In 1978, when the government demonetised high denomination notes (Rs.1000, Rs.5000 and Rs.10000), they were really high denomination notes: Rs.1000 in 1978 is the same as Rs.18000 today. They were a minuscule portion of currency in circulation, and only a handful of people held those notes. In contrast, today, Rs.500 and Rs.1000 are small bills. About 86 percent of currency in circulation is in these denominations. The action has consequences for almost all citizens who have cash in their wallets. How should public policy analysis think about this action? It's useful to break this up into three components:

 

 

In the use of the coercive powers of the State, a sound policy process starts with a good ex-ante analysis, using all the facts and logic available at the time, and then ensures proper execution. The policy process then looks back and introspects, and wonders if things could have been done better. All of us as citizens are part of this three-part thought process.

 

Ex-ante analysis of costs and benefits

 

One way to analyse a policy proposal before it is implemented is to compare its expected benefits with its expected costs. The process of doing this thinking often throws up many good ideas for how to improve an initiative, by discovering ways to achieve the same objective at a lower cost.

 

Expected Benefits

 

To consider the expected benefits of the cash ban, let us begin with the government's stated objectives. The decision is supposed to be an attack on three problems:

 

1. Fake currency: the objective is to eliminate fake currency. This is achievable by this decision, as the notes are to be deposited in bank branches and post offices, and as long as these establishments run a check on those notes being deposited, this objective will be achieved. Losses will be imposed upon those unwittingly holding fake currency. The factories making the fake 500/1000 rupee notes will close down. However, (a) The pressure of work at bank branches and post offices may mean that many bad notes slip through and (b) There is no magic bullet in the fight against counterfeiters, who will now rejig their factories to counterfeit the new designs of notes. It is being reported that the new designs do not have any additional security features, in which case the counterfeiters would soon be back in business.

 

2. Black money: It is useful to distinguish between black money creation, and black money storage. Black money may be created through illegal activities (eg. bribery), or through legitimate activities (eg. trading business) where all taxes were not paid. Black money is accumulated over a period of time and stored into a range of financial instruments. We should understand the likely impact on this unaccounted wealth and on future black money creation:

 

Impact on unaccounted wealth: Unaccounted wealth may be stored in many ways: rupee cash, gold, real estate, foreign currency cash, accounts in foreign countries, and other instruments. Of these, rupee cash is relatively unattractive as it earns a negative real rate of return. A 2012 study by an institute of the Ministry of Finance found that cash is the least preferred instrument for storing unaccounted wealth. Reports from raids by enforcement authorities also suggest that cash is only a small part of unaccounted wealth they find; the instruments of choice are benami properties, gold, diamonds, shares, etc. Moreover, part of unaccounted wealth is laundered, and becomes more difficult to identify. As with the white economy, only a small fraction of the total unaccounted wealth is kept as cash.

 

The November 8 decision will not result in most of these persons being caught by law enforcement agencies, as they have been forewarned: everyone knows that deposits in banks of above Rs.250,000 will result in an investigation. However, persons holding unaccounted cash may have to quietly and privately bear losses. The extent of these losses depends on the rate charged by the market for laundering cash. Here is an example. Person X has Rs. 100 crore in unaccounted wealth, of which Rs. 10 crore is in cash form. Even if person X loses all the cash, much of the unaccounted wealth would still be with him. Since utility of additional wealth diminishes with increasing wealth, the net impact on his well-being would be small. However, he may sell the cash to person Y for Rs. 6 crore in new notes (to be delivered immediately or later). The person selling has taken a 40 percent loss, but still gets to keep 60 percent of the cash, and more importantly, continues to own the other instruments of unaccounted wealth. The person purchasing the cash would now launder the money through a network of persons willing to deposit it in their accounts (relatively safe for small deposits) or exchange it in banks (up to Rs. 4000) for a commission. In this example, black money only changes hands from one corrupt person to another corrupt person, but is not destroyed. It is a test of the character of many people in our society: whether they will fall for the get-rich-quick temptation to help launder some money in exchange for a cut.

 

Some analysts, including the government's Chief Economic Advisor, have argued that this decision would lead to a transfer from black money holders to the central bank and to the government. This assertion raises important questions about what a central bank and government can and should do. If some cash does not return at the end of March, 2017, should the RBI simply decide to reduce its liabilities by that amount, and create a windfall profit? This seems very draconian, and makes many simplistic assumptions about the way the world works. The RBI Governor's promise on the note reads, "I promise to pay the bearer --- rupees." It does not add conditions on this promise. It is alarming to read about people in government and outside talking about confiscating private wealth so brazenly. If government and central bank just give up their liability with a 4.5 month notice, it might erode faith in the currency. It would be unwise to do so. RBI should carry the liabilities on its balance sheet for the foreseeable future, and keep limited windows open for people to return their hard-earned cash. They cannot announce this in advance, but in March-end, they should consider modifying the notification to this effect.

 

Impact on black money creation: Is this decision likely to have lasting impact on the working of the black economy? That depends on the deterrence it creates in the minds of participants. Deterrence is created by one's sense of the probability of getting caught and punished. Since this decision is not likely to ensure large-scale identification, prosecution and punishment of those creating black money, it may not have a significant effect on future black money creation. One may argue that this decision creates a deterrence by limiting the opportunities for black money storage. Others may argue that this decision would force people holding black money to move to other instruments that may be easier traced. Neither of these arguments hold much water. Since the decision is expensive (discussed later), the newly issued Rs. 2000 and reintroduced Rs. 500 notes would, for the foreseeable future, serve as reliable stores of unaccounted wealth. Transactions in the black economy may move to other means, such as USD cash, gold or bitcoin, but these transactions will not go away.

 

3. Terrorism and criminality: The government has pointed to the use of fake currencies by terrorists. Other commentators have highlighted the use of cash by criminals. For a short time, we may see some drop in terrorism and criminality, to the extent that terrorists and criminal organisations are organised around transactions in cash. However, high denomination notes will be back, and criminals will also change their tactics. We have not created new deterrence against terror and crime.

 

To summarise, the major direct benefits of this decision are likely to be: eliminating fake currency, and inflicting one-time losses upon black money stored as cash. We will ignore indirect benefits, such as the increased use of digital money, as there are other, vastly less disruptive ways to pursue those objectives.

 

Expected costs

 

We now turn to toting up the expected costs to society, where three key elements appear:

 

1. Costs of low economic activity during transition period: As the means of exchange, money is the lubricant of economic activity. A sudden monetary shock disrupts the smooth working of the economy. 

 

Barring big cities, the adoption of electronic payment instruments is slow, and the infrastructure is weak. We have been seeing high growth rates of electronic activities, but in absolute terms, this is still very small. A 2013 survey of 3066 households in rural and urban India found that only 3.6 percent of respondents had ever done a non-cash transaction. In June 2016, there were about 75 crore debit cards, but only 12.9 crore payment transactions, which suggests that most people who are issued a debit card never use it for a purchase. Given the lack of financial development, most transactions in India use cash. The action of the government has given an 86% decline in cash in circulation. Using a broader money supply measure (currency plus demand deposits), this is a 55% shock to money supply.

 

As a consequence, in the next few weeks, economic activity is likely to be sluggish, as the main means of exchange would be unavailable. There are many reports on sluggish activity in factor markets and product markets. In informal labour markets, daily labourers are not able to get work. There is a decline in registration of land and other property deals. Many product markets that rely on full or partial cash payment are affected. Anecdotally, we are hearing of a 50 percent decline in sales in many establishments in the last five days. Most of us have heard stories of people not able to pay for weddings, local travel, food, etc. We have also heard about tourists stuck in various destinations with cash they cannot use. Each day this situation persists, the costs will increase. The time spent by individuals in dealing with the crisis of cash is time that could have been used to produce GDP or to obtain utility.

 

These costs will show up in the form of lower GDP growth for the affected period. It is important to note that GDP counts all output, whether the production is done with tax evasion or not. The entire impact should be considered as a cost. Money is not black unless tax has already been evaded. So, even losses imposed on black money holders due to their legitimate economic activity must be included as costs and not as benefits. At the time of production, one cannot say whether tax will be evaded on the activity.

 

2. Cost of printing and delivering cash: About 22 billion printed notes would be destroyed. These would be partially replaced with new Rs.2000 notes, the redesigned Rs.500 notes, more of the old Rs.100 notes, and digital money. The process of collecting and destroying old notes, and issuing new notes, involves costs:

 

 

3. Cost of harassment of honest taxpayers: Government announcements seem to indicate that they will use leads from large cash deposits (greater than Rs. 2,50,000) in banks to conduct inquiries and investigations. Many legitimate cash-intensive businesses would be caught in the net, and may now face months and years of harassment. Since cash is neither black nor white, only thorough investigation can reveal its genealogy. This means there may be many fruitless investigations leading to harassment of individuals and businesses, and avoidable costs to exchequer.

 

4. Other costs: We are hearing news about increased criminality in many places, as some petty traders and shops are offering to accept the old notes at discounted rates. It is too early to say whether this will have a lasting impact as increased criminality. Second, the decision has placed a large amount of cash in transit. This exposes many people to petty crime, reports of which have been trickling in. In desperation, ordinary people are getting duped in many ways. Third, there are reports of health emergencies and other problems that are getting exacerbated because of lack of money. Fourth, there are social practices that have been disrupted by the decision. For example, some people keep stash of savings hidden from their family members. Now, they have to reveal these savings to other family members, leading to emotional and social costs. Finally, there may be a decline in faith in the Indian currency, and we may see people use other store of value and medium of exchange.

 

To summarise, we should analyse:

 

 

Comparing benefits with costs

 

A simple cost-benefit analysis would merely quantify all costs and benefits and look at which is larger. But in this situation, there is a crucial factor to be considered - how a rupee of loss inflicted upon those with black money should be compared with a rupee of cost imposed on those with no black money. If Rs. 10 crore is loss incurred by those with black money, and Rs. 10 crore is the aggregate loss incurred by others, should they be treated as same? I think that would be unwise, and it is important to assign different weights to benefits and costs. Here is one way to assign weights to costs and benefits. Suppose someone comes to you and makes a proposition:

 

Assuming that your giving up some money will lead to a corrupt person losing some money, how many rupees would you be willing to lose to inflict a loss of hundred rupees upon an unknown corrupt persons?

 

Different individuals may offer different values based on factors such as experience of corruption; opinion about retributive justice; belief in reform by punishment; etc. This trade-off that people are willing to take tells us something about the weights that should be assigned to nominal values of costs and benefits of the decision.

 

In order to discover numerical values, I obtained 12 data points by posing the above proposition to 12 persons. The answers ranged from Rs. 0 to Rs. 10, and the average was Rs.2.46. The average respondent was willing to pay Rs.2.46 to inflict Rs. 100 of loss on unknown corrupt persons.

 

This is hardly a precise estimate. The sample was not random (though I did try to maximise the diversity within the 12 persons). Since the respondents knew me, some might be signalling virtue or posturing. We should view this as a very crude estimate. A question using `one hundred rupees' as the hypothetical loss just captures one point. My sense is that the utility function is logarithmic, with people willing to pay progressively lower amounts for each additional rupee of loss inflicted on the corrupt. For example, a person willing to pay Re. 1 to inflict Rs. 100 of loss on the corrupt may not be willing to pay Rs. 1 crore to inflict a loss of Rs. 100 crore. As the loss inflicted increases, the ratio between the cost one is willing to pay and the loss inflicted would increase, and at some point, a person would not be willing to pay any further amount.

 

It is also important to consider distributional consequences. Most of of the economic lives of the poor are cash-based. Daily wage earners are usually paid in cash and they spend in cash. At the end of the month, they may have no savings, but they do vigorous cash transactions during the month. They will be hard hit: each rupee has high utility value at low income levels; access to bank accounts is low; transport cost and opportunity cost of time is a larger percentage of the cash to be exchanged; and so on. Every rupee of cost imposed upon the poor should get a higher weight than that imposed on the non-poor.

 

Now, an obvious question is: should citizens' "willingness to pay" be the determinant of what government can ask them to pay? For instance, this is not the basis for setting tax rates. However, I would argue that this method of assigning weights by asking people their willingness to pay for loss inflicted on black money holders is quite relevant here. The cost of this decision is to be incurred over and above the direct and indirect taxes we are paying. Those taxes are meant for, among other things, catching corrupt people and inflicting losses upon them.

 

Without complicating this unnecessarily, we should know that people can be expected to incur only a small percentage of the losses inflicted upon those holding black money. Taking this into account, and making allowances for an elected government's power to take such decision, I think the cost that honest people can be expected to incur should be weighted up by at least ten times before comparing them with the benefits. So, if the losses inflicted upon those with black money are Rs.2 trillion, the costs for others should not be more than Rs.0.2 trillion. It seems probable that costs of this decision will outweigh its benefits, or will be quite close. What this means is that, the probability of costs outweighing the benefits or being close to the benefits may be larger than the probability of benefits outweighing the costs.

 

Many people seem to be proudly accepting the costs of the government's decision. They hold the assumption that their sacrifices are leading to huge losses for the corrupt. This makes it all the more important to analyse and report the full benefits and costs of the decision, so that people can understand the consequences of what is happening.

 

Concurrent assessment of implementation

 

The cost/benefit analysis is influenced by implementation. If there are a hundred ways of implementing the decision, each would induce a different set of realised benefits and costs. Government has a duty to ensure that a decision is implemented in the best possible way using all the levers available to government, which maximises benefits and minimises costs. Consider just three sets of variables:

 

Timing of the decision: The decision was announced on November 8th, with almost immediate effect (from midnight). The timing is a variable in government's control. No timing is perfect, but this timing seems flawed. First, this is one of the peak wedding seasons of the year, and weddings being cash-intensive events, the ban is causing much misery. Unless there is clear evidence to show that most of the money spent on weddings is black money, this timing is somewhat cruel. Second, one of the ways to launder money is to report it as agriculture income, which being non-taxable is all white money, unless an investigation finds over-reporting. Kharif crops were harvested in October. On the one hand, this opens possibilities for laundering, and on the other hand, it imposes losses on farmers looking to sell their produce. Agriculture wholesale markets use cash and may be unduly affected. Third, this being the festive season, many legitimate businesses are also flush with cash, which means a higher cost imposed on such businesses. If objective was to target those with large accumulated stock of cash, it could have been met even in lean season. Finally, 8 November was the date when there was a possibility that Donald Trump would become President of the US, which is a new set of risks on international scale.

 

Limits on exchange and withdrawal: The government has placed certain limits on exchange of old notes and withdrawal of new high denomination notes and old low denomination notes. These limits are the same for everybody - individuals and businesses. Did these limits need to be so low? Could the government have allowed higher limits to reduce the costs for those not holding black money? For example, if the exchange limit was Rs.10,000 and the withdrawal limit from ATMs was Rs.5000, it may have helped most households with cash required for their day-to-day activities, while still being too low for hoarders of black money. The limits of Rs.4000 and Rs.2000, respectively, have increased the costs for regular households and businesses. Moreover, government should have allowed higher limits for registered businesses on transactions in their current account, to minimise costs to businesses. Some businesses with unusually high transaction volumes could be investigated later, but it would have protected legitimate businesses.

 

Readiness of the banking and postal system: The entire banking and postal system was caught unawares. Government says that it will now take three more weeks to configure ATMs. One can understand that configuring ATMs for the new Rs.500 notes before the decision would have run the risk of giving out information that could be used to predict the government's decision. Perhaps the government could have ensured the ATMs are configured in advance for the newly issued Rs.2000 notes, which were in the works for long, and also ensured greater supply of Rs.100 notes. The situation is so bad that in small towns, most ATMs are still not dispensing cash, and branches easily run out of cash.

 

These are factors that are substantially in government's control, and it seems that the planning ahead of the event can be faulted.

 

Considering alternatives

 

Let us go back to the discussion on benefits and objectives. First, consider the objective of eliminating fake currency. Governments everywhere grapple with counterfeiters, and there is a standard armoury of stratagems for dealing with them. Banning notes overnight is not the usual way to fight fake currency. If eliminating fake currency is the main objective, government could have announced a time period (say, 3 months) to exchange old notes for new notes. This would have eliminated fake currency, and avoided most of the costs of an overnight ban. The overnight ban decision cannot be justified by the objective of eliminating fake currency.

 

On black money, as discussed earlier, the main objective that this decision can reasonably achieve is to inflict losses upon those with black money stored in cash rupees. However, if we do want to solve the problem of black money, it is more important to plug avenues of tax evasion, and to go after more preferred instruments for storing unaccounted wealth (eg. real estate). All said and done, based on what we know, this decision is likely to have only a small impact on those holding black money, and almost no impact on the creation of black money.

 

Should governments run such experiments: Concerns about high modernism

 

Governments do not take the Hippocratic oath. They can inflict harm. However, when they do so, we as citizens have a right and a duty to question them. Large actions which impose more harm should elicit more questions. The government has suddenly decided to squeeze out 86 percent of currency in circulation and 55 percent of liquid money. Costs are being imposed on innocent people, and those with black money might lose just a portion of their unaccounted wealth. Is this a good bargain in economic terms? In my view, the costs may outweigh the benefits, especially if you assign proper relative weights to benefits and costs.

 

When State capacity is limited, we should be mindful of the load that we place upon public administration. The load, of replacing 86% of currency notes with new ones, seems to be a case of premature load bearing, and has given an organisational rout. It is better to cross the river by feeling the stones, with a large number of small moves within an overall strategy.

 

There are so many known unknowns, and unknown unknowns, that we cannot accurately predict the outcomes, nor can we ensure sound implementation amidst such large scale disruption. It is in the nature of governments to process information that is spread out across the economy, slowly and imperfectly. By the time feedback comes in and leads to change in policy, it is too late. Even the best public policy process will be limited by availability of information and estimation methods. This does not mean that governments shouldn't do anything. It only means that we should internalise these constraints and be modest in our aspirations for government. Complex plans of a certain nature should not be made. This is especially true in a complex, multi-layered society like India. 

 

Governments should build on what they know from experience, and then hand the baton to the next incumbent. Trying to solve the entire problem through shock therapies is too risky. It may have huge unintended consequences. Consider an example: cash that does not come back by March-end. Government will claim that all of it was black money. We cannot assume that. In a country where citizens are generally wary of state, and steer clear of it, even those people who have, over a period of time, tucked away a lakh in cash for rainy day or a future plan might eat some losses due to fear of penal action by the State. Some people living abroad may not consider it worth their while to come back or to get to Indian bank branches in their country to exchange a few thousand rupees they may have. This is an example of known unknowns. The unknown unknowns are the unpredictable conspiracy theories and swindle schemes that might flourish on the fertile soil of ongoing chaos.

 

Nuclear options should be used only when other options have been tried. Experiments of this scale can be run if there are no alternatives. India still has a long way to go in strengthening the administrative structures that allow and sometimes abet black money creation. Since most of the unaccounted wealth is likely to be with those creating black money, strengthening these structures would not just plug new black money creation, it would also help confiscate unaccounted wealth. To address the problem of black money, we must focus on compliance management to plug creation of black money, and improved enforcement to crack down on unaccounted wealth. Draconian, disruptive decisions seem to admit that reform of the basic institutions of administration has failed. Has the government done all it could on those fronts? For instance, have the recommendations of Tax Administration Reforms Commission been considered and implemented? On black money, as on other issues, the best thing that a government in India can do is to first get the basics right.

 

Politics of black money

 

Why did the government take such a drastic step? To understand the answer, we must consider the context in which this decision is being implemented.

 

In India, there is this mass hysteria about black money. This has been created by a combination of factors: high ticket size corruption scandals; release of information showing nexus between big business and politics; widespread petty corruption; anti-corruption social movements; political mobilisation around the issue of corruption; and so on. The electorate has signalled that this is an issue that they care about. This is the context of this decision. Some have accused the government of taking the decision because of electoral considerations, while supporters of the government say this is a genuine attempt to solve a big problem. Both sides agree on the scale and seriousness of the problem, even while they disagree on intentions and means. One cannot help but wonder whether the issue has been blown out of proportion.

 

Most of the failures of the Indian State are being blamed on corruption. Capacity is being confused with character. It serves as a simple explanation for all the ills of our governance system. The reality, however, is much more complex.

 

Estimates of black money in India compare favourably with similar developing countries. In 2010, World Bank had estimated "Shadow Economies" of 162 countries. The weighted average size of the shadow economy of these countries in 2007 was 31 per cent of the official GDP, down from 34 per cent in 1999. For India, it had come down from 23.2 percent to 20.7 per cent. Some would argue that it may have increased since then, but it would still comparable favourably with most comparable countries. Just because we are not doing very badly compared to other countries does not mean that we should be complacent, but this should give us cause to reflect on how big a problem this is, and how we should solve it. This is not an emergency situation, but people perceive it to be so. Where do we go from here?

 

Black money creation is a problem, although perhaps not as big as it is believed to be. A bigger problem is how India can deliver high GDP growth for at least two decades, before we get into a demographic nightmare scenario. Solving both these problems - black money and sustained high growth - will require considerable improvement in the capacity of the state to perform its basic functions, which includes asserting itself on residents to take its due. If our government thinks that the black money issue is a major justification for doing difficult things, it should use it to take tough decisions that enhance administrative and regulatory capacity. It could be used to justify reforms of tax administration, government expenditure, investigation agencies, regulatory agencies, and even courts (under supervision of Supreme Court). Getting these basics right would serve us better than a large scale disruption that comes at a potentially big economic cost.

 

These themes are found in many fashionable problems, such as black money, corruption or inequality: they seem to get blown out of proportion and simplistic immediate solutions are then demanded. Statesmanship lies not in pandering to public opinion, but in reshaping it and channeling it towards worthy objectives that strengthen the foundations of our Republic.

 

Conclusion

 

The biggest problem in India is the lack of State capacity. State capacity lies in systematic, professional thinking. This article shows the kind of policy thinking that would have been useful in analysing the question of de-notifying the 500/1000 rupee notes. These decisions have far reaching consequences, and it would behoove us to subject them to thorough intellectual analysis, ahead of time. As we are seeing, no amount of frenetic field activity can overcome a bad policy.