The hypothesis at Wörgl was about speed. Money that cost something to hold, Gesell said, would be spent, and money that was spent would move faster than money that was kept, and the faster movement would do the work that the missing money could not. Of all the claims made in the town in 1932 and 1933, that was the one a town of four thousand people could have tested by counting. This piece follows the number everyone quotes back to where it came from.

Twenty times a month

The first figure came from the mayor. Irving Fisher sent a man from the Geneva office of his index-number service, Hermann Scheibler, to Wörgl at the end of 1932 to question the mayor, the bank, the merchants and the workmen, and in the chapter of Stamp Scrip that resulted Fisher wrote: “As to the rate of turnover, the mayor reports that the amount of the first wages paid has returned to the city 20 times a month.” He printed the figure again in the manual for American towns that makes up the second half of the book, in a list of what was known about the speed of scrip in the places that had tried it: Russell, Kansas, once every three days; Rock Rapids, Iowa, once every four; “Scrip turnover in Woergl (dated type) was estimated at 20 times a month; but what had been the Woergl turnover for conventional money, I am not informed.” Fisher’s own general reasoning, earlier in the book, was that a scrip certificate would be passed on at least twice a week, four times the speed of a dollar in normal times, and that since in a depression a dollar moved at about a third of its normal speed, twice a week was “twelve times the depression average for conventional money”. That was an argument, not a count, and he said so.

Twice a week

The second figure came from an opponent. Alfred Hornung, an official in Innsbruck, went through the Wörgl accounts for a study published in 1934 in the series Tiroler Studien, under the title “The result of the Wörgl shrinking-money experiment” and the subtitle “Is Wörgl a free-money experiment?” He answered the subtitle’s question no, but his count is the only one that anyone made from the books. The emergency money, he wrote, “was for the most part back in the municipal treasury within a few days and could be used again by the commune for payments,” so that “one could reckon that the 1,600 schillings of emergency money flowed through the municipal treasury about twice a week.” Twice a week is 104 times a year. Thomas Wendel, who went through the same material for a thesis in the 1990s and whose figures the Unterguggenberger Institut uses, reckoned the passes through the treasury at 36.43 a year, a third of Hornung’s; the two were counting different things over different months, and neither count has been published with its workings.

Both were counts of arrivals at the one desk in Wörgl that kept a ledger. Between leaving the treasury as wages and coming back as taxes a note might change hands once or ten times, and nobody wrote that down.

Four hundred and sixteen

The third figure is Fritz Schwarz’s, from the book he wrote for the Freiwirtschaft congress that met at Wörgl at Whitsun 1951. He took Hornung’s two passes through the treasury a week, and then wrote, in my translation: “If in between they changed hands three more times, that is six times a week, that makes for every schilling six more payments, together eight a week, or 52 times 8 a year = 416 a year….” He went on: “The 5,490 schillings,” the average amount in circulation, “if they change hands 416 times a year, which can be right, mediate in 13.5 months each 464 schillings’ worth of exchange of goods or of labour or of payments. That makes for the 5,490 schillings in free money 2,547,360 schillings!” He set this against the estimate of the Swiss National Bank’s president that an ordinary franc changed hands about 200 times a year, and concluded that the exchange work of the Wörgl money “must be estimated at more than twice that of a normally circulating franc”.

The chain is worth laying out. Hornung observed that the notes came back to the treasury about twice a week. Schwarz assumed three hand-to-hand payments between each return. He multiplied by fifty-two, called the result something that “can be right”, and multiplied again by the average circulation to get a figure for total business done. In 1998 the mayor’s son, Silvio Unterguggenberger, carried the 2,547,360 schillings forward to the purchasing power of 1998 and arrived at 63,684,000 schillings. The Institut’s pages give 416 a year and 464 in thirteen and a half months with Schwarz’s name attached; most retellings drop the name and keep the number, and some attach it to Fisher, who never printed it.

What was not measured

The hypothesis needed two numbers, and the town recorded neither. The first was how often a certificate changed hands in all. The design made that nearly measurable: every note was numbered, every stamp sold recorded a month of someone holding a note, every redemption at the bank was a note leaving the circuit, and the treasury could log every return. Nobody was asked to keep the log, and the shopkeepers, who handled the notes most, kept no count. The second number was the speed of the ordinary schilling in Wörgl in the same months, against which the certificates were supposed to be fast. The Institut’s modern pages say the notes moved “nine to ten times faster than the national currency”, and Wendel’s thesis is cited for the figure that each official schilling sold 8.55 schillings’ worth of goods in the period while each schilling of free money turned over up to 73 schillings’ worth a year; but the schilling figure is a national average divided by a national note issue, and the free-money figure is Schwarz’s 416 again in another form. Fisher’s sentence of 1933 still stands: what the turnover of conventional money in Wörgl had been, nobody was informed.

None of this means the notes were slow. Everyone who watched them, the mayor, Muralt, Bourdet, Pfister’s shopkeepers and Hornung himself, agreed that they moved, and the treasury’s receipts are evidence that they moved towards the treasury. What it means is that the one measurement the experiment was built to make was never made, and that the number which has stood in for it since 1951 is a count at one desk multiplied by an assumption.

Sources

• Irving Fisher, assisted by Hans R. L. Cohrssen and Herbert W. Fisher, Stamp Scrip, New York: Adelphi, 1933, chapters III, IV and VI. The mayor’s twenty times a month, the twelve-times argument, and “I am not informed.” userpage.fu-berlin.de
• Alfred Hornung, Das Ergebnis des Wörgler Schwundgeld-Versuches: Ist Wörgl ein Freigeld-Experiment?, Tiroler Studien 2, Innsbruck 1934, as quoted in Fritz Schwarz, Das Experiment von Wörgl, Bern 1951, which also gives Schwarz’s own arithmetic and the comparison with the franc. userpage.fu-berlin.de
• Unterguggenberger Institut Wörgl, “Wörgler Freigeld historisch” and “Fakten zum Wörgler Freigeld”, for Wendel’s 36.43, the 8.55 and 73 schillings, the “nine to ten times”, and Silvio Unterguggenberger’s 1998 recalculation. unterguggenberger.org
• Gebhard Ottacher, Der Welt ein Zeichen geben: Das Freigeldexperiment von Wörgl 1932/33, Kiel 2007, pp. 43 and 60 to 61, for Wendel’s figures. silvio-gesell.de