In this article, Amartya Lahiri, Professor of Economics at the University of British Columbia, argues that all public policy must rely on a clear-headed cost-benefit analysis and the recent demonetisation move fails the test.

It has now been over three weeks since the demonetisation of 86% of India´s circulating cash. The main logic that has been given for the measure is that it is a decisive blow to holders of undeclared income, that is, black money. Black money has both a stock and a flow aspect. To assess the impact of the demonetisation we need a quantitative estimate for both of these aspects.

A recent World Bank survey suggests that the parallel economy is around 25% of GDP (gross domestic product), which gives us an estimate of the flow share of the underground economy. The wealth share (stock) of the underground economy is more difficult to estimate. A recent Credit Suisse study estimates the wealth to GDP ratio in India to be around 2. Assuming that wealth creation is similar for both declared and undeclared income, this would suggest that black wealth is about 50% of GDP. It is likely larger since saving rate out of undeclared income is probably greater than that out of declared income. Nevertheless, armed with these two estimates, let us go through the impact of the demonetisation move.

How much can the demonetisation mop up? One measure would be how much is exchanged as a share of what was demonetised. Note that this would be an overestimate since even I have some rupees that can´t be converted as I am not in India right now. The demonetised money is about 10% of GDP. Even if the entire amount was left un-exchanged it would amount to around 40% of the underground economy (or black income) and 20% of black wealth. My guess is that about 85-90% of the demonetised cash will be exchanged (note that as of 28 November 2016 around 60% of the demonetised cash had already been turned in). Hence, the amount that this move will net will be around 2-3% of the black wealth in India (or 4-6% of black income).

These estimated numbers are rather small when compared with the overall stock of undeclared income held in other forms and in view of the disruption the move has created. This is primarily because most undeclared income is not held in cash but rather in gold, jewellery, real estate, foreign currency, or quickly laundered through round-tripping¹.

What about the costs of the move? Let me start by saying that it is very unlikely that the move will have serious long-term implications, save for one concern that I will highlight below. The short-run costs though are unlikely to be insignificant. As has been pointed out by a number of different commentators and analysts, the Indian economy is heavily cash-dependent. The informal sector in particular, depends almost exclusively on cash for transactions. Given that this sector accounts for 80% of employment and 45% of GDP, the costs can clearly be large.

Assessing the size of these costs can be difficult. Given that the disruption is primarily on people not having cash to make payments, one way to make progress is by using a macro/monetary model that is based primarily on the transactions motive for holding cash. This model is fairly rudimentary and is known as the ‘Quantity Theory of Money’. The quantity theory is based on one equation: MV = PY, where M denotes the quantity of money in circulation, P is the price level, Y is GDP and V denotes the velocity which captures the speed with which money circulates in the economy. The equation, based on the idea that money is used to buy up goods and services in the economy, predicts that for a given velocity, a fall in M would translate into a fall in nominal income which is PY.

While the quantity theory seems easy to implement, to use this equation to forecast the effect of a monetary contraction on real GDP, however, one needs to make some assumptions regarding the size of the fall in M, whether velocity will remain unchanged at all, and how much of the change in M will be accommodated by a change in prices P rather than in Y. As an example, under the assumption that 85% of the demonetised cash is exchanged for either new notes or bank deposits, the resultant fall in M1 (sum of currency in circulation and demand deposits with banks) over the months of November and December would be around 8%. Based on this and the fact that nominal monthly GDP is about Rs. 11.5 trillion, the quantity theory would predict that the loss in nominal GDP during these two months would be around Rs. 1.8 trillion. Hence, nominal growth over the fiscal year (FY) 2016-17 would be lower by Rs. 1.8 trillion. If we were expecting 7.5% real GDP growth over this fiscal year and 4% inflation (or an 11.5% nominal growth), then this output reduction due to demonetisation would imply a 1.7 percentage-point reduction in nominal growth. As long as inflation remains as expected, this would imply that real GDP growth for FY16 would be 1.7 percentage points lower than expected.

The calculation above should make clear that there are a number of uncertainties in this projection. The amount of demonetised cash that is returned could be higher or lower, velocity of M1 could decline as people change their behaviour in response, inflation could come in higher or lower, or M1 could fall less as the money multiplier itself rises due to a decline in the cash to deposit ratio. Indeed, varying these assumptions, one could easily come up with real GDP growth rate cuts for FY16 that vary from a low of 0.5% to a high of 3.5%. One shouldn’t be surprised if the cost ultimately comes in somewhere between 1.5 and 2% of GDP growth. That is high.

There is another aspect to the demonetisation move that is worth stressing. The conditions under which it has been done implies that the government has converted its uncontingent liabilities into contingent liabilities. The demonetised currency was originally issued (and held) without any contingencies. Now the government has announced that these liabilities will only be redeemed if the holders can meet various conditions such as satisfactorily documenting the source of the cash, and exchanging them within a given time period. These conditions were not part of the original compact when the Reserve Bank of India (RBI) issued these liabilities. This is akin to debt repudiation. Countries that have in the past defaulted on their liabilities either through demonetisation or outright debt default have had a heavy price to pay in their future ability to issue similar liabilities due to a severe loss of credibility. While India might escape the reputational hit that was taken by some of the countries that have done this in the past, one nevertheless worries about the potential loss of credibility of the RBI and the government due to the episode.

In terms of the flow creation of black money, demonetisation by itself, will do nothing. It is the tax system and getting people to pay taxes that is the issue. The RBI is now issuing Rs. 2,000 notes which, in fact, will make it marginally simpler to use cash to store undeclared income, in as much as people hold such income in cash. Forcing people to use banks and provide tax identification numbers for large transactions will help in reducing the flow creation of black money. But these measures are, and can be taken, independent of demonetisation.

The move did help in dealing with the counterfeit currency problem. It also dealt a blow to terrorist cash holdings. But these two were not that large to start with and have not been trotted out by the government as the main reasons for the move.

The long-term goals of reducing the cash dependence of the economy, encouraging more financial inclusion, and mainstreaming of electronic payment methods, are good. So is the desire to eliminate black money. But this is a sledgehammer move given the costs and disruptions. It probably does signal intent of the government which, if credible, could alter private-sector behaviour and succeed in shrinking the share of the underground economy. But that is a big unknown. All public policy must rely on a clear-headed cost-benefit analysis. This one fails that test.

Note:

1. Round-tripping refers to circular movement of capital. In the context of black money, capital leaves the country through various channels such as inflated invoices, payments to shell companies overseas, the hawala route, and so on. After cooling its heels overseas for a while, this money returns in a freshly laundered form; thus completing a round trip.