Accordinglyhegivesnoticethat，forthepurposesofhisbook，hewillassumethistobethecase，andmoneytobe’invariableinvalue。’31Wecanthus，ontheonehand，comparevaluesatdifferentperiods。Athinghasthesamevalueatalltimeswhichatalltimesrequires’thesamesacrificeoftoilandlabourtoproduceit，’32The’sacrifice’measuresthe’utility，’andwemayassumethatthesamelabourcorrespondsinallagestothesamepsychologicalunit，But，ontheotherhand，atanygivenperiodthingswillexchangeinproportiontothelabourofproducingthem。
ThisfollowsatoncefromRicardo’spostulates。Giventhesinglerateofwagesandprofits，andassumingthecapitalemployedtobeinthesameproportion，thingsmustexchangeinproportiontothequantityoflabouremployed；forifIgotthesamevaluebyemployingonelabourerasyougetbyemployingtwo，myprofitswouldbehigher。Ricardo，indeed，hastoallowformanycomplexitiesarisingfromthefactthatverydifferentquantitiesofcapitalarerequiredindifferentindustries；butthegeneralprincipleisgivenbythesimplestcase。Hencewehaveameasureofvalue，applicableatanygiventimeandincomparingdifferenttimes。Itimplies，again，whatM’Cullochsumsupasthe’fundamentaltheorem，’thatthevalueof’freelyproducedcommodities’dependsonthequantityoflabourrequiredfortheir’production。’Whatismadebytwomenisworthtwicewhatismadebyoneman。ThatgiveswhatM’Cullochcallsthe’cluetothelabyrinth。’
Thedoctrineleadstoapuzzle。
IfIcanmeasurethe’sacrifice，’canImeasurethe’utility’whichitgains?The’utility’ofanounceofgoldisnotsomething’objective’likeitsphysicalqualities，butvarieswiththevaryingwantsoftheemployer，Ironorcoalmaybeusedforaninfinitevarietyofpurposesandtheutilitywillbedifferentineach。Thethingmayderivepartofits’utility’fromitsrelationtootherthings。Theutilityofmyfoodisnotreallyseparatefromtheutilityofmyhat；forunlessIeatIcannotwearhats，mydesireforanyobject，again，ismodifiedbyallmyotherdesires，andevenifIcouldisolatea’desire’asapsychologicalunit，itwouldnotgivemeafixedmeasure。Twicethearticledoesnotgivetwicetheutility；adoublestimulusmayonlyaddasmallpleasureorconvertitintoagony。Theseandotherdifficultiesimplythehopelessnessofsearchingforthischimericalunitof’utility’whenconsideredasaseparatething。Itshiftsandescapesfromourhandsdirectlywegraspit。Ricardodiscussessomeofthesepointsinhisinterestingchapteron’ValueandRiches。’Gold，hesays，maycosttwothousandtimesmorethaniron，butitiscertainlynottwothousandtimesasuseful。33Suppose，again，thatsomeinventionenablesyoutomakemoreluxuriesbythesamelabour，youincreasewealthbutnotvalue。Therewillbe，say，twiceasmanyhats，buteachhatmayhavehalfitsformervalue。Therewillbemorethingstoenjoy，buttheywillonlyexchangeforthesamequantityofotherthings。Thatis，hesays，theamountof’riches’varies，whiletheamountofvalueisfixed。This，accordingtohim，provesthatvaluedoesnotvarywith’utility。’’Utility，’ashedeclaresinhisfirstchapter，is’absolutelyessentialtovalue，’butitis’notthemeasureofexchangeablevalue。’34Asolutionofthesepuzzlesmaybesoughtinanymoderntext-book。Ricardoescapesbyanapparentlyparadoxicalconclusion。Heisundertakinganimpossibleproblemwhenhestartsfromthebuyers’desireofan’utility，’thereforeheturnsfromthebuyerstothesellers。Thesellerhasapparentlyameasurableanddefinablemotive——thedesiretomakesomuchpercentonhiscapital。35Ricardo，unfortunately，speaksasthoughthetwopartiestothebargainsomehowrepresentedmutuallyexclusiveprocesses。’Supplyanddemand’determinethevalueof’monopolisedarticles，’butthecostofotherarticlesdependsnot’onthestateofdemandandsupply，’but’ontheincreasedordiminishedcostoftheirproduction。’36Why’not’and’but’?Ifsupplyanddemandcorrespondstothewholeplayofmotiveswhichdeterminesthebargain，thisislikesaying，accordingtotheoldillustration，thatwemustattributethewholeeffectofapairofscissorstoonebladeandnottotheother。Hisviewleadstotheapparentconfusionoftakingforthecauseofvaluenotourdesireforathing，butthesacrificewemustmaketoattainit。Bentham37said，forexample，thatRicardoconfused’cost’with’value。’Thedenialthatutilitymustinsomesenseorotherdeterminevalueperplexesanintelligibleandconsistentmeaning。
Itisclearlytrue，uponhispostulates，thatthevalueofgoods，otherthan’monopolised，’mustconformtothecostofproduction。Hespeaksasifheconfoundedanecessaryconditionwithan’efficientcause，’andasifoneoftwocorrectiveprocessescouldbeexplainedwithouttheother。