Ideal Money

Author(s): John F. Nash, Jr.

Source: Southern Economic Journal, Vol. 69, No. 1 (Jul., 2002), pp. 4-11
Published by: Southern Economic Association
Stable URL: http://www.jstor.org/stable/1061553 .
Accessed: 24/07/2013 01:35

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
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.

Southern Economic Association is collaborating with JSTOR to digitize, preserve and extend access to
Southern Economic Journal.

http://www.jstor.org

This content downloaded from 205.133.226.104 on Wed, 24 Jul 2013 01:35:40 AM

All use subject to JSTOR Terms and Conditions
Southern Economic Journal 2002, 69(1), 4-11

DistinguishedGuest Lecture

Ideal Money
JohnE Nash, Jr.*

Moneycan be recognizedas a technological development comparableto the wheel and of

similarantiquity.Amongthemorerecentdevelopments in thetechnology thatfacilitates
trans-
fersof utility(in thesense of game theory)are systemslike thoseof EZ Pass, by meansof
whichvehiclestraversing tollbridgesor tollhighways can pay theirtollfeeswithout stopping
fortheattention of humanpersonnel manning thetollbooths.In thislecture, remarks
I present
aboutthehistory of monetary systemsand aboutissuesof comparative qualityor merit,along
witha specificproposalabouthowa systemor systems of "ideal money"mightbe established
and employed.In addition,I criticizetheKeynesianpsychology in relationto thehistoryof
theinfluence of Keynesianson thepracticalcharacteristicsof nationalcurrencies.

1. Introduction

The special commodity,or medium,thatwe call money has a long and interesting history,
and since we are so dependenton our use of it and so much controlledand motivatedby the
wish to have more of it or not to lose what we have, we may become irrationalin thinking
about it and fail to be able to reason about it like we do about a technology,such as radio, to
be used more or less efficiently. Therefore,I wish to presentthe argumentthatvarious interests
and groups,notablyincludingKeynesian economists,have sold to the public as a quasi doctrine
thatteaches, in effect,that "less is more" or that(in otherwords) "bad money is betterthan
good money." Here we may recall the classic ancienteconomics sayingcalled Gresham's law:
"The bad money drives out the good." This saying of Gresham's is of interesthere mainly
because it illustratesthe old, or "classical," concept of bad money,which is not in line with
the thinkingof Keynesian economists.

2. Money,Utility,and Game Theory

In the sortof game theorythatis studiedand applied by economists,the concept of utility

is very fundamentaland essential. Von Neumann and Morgenstern(1953) give a notablygood
and thoroughtreatmentof utilityin theirbook on game theoryand economic behavior.The
concept of (mathematical)utilitydoes indeed predatethe book of Von Neumann and Morgen-

* Room 910, Fine Hall, Princeton

University,Princeton,NJ 08544-1000, USA.
ProfessorJohnE Nash, Jr.,presentedthe DistinguishedGuest Lecture at the 2001 Annual Meeting of the Southern
Economic Association in Tampa, Florida. He received a Ph.D. in mathematicsfromPrincetonin 1950 and taughtat MIT.
In 1994, he won the Nobel Prize in Economic Science forhis seminal contributionsto game theory.He is now affiliated
with the Departmentof Mathematicsat PrincetonUniversity.

This content downloaded from 205.133.226.104 on Wed, 24 Jul 2013 01:35:40 AM

All use subject to JSTOR Terms and Conditions
 Distinguished Guest Lecture 5

stern.For example, as a concept, mathematicalutilitycan be traced back to a paper published

in 1886 in Pisa by G. B. Antonelli(Antonelli 1971).
When one studies what are called cooperative games, which, in economic terms,include
mergersand acquisitions or cartel formation,it is found to be appropriateand is standardto
classify these games into two basic groups: games withtransferableutility(type 1) and games
withouttransferableutility(type 2). In the world of practicalrealities,it is moneythattypically
brings into existence games of type 1 ratherthan games of type 2; money is the "lubricant"
thatenables the efficienttransferof utility.Moreover,in general,if games can be transformed
fromtype 2 to type 1, thereis a gain, on average, for all of the players in termsof whatever
the expected outcome mightbe.
However, money's functionin generallyfacilitatingthe transferof utilitywould seem to
be performedas well by the currencyof Thailand as by thatof Switzerland.One can ask the
question,How do "good money" and "bad money" differ,if at all, withregardto the valuable
functionof facilitatingutilitytransfer?But if we considercontractswith a relativelylong time
axis, thenthe differencecan be seen clearly.
Consider a society in which the money in use is subject to a rapid and unpredictablerate
of inflationsuch thatmoney worth100 now mightbe worthbetween 50 and 10 by a year from
now. Who would want to lend money forthe termof a year? In this context,we can see how
the qualityof a money standardcan stronglyinfluenceareas of theeconomyinvolvingfinancing
with longer-term credits.

Keynesians

The thinkingof J. M. Keynes was actually multidimensional,and consequentlythereare

quite differentvarietiesof persons at the presenttime who follow,in one way or another,some
of the thinkingof Keynes. Of course, some of his thinkingwas scientificallyaccurate and thus
not disputable.For example, an early book writtenby Keynes (1921) was themathematicaltext
A Treatise on Probability.
The label Keynesian is convenient,but to be safe, we should definethistermto constitute
a partythatcan be criticizedand contrastedwith otherparties.Thus, let us defineKeynesian
to be descriptiveof a school of thoughtthatoriginatedat the time of the devaluations of the
pound and the dollar in the early 1930s. Then, more specifically,a Keynesian would favorthe
existence of a manipulativestateestablishmentof centralbank and treasurythatwould contin-
uously seek to achieve "economic welfare" objectives with comparativelylittleregardforthe
long-termreputationof the national currencyand its associated effectson the reputationof
financialenterprisesdomestic to the state.
And, indeed, a very famous saying of Keynes's was "... in the long run we will all be
dead .... "

3. HistoricalObservations

The historyof the gold standardis ratherinteresting.It can be tracedback to 1717, when
Isaac Newton,as "Master of the Mint" in London, set a standardquantumof gold to correspond
to the currencycalled thepound sterling.In 1931, this standardfinallyfailed to be supported

This content downloaded from 205.133.226.104 on Wed, 24 Jul 2013 01:35:40 AM

All use subject to JSTOR Terms and Conditions
6 JohnF. Nash,Jr.

any longer,and this happened at a time when the London governmenthad shiftedto the left
politicallyand at a time of global economic stresses.
The U.S. dollar had itselfbeen imitativelyput on a fixed relation to gold, like various
othercurrenciesat various times. That relationwas not supportedafter1933, similarlyto the
case for the Britishpound. Anothersimilarlylinked currencywas the Swiss franc.In its case,
the original standardrelationto gold was discontinuedin 1936.
it could be said that therewere times and places
In later times, viewed retrospectively,
where it perhaps seemed as if therewere a gold standardbut therereallywasn't. In 1971, there
came, underthe presidencyof Nixon, the finalcompletedisavowal of the concept of a linkage
between the U.S. dollar and gold. This developmentnaturallyalso markedthe beginningof a
period of stronginflationin termsof dollar prices.

Central Banks as "Pardoners" and as Printersof Money

A debt pardoner is an agency or authoritythatcan pardonthe "sins" of the overindebted

or of those who otherwisewould go bankrupt.Banks can thussometimesbe saved fromfailing.
The governmentof a countrythat issues the locally accepted currencyalso thus pardons its
own "sins."
If the euro currencybecomes established,then Rome and Paris will no longer be able to
play the same roles as pardonersas they have in the past. National pardonerswill become
ineffectivewithinthe area of the euro unless theysecede fromthe euro currencybloc. Instead,
pardoningwill depend on the actions of a sortof "Holy Roman Emperor" in Frankfurt, which
would be able to generallypardon all of the debt-sinnersof the euro-moneybloc.

The Britishand the Euro

The inhabitantsof the United Kingdom,havingin the past had the best and mostdominant
of all currencies,as well as the biggest colonial empire,may feel so reluctantto accept the
comedown thatagreeingto the rule of a (collective) Holy Roman Emperorin Frankfurt would
entail thattheywill decline to accept membershipin the euro-moneybloc. Technically,it would
be possible forthemto arrangeto have a money of theirown whose quality would be at least
equal to thatof the euro. (Presumably,this would be the Swiss strategy.)

Ideal Money

The euro-moneyprospect opens up many interestingpotentialitieswith regardto games

of alignment.In general,membershipin a social club is desirableonly if thereare nonmembers.
A good alliance can be reduced to an absurdityif it becomes too broadly inclusive. However,
a global money standardcould have a value similarto thatof standardmeasures such as those
of the metricsystem.
There is tremendousvalue in simplyhavingpricesquoted conveniently.Witha euro-money
standard,a manufacturerin Spain can give informationsimply to prospective customersin
France or Germany.These conditionswould be comparable to conditionsin the United States,
where items can be orderedfromcatalogs and shipped interstatewithonly the complicationof
the stateretail sales taxes sometimesbeing involved.

This content downloaded from 205.133.226.104 on Wed, 24 Jul 2013 01:35:40 AM

All use subject to JSTOR Terms and Conditions
 GuestLecture
Distinguished 7

4. States withoutSin

It was observed thatthe statesof Luxembourg,Andorra,Liechtenstein,San Marino, Mon-

aco, and Vatican City have been constitutionallywithoutsin in the sense of the defaultingof
bankrupts,and they have also not themselves pardoned any money debtors. This is simply
because theyhave not printedany money forwhich theywould also determinethe value. Each
of these states would eitherissue a currencywith a value definedby the value of the currency
issued by some larger state or simply not issue any currency.Outside of Europe, Panama and
Liberia can be named as examples of statesin the same category.

Economics and Money Quality

How does the traditionalquality of a state's currencycorrelatewith that state's typical

level of economic advancementand the standardof living?The correlationseems to be positive,
since Switzerlandhas a notablyhigh standardof prosperityand notably "good" money,and,
of course, the whole spectrumof prosperityand money quality of states must be studiedfor a
scientificconclusion.

How Could "Good" Money Become a Standard?

The historicalfact seems to be thatthe gold standardwas, in its time,a basis thatfavored
the prosperityof the United Kingdom and of otherstates,like theUnitedStates and Switzerland,
thatadopted the concept of the standard.Nowadays, however,few would propose a returnto
the actual use of simplythe metal gold as a standard,forthe followingreasons. (i) The cost of
mining gold effectivelydoes depend on the technology.Recent cyanide leaching techniques
have made it possible again to profitablymine gold at formerlyabandoned sites in the U.S. so
thatit is now a big producer.However, the unpredictability of the cost is a negativefactor.(ii)
The location of potentialgold-mininglocations may not be "politically appealing," so it would
seem undesirable to make a political choice to enhance the economic importanceof those
particularareas. (iii) There is some negative psychologyabout gold such thateven if it were
the most logical choice afterall, the unpopularityof the idea could be very obstructive.
But, a modernalternativeis possible, one thatwould provide a good standardindependent
of statepardoners.This idea occurredto me fairlyrecently.
However, the possibilities with regardto actually establishinga norm of money systems
thatcould qualify as "ideal" are dependenton the political circumstancesof the world. If the
world had in fact become a single empire with a centralgovernment,then what is now inter-
national trade,with shippingon the oceans throughareas considered the propertyof no state,
would be replaced by the equivalent of domestic commerce withinthe United States. This
development would profoundlymodify the circumstancesrelevant to the establishmentof
"good" or "bad" systemsof money. What I have to suggest is not appropriatefor the world
empire context.
It can also be remarkedthat "bad" money,or the inverse of "good" or "ideal" money,
is basically a consequence of deficiencieson the part of governmentsand politiciansof a sort
relating to morals, virtue,or ethics. Thus, the phenomena of "bad" money are essentially
understandablevia Machiavellian studies.
We of Terra could be taughthow to have ideal monetarysystemsif wise and benevolent

This content downloaded from 205.133.226.104 on Wed, 24 Jul 2013 01:35:40 AM

All use subject to JSTOR Terms and Conditions
8 JohnF. Nash,Jr.

were to take us in hand and administerour nationalmoneysystemsanalogously

extraterrestrials
to how the Britishrecentlyadministeredthe currencyof Hong Kong.

5. Natural Comparisonsof Value

In most states with an advanced economy,the authoritiespreparestatisticsthatare com-

parable to the U.S. consumerprice index (CPI). (In the United States this statisticgoes back to
when the dollar was indeed a gold standardcurrency.)As inflationhas become more of a
standardand expected phenomenon,the CPI has been used and interpreted as the mostrealistic
and practicalmeasure of the actual rate of inflation.When it is at the 2-3% level, it is currently
fashionableforall economic and financialcommentatorsto say that"inflationis not a problem"
or "inflationis undercontrol." This, of course, involves a sortof psychology.Over the current
expected human lifetimeof 70 years,the value of one unitof currencyat the timeof a person's
birthwould exceed fourunits of its value at the time of thatperson's expiration.

A Nonpolitical Value Standard

A possible nonpoliticalbasis for a value standardthatcould be used formoney would be

a good industrialconsumptionprice index (ICPI) statistic.This statisticcould be calculated
fromthe internationalprices of commoditiessuch as copper,silver,tungsten,and so forththat
are used in industrialactivities.
Here we can returnto the understandingthat money has the practical value of creating
games fortraders.These are games withtransferable utility,but if the moneywere notavailable,
the game of the traderswould be a game withouttransferableutilityand thusnaturallya game
with less efficiencywithregardto the possibilitiesforthe participantsto maximize theircom-
bined gains.
If we thenconsider which commoditieswould be optimallysuitable forprovidinga basis
fora means of transferring utility,and if we specificallyconsiderthe possibilitythatthe trading
partnersmay be located in different nations and perhaps on differentcontinents,thenthe suit-
ability of such commodities with regard to the ideal functionof facilitatingutilitytransfer
depends on the extentto which such a commodityseems to have a value independentof its
geographicallocation.
Clearly,in termsof thisgeographicalperspective,gold has historicallybeen optimal,large-
ly because the labor cost of moving it over great distances is so small relativeto the value of
what is being transported.Thus, gold formeda very efficiently movable mediumforthe trans-
portationof a value exchangeable for othervalues, ultimatelyderiving,in one way or another,
fromhumanlabor (withthe achievementsof warriorsherealso being viewed as involvinglabor).
However, rightnow platinumwould be even betterthan gold, because it has more value per
unit of weight.
Crude petroleumcould also be used forbartertransactions,and in view of thepresentstate
of the global economyit would seem a propercomponentof an index of prices of internationally
traded commoditiesthatenterinto the costs of industrialconsumption.We can see thattimes
could change, especially if a "miracle energysource" were found,and thus if a good ICPI is
constructed,it should not be expected to be valid as initiallydefinedfor all eternity.It would
instead be appropriatefor it to be regularlyreadjusteddependingon how the patternsof inter-

This content downloaded from 205.133.226.104 on Wed, 24 Jul 2013 01:35:40 AM

All use subject to JSTOR Terms and Conditions
 GuestLecture
Distinguished 9

national trade would actually evolve. Here, evidently,politicians in control of the authority
behind standardscould corruptthe continuityof a good standard,but dependingon how things
were fundamentallyarranged,the probabilitiesof serious damage throughpolitical corruption
mightbecome as small as the probabilitiesthatthe values of the standardmeterand kilogram
will be corruptedthroughthe actions of politicians. Moreover, commodities with easily and
reliably calculable prices are most suitable,and relativelystable prices are very desirable.An-
otherbasic cost thatcould be used would be a standardtransportation cost, the cost of shipping
a unitquantityof somethingover long internationaldistances.
Hence, it seems that such an ICPI could be calculated in an essentiallyscientificfashion
aftersome practicalinitialchoices were made. Moreover,this standard,as a basis forthe stan-
dardizationof the value of the internationalmoney unit,would remove the political roles of the
"grand pardoners," the state authoritiesthat can forgive the debts of debtors,including,in
particular,those of themselves.(The national debt of a statecan, in principle,be trivializedby
a sufficientamountof inflation.)

Euro, Frankfurt,Standard

If such an objective and nonpoliticalstandardwere used in the Frankfurt

of a euro-money
system,thentherewould not be a "Holy Roman Emperor" (althoughof a collective structure)
with the power to pardon or not pardon the debt-sinsof all European debtors(of the zone) in
general.Thus, therewould be nothingto argue about,such as whetheror notthosemostdesirous
of Keynesian pardoningshould be favored with a general pardoningof past debts by the fiat
of inflation.

RefinedIndices

If the technical problem of designingan index of prices to serve as a basis for a money
of standardvalue is consideredin a more elaborate fashion,it seems thatit is possible to define
the sortof index thatwould vary smoothlyand yet would also vary in an appropriateway over
longerperiods. Here, the apparentproblemis thatthe prices of certaincommoditiesthatwould
be ideally suited to the measurementof long-termchanges in the costs of industrialproduction
may tend naturallyto be volatile in theirvariationsdepending on business cycles. Moreover,
the prices of othercommodities,services, and so forthmighttendto vary much moregradually
or smoothlybut not be reliable in termsof long-termconsiderationsforone reason or another.
For example, the prices of copper and nickel mightvery well represent,over long periods,
the actual costs of industrialproduction,while the prices of silver and gold mighttend to vary
comparativelymuch more smoothlythan those of the baser metals. It is possible to construct
a price index based on moving averages thatwould have the smoothnessof the prices of the
gold and silver and yet, over longer periods, would basically follow the values of the baser
metals. This index could be constructedby computinga moving average of the index for the
base metals computedby pricingthemmodulo the index of the precious metals.
In actual application,it would not be a matterof base and precious metals but ratherof a
varietyof commoditiesthatwould be selected fortheirsuitabilityin one sense or another.Also,
for the index formedon the basis of thingswith naturallysmoothlyvaryingprices, it seems
thatit would be intrinsicallyquite feasible to make use of costs of services, energy,or prices
thatdepend on the national location of the definitionof the commodity,service, or asset being

This content downloaded from 205.133.226.104 on Wed, 24 Jul 2013 01:35:40 AM

All use subject to JSTOR Terms and Conditions
10 JohnF. Nash,Jr.

priced. Hence, by using this approach, the temptationto include thingsthat would otherwise
seem inappropriatejust to obtain stabilityor smoothnesscan be avoided.
Of course, the fundamentalprincipleremains thatif a political basis existed for changes
in a standardindex, it is not unlikelythat a formof corruptionwould appear. This issue is
comparable to the issue raised recentlyin the United States when certaininterestswished to
devalue the originalCPI computedby the Labor Departmentso as to have more federalbudget
money available to reduce taxes or for otherpurposes, with Social Securitybeneficiariesre-
ceiving less.

Practical Considerations concerningLong-Term Value Trends and the Safe-

Deposit Box Singularity
There is a problem for the issuer of a currency,whetherin coinage, paper,or electronic
form,that if this currency(or money) is too good, then it could be exploited by all sorts of
parties and intereststhat might simply wish to safely deposit a store of wealth or even to
conservativelyinvest some assets forfuturegood value. (The word good is used here in terms
of comparativevalue trends,as in good investments.)Then, underextremeconditionsthe cur-
rency issued by one state could be exploited by parties not of that state as a sort of "safe-
deposit box" on which they would not need to pay any rentalfees or fees like those paid to
the managersof mutualfundsforinvestment.
But, simplyto improvetheconditionsunderwhichagreementsregardinglong-termlending
and borrowingwould be made, a money would be more or less equivalentlygood if it had a
completelysteady and constantrate of inflation.Then this inflationrate could be added to all
lending and borrowingcontracts.Hence, the problem of a money that would be too good is
avoidable. However, this has not been known as a problem in the past historyof money. In
fact,thereis a naturalcharacterof the sortof materialthingsto which money standardshave
been relatedthatmakes themnaturallytendnotto provide "too good" a channelforinvestment.
It has simply not been profitablein the past for any economic parties to use any varietyof
money as a channel for investment,although,of course, therehas been much currencyspecu-
lation,which is a different thing.Of course, the issuer of a currencyalso needs to be properly
prepared for the possibilityof speculationon the part of interestsdomiciled in foreignstates,
etc., etc.

Natural Value Trends

The long-termtrendof the value of any index of priceswill depend,sometimespredictably,
on the choice of the compositionof thatindex. It is a coincidentalfactthatthe inherentnature
of miningand miningtechnologymakes it possible for the prices of certaincommoditiesthat
are produced as a resultof the devotionof labor and capital to the effortof miningto increase
less (or decrease more) than mightbe expected. There is a "dimension paradox": Agricultural
products are produced by using the two-dimensionalresource of the earth's surface, so the
"disappearingfrontier"createsa limitation.In contrast,some mining,particularlyforelemental
metals, can essentiallybe done in threedimensions,although,of course, thereare increasing
costs for deep digging. So, really thereis lots and lots of gold, silver,platinum,tungsten,and
so forthout thereand more can be foundby diggingdeeper.
Thus an index can be chosen so that its value does not rise, say, like that of a typical
Rembrandtor like the ratio of the human population to the total surface area of the earth.If

This content downloaded from 205.133.226.104 on Wed, 24 Jul 2013 01:35:40 AM

All use subject to JSTOR Terms and Conditions
 Distinguished Guest Lecture 11

the value trendof a currencyis such thata naturalinterestrate is not negative,then it is not
an unattractivetask for a centralcurrencyauthorityto mintor printthe physical currencythat
would circulate.Then, the issuer of currencywould be partiallyin the position of a borrower
not paying intereston borrowedmoney.

6. Author'sNote

The above textoriginallyderives frommy outlineforthe lecturesgiven at various specific

locations of the European School of Economics in Italy in October 1997. Subsequent to that
time, afterconsultingwith some of the economics facultyat Princeton,I learned of the work
and publicationsof Friedrichvon Hayek. I mustsay thatmythinkingis apparentlyquite parallel
to his thinkingwith regardto money and particularlywith regardto the nontypicalviewpoint
regardingthefunctionsof the authoritiesthatin recenttimeshave been thesources of currencies
(earlier coinage).

References
Antonelli,Giovanni Battista. 1971. On the mathematicaltheoryof political economy. In Preferences,utility,and demand:
A Minnesota symposium,edited by JohnS. Chipman, Leonid Hurwicz, Marcel K. Richter,and Hugo F. Sonnen-
schein. New York: HarcourtBrace Jovanovich,pp. 333-64.
Keynes, JohnMaynard. 1921. A treatise on probability. Temecula, CA: Best Books.
Von Neumann,John,and 0. Morgenstern.1953. Theoryof games and economic behavior.New York:JohnWiley & Sons.

This content downloaded from 205.133.226.104 on Wed, 24 Jul 2013 01:35:40 AM

All use subject to JSTOR Terms and Conditions