The Question About Adarsh No One Is Asking

Nikhil Service Station is one of the more popular petrol bunks in South Bangalore. If you wonder why you have never heard of it, however, it is because nobody refers to the petrol bunk by its real name. The bunk is owned by Anitha Kumaraswamy, wife of former Karnataka Chief Minister HD Kumaraswamy and daughter-in-law of former Prime Minister HD Deve Gowda. The service station came into service in the early 2000s. It had been allotted by Indian Oil in the “Kargil martyrs” quota. It is now a landmark in South Bangalore, and popularly known as “Deve Gowda Petrol Bunk”.

The reason I’m bringing up the issue of the petrol bunk is to draw a parallel with the scam-ridden Adarsh Cooperative Housing Society in Mumbai, which was again ostensibly built for “serving and retired army personnel”. The Adarsh scam is in the news once again due to the rejection of the report by the Maharashtra cabinet and the connection with arrested diplomat Devyani Khobragade.

The question about Adarsh that nobody is asking is this – why is it the government’s business to construct housing for “serving and retired army personnel”? Why is it that the government should compensate families of martyrs with petrol bunks and LPG dealerships and not cash? Aren’t these structures designed to be scammed?

Nobody argues that the army must be paid well. Nobody argues, either, that army personnel should be generously insured and compensated, given the hazardous nature of their jobs. My argument, however, is that this insurance and compensation should be universal and standardized. Allotments such as housing and LPG dealerships are discretionary by nature, and that makes them prone to abuse.

Consider for example, a housing society the government constructs for “serving and retired army personnel”. Let us say that the society has 500 apartments. How does the government choose who gets these apartments? And in what way are the 500 such chosen personnel different from those that did not get the allotment? Does this discretionary allotment not leave the system to abuse? Does this not lead to unhealthy competition among the “serving and retired army personnel”? Do we want that in our armed forces?

On a similar note, after each railway accident, we have the railway minister announcing a discretionary compensation for the dead and injured. The question is why this should be discretionary. Cannot the railways simply buy group insurance for all its passengers, which is automatically paid out upon an accident?

The argument I’m making is that some of  the processes we follow are designed to be scammed. In the time of tragedy, either in an accident or in battle, what we need is a standardized and predictable response on behalf of the government agencies. By not putting that in place, the system is prone to abuse.

ATMs and their security

Paul Volcker, former chairman of the US Federal Reserve and proponent of the Volcker Rule following the financial crisis of 2008 once remarked that the only useful financial innovation in the last twenty years is the ATM. The biggest advantage of the ATM is that because you can get money at any point of time on demand, you don’t need to keep too much of an “emergency stash” at home. For example, you are now extremely unlikely to find more than five thousand rupees in hard currency in my house at any point of time (including my wife’s and my wallets, and our “emergency stash”). In the pre-ATM era, when we would have to wait to visit a bank branch to withdraw money, we would have to keep a much larger sum at home as an emergency fund.

So how does this help the economy? Lesser cash in people’s homes and wallets means more cash in the banking system. Which means that at any given point in time, the banks have more money to lend out, and so the supply of credit is higher, reducing the cost of credit. Reduction in the cost of credit improves investment and thus leads to higher economic growth – which is good for everybody. The ATM is thus pareto-positive in stimulating economic growth.

And this is not all. The presence of the ATM has meant that one of the basic activities for which people would visit bank branches – to withdraw money – has now declined massively. Thus, it is possible for banks to run with much leaner branch infrastructure and this again pays back to the general public in the form of a lower “spread” between the cost of a deposit and the cost of a loan. I read on twitter yesterday (unable to find link now) that the average cost of servicing a customer at a teller counter is Rs. 176 while at an ATM it is Rs. 6. This order of magnitude difference is hard to ignore.

And we are not done yet, for we haven’t yet factored in the ease of drawing money now in the age of the ATM. Ten years back I remember having to wait at the bank branch at IIT for about twenty minutes to withdraw cash. I would have to fill up and submit a form, collect a token and wait till my number was called before I was handed my money. The transaction cost (for the customer) of withdrawing money was way too much. And one had to go during the branch timings. It is all so different now!

Now that we have established that ATMs have a socially and economically useful purpose, let us get to their security. On Tuesday this week in Bangalore a woman was mugged at an ATM when she had gone to withdraw money. The assaulter threatened her with a pistol and a machete, and assaulted her anyway and decamped with her money. The event was caught on the CCTV camera at the ATM and the footage was played out on national television.

http://www.timesnow.tv/India/Woman-attacked-inside-ATM-in-Bangalore/videoshow/4441948.cms

Following this incident the Home Minister of Karnataka has given a directive that banks appoint security guards at ATMs or shut down the ATMs. Initially he gave an ultimatum of three days to implement this rule, but then the impracticality of the suggestion dawned on him and the deadline has now been extended. The question, however, arises on who is responsible for safety of the ATM.

There are two components to safety at an ATM – safety of the cash and safety of the customers who visit it. The cash at the ATM is the bank’s private property, and the bank has chosen to put the cash there (and not somewhere else), so it can be argued that security of the cash inside the ATM machine is the bank’s responsibility. I don’t think there needs to be too much debate on this.

What is debatable, however, is the responsibility of security of people visiting the ATM. The question is if it is the responsibility of the bank or as a public good it is the responsibility of the government. Let us draw an analogy. Let’s say you are visiting my house, and at exactly the same time a robber happens to pay a visit. In the course of the robbery you get injured. Can the state hold me liable for your injury for not securing my house enough against the robbers? Isn’t it the state’s responsibility in the first place that the robbers were on the prowl and they just happened to rob my house when you were visiting?

Public safety is a public good. To get technical, it is non-rival (by keeping the streets safe for you, the streets are also kept safe for me) and non-excludable (having kept the streets safe, you cannot exclude me from enjoying the safe streets). And it being a public good, it is the responsibility of the state to provide it. It also means that it is the responsibility of the state to provide public safety everywhere – be it private or public places. Arguing that the ATM, since it belongs to the bank, is not a public space and hence the state is not responsibility for security there is thus wrong. So the state has no right to demand that banks employ private security guards to guard the ATMs.

So if it is the state’s responsibility to keep ATMs safe does it mean that police be appointed to guard the ATMs? Of course not, for the police’s job is not to guard private property that is the ATM – their job is to ensure public safety. Effective policing would mean that the thug who attacked the woman at the ATM wouldn’t be in business at all, and that he wouldn’t have thought of committing this crime.

So if we don’t have private security guards or cops guarding the ATMs how are we going to keep them safe? I argue that it is a matter of design. If you were to watch the video above, you will notice that the first thing the thug does on entering the ATM behind the victim is to pull down the shutters – thus the happenings of the ATM is shielded from the public eye. If ATMs are by definition perennially open what is the purpose of the shutter? You might also notice in the video that the thug pulled down the shutter once again while exiting. Consequently the victim was found only three hours later and that might have had serious consequences in terms of her health. Would the ATM not be better off without that shutter?

Then, there is the question of whether we need a room at all to house the ATM. Here in India, everywhere except in malls, ATMs have their own rooms, and it was in one such room that the mugging happened on Tuesday. On my few visits abroad, however, I’ve noticed that ATMs there never have their own rooms – they are simply holes in walls on the street from which you can get cash. That automatically puts the ATM in a public space and makes them safer (especially if they are on busy streets). The ATM rooms only provide a false sense of security and can prove counterproductive like in the case we just saw.

As we saw in the first part of this piece, ATMs perform a socially valuable function and it is in the interest of banks to encourage customers to use them. That, however, doesn’t mean that banks appoint guards to all ATMs – there might be an alternate solution that might be cheaper and easier to implement, and it is for the banks to find it. It is NOT the state’s business to mandate how the banks get customers to use their ATMs – the state has to concentrate on maintaining public safety.

In June last year the Reserve Bank of India allowed non-bank entities to run “white label ATMs” – cash dispensing machines that are not affiliated to any banks. The first such ATM came up earlier this year. I’m hopeful that some of these ATM companies will gain enough scale that they can solve the ATM design issue and make them safer and more customer friendly.

Pricing fines for ticketless travel

In large mass transit systems such as those in Mumbai (or even Chennai), ticket checking turnstiles can significantly slow the flow of human traffic. The sheer number of passengers that use these transit systems daily makes it impossible to check the ticket of each and every traveler. Hence, the Railways, rather than checking the tickets of every passenger, instead relies on random checks. During these random ticket checking efforts, people traveling without a ticket are asked to pay a fine. This, the Railways hope, will be deterrent enough for people to purchase tickets before travel.

However, rather than ensuring deterrence, what this system has resulted is in an informal “ticketless travel insurance” economy. The concept is simple – rather than buying a ticket from the official ticket counter, you instead buy protection from an “informal insurance provider”. For a nominal “premium” (believed to be in the range of Rs. 100 per month) these providers insure you against ticketless travel. In other words, in case you get caught by the ticket checkers during the course of your “insurance”, these “insurance providers” step in to pay your fine! Check out this article in The Hindu about how these insurance providers work (WARNING: The link isn’t working too well for me, and is taking me to a third party site a few seconds after loading The Hindu page, so be careful before clicking through).

The very existence of this market, however, implies that fines for ticket less travel are not being priced properly. The math is fairly simple: if the price of the ticket is p and the probability of your ticket getting checked is frac {1}{N} , then the fine for ticketless travel should be strictly greater than Np. If not, it works out cheaper for your to pay the fine each time you are caught rather than buying the ticket.

So what role is being played by these “informal insurance companies”? Risk management! People don’t like risk. While on an average your ticket might be checked only once in 30 days (number pulled out of thin air), there is no reason that you will not be pulled up for ticket less travel multiple times in a month. By outsourcing the risk to a central party who pools the risks (from several commuters), you have a steady cash out flow and are hedged against getting caught multiple times (you might get caught but your insurer pays the fine). In fact, this is how insurance works in other sectors also.

What should the Indian Railways do to drive these “informal insurance companies” out of business? Currently, if the fine is S, S le Np. From this equation, you can see that the Railways can do one of three things so that this inequality gets reversed – the price of a ticket can be reduced – but that would be equivalent to cutting off the nose to spite the face, for it would have significant adverse impact on the railways’ revenues. Next, N can be reduced, or in other words the frequency of surveillance be increased. This, too, is not easily implementable since the Railways will have to invest in additional resources to check tickets. The last option is to increase S, and there is nothing that prevents the railways from doing this!

How will this work, though? By raising the cost of fines for ticketless travel while keeping the frequency of ticket checking constant, the “premium” a commuter will have to pay to these insurance companies will increase. If the fine amount is increased to a certain level, the premium a commuter will have to pay to buy ticketless travel insurance will exceed the price of buying tickets! And the insurance market will implode.

While this seems like a simple solution in theory, I’m not confident of it being implemented any time soon. Who knows – one might have to go to the Union cabinet to increase the level of fines in local trains. That’s how our railways is structured.

Why Cash Transfers Should Not Replace Midday Meals in Schools

Admin Note: This is not a typical RQ post, in that this has no numbers. Yet, since this is policy related I think it makes sense to put it here

I’m normally a big fan of cash transfers. I’m glad that the Indian government has started implementing it for things like fuel subsidies and certain other benefits. After all, by simply providing the subsidy in cash (market price minus intended “subsidized price”) the government achieves the subsidy while not really having to bother about managing the supply chain. I would have been less unhappy with the Food Security Bill had it been designed as a cash transfer scheme, rather than giving further responsibility to the much-maligned Public Distribution System. With the midday meal system in schools, though, I make an exception.

Following the tragedy in the Bihar school this week, people have called for the government to scrap the midday meal scheme in schools and provide students a cash subsidy instead. Some people have argued against it quoting economies of scale (for example, ISKCON, under its Akshaya Patra scheme, provides midday meals to children in Bangalore schools at the cost of Rs. 6 per child per day, and that amount cannot but much food in the market). That aside, there is a fundamental economic argument against providing for children’s midday meal in the form of cash.

Every year, during Christmas time, journalist Tim Harford (of Financial Times, BBC Radio 4, etc.) writes an article that states that gifting induces a net weight loss, and the economically ideal way of gifting is to gift cash. For example, if I give you Rs. 100 in cash, you can do whatever you want with that cash. Instead, let us say that I use the Rs. 100 to buy you a gift (let’s say a pen). Now, irrespective of how much value you see in the pen, you don’t have the option any more of spending that Rs. 100 on anything except the pen I’ve got you. So you are in effect poorer than you would be had I simply given you the Rs. 100 rather than buying you the pen.

The question now is whether I want you to buy more pens or less. If for some reason I believe that buying more pens is good for your health, I can do my bit in encouraging that behaviour by gifting you pens rather than gifting cash. If on the other hand I don’t have a view on whether pens are good for you, I will gift you cash.

Government subsidies work the same way. If the government wants to encourage consumption of a particular good or service, it subsidizes it directly. If, on the other hand, the government doesn’t have a particularly strong view on whether a citizen should consume more or less of a particular good, but only wants the citizen to be able to afford that particular good, it provides cash. With the current form of the food security bill (where the government has promised to give foodgrain at subsidized prices) the government is implicitly stating that it wants to encourage people to eat more foodgrain (which flies in the face of data which shows that most Indians already eat too much cereal and too little of other nutritious foods). If the intention were only to ensure that people can afford food grains, a cash transfer would have sufficed. Similarly, by moving to a cash transfer scheme for cooking fuel, the government has signaled that it doesn’t particularly encourage the use of cooking fuel, but it simply wants to make it affordable for whoever wants to consume it (without distorting markets).

While the stated aim of many states in implementing the free mid-day meal in schools is to encourage attendance, there is a more fundamental reason to it. It is in the country’s interest to ensure good health and nutrition of children, in order to enhance their possible contribution to the economy when they grow up (studies have shown that malnutrition and poor health leads to lower educational attainment which leads to lower capacity to contribute to the economy). In this light, it is in the country’s direct interests that children are well fed, and the school is a location where children gather and can be fed (this is where the economies of scale bit comes in). That it encourages attendance is only a positive externality.

Now that it has been established that it is in the country’s interest for the child to be well fed, and that the midday meal in school is a good opportunity to thus feed the child, this presents a classic case for giving a “direct subsidy” rather than a cash transfer. That meals served within school premises are not tradable goods and hence won’t distort markets is a bonus.