efore Jackson left office. This ord er directed the U.S. Treasury to accept no cu rren cy other than specie in pay m ent of debts to the federal governm ent.112 The second policy initiative was the Independent T rea sury Act, a product of the Van Buren Administra 110For accounts of the Bank War, see Catterall (1902), Ham mond (1957) and Schlesinger(1953), among many others. 111For a description of the Loco Focos (formally, the Equal Rights Party), a party of antibanking radicals which arose during the Bank War and was ultimately instrumental in the rise of Free Banking and the Independent Treasury, see Redlich (1951), pp. 188-90, Hammond (1957), pp. 493-99, and Schlesinger (1953), chapters XV-XVI. " 2For analyses of the motivation behind, and/or the impact of, the Specie Circular, see Hammond (1957), pp. 455-57, Schlesinger (1953), pp. 129-31, Smith (1953), p. 185, Temin Digitized for FEDERAL FRASER RESERVE BANK OF ST. LOUIS tion. This Act w ithdrew the U.S. governm ent’s cash deposits from the large state banks (some times called pet banks) w here they had been placed after the demise of the U.S. Bank. H enceforth, the Treasury would act as its own banker. The net effect of these tw o actions was profoundly anti banking. The federal governm ent would no longer deal with the state banks, or encourage (or even recognize) their note circulation.113 The states’ response to the problem o f bank regulation reflected tw o features of public atti tudes tow ards banking w hich had grow n out of the Bank W ar: suspicion of concentrated financial pow er, and p referen ce fo r "hard m oney.” Under the bank chartering system that existed in most states prior to the late 1830s, the issuance of a ch arter required a special act of the state legisla ture. T he process of securing legislative assent was lengthy, cum bersom e, uncertain and occa sionally corrupt; in addition, banks that already possessed ch arters generally lobbied vigorously against the issuance of new ones. The upshot was that most states had a relatively small num ber of banks, and that these banks possessed, or w ere believed to possess, considerable m arket p ow er.114 The "fre e ” in the free banking laws reflected the desire to reform the chartering process in the direction of free entry. Each state would form u late a standardized ch arter, and any individual or group w hich was able and willing to m eet the term s of this ch arter could organize a bank in that state. The free banks would be regulated by the state auditor, or by a state banking agency created for that purpose; the legislature would be involved only indirectly. This system was intended to greatly increase the num ber of state banks.115 T h ere is another sense in which the term "free banking” is a m isnomer, how ever. Free banking was not in any sense la issez -fa ire or unregulated banking. The standardized ch arters imposed num erous and relatively stringent restrictions on the banks’ capitalization and reserves, the condi- (1969), pp. 120-28, Timberlake (1960) and Timberlake (1978), chapter 5. 113For discussions of the role of the Independent Treasury see Hammond (1957), pp. 542-45, Taus(1943) and Timberlake (1978), chapter 6. 114See Hammond (1963), pp. 7-8, (1957), pp. 577-80, Redlich (1951), pp. 188-90, and Schlesinger (1953), p. 286. 115For discussions of the rationale behind Free Banking see Redlich (1951), pp. 187-204, and Hammond (1957), pp. 572-80. 55 tions under w hich they could issue notes, and the types of assets they could hold. These restrictions were designed, in the “hard money” spirit, to ensure that the banks' notes would always be convertible on demand, and that they would be relatively immune from losses associated w ith declines in the value o f the banks' assets. One restriction was so common that it has com e to be regarded as characteristic of free banking: the notes of the free banks had to be 100 p ercen t backed by hold ings of state or federal governm ent secu rities.116 Free banking experienced some problem s, espe cially during its early years. It seems in particular to have been characterized by a relatively high rate of bank failures. These failures, and a few notorious instances of fraud, have given the sys tem a bad reputation among historians.117 There w ere also many complaints that the need to keep track of the m arket values of the many different types of bank notes in circulation—some of which w ere counterfeit or issued by failed or insolvent banks—materially reduced the effectiveness of bank curren cy as a medium of exchange.118 Recent research has revealed that the losses to noteholders associated with free bank failures w ere actually quite small on average.119 Ironically enough, many of the losses and failures that did occu r w ere caused by defaults on state govern m ent bonds, w hich w ere w idespread during the 1840s and 1850s.120 Entrepreneu rs responded to the diversity of bank curren cies by publishing “b an k n o te r e p o r te r s ” in w h ich th e value o f different bank curren cies w ere recorded, insol vent banks identified, and the appearance of com m on counterfeits described.121 The system w orked well enough that, until the Civil W ar broke out, th ere seem s to have been little political support for any federally dominated alternative. Additional evidence that contem porary legislators 116See Rockoff (1975), pp. 2, 81-87, and Rolnick and Weber (1984), (1988). l17For unsympathetic descriptions of the record of Free Banking, or references to such descriptions, see Cagan (1963), pp. 19-21, Friedman (1960), p. 6, Hammond (1957), pp. 723, 741-42, (1963), Rockoff (1975), pp. i-ii, and Rolnick and Weber (1983). 119See Cagan (1963), pp. 19-20, Hammond (1957), pp. 722-23, (1963), p. 14, and Rockoff (1975), pp. 26-29. ' 19Rockoff (1975) and Rolnick and Weber (1983,1988) use data drawn from state auditors’ reports to provide careful esti mates of the costs of free bank failures. Rockoff estimates that the total losses endured by free bank noteholders from 1836 through 1860 were less than $2 million—about the cost, viewed free banking as viable is provided by the fa c t th a t n ew states co n tin u e d to adopt fre e banking laws through the early 1860s. Rockoff (1975) notes that “on the eve of the Civil W ar over half the states, including the most populous states, had free banking law s.”122 T h e Civil W ar a n d th e D em ise o f State B a n k C u r re n c y S y n o p s is : The changes in the American cu r rency system produced by the Civil W ar w ere almost as profound as those produced by the Rev olutionary W ar. W hen the Civil W ar began, the U.S. had a relatively decentralized system in which paper cu rren cy was issued by state-chartered and -regulated banks; the federal governm ent had no role in the provision of paper money, and there seemed little prospect that it would acquire one. W hen the w ar ended, the nation had a relatively centralized system in w hich paper cu rren cy was issued by federally ch artered and regulated banks; it was no longer possible for state banks to issue paper money. In addition, the federal governm ent had acquired and used the power to issue paper money and to m ake it legal tender. This huge, rapid transform ation was possible for tw o rea sons. First, the w ar produced a dram atic shift in the balance of pow er betw een the m ajor political parties, and betw een state and federal govern ments. Second, the w ar produced an unp rece dented need for federal governm ent revenue. The demise o f free banking, and m ore broadly of the system of decentralized, state-regulated provision of curren cy, was caused less by any problem s this system may have experienced than by the outbreak of the U.S. Civil W ar. This extra ordinary political event created a pair of peculiar circum stances w hich helped determ ine the future in 1860, of a single year of 2 percent inflation. The estimates provided by Rolnick and Weber are slightly higher (see their 1983 paper, p. 1089). Cagan (1963), by contrast, reports with apparent endorsement an estimate by Jay Cooke that the losses were $50 million per year. Cagan’s use of this esti mate in his influential article served to reinforce the conven tional view that free banking was a national disaster. 120See Rockoff (1975) and Rolnick and Weber (1982,1983,1984,1985,1988). 121See Hammond (1963), p. 14, Temin (1969), p. 50, and Rockoff (1975), pp. 23-25. 122Rockoff(1975), pp. 2-4. SEPTEMBER/OCTOBER 1991 56 evolution of the U.S. cu rren cy and banking sys tem. The first was that it gave the Republican party, the political successor of the old Whig party, almost com plete control over the federal govern ment. The m ajority o f the Dem ocratic congress men and senators w ere southerners who abandoned their seats and defected to the Confederacy at the outbreak of hostilities. Since the Republicans had inherited from the W higs a p referen ce for m one tary and financial centralization, while the Demo crats rem ained the party of decentralization, the defection o f the D em ocrats greatly increased the prospects fo r centralizing change in the m onetary system. The second determ ining circum stance was the federal governm ent’s need for enorm ous new sources of revenue. Though taxes played some role in Civil W ar finance, particularly in the w ar’s later years, it was clear almost from the beginning that the bulk of the expenditure burden would be m et by borrow ing.123 As the w ar’s cost mounted, how ever, the federal governm ent began to experi ence difficulty obtaining the sums required on the open securities m arket, and turned for assistance to the large Eastern banks.124 Unfortunately, the sum required by the governm ent far exceeded the banks' aggregate specie holdings; when the specie borrow ed by the governm ent failed fo r various reason to retu rn to the banks as rapidly as they had anticipated, they w ere forced to suspend spe cie paym ents.125 T he suspension seemed to close o ff the possibility of fu rth er bank loans, leaving the federal governm ent desperate for new sources of funds and for new ways to increase the demand for its debt. One way in w hich the governm ent might "b o r row ” was by issuing paper currency. Early in 1862, at the urging of Secretary of the Treasury Salmon 123For the government’s reluctance to raise taxes during the early years of the war, see Mitchell (1903), pp. 16-19, p. 37, pp. 72-73. 124The federal government’s financing problems were greatly exacerbated by its reluctance to sell its securities, which typi cally returned interest at the rate of 6 percent of face value, at a discount—its reluctance, that is, to borrow at interest rates in excess of 6 percent. See Mitchell (1903), pp. 48-50, 64-65. 125For accounts of the suspension see Mitchell (1903), pp. 37-43 and Hammond (1970), pp. 150-59. 126The provisions of the first and the two subsequent Legal Tender Act(s) are summarized by Mitchell (1903), pp. 44-118. 127ln 1875,10years after the end of the war, Congress enacted legislation making the greenbacks convertible in specie, according to the prewar definition of the dollar, beginning in 1879. This legislation became known, somewhat mislead Digitized forFEDERAL FRASER RESERVE BANK OF ST. LOUIS P. Chase, Congress passed the first Legal Tender Act. The Act authorized the T reasury Departm ent to issue $300 million in paper cu rren cy .126 This curren cy was not convertible in specie, nor re deem able in specie at any fixed future date; it was, however, made legal tender in paym ent of public a n d p r iv a te d e b ts .127 L a te r in th e w ar, c o n s id e r ab le ad d itional q u an titie s of th e se g re e n b a ck s w ere issued. The Legal Tender Acts m arked the first time in (post-revolutionary) U.S. history that the federal governm ent had issued fiat cu rren cy —curren cy which was entirely irredeem able, and was legal tender for private debts. The Constitution did not give the federal governm ent any explicit right to issue paper cu rren cy (fiat or otherwise), or to make paper cu rren cy legal tender; indeed, it con tained language which was widely interpreted as implicitly denying the governm ent these rights.128 This made the legality of the Acts seem very doubtful. In the event, how ever, no attem pt to challenge them managed to reach the Suprem e Court until several years after the Civil W ar had ended. T h ere ensued one of the m ore bizarre episodes in U.S. m onetary history. Chase, who was now Chief Justice, voted with the cou rt m ajority to strike down legislation whose form he had ap proved, and whose passage he had recom m ended, w hen he was Secretary of the Treasury! The greenbacks w ere saved w hen Congress, w hich rem ained dominated by the Republicans, voted to increase the num ber o f Supreme Court justices by two; President Grant acted quickly to fill the re sulting vacancies with judges who supported the Acts' constitutionality. The enlarged court voted 5-4, w ith Chase dissenting, to uphold the A cts.129 This decision set a precedent that was later used to justify fu rth er steps on the part of the federal governm ent to regulate or control the curren cy ingly, as the “ Resumption Act.” [The history and provisions of the Resumption Act are summarized by Friedman and Schwartz (1963), pp. 44-50. See also Timberlake (1978), chapter 8.] 128For discussions of the constitutionality of legal tender federal currency, see Breckinridge (1903), pp. 114-37, Mitchell (1903), pp. 51-55, Hurst (1970), pp. 182-86, Christainsen (1988), Timberlake (1989), (1991), chapters 8-10. ' 29See Breckinridge (1903), pp. 127-137, Friedman and Schwartz (1963), pp. 46-47, Mitchell (1903), pp. 71-74, and Timberlake (1989), (1991), chapter 9. 57 sy ste m —n otably, the c re a tio n o f th e F ed eral Reserve System. The greenbacks m arked a change in the U.S. curren cy system which w ent beyond governm ent issuance of currency. As has already been noted, the fiscal crisis of late 1861 caused the private banking system to suspend specie payments. On earlier occasions w hen there had been a national or regional suspension, most of the surviving banks had refused to redeem their notes in any way until they could resum e redeem ing them in specie on demand. The fact that the greenbacks w ere legal tender, how ever, gave bank debtors the legal right to use them to repay their debts. Since greenbacks began trading at substantial discounts shortly after they w ere first issued, m oreover, debtors hastened to take advantage of this opportunity. The banks consequently felt enti tled, and indeed compelled, to redeem their notes in greenbacks.130 T h ereafter, the banking system redeem ed its notes and deposits almost exclusively in legal tend er paper currency, regardless of w hether it was convertible in specie. This ensured that goods and assets would be priced in legal tend er paper—in other words, that governm ent cu rren cy would replace specie as the nation’s unit of account. Because neither Congress nor the Administra tion was willing to risk a repetition of the Revolu tionary hyperinflation, greenbacks could be used to finance at most a small portion of the w artim e deficit.131 T he balance of the deficit had to be financed by the issuance of conventional, interestbearing debt.132 Secretary Chase responded to this situation by developing a strategy fo r m onetary reform w hich promised to simultaneously achieve both cu rren cy centralization and debt demand enhancem ent. This program , w hich was ultim ate ly embodied in the National Banking Act, called for the creation of a "National Banking System ” (NBS)—a system of private, federally chartered banks which would be nationally regulated ana logues of the state-chartered “free banks.” The Act imposed reserve, capital, convertibility and other requirem ents that w ere generally similar to those >30See Mitchell (1903), pp. 144-49. 131The limited financial resources of the southern states made it difficult for the Confederate government to borrow large sums. As a result, it was forced to cover a very substantial fraction of its deficit through currency creation. This strategy ultimately produced a hyperinflation. See Lerner (1956), and Timberlake (1978), pp. 102-03. 132Mitchell (1903), pp. 119-31, discusses the reasons the federal government issued no greenbacks from 1863 to the end of the Civil War. imposed on the free banks. These requirem ents w ere to be adm inistered, and the national banks regulated, by a new federal agency called the O ffice of the Com ptroller of the C urrency.133 For the first time in U.S. history, the federal govern m ent had moved to create a system of private banks (rather than a single, centrally administered private bank) under its direct regulatory control. From the perspective of the cu rren cy system, the key features o f the National Bank Act involved the notes the national banks w ere to issue. These notes w ere to be printed by the Treasury Depart ment and issued to the banks, rath er than printed by the banks directly; they w ere to look entirely uniform , except for an indication of the identity of the issuer. In order to obtain a given value of bank notes, a national bank had to deposit U.S. govern ment securities of essentially equal value (state governm ent securities would not do) with the Com ptroller of the Currency. Thus, national bank notes w ere to be 100 percent backed by U.S. gov ernm ent securities. The requirem ent th at the notes be backed by fe d e r a l governm ent securities was designed to create a “captive” demand fo r federal debt on the part of banks of issue. Since notes w ere the prin cipal liabilities of contem porary banks, and since the fram ers of the Act evidently expected most of the state banks to apply for federal charters, there was every reason to expect that the Act would force the banking system to purchase Treasury securities in large quantities. This, it was hoped, would materially ease the federal governm ent’s borrow ing problem s.134 Congress, anticipating heavy demand for nation al bank notes, included provisions in the Act estab lish in g a m axim um q u an tity w h ich could be issued and allocating this quantity across the various regions.135 Contrary to expectation, how ever, during the year or so after the Act was passed the num ber of ch arte r applications was small, and the volume of U.S. bonds deposited as note backing was far low er than anticipated. Congress 133Forthe history and provisions of the National Banking Act, see Redlich (1951), pp. 99-113 and Hammond (1970), chapters 10-11. 134Discussions of the budgetary motivations for the National Banking Act can be found in Hammond (1957), p. 725, (1970), pp. 135-36, 286-95, and Mitchell (1903), pp. 37, 44-45, 103, 109. 135These provisions are summarized by Redlich (1951), p. 118. SEPTEMBER/OCTOBER 1991 58 responded by amending the Act to impose a puni tive tax on state bank notes—a tax rate so high (10 percent) that it made note issue by state banks entirely u nprofitable.136 This decision eliminated state bank notes from circulation, and m arked the final demise of state curren cy systems. Of course, a system under w hich any curren cy that was not issued directly by the federal govern m ent was printed by the federal governm ent, fully backed by U.S. Treasury securities, and issued in quantities and locations closely regulated by the governm ent, might be said to have differed very little from a system under w hich the federal gov ernm ent directly issued all paper cu rren cy .137 Indeed, it could be argued that the only really significant differences betw een the NBS and a direct note issue system w ere that under the NBS, the governm ent (1) had little short-run influence over the total quantity of notes (which indeed proved relatively unresponsive to short-run influ ences of any kind), and (2) assigned the responsi bility fo r clearing the notes (and thus fo r ensuring their convertibility in specie) to the issuing banks. W hen 50 years o f experience with the system seemed to suggest that these features w ere serious liabilities, the federal governm ent used the broad m onetary pow ers it had acquired during the Civil W ar to establish a system of direct issue—the Fed eral Reserve System. CONCLUDING REM ARKS The changes in the U.S. curren cy system that resulted from the Legal T en d er and National Bank ing acts stand, along w ith the m onetary clauses of the U.S. Constitution, as classic examples of cases in which the basic stru ctu re of the system was strongly influenced by extraordinary political events with largely non-econom ic (or at least, non monetary) causes. If, as we have seen, the American colonies could have obtained their independence from Great Britain without fighting a long, expensive and divisive revolutionary war, the monetary his tory of the next 90 years might have been very different: historical evidence suggests that the 136See Friedman and Schwartz (1963), pp. 18-19, Hammond (1957), pp. 732-34, and Redlich (1951), p. 113. 137This point is made by Friedman and Schwartz (1963), p. 21. 138See note 89 above. 139Redlich (1951) comments that had Secretary Chase promoted the National Banking Act less vigorously, the early Digitized for FEDERAL FRASER RESERVE BANK OF ST. LOUIS states might have retained the right to issue their own currencies, and that these curren cies might not have been convertible in specie. Subsequently, if the American states could have resolved their sectional disputes without fighting a long, expen sive and divisive civil war, the m onetary history of the ensuing 50 years might also have been very different: historical evidence suggests that the states might have retained the right to ch arter banks o f issue, and that the federal governm ent’s role in the U.S. cu rren cy system might have rem ained relatively limited. It should also be remembered that the federal government chartered the second United States Bank in response to financial dislocations associated with the W ar of 1 8 1 2 .138 In the absence o f Second Bank opposition, the inconvertible banking sys tem s that arose in the w estern and southern states after the Panic of 1819 might have survived and becom e entrenched; without a Second Bank for the Jacksonians to fight, the "hard m oney" prin ciples of the Free Banking Era might never have becom e popular. Here again a sequence of essen tially political disputes played a key role in dictat ing the evolution of U.S. currency arrangements.139 C u r re n c y S y stem E v o lu tio n : An A lterna tiv e View As we have seen, U.S. m onetary history has been punctuated by a sequence of rath er abrupt transitions from one curren cy system to another with very different features. These transitions are often in terp reted as p a rt of a process of Darwinian advancement—a process, that is, through which old and relatively inefficient systems w ere replaced by new and m ore efficient successors. The modern cu rren cy system em erged out of this process as the most efficient system yet devised. W hile this historical interpretation certainly sounds plausible, it is one that we should accept or reject on the basis of evidence concerning the rela tive efficiency of past and presen t cu rren cy sys tems. Unfortunately, the prestige of Darwinism Civil War suspension also might have led to the development of a currency system based on inconvertible bank notes (p. 95). 59 has becom e so great that econom ists tend to reverse the logical process by using the various system s’ orders in the historical sequence as the basis for efficiency com parisons. (If only the fittest systems survive, then the systems that survived at each stage must have been the fittest.) The disap pearance of older systems is regarded as com pel ling evidence that they w ere less efficient than their successors. The claim that the cu rren cy system, if left to itself, tends to progress (slowly) in the direction of greater efficiency is not disputed in this article; indeed, several examples of this sort of progres sion have been presented above. W hat the article has argued is that the U.S. cu rren cy system has not been left to itself, and that its evolution has been anything but an orderly and inevitable progression tow ard econom ic efficiency. Instead, it has been dominated by political decisions that w ere largely uninfluenced by efficiency consider ations. Many of these decisions w ere made in response to political pressures of a particularly urgent sort—pressures growing out of the U.S. governm ent’s (and earlier, the British and/or colonial governm ents’) involvement in prolonged and expensive wars. It is, of course, possible that we have been fortunate, and that the political process has given us a cu rren cy system that is very efficient, or at least m ore efficient than the historical alterna tives. It is also possible that w e have not been quite so fortunate; the question is a complex one, and cannot be answ ered here. This article is content to point out that the m odern cu rren cy system has not developed because of any clear advantage in efficiency. The possibility that history provides attractive alternatives cannot be ruled out, and the question of which system is best rem ains both open and interesting. REFERENCES Arnold, Samuel G. History of the State of Rhode Island and Providence Plantations, vols. I and II, (D. Appleton, 1878). Bancroft, George. “ Bank of the United States,” North American Review (January 1831), pp. 21-64. Bates, Frank Greene. “ Rhode Island and the Formation of the Union,” Columbia University Studies in History, Economics and Public Law, vol. X, no. 2 (The MacMillan Company, 1898). Berry, Thomas S. Western Prices Before 1861 (Harvard University Press, 1943). Billias, George A. “ The Massachusetts Land Bankers of 1740,” University of Maine Studies, Second series, Number 74 (Maine University Press, 1959). Breckinridge, S. P. Legal Tender: A Study in English and American Monetary History (University of Chicago Press, 1903). Brock, Leslie V. The Currency of the American Colonies 1700-1764 (Arno Press, 1975). Bullock, Charles J. Essays on the Monetary History of the United States (The MacMillan Company, 1900). Cable, John R. “ The Bank of the State of Missouri,” Studies in History, Economics and Public Law, vol. 102 (Columbia University Press, 1923). Cagan, Phillip. “ The First Fifty Years of the National Banking System—An Historical Appraisal,” in Deane Carson, ed., Banking and Monetary Studies (Richard D. Irwin Inc., 1963). Calomiris, Charles W. “ Institutional Failure, Monetary Scarci ty, and the Depreciation of the Continental,” Journal of Economic History (March 1988), pp. 47-68. Calomiris, Charles W., and Charles M. Kahn. “ The Role of Demandable Debt in Structuring Optimal Banking Arrange ments,” American Economic Review (June 1991), pp. 497-513. Carothers, Neil. Fractional Money (Augustus M. Kelley, 1930), (reprinted 1967). Catterall, Ralph C. H. The Second Bank of the United States (University of Chicago Press, 1902). Checkland, S.G. Scottish Banking: A History, 1695-1973 (Glasgow: Collins, 1975). Christainsen, Gregory B. “ Fiat Money and the Constitution: A Historical Review,” in Thomas D. Willett, ed., Political Busi ness Cycles (Duke University Press, 1988). Clapham, Sir John. The Bank of England, A History, vols. I and II (Cambridge University Press, 1944), (reprinted 1958). Clough, Shepard B. and Charles W. Cole. Economic History of Europe (Heath, 1941). Cole, Arthur H. Wholesale Commodity Prices in the United States, 1700-1861 (Harvard University Press, 1938). Davis, Andrew McFarland. “ Currency and Banking in the Province of Massachusetts-Bay,” American Economics Association Publications, Third series, Part I: “Currency,” Volume I, Number 4, December 1900, Part II: “ Banking,” Volume II, Number 2, May 1901 (The Macmillan Company). Dewey, Davis R. State Banking before the Civil War, National Monetary Commission, (U.S. Government Printing Office, 1910). Dorfman, Joseph. The Economic Mind in American Civiliza tion, 1606-1865, vols. I and II (London: George G. Harrap & Company LTD., 1947). Dowd, Kevin. “ Option Clauses and the Stability of a Laissez Faire Monetary System,” Journal of Financial Services Research (December 1988), pp. 319-33. Dowrie, George W. “The Development of Banking in Illinois, 1817-1863,” Studies in the Social Sciences, vol. XI, no. 4 (University of Illinois Press, December 1913). Dwyer, Gerald P., and R. Alton Gilbert. “ Bank Runs and Private Remedies,” this Review (May/June 1989), pp. 43-61. Ernst, Joseph A. Money and F>olitics in America, 1755-1775 (The University of North Carolina Press, 1973). Felt, Joseph B. An Historical Account of Massachusetts Cur rency (Bun Franklin, 1839). Fenstermaker, J. Van. The Development of American Com mercial Banking: 1782-1837 (Kent State University Press, 1965a). _______ . “ The Statistics of American Commercial Banking, 1782-1818,” Journal of Economic History (1965b), pp. 400-13. SEPTEMBER/OCTOBER 1991 60 Ferguson, E. James. “ Currency Finance: An Interpretation of Colonial Monetary Practices,” William and Mary Quarterly, 3rd series, vol. X (1953), pp. 153-80. ________The Power of the Purse (University of North Caroli na Press, 1961). Feavearyear, Sir Albert. The Pound Sterling: A History of English Money (Oxford at the Clarendon Press, 1963). Flanagan, Beecher. “A History of State Banking in North Carolina to 1866,” Doctoral dissertation, George Peabody College for Teachers, 1935. Fraas, Arthur. “ The Second Bank of the United States: An Instrument for an Interregional Monetary Union,” Journal of Economic History (June 1974), pp. 447-67. Franklin, Benjamin. Essays on General Politics, Commerce and Political Economy, Volume II, Part II of Jared Sparks, ed., The Works of Benjamin Franklin (Augustus M. Kelley, 1971). Friedman, Milton. A Program for Monetary Stability (Fordham University Press, 1960). Friedman, Milton, and Anna J. Schwartz. A Monetary History of the United States: 1867-1960 (Princeton University Press, 1963). Gallatin, Albert. The Writings of Albert Gallatin, Henry Adams, ed., vol. Ill (J. B. Lippincott & Company, 1879). Gouge, William M. A Short History of Paper-Money and Bank ing in the United States (T. W. Ustick, 1833). Gould, Stephen Jay. Wonderful Life: The Burgess Shale and the Nature of History (W. W. Norton & Company, 1989). Govan, Thomas P. “ Banking and the Credit System in Georgia, 1810-1860,” Doctoral dissertation, Vanderbilt Univer sity, 1937 (Arno Press, 1978). Greene, Jack P., and Richard M. Jellison. “ The Currency Act of 1764 in Imperial-Colonial Relations, 1764-1766,” William and Mary Quarterly, Third series, Volume XVIII (1961), pp. 485-518. Hammond, Bray. “ Long and Short term Credit in Early American Banking.” Quarterly Journal of Economics, vol. XLIX (1934), pp. 79-103. ________Banks and Politics in America from the Revolution to the Civil War (Princeton University Press, 1957). ________“ Banking Before the Civil War,” in Deane Carson, ed., Banking and Monetary Studies (Richard D. Irwin Inc., 1963), pp. 1-14. ________Sovereignty and an Empty Purse: Banks and Poli tics in the Civil War (Princeton University Press, 1970). Heath, Milton Sidney. Constructive Liberalism: The Role of the State in Economic Development in Georgia to 1860 (Harvard University Press, 1954). Hepburn, A. Barton. A History of Currency in the United States (Augustus M. Kelley, 1903), (reprinted 1967). Holder, Brantson B. “The Three Banks of the state of North Carolina: 1810-1872,” Doctoral dissertation, University of North Carolina, 1937. Holdsworth, John T., and Davis R. Dewey. The First and Second Banks of the United States (U.S. Government Printing Office, 1910). Horsefield, J. Keith. British Monetary Experiments, 1650-1710 (Harvard University Press, 1960). Hoyt, Elizabeth S. “ Reactions in North Carolina to Jackson’s Banking Policy, 1829-1832,” North Carolina Historical Review (1948), pp. 167-78. Huntington, A. T., and Robert J. Mawhinney. Laws of the United States Concerning Money, Banking, and Loans, 1778-1909, National Monetary Commission (U.S. Government Printing Office, 1910). Digitized forFEDERAL FRASER RESERVE BANK OF ST. LOUIS Hurst, James W. A Legal History of Money in the United States 1774-1970 (University of Nebraska Press, 1973). Kemmerer, Donald L. “The Colonial Loan-Office System in New Jersey,” Journal of Political Economy (December 1939), pp. 867-74. King, Robert G. “ On the Economics of Private Money,” Journal of Monetary Economics (July 1983), pp. 127-58. Klein, Benjamin. “The Competitive Supply of Money,” Journal of Money, Credit and Banking (November 1974), pp. 423-53. Knox, John Jay. A History of Banking in the United States (Rhodes & Company, 1903). Lerner, Eugene M. “ Inflation in the Confederacy, 1861-1865,” in Milton Friedman, ed., Studies in the Quantity Theory of Money (Uriwerstiy of Chicago Press, 1956). Libby, Orin G. The Geographical Distribution of the Vote of the Thirteen States on the Federal Constitution 1787-8 (Univer sity of Wisconsin Press, June 1894). Macaulay, Thomas B. A History of England from the Ac cession of James II, vols. IV,V (Longmans, Green, Reader & Dyer, 1877). Macon, Hershal L. “A Fiscal History of North Carolina, 1776-1860,” Doctoral dissertation, University of North Caro lina, 1932. Metz, William D. “ Politics and Finance in Massachusetts, 1713-1741,” Doctoral dissertation, University of Wisconsin, 1945. Mitchell, Wesley C. A History of the Greenbacks (University of Chicago Press, 1903). Nettels, Curtis P. The Emergence of a National Economy, 1775-1815 (M. E. Sharpe, Inc., 1962). ________The Money Supply of the American Colonies Before 7720 (University of Wisconsin Press, 1934). Nevins, Allan. The American States During and After the Revolution: 1775-1789 (Augustus M. Kelley, 1924). North American Review. “ Bank of the United States” (April 1831), pp. 524-63. Phillips, Henry Jr. Historical Sketches of the Paper Currency of the American Colonies (W. E. Woodward, 1865). Redlich, Fritz. The Molding of American Banking: Men and Ideas, Parts I and II (Johnson Reprint Corporation, 1951). Rockoff, Hugh.“The Free Banking Era: A Re-examination,” Journal of Money, Credit and Banking (May 1974), pp. 141-67. ________The Free Banking Era: A Re-examination (Arno Press, 1975). ________ “ Institutional Requirements for Stable Free Bank ing Regime,” Cafo Journal (Fall 1986), pp. 617-34. Rolnick, Arthur J., and Warren E. Weber. “ Free Banking, Wildcat Banking, and Shinplasters,” Federal Reserve Bank of Minneapolis Quarterly Review (Fall 1982), pp. 10-19. _______ . "New Evidence on the Free Banking Era,” Ameri can Economic Review (December 1983), pp. 1080-91. _______ . “ The Causes of Free Bank Failures: A Detailed Ex amination,” Journal of Monetary Economics (November 1984), pp. 267-91. ________“ Banking Instability and Regulation in the U.S. Free Banking Era,” Federal Reserve Bank of Minneapolis Quarterly Review (Summer 1985), pp. 2-9. ________“ Explaining the Demand for Free Bank Notes,” Journal of Monetary Economics (January 1988), pp. 47-71. ________“ Gresham’s Law or Gresham’s Fallacy?” Journal of Political Economy (February 1986), pp. 185-99. 61 Rothbard, Murray. The Panic of 1819 (Columbia University Press, 1962). Russell, Steven. “ Some Implications of the Colonial Currency Experience,” Working paper, April 1988. ________‘“ Free Banking’ as Inconvertible Banking: North Carolina, 1819-1828,” Working paper, September 1989a. _______ . “ Inconvertible Banking in the Early Nineteenth Century United States: A Legal Restrictions Approach,” Working paper, October 1989b. _______ . “ The Pricing of North Carolina Bank Notes, 1818-1827: An Empirical Investigation,” Working paper, May 1989c. Santoni, G.J. “A Private Central Bank: Some Olde English Lessons,” this Review (April 1984), pp. 12-22. Schlesinger, Arthur M. Jr. The Age of Jackson (Little, Brown and Company, 1953). Schweitzer, Mary M. “ State-Issued Currency and the Ratification of the U.S. Constitution,” Journal of Economic History (June 1989), pp. 311-22. Scott, William Robert. The Constitution and Finance of English, Scottish and Irish Joint-Stock Companies to 1720, Three volumes (Cambridge: The University Press, 1910-12). Smith, Adam. The Wealth of Nations, Edited and introduced by Edwin Cannan (Random House, Inc., 1937, originally published in 1776.) Smith, Bruce D. “ Money and Inflation in Colonial Mas sachusetts,” Federal Reserve Bank of Minneapolis Quarterly Review (Winter 1984), pp. 1-14. ________“American Colonial Monetary Regimes: The Failure of the Quantity Theory and Some Evidence in Favour of an Alternative View,” Canadian Journal of Economics (August 1985a), pp. 531-65. ________“ Some Colonial Evidence on Two Theories of Money: Maryland and the Carolinas,” Journal of Political Economy (December 1985b), pp. 1178-211. Smith, Walter B. Economic Aspects of the Second Bank of the United States (Harvard University Press, 1953). Smith, Walter B., and Arthur H. Cole. Fluctuations in American Business, 1790-1860 (Harvard University Press, 1935). Sumner, William G. A History of Banking in All the Leading Nations, Volume I (The Journal of Commerce and Commer cial Bulletin, 1896). Sylla, Richard, John B. Legler, and John J. Wallis. “ Banks and State Public Finance in the New Republic: The United States: 1790-1860,” Journal of Economic History (June 1987), pp. 391-403. Taus, Esther Rogoff. Central Banking Functions of the United States Treasury, 1789-1941 (Columbia University Press, 1943). Taylor, Robert J. Western Massachusetts in the Revolution (Brown University Press, 1954). Temin, Peter. The Jacksonian Economy (W. W. Norton & Company, Inc., 1969). Thayer, Theodore. “The Land-Bank System in the American Colonies,” Journal of Economic History (Spring 1953), pp. 145-59. Timberlake, Richard H. Jr. “ The Specie Circular and Distri bution of the Surplus,” Journal of Political Economy (April 1960), pp. 109-17. _______ “ Denominational Factors in Nineteenth-Century Cur rency Experience,” Journal of Economic History (December 1974), pp. 835-50. _______ . The Origins of Central Banking in the United States (Harvard University Press, 1978). ________“ The Significance of Unaccounted Currencies,” Journal of Economic History (December 1981), pp. 853-66. _______ . “The Government’s License to Create Money,” Cafo Journal (Fall 1989), pp. 301-21. ________Gold, Greenbacks, and the Constitution (George Edward Durrell Foundation, 1991). U.S. Constitution. Viner, Jacob. Studies in the Theory of International Trade (Harper & Brothers, 1937). Wallace, Neil. “A Legal Restrictions Theory of the Demand for ‘Money’ and the Role of Monetary Policy,” Federal Reserve Bank of Minneapolis Quarterly Review (Winter 1983), pp. 1-7. White, Lawrence H. Free Banking in Britain: Theory, Ex perience, and Debate 1800-1845 (Cambridge University Press, 1984). ________“Accounting for Non-Interest-Bearing Currency: A Critique of the Legal Restrictions Theory of Money,” Journal of Money, Credit and Banking (November 1987), pp. 448-56. West, Robert Craig. “ Money in the Colonial American Econo my,” Economic Inquiry (January 1978), pp. 1-15. Wilburn, Jean Alexander. Biddle’s Bank: The Crucial Years (Columbia University Press, 1967). SEPTEMBER/OCTOBER 1991 F e d e ra l R e s e rv e B a n k of St. L o u is Post Office Box 442 St. Louis, Missouri 63166 The R e v ie w is published six times p er year b y the Research and Public Inform ation Department o f the Federal R eserve Rank o f St. Louis. Single-copy subscriptions are available to the public f r e e o f charge. Mail requests f o r subscriptions, back issues, or address changes to: Research and Public Information Department, Federal R eserve Rank o f St. Louis, P.O. Roy 442, St. Louis, Missouri 63166. The views expressed are those o f the individual authors and do not necessarily reflect official positions o f the Federal R eserve Rank o f St. Louis o r the Federal R eserve System. Articles herein may b e reprinted provided the source is credited. Please provide the Rank’s Research and Public Inform ation Department with a copy o f reprinted material.