In 1341, Mohammed bin Tughlaq suffered a heavy drain upon the treasury from his munificence in dealing with the distress caused by his decision to move the capital to Daulatabad and back again, as the new capital did not have an adequate water supply. With many perishing on the two forced marches, the sultan opened the fisc to mitigate the distress of his subjects. To meet this fiscal drain, the sultan made his famous experiment of instituting a token currency of brass and copper to replace the silver and gold coinage. But he forgot that the success of this scheme depended upon a state monopoly in issuing the token currency. The result says contemporary political thinker Ziauddin Barani “turned the house of every Hindu into a mint, and the Hindus of the various provinces coined crores and lakhs of copper coins”. Their value relative to the silver-gold old currency soon fell so low “that they were not valued more than pebbles or potsherds”. With the ensuing disruption of trade “the sultan repealed his edict, and in great wrath he proclaimed that whoever possessed copper coins should bring them to the treasury and receive the old ones in exchange…. So many of these copper tankas were brought to the treasury that heaps of them rose up in Tughlakabad like mountains.” How this run on its reserves was met is unexplained, but as the 19th century expert on Indian numismatics E Thomas pointed out “if good money was paid for every token, true or forged, the sultan’s temporary loan from his own subjects must have been repaid with more than even oriental rates of interest” (Stanley Lane-Poole: Mediaeval India, pp. 135-136). This was the first conspicuous Indian example of what Milton Friedman described in his book entitled Money Mischief.
On November 8, 2016, Prime Minister Narendra Modi announced the demonetisation of 86 per cent of Indian currency and its replacement by new notes. But the remonetisation, which ideally needed to occur simultaneously to prevent any contraction of output, was dilatory. Given the need for secrecy, the replacement notes could not have been printed and stored domestically before the announcement. After November 8, printing presses at Indian mints were estimated (in an editorial in this newspaper) to take about a year to print the replacement currency. That this was correct, see Figure 1 showing the broad money supply in India from 2010 to 2017. There was a massive monetary contraction in the last three months of 2016 which has not still been fully reversed (on the basis of the annual rate of money growth).
As the editorial in this newspaper also noted, given the constraints on the speedy replacement of notes through domestic mints, India could have used foreign facilities for printing currency. In the mid 1990s, I was a visiting fellow at the International Centre for Economic Research in Turin. My neighbour in an adjoining office was the last governor of the Somali central bank before his country descended into chaos and he was exiled. He showed me a graph of the inflation rate in lawless Somalia. It was flat, he explained, because Somali currency had been securely printed by the Swiss firm De La Rue and shipped by air to Somalia. The last consignment had arrived just before Somalia collapsed and he fled. As there was no government to pay for any further printing and shipping of currency from Switzerland, the Somali money supply remained fixed, and hence there was no inflation!
So we would expect the monetary contraction lasting nearly a year after demonetisation would have led to a sharp contraction in nominal GDP. The Economic Survey 2016-17, Volume 2, in Box 3, notes that instead growth of nominal GDP accelerated with the substantial decline in broad money. They call this is a puzzle. It is not. As the informal sector is primarily cash-dependent, it would have contracted disproportionately more than the organised sector. But, a continuing lacuna in Indian national income accounting is that informal sector output is only reliably surveyed in quinquennial NSS surveys. In the intervening years it is estimated as a fixed proportion of the organised sector. So there is no firm basis for judging the effects of demonetisation on nominal GDP. Moreover, there is no long-term survey data on monthly or quarterly employment in the informal sector to judge the employment effects after demonetisation, which may not be reversible.