Vous êtes sur la page 1sur 15

Economic History Association

Monetary Theory and Roman History

Author(s): Marcello de Cecco
Source: The Journal of Economic History, Vol. 45, No. 4 (Dec., 1985), pp. 809-822
Published by: Cambridge University Press on behalf of the Economic History Association
Stable URL: http://www.jstor.org/stable/2121880 .
Accessed: 17/05/2014 02:36

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Cambridge University Press and Economic History Association are collaborating with JSTOR to digitize,
preserve and extend access to The Journal of Economic History.


This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
Monetary Theory and Roman History

In the study of Roman money Theodore Mommsenremains 135 years after his
work a toweringfigure,more pragmaticthan theoreticalin his economics, yet still
sound. He saw the politics in monetaryhistory, and especially its connectionwith
the strengthof the state. His view is more penetratingthan MV = PT, fashionable
in twentieth-centuryscholarshipon Rome. And it is bettereconomics thanoffered
by the Polanyi School as the alternativeline of analysis. The Polanyists infer an
absence of a Roman monetary system from the failure of some part to be as
sophisticatedas the best. On the contrary,the Romanmonetarysystem does not
look so differentfrom that of Europe since Mommsenwrote, uneven in its use of
monetarydevices, but sensibly so.

A history textbook used in the fourth grade of Italian elementary

schools, a living testimonialto the influenceof Benedetto Croce on
Italian education, says, "In the study of history we take from the past
only what is of interest to us." Modern historians of the Roman
monetary system and its vicissitudes have looked at Roman monetary
history throughthe filter of one or anotherversion of modernmonetary
theory. Roman monetary history, as a separate discipline, was more or
less invented by Theodore Mommsen, who published his Geschichte
des Romische Munzwesenin 1860.1Mommsen's work (which I quote in
the French translationof 1865)deserves the most attentive and careful
treatment, for the display of scholarship it presents, and because its
author demonstrates that he is not taking one or another monetary
theory for granted. He subjected monetarytheory as he had received it
to the test of Roman monetary history, and was capable even of
transcendingthe monetary knowledge of his time. With Mommsen we
retreatfrom viewing the Romans as primitives,and returnto a previous
era, when intellectuals approached Roman history with a feeling of
Journal of Economic History, Vol. XLV, No. 4 (Dec. 1985). ? The Economic History
Association. All rightsreserved. ISSN 0022-0507.
The authoris Professorof Economicsat the EuropeanUniversityInstitute,Florence,Italy, and
of MonetaryEconomics at Universitaidi Siena.
This paper was written while the author was a memberof the Institutefor Advanced Study,
Princeton.It would be perhapsmore appropriateto say thatit was writtenbecause the authorwas,
in the academic year 1983/84,a memberof the Institute. It was conceived as an experimentin
interdisciplinarywriting,to be subjectedto discussionsand criticismsby people who, specialistsin
fieldsas diverseas classics andeconomics, were presentat the Instituteat the sametime. For these
reasons, the paperwas read, at various stages of completionby practitionersof differentarts, like
Glen Bowersock, Albert Hirschman, Fergus Millar, Axel Leijonhufvud,Paolo Baffi, Sir John
Hicks, Stanley Fischer, Donald McCloskey, Carlo Cipolla, HartmutGalsterer. Most of these
learnedgentlemengave the authormany very useful comments, but they ought in no way be held
responsiblefor mistakeswhich remainin the paper,whichare to be attributedsolely to the author.
' Theodore Mommsen, Geschichte des Romische Mfinzwesen (Leipzig, 1860); French transla-
tion, Histoire de la Monnaie Romaine (Paris, 1865).


This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
810 De Cecco

Mommsen's preface expounds the main tenets of his monetary

theory. If we give the preface the attentionit deserves, and link it to the
numerous theoretical remarks scattered through the work, we can
dispel one misapprehensionabout Mommsen. He is supposed to have
been a convinced metallist. Sture Bolin, in his State and Currencyin the
Roman Empire, characterizes Mommsen's monetary theory so:
The Roman monetary system, as Mommsen envisaged it, resembles the system of
coinage based on a full metallic standardwhich prevailedin his own lifetime, the age of
classical Liberalism-or, rather, it is the same system. Mommsen has, quite simply,
shifted the monetary system of his own day two thousandyears back in time and, in
testing various ancient currencies,he has based his analysisand madehis classification
on principleswhich were valid for the coinage of his own time.2
Bolin was radicallywrong. In his preface, Mommsenfirmlyexpresses
the idea that it is the state that has the rightto establish "in the interest
of all, what we could willingly call a derogation to Natural Law, by
exclusively attributingto a particularsubstance the special privilegeof
representing the value of all others; only the State can compel all
citizens to accept the choice it has made, and it keeps the rightto change
its choice according to changed circumstances."3That in the course of
history the choice has fallen on gold, silver, and copper, afterall sorts of
other primitive substances had been used as money, Mommsenaffirms
soon after, but his is not a blind declarationof faith in metallism. The
importance of the state in the process could not be underlined more
Mommsen, however, is no simple Aristotelian, no blind believer in
the Staatlische Theorie des Geldes, no pure chartalist. He is convinced,
as he writes soon after in the preface, that "commerce and usage have
always preceded legislation" and "the changes introduced in the
currencies have always been among the main causes of trouble which
have agitated the domestic life of peoples." But, he says, "these
revolutions have never been the result of any positive law; but laws had
necessarily to intervene to put an end to crisis and consecrate to change,
so that we can say that the legal consecrationsof those revolutionsis for
history the only proof that they took place." Thus, "Romantradershad
been countingfor a long time in silver pounds, while copper was still the
only currency legally recognized by the State."4 And, he adds:
modificationsof this kind in the legislations were constantly motivatedby a powerful
impulsecomingfrom outside, and can always be consideredas the legal consecrationof
an accomplished fact. In Rome, the first issue of the silver denarius is linked to the
conquest of Italy, and the gold currency dates from the time when the Roman
governmentexchanged the dominationof Italy for sovereigntyover the world.5
2 Sture Bolin,
State and Currency in the Roman Empire (Stockholm, 1958), p. 12.
3 Mommsen,Histoire, p. xvi.
4 ibid., pp. xvii-xviii.

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
Roman Monetary Theory 811

Mommsenshows the differencebetween a unit of account, which can

be a metal ingot with a governmentseal, and a piece of currency, where
the state seal is not attributedindividuallyto each coin but to the whole
issue. The public must accept each coin without checking whether it
corresponds to what is declared on its face. The essence of money is
therefore, according to Mommsen, in its having a fixed nominal value.
And he says that, among Roman coins, only the Victoriatus is a piece of
coined metal circulating without a fixed nominal value. Mommsen, in
other words, is well aware of fiduciarycurrency, and knows, too, that a
mixed or fiduciary currency must be overvalued to remain in circula-
tion. Mommsen is convinced that the use of the power to issue coins
replaced the use of taxation. He is so conscious that the two financial
expedients are equivalent that he asserts on scant evidence that the
issuing of money was treated in the Roman constitution on the same
footing as taxation-that it was as limited, and was subjected to the
authorityof the Comitia Tributa. The same can be said of his treatment
of the origins of the tresviri monetales. And the fiscal function that
coinage performs in the Roman economic system through the perma-
nent overvaluationof Romancurrencyand the recurrentdebasingof the
same are seen by Mommsen, as we already noted, as a condition which
does not lead to inevitable ruin. If well managed, it can last indefinitely
and constitute the very essence of ancient economic policy, one
instrumentof ancient statecraft.
The fiscal approach to monetary policy is definitely not consistent
with the doctrine of nineteenth-centurypure laissez faire. During the
Punic Wars the Republic had forced more and more the circulationof
the copper coins. At the beginning of the Hannibalic War, noted
Mommsen, this de facto debasement was legalized by establishing a
1:112 silver-copperratio, comparedto the previous 1:250. Had Momm-
sen been a doctrinaire contractualist, who read Roman monetary
history throughthe spectacles of his own age, he would not have written
as follows: "The financial aim of this measure was to give to coined
copper, which was exchanged for silver in trade, a value double its
metallica value. This result . . . was adequate to the exceptional cir-
cumstances in which the State found itself at that moment: its finances
were damaged, but its credit remained still intact."6
Mommsen was equally aware of the redistributiveeffect of monetary
policies. He noted the revolutionaryeffect of the Lex Valeria of 84 B.C.,
by which the As Librale worth 1/10 of a denarius was suppressed and
replacedby anotherAs worth 1/16 of a denarius. Debtors were allowed
to repay old debts with the new As. This measure, he writes, was the
equivalent of a discount of 75 per 100 on the capital that had been lent
out, and was broughtabout merely by changingthe unit of account. "A
6 Ibid., vol. 2, p. 68.

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
812 De Cecco

legalized bankruptcy was thus consummated," he wrote, "But this

state of affairs was too violent and arbitraryto last. Sulla quashed the
Lex Valeria and reinstated the old unit of account."7
Mommsen's theoretical syncretismis apparentalso when he comes to
dealingwith the most interestingpartof Romanmonetaryhistory: "The
monetary crisis of the III century" and "The reorganizationof Money
in the IV century." These are the chapter headings he chooses for this
period, and his interpretation has withstood the test of time. The
concept of monetarycrisis is, for Mommsen, a completely financialone.
The Roman Empire in the thirdcentury experienced a growinginability
to match revenues and expenses. The fiscal crisis was solved by
increased recourse to monetary debasement. None of the three coined
metals escaped overvaluation. But it was the silver and copper coins
that were most debased. And it is the Antoninianus of Caracallawhich
can, according to him, be called the "Assignat of this period." Yet
Mommsen is not interested in inflation,and has no analysis of the effect
of monetary oversupply on prices. He concentrates instead on describ-
ing the gradualdisintegrationof the Roman monetary system.
When it comes to analyzing the monetary reorganizationconducted
by the Illyrian emperors in the fourth century, he is quite clear about
setting the beginningof the reorganizationin the reign of Aurelianus,in
the last quarterof the third century. Aurelianushad closed the munici-
pal mints and repressedthe fraudsof the monetarii, the mintemployees,
with measures that had led the monetarii to riot in the streets of Rome,
repressed in a bloodbath. About the reform of Diocletian Mommsen
again establishes the terms of future discussion. He notes that Diocle-
tian's policy was self-contradictory. On the one hand, good gold and
silver coins were issued; on the other, an immense quantityof debased
copper coins was put out, which nullifiedto a great extent the effect of
the reorganizationof the gold and silver coinage.
Writingon the furtherreformof the coinage realizedby Constantine,
Mommsenexpresses what I consider his most subtle remarkson Roman
monetary history. He writes:
We know that under the reign of Constantineall payments in gold coin were made
according to weight and that, moreover, gold ingots were accepted in payment by
weight, and that their weight was controlledregularly.The governmenthad manufac-
tured and placed in the main cities special standards,to facilitate the control of the
weight of the gold coins, and had ordered that special employees had to effect these
controls upon requests of the citizens. . . . This is what explains to us the immutability
of the gold coin which, in a state like the ByzantineEmpire,could seem extraordinaryif
the gold coins had been a currency in the strict sense of the word. But this currency, in
these peculiar circumstances, was more enunciative than dispositive, nobody was
compelledto accept it by its nominalvalue, which was the most effectiveway of putting
a stop to all debasementbecause, once weighinghad been permitted,the reductionof
weight did not bringany profitto the state; the seal impressedon the coins was no more
' Ibid., p. 75.

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
Roman Monetary Theory 813

thana sort of controlon the qualityof the metal, and thingshadgone back, by the force
of events, to the primitive system, where a seal was impressed on copper bars. . . . It
was inevitable to get to the point where the Romans had got, that is, to demonetize
money and to consider it only as a precious metal; because private citizens take back
theirrightnot to accept currencies,when the latterhave no longerweightand title fixed
by law.8

One finds here the essential elements of Mommsen's monetary

theory. It is certainly impossible to define it as laissez faire, liberal,
metallist, anachronistic, as Bolin has done. It seems to me, on the
contrary, pragmaticand relativist in the extreme.
The twentieth century can be called a century of inflationand of great
monetary experiments. It is not the only century in modern times to
deserve this label, but it is an appropriateone. Being, like the rest of us,
men of their times, Roman historians have fallen in our century easy
prey to the temptationto interpretRoman monetaryhistory in terms of
contemporaryhistory and contemporaryeconomic theory.9Our centu-
ry, moreover, has seen the triumph of anthropology and economics
among the social sciences. Histories of Rome cound not be unaffected
by these cultural trends. Roman monetary history, however, has been
duringthe whole course of the century heavily influencedby the great
debate between primitivists and modernists. We can see now that the
debate was no more than a continuationof the polemic conducted in the
eighteenth and nineteenth centuries by people like Mengotti, Ferguson,
Ferrara, and Dureau de la Malle. The debate was enhanced by the
blossoming of the German historical school, emphasizing historical
relativism. It involved monetary history from the beginning: the pro-
pounders of "primitivism" emphasized the distinction between the
natural economy and the monetary economy, stressing the predom-
inance of naturaleconomy in the classical world.'0
8 Ibid., vol. 3, p. 157.
9 The "contemporary"inspirationof much Roman historiographyhas been noticed by A.
Momiglianoin his appreciationof Rostovzev's work.
'0 A most originalattemptto place the phenomenonof money in ancientRomewithinthe context
of Romaninstitutionswas made by H. von Scheel, in his short article translatedas "I concetti
fondamentalidel 'Corpusluris civilis"' in V. Pareto,Bibliotecadi StoriaEconomica(Milan,1907),
vol. 2, but which originallyappearedin Germanin the 1860s.Von Scheel shares in his articlethe
antipathyMomiglianohas noticed in seventeenth- and eighteenth-centuryItalianjurists for the
Romanworld and Romanlaw in particular.Von Scheel's thesis is that the Romans,unlike what
they were made to appearby Rodbertusand his followers, had an exclusivelymonetaryeconomy,
because they had no economic production, no modern property relations, no concept of the
individualas separatefrom the state. As they had no free-laborproduction,they had no value in
use and only value in exchange. Hence the dominanceof monetaryrelationsin theireconomy. Von
Scheel sees the modern world as originatingfrom the barbarians'institutions. "While German
propertyis a consequence of labor, Roman propertyis a consequence of possession; while the
GermanState grew as a graduallimitationof individualpropertyrights,the RomanState is itself
the source of propertyrights" (p. 247). The extensive monetizationof the Romaneconomy in the
imperialperiod, even outside the cities, and in the more agro-pastoralcontext, was confirmed
recently by Fergus Millar'selegant textual analysis of Apuleius'novel. See his "The worldof the
golden Ass" in Journal of Roman Studies, 71 (1981).

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
814 De Cecco

As I noted above, Mommsen's monetaryhistory did not deal with the

connection between money and prices. There is evidence that Momm-
sen was fully aware that the monetary crisis of the third century had
inflationaryconsequences. He wanted, however, to stress that it had
been above all a fiscal crisis and a crisis of State authority, which the
Roman State had survived only by abdicatingits sovereign prerogative
to tax by issuing money, or as he put it, by a "demonetisation of
money." Roman monetaryhistorianswho wrote after the great Europe-
an inflations following the First World War turned to interpretingthe
decline of the Roman Empire as being chiefly determinedby inflation.
They were helped in this revision by the perfecting, by economists,
around the turn of the century, of the modern versions of the quantity
theory of money.
The theory was in full bloom in the twenties and thirties. Roman
monetary historians were well documented, and were becoming even
better documented, on Roman monetary vicissitudes in the third and
fourth centuries. They had not been much interested in the dynamicsof
prices in the same period. They now dedicated themselves to the tasks
of finding evidence that there had been inflations in Roman times
(especially in coincidence with the decline of Rome) and, that the
inflationshad been caused by the debasements in the third century.
Fritz Heichelheim is the Roman historianwho has done most to lend
credibilityto the notion that inflationdeterminedthe decline of Rome."
In his work, which appearedat the beginningof the thirties, he strove to
demonstrate that state expenditure, especially soldiers' pay, had
brought about the monetary crisis, which had led through successive
dramaticdebasements to raginginflation. It is worth rememberingthat
this was precisely the view (if we do not put the emphasis on soldiers'
pay, but on general state expenditure)that the British Governmenthad
of the great German inflation, a view which Bresciani Turronidefined,
in his famous book on the Germaninflation, as the "Inter-Ally view."
In its applicationto Britishdomestic economic policy, it became known
as the "Treasury View."
But we do not have to concentrate on Heichelheim. The quantity
theory, and in particularIrving Fisher's famous equation, acquired a
surprisingpopularityamong Romanhistorians. Fisher's equationstands
out conspicuously in the pages of Santo Mazzarino's "Aspetti sociali
del Quarto secolo." It is used by Edmond Frezouls to prove that
Diocletian's monetary reform might have actually been successful, in
the sense of bringingthe Roman public back from the naturaleconomy
where it had taken refuge, to the monetary economy, with disastrous
" The mainreferenceis to his "ZurWahrungskrisisder romischenImperiumsin 3. Jahrhundert
n. Chr.," in Klio, 26 (1932/33), although the same theory is still advanced in his An Ancient
Economic History (Leyden, 1970), vol. 3. The monetarycrisis of the Roman Empirehad been
likenedto the collapse of CentralEuropeanmonetarysystems afterthe GreatWarby A. Sepr6, in
Circolazione Monetaria e Prezzi (Rome, 1922).

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
Roman Monetary Theory 815

consequences on prices. But the equation also looks at us from the

pages of Mario Mazza's "Lotte sociali e restaurazione autoritaria."
And C. R. Whittakerwheels it out, as recently as 1980,thoughhe uses it
to criticize the "great inflation" theory.'2
The emphasis, one may say the over-emphasis, given by Roman
historians, especially in the interwarperiod, to the theme of monetary
crisis leading to inflationary decline, could not fail to engender a
reaction. This duly came in the postwar period (in the writingsof Marc
Bloch in the early thirties one can see that not everyone hadjumped on
the bandwagon, even when it was rolling most noisily). The theory of
the inflation-induceddecline has been criticized from differentangles.
Some writers concentrated on an examinationof available evidence, to
distill from it, for instance, that after all there had not been much price
inflationin the third and fourth centuries.13Others have questioned the
meaningof inflation,by pointingout that in a system where prices were
expressed in gold, silver, and copper, one has to be clear in what unit
one is counting. Still others, like Julien Guey, have vented the hypothe-
sis that the Romans might have been successful in their devaluation
policies, that is, that they had managed to effect debasements not
followed by price inflation. It is interesting to note that Guey's article
was published in 1962, soon after the French franc had been successful-
ly devalued by Antoine Pinay.'4 Others, in particularMireille Corbier,
have attempted to apply Keynes's criticism of the quantity theory, in
order to prove that an increase in the Roman money supply could have
induced a rise in output, ratherthan in prices.'5
A whole new school of Roman historians has criticized the theory of
inflation-induceddecline from a much more fundamentalpoint of view.
These are the ancient historians who derive their inspirationfrom the
criticism Karl Polanyi leveled at the application of modern monetary
theory. Polanyi believed that in primitive economies money is a series
of specialized instruments, while modern economies are equipped with
a so-called "general" money.'6 The attributionto Polanyi is peculiar
considering that Max Weber gave a clear and articulateversion of it. "

12 Frezoul's article, "A propos de la hausse des prix sous Diocletien," is contained in the
Melanges Carcopino(Paris, 1965). C. R. Whittaker'sarticle, "Inflationand the Economy in the
Fourth Century A.D.," is published in C. E. King, ed., Imperial Revenue, Expenditureand
MonetaryPolicy (Oxford, 1980)(BAR Series 76).
3 See Whittaker,"Inflationand the Economy."
'4 JulienGuey's article, "L'Aloi du dernierromainde 177a 211 apres J.C.," was publishedin
the Revue Numismatique, 4 (1962).
Is See her "Fiscalitdet monnaie, problemesde methode," in Dialoghi di Archeologia(1976/77)
and "Devaluationset fiscalit6"in AA.vv., Les Devaluationsci Rome (Rome, 1975).In the firstof
these articlesCorbierpointsout that Heichelheimhadbeen influencedby the monetarysituationof
the Europeof his time, but fails to repeat the procedureof what concerns Guey and herself.
16 The standardreferenceis K. Polanyi's "The Semanticsof Money Uses" (1957),reprintedin
his Primitive, Archaic and Modern Economics, G. Dalton, ed. (Boston, 1968).
17 M. Weber, General Economic History (a course of lectures delivered in 1920, edited by
Hellmanand Palyi) (New Brunswick,N.J., 1981), pp. 236ff.

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
816 De Cecco

In the relevant literature on Roman money, however, the canonical

references are to Polanyi, not to Weber.
There are, however, Roman monetary historians who ought to be
counted among the forerunnersof the "substantive" school, who did
not know, like M. Jourdain,that they had been speaking prose. I refer
here to Sture Bolin, who aimed, in the whole of his book on state and
currency, to show that money in Rome had been above all an instrument
of taxation. He took great exception to what he considered a common
misapprehension, that, as he put it, "objects of a never varying type
have had never varying functions."'8 I cannot find in Bolin's book
references to either Weber or Polanyi. It is a pity that he chose as his
intellectual adversary Mommsen, who was perfectly aware of the
predominantlyfiscal function that money had in Roman times.
What is remarkablein Bolin's analysis is that, after having stated his
evolutionary theory of money, and his conviction that money was
always circulatingat a value well above its intrinsicvalue, he attempts
to prove that overvalued coins remained in circulation because the
Roman monetary authorities knew exactly how to apply the modern
theory of the "gold points." He uses an elaboratestatisticaltest. One is
left wonderingwhat kind of evolution can possibly take place after such
sophistication had been reached.
The explanation for Bolin's emphatic reiterationof the thesis is the
obdurate defense, by people like Mickwitz and Johnson and West, of
pure metallist theories. The obduracyis understandable.The preface to
the book by Johnson and West, Currency in Roman and Byzantine
Egypt, thanks Erwin Kemmerer and Frank Graham for advice on
monetary theory, and economists know well what convinced metallists
those two were.'9 Bolin's thesis of the fiduciarynatureof Romanmoney
(which had also been Mommsen's thesis) does not seem to have
convinced the profession. Recent literature,like JulienGuey's "De l'or
de Daces au livre du Sture Bolin" (MelangeCarcopino,see footnote 12)
and Claude Nicolet's review of Roman economic thought, still ex-
presses the conviction that Roman money was mostly accepted at its
metallic content, in spite of Bolin's efforts to prove the opposite and in
spite of the whole corpus of Roman legal evidence.20 Francesco de
Martino, in his massive review of Roman economic history (Storia
Economica di Roma [Florence, 1980])guardedly accepts the metallist
Bolin, State and Currency, p. I 1.
9 L. C. West and A. C. Johnson, Currency in Roman and Byzantine Egypt (Princeton, 1944).
20 C. Nicolet, "II pensiero economic romano," in L. Firpo, ed., Storia delle idee politiche,
economiche e sociali (Torino, 1982), vol. 1.
21 disregardhere one of the loci classici of debateof Romanmonetaryhistory,the controversy
about Pliny's "gold drain." Edward Gibbon's treatmentof the subject is so convincing that
referenceto it suffices treatmentfor this subject. "It is no easy task to confine luxurywithinthe
limits of an Empire," says Gibbon. Importsflowed into Rome and the Romans paid, he says,

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
Roman Monetary Theory 817

Apart from Sture Bolin, who seems to have gotten there all by
himself, other Roman monetaryhistoriansacknowledgethe influenceof
the "substantivist" school on their thought. Michael Crawford, for
example, an eminent Roman monetary historian and acknowledged
Polanyist (an inspirationwhich he shares with other Cambridge-based
ancient historians and which he probably derives from Moses Finley),
set himself in a closely argued article of 1970 a very ambitioustask.
A wide variety of objects may function as money in the different uses which this
possesses, for payment, for storing wealth, for measuringvalue, and as a means of
exchange. ... In the Roman world coined money was clearly dominantover other
forms of money in the first three uses, and I want here to explore the extent to which it
served as a means of exchange, partly because this is the most distinctivefunction of
money and one for which coined money or a token substituteis essentialto achieve any
greatversatilityand partlybecause the problemsinvolved seem particularlycomplex.22

Serving as a means of exchange is "the most distinctive function of

money. 23
It so happens, however, that the type of monetary theory Crawford
prefersis the same preferredby generalequilibriumtheorists. It has, for
a number of decades, been argued by all sorts of critics of general
equilibriumtheory, that it is in fact impossible for general equilibrium
theorists to include within their analyticalframe a type of money which
is at the same time a means of exchange and a store of value. General
equilibriumtheory cannot survive it. One may note, for instance, that
this is the core of the Keynesian criticism of traditional monetary
theory. What one has to ask is whether one can logically conceive of a
money which plays the means of exchange role without at the same time
being a store of value. This can only happen if, after having served as a
means of exchange, in a round of exchange, the money promptly
disappears,and nobody ends up with a stock of money. There must be,
in other words, no intertemporaltransactions. People are only buying
today's goods and selling today's goods.
But if this is true, one does not know whether money, in such a
system, can exist at all. Its convenience as a means to remove the

quotingPliny, with preciousmetals. However, since the gold/silverratiohad risen betweenPliny's

time and the reignof Constantine,Gibbonconcludesthat the supplyof silver, from Spanishtimes,
must have been so abundantto more than make up for the drainto the East. FollowingGibbon,I
fail to be worriedby the Roman Empire'sbalanceof paymentsproblems.
22 M. Crawford,"Money and Exchange in the Roman World,"Journal of Roman Studies, 60
(1970).I have selected this articlefor carefulanalysis because the mainopinionshe expresses in it
are repeatedin his numerousmore recent works.
23 AnotherCambridgeancient monetaryhistorian,PhilipGrierson,in an articleon "The Origins
of Money" in G. Dalton, ed., Research in EconomicAnthropology(Greenwich,1978),maintains
that serving as a means of exchange is the function anthropologistsusually indicateas money's
most importantfunction. Griersonhimselfprefersto indicatethe "measureof value" as the main
function money performs. He reproaches Sir John Hicks for preferringthe "store of value"
function.He would, I presume,not agree with what I say (followingSir John'steachings)a propos
of Crawford.

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
818 De Cecco

difficultiesinherentin double coincidence is undoubtedbut, in a system

without past and future, how is money to come about and how is it to
disappear?In other words, how do you create it, at the start, and how
do you destroy it, at the end of each round, so that nobody will be stuck
with it?
A timeless economy, or one composed of noncommunicatingperiods,
is the only one which does not need a money. Actual economies, which
have a past, a present, and a future, all need and have money, that is to
say, a commodity more saleable than others, which will transportvalues
across time and space and which, for that capacity, will be trusted as a
means of exchange. Granted this, one has to ask whether an actual
economy needs coined money. From the pure contractualistpoint of
view, the answer is that it does not. Coined money requires the
existence of the state. But there have been states which have gone on
for centuries without resorting to coinage. Usually, in the most devel-
oped among those coinless economies, tradeavailed itself of metal bars,
graded to control their weight and purity, by respected merchants.
Those economies had moneys functioning as media of exchange be-
cause they were good stores of value, but the state did not take part in
the process.
With a successful coined money system, the state takes over the
money supply function. It is clear that coined money, to be successful,
must be superior, as a means of exchange, to privately graded metal
bars. And, to be a good means of exchange, as we have noted, a coin
has to be a good store of value. Of course, there are trade-offsbetween
these functions. People continued to use dollars in the 1970s, which
involved holding them, because they were the most widely known and
used currency, even when their ability to serve as store of value
diminished. They looked for a substitute only when the United States
government did not want to give a higher interest rate to people who
held them. The capital loss envisaged on dollars exceeded the conve-
nience yield accruingfrom using dollarsin tradingand in carryingvalues
throughtime and space. Crawford,in separatingthe means of exchange
function from the others, in particularfrom the store of value function,
negates the possibility that a coin keep its function as money. In other
words, he envisages the possibility that a coin can be a store of wealth,
measure of value, means of payment, but not a means of exchange.
But is it really possible that an instrumentserving all these functions
does not serve the last one? We have alreadyseen that the reverse is not
possible, that is, that in order to be a means of exchange, a money must
be a store of value and a means of payment. I think that if we were not
consideringa coin, but a privatelygradedmetal bar, it would be possible
to envisage that such an abode of purchasing power would be only
infrequentlyexchanged. But in Crawford'sexample, we are considering
coins of small face value. And he proves in his article that, since he fails

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
Roman Monetary Theory 819

to findthem in hoards in the countrysideand in the northwestpartof the

Empire, coins in the four centuries centering on the birth of Christ did
not function as a means of exchange. From that he infers that, since
money's most importantfunction was not served, coined money, as it
existed in Rome, did not serve all functions. Thus it was specialized
The reasoning entails a common misapprehension,which it is neces-
sary to dispel. A monetary system is composed, even today in highly
developed economies, of differentmonetaryinstruments,each servinga
specialized purpose. They are all part of a system because they are
fungible with one another, even if at a cost. And the subset of goods
comprisingall the money instrumentsthat form the monetarysystem is
as a whole composed of goods more saleable than any goods in the
subset of nonmonetarygoods. This does not imply that all the goods in
the monetary subset are the same, from the functional point of view.
Some of them are better suited to serve as stores of wealth, others are
better suited to serve as means of payment. Small transactions, for
instance, are usually effected with coins of small denomination.A large
denomination instrument cannot always be used to finance a small
transaction. Milton Friedmanused to say in his lectures in the golden
sixties that twenty-dollar bills were not money because one could not
use them to pay for a small purchase in a supermarket.
Large denomination instruments are used for storing wealth in the
countrysideand among the poor even today. At the same time, there are
regions of less intense monetizationeven today in developed countries.
This does not mean that the whole economy is only imperfectly
monetized, or that money does not function as a means of exchange. As
long as some of the instrumentsformingthe monetarysubset are used, a
monetary system can be said to exist, and the economy can be called a
monetaryeconomy, even if some of the monetaryinstrumentsare used
for some functions and some for other functions. That things are so is
the idea of a system.
What Crawford would have to show to prove his point is that, if a
Roman tried to finance a purchase by any of the instrumentsthat were
partof the Roman monetarysystem his paymentwould be refused. But,
all he in fact shows is the absence of certain coins in certainparts of the

24 See von Scheel, "I concetti fondamentali."Here von Scheel expressed, on the contrary,the
opinionthat in Rome money's main functionwas to serve as a meansof exchange. "In the pecunia
thereis firstof all the concept of the universaland publiclyrecognizedmeansof exchange.. . . The
aim of money is to constitutea meansfor the circulationof goods which, throughits intromission,
and the juxtaposition of merx and premium,becomes a special negotium, the emptio-venditio" and
"forjuridicalnegotia we cannot thinkof any other measurethan what is representedby pecunia"
(p. 736). Crawfordtends to play downjuridicalevidence, butone cannotfail to be impressedby the
remarkablewealth of quotes from the corpusjuris von Scheel is able to assemblein orderto back
his statementon the main functionof money in Rome. The periodhe studies coincides with that
studiedby Crawford.

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
820 De Cecco

Romaneconomy, and the persistence of an agio between Romancoins,

which allowed bankers and moneychangers to earn a living. The
absence of small coins in the countryside and in the more rural
northwest is hardly surprising. It makes sense that people would be
more self-sufficientin the countryside, and that they sold crops once a
year and were paid for those large transactions in silver, which they
hoarded. Nor is the agio that persisted between coins of the Roman
monetary system surprising. Consider the money-changingmachines
one finds in airports,where one puts in a dollarbill and receives 75 cents
in small change. Try to change a personal check in a part of Britain
where one is not known, or try to pay for a small purchasewith a large
banknote in a place where counterfeitingwas widespread. Upon being
refused those particular monetary instruments, would one conclude
about the American or British system what Crawfordhas concluded
about the Roman monetary system?
One of the pieces of evidence Crawfordquotes to prove his thesis is
worth consideringin some detail. He tells us that, in spite of attemptson
the part of Roman legislators, variable exchange rates existed among
Roman coins. The financial life of ancient Rome is one of the least
studied parts of Roman life. From the wealth of material, legal, and
literaryevidence available, this should not be so. But it is. Startingfrom
ignorance, therefore, we can only conjecture about the meaningof the
phenomenon Crawfordbrings to our attention. A differentialbetween
the rate of exchange of differentmonetaryinstrumentsin differentparts
of the same monetary area exists even today, in very developed
countries. Given the specialized functions of the various instrumentsin
the monetary system, it is altogether possible that demand and supply
conditions for each of them may not be the same everywhere in the
monetaryarea. The instrumentused for saving may be in great demand
in a high saving partof the country, while the instrumentused for paying
the poor may be in great demand, and in short supply, in parts of the
area where many payments of that kind take place. In developed
countries today, there are regions where the interest rate banks pay on
deposits is much higherthan elsewhere, or much lower than elsewhere.
The differentials persist despite excellent communicationsbecause of
market imperfection and segmentation, and in spite of the activity of
central banks and other monetary authorities. Crawford concludes,
from the persistence of differential exchange rates between Roman
coins in different parts of the Empire, that there obviously was no
conscious monetary policy or, as he puts it, no economically motivated
monetary policy. Would he conclude, upon noting a persistent deposit
rate differentialbetween Norwich and Manchester, that Britainhad no
economically motivated monetary policy?
It would be fairer, however, to compare conditions in the Roman
Empireto those prevailingin the last quarterof the nineteenthcentury.
Right up until the First World War the Bank of England experienced

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
Roman Monetary Theory 821

what were called "internal drains," which were seasonal and thus
foreseeable, and which it could not, contra Crawford,offset completely.
To those we may add the equally seasonal "autumn drains," which
drew gold reserves, every year, from the coffers of the Bank of England
to the plains of the Midwest. Americanfarmershad to be paid for their
crops, and they wanted gold coins. Gold had to come from New York
and from London, attractedby a seasonal interest rate differential.The
differentialreversed itself when farmersreturnedthe coins, by purchas-
ing in the shops, by paying taxes and other debts.
In the same period, in spite of the presence in British India of an
administrationof exceptional economic sophistication, the phenomena
that Crawfordnotices in the RomanEmpirewith referenceto the money
market, all existed, and persisted. They were studied by scores of
specialists, who wrote admirable reports on them. But suppose this
writtenevidence had perished, and only evidence of the existence of the
phenomenahad survived. Would Crawfordthinkthe monetarypolicy of
the Raj had no "economic motivation"?
What an economist sadly misses when reading Roman monetary
history is a sense of the development of the monetary and financial
system in the millennium spanned by the Roman state. The centuries
Crawfordrefers to, for instance, saw the transformationof Rome from a
relatively unimportantrepublic into a great world power. Yet he writes
that in that period "The Roman coinage system underwentno sudden,
majorchanges." But, in that very period, the Romans developed their
monetarysystem from the relative simplicityof a predominantlycopper
currency to the extreme sophistication of a trimetallic system, and
extended it to apply to the greaterpartof the lands they conquered,thus
forming a domestic monetary area larger than any the West has ever
known. Crawford's statement is equivalent, in historical terms, to
sayingthat the Britishmonetarysystem had undergoneno suddenmajor
changes from the time Sir Isaac Newton established the sterling/gold
parity to the outbreak of the First World War.
But no satisfactory description exists of this evolution, which a
practitioner of money and banking finds very remarkable indeed.
Roman monetary authoritiesmay have known finance but not econom-
ics, as Finley and Crawfordtell us, but they did manage to supply the
Empire with coins for many hundreds of years. The system of mints
they established and the division of power between central and local
authorities on coinage supply was highly functional.25 As to the
motivation of money issues, be it to enrich themselves, as Finley
asserts, or to pay their soldiers, as Crawfordmaintains,Roman emper-
ors do not seem all that remote from our own statesmen's motivations.

On the supply of money and workingof the Roman monetarysystem, compareCrawford's
pessimismwith what is, in my view, a morerealisticassessment, containedin K. Hopkins,"Taxes
and Trade in the Roman Empire,"Journal of Roman Studies, 70 (1980).

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions
822 De Cecco

In a world that knew no rapid transportation system, new coins

reached the farthestcorners of the Empirewith surprisingrapidity.And
when there were discontinuities and interruptions,there seem to have
been enough argentarii, nummularii, foeneratores, oppii, trapezitae, to
regulate the velocity of circulation and reconcile demand and supply
conditions. One should pause to reflect on the very number of words
that were used to describe, in current Latin, financial functions and
financialintermediaries.How did this system acquirethe depth, width,
and resiliency it obviously had reached by the second century after
It is dangerous, in short, to depend on self-descriptionsof economic
behavior in a society where the practitionersdo not seem to have been
bent on describing how the economy worked. Suppose the only
documents about the British financial system were the minutes of the
Courtof Directors of the Bank of England.One could read them and, at
least until the 1920s, find precious little evidence that that august body
had any idea of what Finley and Crawfordcall economics.
Economics, as our two eminent historians use this word, is a
discipline invented by Keynes. But, if knowing economics means to be
aware of the results attendant upon some actions in the economic
sphere, then a careful study of Roman literary and legal evidence will
convince anybody that the Romans knew a lot of it.26Like Montague
Norman, the greatest Governorof the Bank of England,they had never
found it necessary to study it formally to talk about it. The same could
not be said, to be sure, of English bankers in the early decades of the
nineteenthcentury: they were extremely well versed in the new science
of political economy, and have left very originalwritings about it. But
shall we conclude from that that Normandid not know his job as well as
Henry Thornton or Horsley Palmer? Or, should we conclude, more
modestly, that some periods see a greaterintegrationof intellectualand
practicalfunctions than others?
26 A similarviewpoint is expressed by E. Lo Cascio in a very perceptivearticle, "State and

coinage in the late Republicand early Empire," in Journalof Roman Studies, 71 (1981). Claude
Nicolet, in his famous contributionfor the Annales, tconomie, Socidt6,Civilisation(ESC) (1971),
had even contended that the Roman writers had a clear understandingof the relation between
money and prices. He quoted very convincing passages from Roman writers, among which the
famousone from Suetonius, which Lo Cascio quotes againin his article,provesthatthat writer,at
least, knew not only that money has an influenceon prices, but that it also influenceschanges in
outputthroughchanges in the interest rate. For the verbatimquotation,see Lo Cascio, p. 86.

This content downloaded from on Sat, 17 May 2014 02:36:40 AM

All use subject to JSTOR Terms and Conditions