1.International Journal of Community Currency Research VOLUME 17 (2013) SECTION A 45-57COMPLEMENTARY CURRENCY AND ITS IMPACT ON THE ECONOMY Octavio Groppa* Pontificia Universidad Catlica Argentina ABSTRACT The paper aims toshow the impact that a complementarycurrency may have on a national economyfromatheoretical point ofview.Asystemdynamicsmodeliscreatedtodescribethe mechanicsofmoneyissuancein capitalisteconomiesaswell asineconomieswhere thereisno inside money.As an example, the ;irst outcomes of a barternetwork implementedin 2008by the STROfoundation in El Salvador(called PuntoTransacciones) are presentedand analyzed. Finally,usingdatafrom acomplementarycurrencyexperienceinElSalvadorthe spending mul- tiplieriscalculated.The main resultshowsthatthere isa greaterspending multiplierin digital community currenciessystems thaninregularmoney market.AlthoughthemagnitudeofPT networkisstillnegligiblefrom amacroeconomicpoint of view,theresultisadesiredoutcome whichmayhelptocushiontheimpactofmacroeconomicshocks onlabourmarket, contributing tostabilizeaggregatedemand.*Email:octavio_groppa@uca.edu.ar To cite this article: Groppa, O. (2013)Complementarycurrencyandits impact on the economy Interna- tionalJournalofCommunityCurrencyResearch17(A)45-57ISSN1325-9547

2. INTERNATIONAL JOURNAL OF COMMUNITY CURRENCY RESEARCH 2013 VOLUME 17 (A) 45-571.INTRODUCTIONGROPPA2.THEMECHANICSOFMONEYISSUANCEThe expansion of IT in recent years has brought about a numberof innovations inthe monetaryarena. Firmshave beenexperimenting withvouchers, discount coupons,sky miles,points,and the like,inorderto attractcustomers or preserve them throughloyaltyprograms.Inparallel,com- plementarycurrencies(CC), manyof whichuse computer- based exchange systems, have been growing year after year.Theseexamplesshowin agreaterorlesserdegree the emergenceofthequietnewphenomenon ofprivateornon- nationalmoneys.2.1TheCapitalistSystem In our traditional economies money is created through a two-stage process.In the ;irstplace,a central bank issues debt money backed by its assets. It can be done as a means ofmonetizinga low liquiditydebt(typically,whenit purchases governmentssecurities),orelsebygivingcredit tocommercial banks.Thisdebtinstrumenthas a high de- greeofliquidityandabilitytobe transferredtothirdpar- ties, given that it is invested with an unrestricted settle- mentpower.Thisiswhatweknow asthemonetarybase or outsidemoney.A review ofthe literature on the subjectreveals that few studiespresent a theoretical approachtothe issue. Schra- ven (2000) analyzes the functions and bene;its expected from joining in a complementary currencynetwork,buthe doesnotofferanexplanation oftheirmechanism. Stodder (1998; 2009) ;inds a counter-cyclical effect of corporate bartersystemsaswell astheWIRBankscurrency,respec- tively, using econometric techniques. Pja;jar et. al. (2012) useprobitmodelstoprovethatcommunitycurrencies are complementary (not substitutes) of national moneys and that the usage of thesealternative currenciesisassociated withmonetarystability,whatconstitutesa ratherpuzzling conclusion. Schroeder (2002), in his description of the Tauschring, offerssome re;lections on a handful of issues related to CC unemployment, locality, ecology, between others.He does not either construct any formal model or explanation,butrevisitssomeideasofGesellandhis critics, or Lietaer.Krohn and Snyder(2008),in their studyofIth- aca Hours, mention some characteristics of economies where complementarycurrenciesappear, anddescribe the basicfeatures of money. Theyalsorun a model in the US over the 1990s in order to ;ind some effect of a higher growthrateincities where complementarycurrencieshad beendeployed,buttheycant;indsuchevidence.With the moneyreceived,commercialbanksofferloans.In turn, borrowed money isspentbycredit takers (e.g.,com- paniesthatpurchasesuppliesandpaywages).Then, a sub- sequentcirculationofmeansofpaymenttakesplace,which perform a given number of operations. There is, on the other hand, an amount of money that can be hoarded or saved inforeign assets.Conversely, itis alsopossible that previouslyhoardedamountsare re-enteredintothecircuit. Inaddition, there existsanotherportionofthe amountsin circulation thatisreturnedinto the banking system in the form of deposits. In this case, commercial banks keep a portionasreservetomeetwithdrawals,andare allowedto lendthe remainder.In sofarthe originaldepositorcontin- ues using his money writing checks against his current account, that "remainder" provided by the commercial bank is newmoneyput intocirculation,what is knownas inside money and is founded on the described fractional reserve system (asnotedbyTobin1963, strictlyspeaking thissecondaryissue occursin thesystemasa whole,since credit creationbyonebankmaybeabsorbedoffsetas a depositbyanotherone.) The moneymultipliereffect canalsobe seenasadynamic systemwheresecondarycreation isde;inedbyarecursive structure formed by depositsof savers that are put again into circulation as loans, generating expenditure (con- sumption and investment) and new savings deposited in the bankingsystem.Since wearefocusingin the structure ofthemoneycreation process(andnotdeterminingmoney supply), for the sake of simplicity we overlook interest ratesandthe;lowofloan cancellation(whichcouldberep- resented by an arrow from Income to Credit), as well as other types of deposits (saving or time deposits) or even more complex credit instruments like derivatives, swaps and others (which were in the very origins ofthe current crisis).ThesystemisdepictedinFigure1.Most ofthe articles describe the mainfeatures ofcomple- mentary currenciessystems (CCS) in several experiences, either using the case study methodology (see IJCCR2011; Schroder 2006; Lietaer2004; Brenes2011; Ingleby1998) or conducting surveys (Lepofsky 2005; Jackson 1997). Theyusuallydetailthesingularities ofthe implementation ineachcase. Theobjective ofthepresentpaperistoanalyzefroma con- ceptualpointofviewthe impactofCC inthemonetarysys- tem. The ;irst section presents a system dynamics model which describes the core of the structure ofconventional monetarysystemsinoureconomies. Thefollowing section discusses whathappens whena CC is introduced.Thirdly, empirical data from the complementary currency system developed by the Social Trade Organization (STRO) in El Salvador (called PuntoTransacciones)isshown and a ;irst attempt to measure the spending multiplier within the networkismade.The;inalsectionconcludes.46 3. INTERNATIONAL JOURNAL OF COMMUNITY CURRENCY RESEARCH 2013 VOLUME 17 (A) 45-57GROPPAFigure 1. Diagram of money creation under capitalism The diagram showsaninitial ;low(outside money) thatis transformed intocredit.Credit, from theside ofthe credit taker,isincome thatwill bespentorelsesavedorhoarded. The spent portion is another agents income (hence the regressive loop).But,giventhatconsumptionpropensity is less than 1,this "multiplier effect" (of expenditure or de- mand,inthiscase) has diminishingimpact.The statevari- able Saving accumulates savings generatedoneach round ofthe spending multiplier.They are entered in bank sys- tem asdeposits. Banks,after putting aside an amount as reserve,lendthemagain,carryingoutasecondaryissuance that will satisfy demand for credit. Thus, bank reserves perform an entropic function, as well as a positive net hoarding.The variableHoarding isdistinguished from Sav- ings inordertoremark itsdiverse function in relation to the economic system. The ;irstone expressesthe commu- nication withthe"outsideworld",whilethesecondonehas a role in the circulation and internal redistribution of monetary resources (allocating them, for example, to in- vestment). Hoarding operates as a buffer receiving leak- ages from the system. However, given that any hoarding will be consumed at some futuretime,thereis a returning ;lowto the systemthatwill takeplaceinanuncertaindate. The magnitude ofthiscumulative variable dependson the time agentswanttokeeptheirsavingsoutsidethesystem, which is in turn function ofrealinterestratespaidon sav- ings within the system and on the abilitytomaintain the value ofsuch amounts(e.g.,by purchasing external assets, provided,of course,that itsvalue is not overestimated,as in a ;inancial bubble).The model is ideal insofarasit ex- plainsthestructure ofthemoneycreatingprocess,without reference to the behaviour of real variables. That is, the amountofcreditissuedis determinedby demand,whichis here an exogenousmagnitude.Finally,notice thatthissys- tem is de;ined by macroeconomic functions, and not by types of actors or currencies. In other words,those func-tions maybe performedusing cash,e-money orothercur- rencies,whatsoever. Therefore, the system has two instances of recursion (or "multipliers"), whichform bothgeometric series:spending andmonetaryissuance. 2.2FormalExpression Figure 1 canbeexpressedasa simple system ofdifference equations based on the de;inition of the variables that comprise it. Forsimplicity,we assume that net hoarding is zero:(1)Where Yt isincomegenerated withinthemonetarycycle t; is the marginal propensity to consume; t is velocity (with assumedtobe ;ixed),thatis,the numberofoutlays performed by monetarycycle (t) with a same original in- come; Cr iscredit; e1 and e2 are outside money (which at ;irstisdata) andinside money,respectively;andD,depos- its.Thesystemcanbesummarizedas:47 4. INTERNATIONAL JOURNAL OF COMMUNITY CURRENCY RESEARCH 2013 VOLUME 17 (A) 45-57GROPPAplus income from a third party (which, strictly speaking, would be a redistribution), or by creating new money (primary issue or outside money). Insofar there are no incentives to the accumulation of surplus income, almost all income isspent within a given lapse time,andinterest rates can be close to zero. The monetary system is now simpler(see;igure2below).(2)ifonlytwocyclesofissuance areconsidered.Ifit isdesired to extend the time period,sothat there were included w periods of monetary recurrence, the expression takes the followingform:Thismeansthat now =0and,so, Ytin(2)islarger.Asitis obvious, from the other side, to get a similar level of Y, a system withnoinsidemoneyneedsagreaterprimary issue (e1). Itshouldbenoted thata monetarysystemis a structure of information on relative balances of members within an exchangecommunity.What distinguishesone system from anotherare design features, such as the magnitude of ag- gregate imbalances thatwill be tolerated(aswell as their average duration), the magnitude of the imbalances each memberisallowedto have(accesstocredit,itsperiod), the authority empowered to make such decisions(banks, the communityasa whole through anyestablishedprocedure ofsocial choice), criteria underlying those decisions(rate ofROI / NPV, or, rather,in virtue ofvalues ofa different nature social, ecological, etc.),orwhatmoneyrepresents, that is, whatisitsreal reference thethingor,put in an- otherway, whatkindofsupport ithas(the workofmem- bers, a third currency, a commodity, otherkind of goods, assetsormembersproperty).(3)sothatiftandwtendtoin;inity,(4)However, the bene;its of the analyzed design from the theoretical point of view are particularly quali;ied when analyzing the economic performance ofthese subsystems. As it was mentioned, the main objective of these experi- ences is to ensure local labour. Thus, the effect from the macroeconomic standpoint is limited (Vasconcelos Freire 2011,126).Thefactofnotprovidinglargecredit amounts isalso an important barrier for their expansion. Actually, the largest economiccircle, the Swiss WIR, became abank when administrators decided to offer loans beyond the mere redistribution ofsurplusesamong its members, that is, issuing means of exchange over the amount of open credit balances (creating therefore inside money) (Studer 1998).which summarizesthe effects of bothmultipliers. The dis- tributionin time ofthemultipliereffectdepends,therefore, onthevelocityofcirculationofmoney.3.COMPLEMENTARYCURRENCYSYSTEMS 3.1MainFeatures ACCSaimstocreate asubmarket where meansofpayment circulate, preventing leakage from the system (reducing entropy) and hencemaximizinglocal labour.Itoperatesas asortofvirtualmarketprotection. Perhapsanelementthat distinguishes CCS from pure mu- tual credit systems is that the ;irst ones function as an economy with a ;ixed oradministeredexchange rate with reserves to backup legal tender (foreign currency and/or some sort ofcommodity, like gold). This design facilitates the recruitmentof members to the network, since anyone whoenters isable to exit wheneverhe wants, getting the moneyoriginallydepositedback(on paymentofa tax that punisheswithdrawals).3.2MacroeconomicimpactofCCS The introductionofa CCS has manyanglesto be analyzed regarding its impact on the market. We mention particu- larlytheeffecton spending,interest,in;lationand business cycle. Beginning withthe effect ontotalspending,acomplemen- tarycurrencyexpandsaggregate demandvia anincreasein thespending multiplier.Asnoted,bydiscouragingthe stor- age of surpluses without application, marginal propensity to consume grows, producing an increase in consumption and the use oflocal labour. That is, income is completely spent (within a given time lapse),so thatSays law isful- ;illedwithinthesystem.Since there is no ;inancial intermediation, CC conform monetary systems with no deposits (and no bank re- serves); hence, savings can be lent directly from surplus suppliersto creditdemanders.Inotherwords, to perform an economic operation any person can obtain means of payment in twoways: eitherthroughdirectlending ofsur-48 5. INTERNATIONAL JOURNAL OF COMMUNITY CURRENCY RESEARCH 2013 VOLUME 17 (A) 45-57GROPPAFigure 2. Diagram of money creation in a mutual credit system Letusseethisformally.Whenacomplementarycurrencyis introduced,circulationoflegal tenderisdiminished by the amount ofmoney setaside to support the new monetary subsystem. However, total spending is increased because consumption increases in CC, as this has a lower (aggre- gate) saving rate thanlegal tender(or what is the same, a consumptionpropensitythatisonaverageequal ornextto one).eachcurrency hforeveryagentj, weightedbyj,savingsin kwillbede;inedby:(7) wheresubscripth =1indicatesthe proportionofspending on legal tender, while the value 2 is the spending which uses complementarycurrency. However,since 2 tends to zero,giventhatthereisnoincentivetotheaccumulation of savingsinthesubsystem:Atany moment,deposits(di) are a portionofincome (Y). With+=1wecanexpress:(5) wherediare newdepositsmadeatthemomenti,expressed asaproportionofincome(Y) From thenonwards, totalexpenditure will be done in two currencies:Now assume that at time k a portion of moneyincircula- tionisputasidetosupportaCC;consequently,spending on legal tenderwillbereducedonthatamount bydiminishing the average saving rate (), as seen below. Being J the numberofagents, the average savingsrate oftheeconomy istheaverage savingrateofeachmember weightedby the proportionbetweentheirincomeandtotalincome.(8)being the expenditure done in complementarycurrency, and Pm/P the relationshipofpriceindexesineachsystem (legal tender and complementarycurrency, respectively), whichoperatesasaconverterforpurchasepowerparityor realexchange rate.It isexpected thatPm/P 1,given that complementarycurrencyhasalimitedscope.(6)Discriminating two groups depending on the currency used, and de;ining jh as the proportionofexpenditure on49 6. INTERNATIONAL JOURNAL OF COMMUNITY CURRENCY RESEARCH 2013 VOLUME 17 (A) 45-57 On the otherhand, the effect in time of the money multi- plier will condition expenditure (Y) through the following expression:GROPPAsince 2 tendsto zero within the CC subsystem. However, from the period k+2, does not lose the portionreserved asbackup,sothat:(14) (9) Thus,undera bi-monetarysystem with CC, total spending (or income,Y)is (inshortterm) higher thantheonewhich occurs underpurecapitalistsystem because oftheincrease inthepropensitytoconsume.Sothat,if in moment k acomplementarycurrencyisintro- duced,thepreviousexpressioncanbe split in twoperiods, andexpenditureinlegaltenderwillbeconditionedby:Ascan bededuced,introducingaCCoperatesasadecrease in ratesofinterestvia thereductionofthe relative weight ofinterests paid in relation toexpenditure,since interest rate is i=Interest amount/Y, and,aswas argued, Y=G