‘It’s Not Enough To Build Toilets’

His surname means a girl-child in a number of Indian languages. When asked about his family name, he clarifies that it is a version of Ghori or those who came from the Ghurid kingdom that spanned parts of Central Asia and the Indian subcontinent in the 13th century. Varun Gauri, co-director of the  World Development Report 2015 of the World Bank, obtained his doctorate in public policy from Princeton University in 1996 after which he joined the World Bank group. He has worked and conducted research in different parts of the world, including in Chile and in Jaipur, India. He spoke to Paranjoy Guha Thakurta during a recent visit to New Delhi. Here are edited excerpts from an interview that lasted nearly an hour.   

 

 

Economics is a profession which has been evolving over the past 20-30 years. Some economists, in order to understand individual decision-making, looked at human beings almost as if they were supercomputers. They were supposed to be able to assimilate all available information, calculate the consequences of incentives for their own maximum utility, make decisions self-interestedly and use all explicit information. But we have learnt over the past several decades that real human beings are not like that. In some sense we knew that all along. but the contributions of cognitive psychology, social psychology, sociology, anthropology and other disciplines over the past few decades have shown that people are not only 'irrational' but they are 'predictably irrational' -- to borrow the title of a book by the behavioural scientist Daniel Ariely. This means that people depart from standard economic predictions  in systematic ways and once we become cognizant of those departures, we can design development policies differently  so as to  help people make choices in their own interest. So, economics has changed, development economics is changing and it was time to recognize that.

 

 

The idea that there are social, psychological and cultural influences on decision-making has very deep roots in economics. Adam Smith himself was very aware of moral, psychological and social influences. He wrote a famous book, Theory of Moral Sentiments, alongside his more famous Wealth of Nations. Many other economists were quite aware of this throughout the 19th century. When economics became more mathematical it tended towards maximising functions and some of these nuances were lost. 

 

At the same time, there was a view -- perhaps best expressed by Milton Friedman in the 1950s -- that we can model individual decision-making as if people are rational and it should be maximising, even though it may not be. This view has very powerful predictive power and in markets, there are opportunities to take advantage of people who are irrational. The businesses of irrational people will close down eventually. Friedman wrote that a billiards player doesn't really know Newton's laws of mechanics but is still able to execute a billiard shot with great skill. Similarly, a businessman didn't know all the first and second order conditions of profit maximization but acted on that basis anyway. What we know now is that the facts are somewhat more complex. People who are making decisions are subjected to systematic biases. This is especially true for individuals who are living in poverty. That really is the focus of this report. If a businessman doesn't maximise profits and goes out of business, we may think that's simply the result of the choices he made to live in a market economy. But if a poor person goes out of business, that's not a happy story.

 

Many of the key decisions that the report focuses on -- decisions to send one's children to school, take medicines, to make healthy choices, to save enough for health emergencies or for old age -- are in some sense one-shot opportunities or at least, present limited opportunities for an individual to learn. There are few repetitions. If a businessman fails, he faces competitive pressure and can respond. If not, someone else may take his place or he can go into a new business. But if a poor person doesn't take his medicine that has potentially life-altering consequences, there is no opportunity for him to learn. So these are not full market decisions. That's the first point. The second point is that even in large markets some of these systematic biases persist. There was a striking article in the American Economic Review looking at the resale value of used cars in the US. It turns out the car with a speedometer reading of 80,100 miles is systematically and significantly cheaper than a car with 79,900 miles.

 

 

This is applicable for (mileage readings) of 60,000 and 50,000 as well. Psychologists will call this 'left digit bias'. You simply focus on the left digit when estimating values. It's a shortcut. It often suits us very well but the key finding from cognitive psychology is that we use a lot of shortcuts in making calculations. These shortcuts serve us in many respects but also can mislead us. These are systematic biases.  If I ask you, for instance, to take a four digit number and add three to each of the digits: 4612  will be 7945. I ask you to do this five or six times every couple of seconds measured by a metronome, your heartbeat will increase by seven beats a minute and your pupil will dilate 50 per cent and your blood pressure will increase. Thinking actually costs biological energy. So people use shortcuts... and if we all have the same biases and our brains are all similar, these biases will persist in markets. 

 

 

Well, it's really quite common for people in developing countries to find that signing up for government programmes is overwhelming. The problems are exacerbated by the fact that for a poor individual, poverty is in itself a kind of cognitive tax.  We often think of poverty as shortage of resources. But it is also true that when you are poor, your mind is taxed by thinking about imagining yourself selling goods on the streets, about your livelihood and where your next meal will come from. You ask: where do I get my supplies, what if someone harasses me, what about my kids?

 

You and I are fortunate enough to have good jobs. We get a pay cheque and we put it in our bank account. When we save, we have automatic transfers for various purposes to our savings account or brokerage account. A poor person has to figure out how much to save, how they save, where they save. All of this takes up cognitive bandwidth. There is an interesting study by Sendhil Mullainathan at Harvard University and Eldar Shafir and his colleagues at Princeton University, that looks at sugarcane farmers in India. They studied the same farmer before and after harvest when he had fewer financial stresses and found that his IQ (intelligence quotient) level drops ten points during a cash-strapped period. If I give you an IQ test, then keep you up all night and give you the same test tomorrow, your IQ level would be 12-13 points lower. 

 

 

This is the effect of poverty on mental bandwidth and fluid intelligence and hence, the need to support decision-making. But there are a lot of government programmes where everything is in place -- the administration is there, the district collector has set it up -- but people don't take advantage of the programme. This is partly because the costs of signing-up are high because the process is very difficult.

 

 

It's hard to save if you don't know how much to save and when you have a lot of pressure on you to take care of family and kin. It's easy to get distracted. This may be called an intention-action divide. Now one thing we can do to help people is by creating mental accounts. I'll give you one example from Kenya and another from India.

 

In Kenya, individuals were encouraged to save for health emergencies. People were assigned to two pools in a random manner. In one pool, they were simply encouraged to save while those in another pool, were given a safety lock box and a key and were asked to put some money in the box. They were given a passbook with the name of what they were saving for, say, an insecticide-treated mosquito net to prevent malaria. Individuals in the second pool saved 70 per cent more over the course of a year than those in the first pool. It wasn't a real lock box because they themselves had the key. But having a reminder, what economists call a mental account, helped. It's almost like a piggy bank -- a visual, physical reminder to help you to save.

 

We can all learn from how we talk to our children. 

 

In India, a group of individuals working on an infrastructure project were encouraged to save. Originally they were earning Rs 575 per week and they were saving about 0.75 per cent of the amount or Rs 5-6 per week , not a substantial amount. A financial planner came in and looked at their accounts and realised that they could actually save Rs 40 per week and they were encouraged to do so. One group was given their savings in an envelope. Some were given Rs 40 per week and some were given Rs 80. Those in another group were given the same amount of money in two separate envelopes, one of which was marked. It was found that those who were given money in two envelopes saved substantially more than those who were given the same amount in one envelope. Why? When you have two envelopes, you actually experience more guilt in opening the second envelope. It's another kind of mental accounting. So there are number of processes like this one which can help people fulfil their own intentions to save more.

 

 

Let me give you some quick examples which are again on the intention-action divide. Data entry clerks in Mysore were paid in this experiment on the basis of the number of correct data strings they enter in the computer. It was observed that although they were paid for each correct data entry, their productivity increased 8 per cent on pay day or just before pay day and 15 per cent just before festivals. This was not optimal. One should have had steadily high productivity. They were then given an option of committing themselves to higher productivity at a target that they themselves set or suffer a penalty of lower pay for each incorrect data string entered. One-third of them accepted this contract and their productivity increased 6 per cent. To get that increase through an increase in wages, one has to increase wages by 18 per cent. In other words, it was a very cost effective intervention by getting people commit to their own higher targets. Another example involves payment of bonuses. Sometimes, if you give a bonus to people, they can actually work more. But a bonus is particularly powerful when it is unexpected, when it is seen as a gift, rather than something that people are entitled to. 

 

Let me also give an example from Zambia where hair-dressers were selling female condoms to fight HIV/AIDS. They were divided into groups with different rates of commission, ranging from 10 per cent to 90 per cent. One group of hairdressers were given only stars and given recognition in a public ceremony. Condom sales almost doubled in this group. This is not to say that financial rewards had no impact, but that star treatment was particularly effective. The reason it worked in this instance was because the commission for condom sales was a small share of the total income of the hairdressers while social recognition was very powerful and effective. 

 

 

The best example in this context is Swachh Bharat. The government here is embarking on a very significant programme to improve sanitation throughout the country. There will be a substantial effort to increase the construction of toilets. The World Bank may be supporting this endeavour. However, it is not enough to build latrines and toilets. You have to get people to use them. We have to change behaviour. How do you do that? It's not obvious. One set of programmes that use psychological and sociological insights has shown promise in getting people to change their practices. In a community programme like total sanitation, there is a focus on open defecation. When individuals in villages see flies flying back and forth between excreta and food, that makes a big impact. Sometimes, communities come together to change their practices. Once they make such promises, peer pressure is exerted.

 

 

I think the recommendation is that Swachh Bharat is a behaviour-change programme first and foremost. There is no magic bullet, as you rightly pointed out. It's a very challenging problem when you have a large number of people changing lifelong habits, it's difficult and it's embedded in the social system which complicates things. What we need to do is experiment, try things out, measure the impact and measure behaviour changes and then adjust the programme going forward.

 

One idea is to try and create a tipping point, a kind of a cascade. This is how social norms change sometimes. Let me give you an example which involves an anti-drinking campaign in the campus of the Princeton University in the US. I know it seems as if this is a different topic but on the Princeton campus, students drank a lot of alcohol because they felt others were also drinking a lot of alcohol. They typically overestimated the amount of alcohol that others were drinking. When randomly selected students were told that others were in fact drinking less than what they thought, their own drinking behaviour changed and came down. They were trying to fit in. 

 

In the context of the sanitation programme in India, if we can make it clear to people that others are changing their behaviour and that they too should join the bandwagon. When we carefully measure behaviour change and feed that information back to people, they become aware that the norm is, in fact, changing. So monitoring and evaluation is not simply improving the effectiveness of the programme but have become part of the intervention. I think that would be a very powerful approach but, of course, it needs to be tested on the ground.

 

 

Fair enough. Climate changes are an enormous challenge to the world but the adjustment and mitigation strategies should not be borne primarily by poor individuals but by those who are in fact more privileged in this country or in the global North. As a matter of fairness, I tend to agree with this position, Nevertheless, I would say that there are psychological and social insights that can help reduce the consumption of energy and also water on the part of those who already have it, leaving more for others. Those insights include the idea of benchmarking since people typically don't know how much electricity they consume, they don't know what's the right amount. 

 

 

If you tell people how much they consume relative to their neighbours, then they begin to change their behaviour and get back to this idea of peer pressure. A long running study in the US has found that this behaviour persists several years after the benchmarking information is actually removed, leading to potentially long-term changes in habits. It's very cost effective.

 

 

One of the thrusts of the report is that our mental model affects our decision making. Mental models, which are culturally informed understandings of how the world works and our place in the world, does affect decision making. People have mental models about not only about things like health or what causes climate changes, what causes disease and so on but also about themselves and others. Stereotypes have a real affect on behaviour and the experiment you describe is a good example of how that process  works. The good news in some of these studies is that these identities are subject to what cognitive psychologists would  call 'framing', that these are not necessarily unavoidably deep but that circumstances can shift our identities. In fact we have multiple identities and if we can find ways to help people stereotype themselves less, they may be able to improve their performance. 

 

A compelling study was done in California with low income African-American students. These students were asked to perform a self-affirmation exercise in the seventh grade and these exercises improved their grades and their performances a year and a half later. But the interpretation of the experiment is that everyone's grades go down in middle school but if you are subject to self-stereotyping like low income African-American, you tend to think: 'Of course, I will fail; of course, it will happen to me.' But if you insulate people and let them know that other people, in fact, may also have grades that may go down, then they become less subject to the fact that they belong to a socially underprivileged caste. 

 

 

There are at least two counter-intuitive points to be made here. The first is that we typically think that people engage in crime, corruption or anti-social behaviour because they can get away with it and if we increase the likelihood of their getting caught or increase the penalties, they are less likely to engage in anti-social behaviour. This is Gary Becker's model of crime. But, first of all, people don't seriously calculate such probabilities, right? They don't really know how often they are going to get caught. Secondly, lengthening or making the punishment more severe, may or may not change behaviour. If I go from a ten-year jail sentence to a 30-year jail sentence, it kind of all seems the same to me. This is because we tend to discount the longest things in the future. So that's a bit counter-intuitive when we think about how to stop anti-social behaviour. It's also true that sometimes punishment can exacerbate a problem. There is research that is preliminary but compelling which shows  that when you audit individuals for tax evasion, people who are audited behave worse. There are  studies done on small businesses and also on individuals. The literature on individuals is mixed, but many behave in a worse manner for a couple of reasons. First, once they have been caught they might think that next year, the probability of my being caught is very low. Secondly, they kind of get angry that they have been audited and behave worse in the following year. So our standard approach to curbing corruption, which is to use punishment and increase the severity and probability of being punished, may not be entirely sufficient.

 

The point to note here is that most of us are not scoundrels; most of us are, in fact, morally lazy. In other words, we are not hell bent on being anti-social but we think that it's just not that important. If we can frame honesty, put it in front and in the centre, we may be able to eliminate the problem. There is a study which shows that if you are filing your expense report, people are more likely to exaggerate when they sign at the very end of the report instead of the beginning of the report. If you sign at the beginning of the form, you are sort of reminded about the importance of honesty. If you wait till the end, it's less powerful. So that's an example of how psychological factors can be used to promote honest behaviour.

.

 

I don't know enough details about the scheme to comment on it. But what I will say is that social norms are important in tax behaviour. People tend to do what others are doing. The behavioural insights team in the UK is an entity which was part of the cabinet and is now a public-private entity. In large studies of British taxpayers, it has been shown that small changes in the fine print can increase tax compliance. When a sentence is written that nine out of ten taxpayers pay their taxes on time, those who don't pay taxes on time realise that they are in a minority and this realisation can help increase tax collections by between two per cent and five per cent. The larger lesson is that people follow others when they hear about what they are doing...

 

 

Well, what the private sector and what marketing professionals have been doing for some time is to pay very close attention to the contextual influences on producing decisions, They don't necessarily say that if I make the fact that I discounted the product by ten per cent out there in the open, people will necessarily act on it. They instead make this information very salient to people, they put it in big letters, in bright colours, they take advantage of the fact that people are influenced by their peers and use role-models in selling products.

 

 

They'll tap into the cultural theme, like 'this is what it means to be Indian'. 

 

 

I think that the private sector has not gone in for a fairly rigid model that says that individuals are profit maximizing and utility maximizing at all times, that they are basically selfish and that they're, in fact, using all the information they have. This micro-economic approach has been dominant in the public sector and it has an important place. But, at the same time, we need to pay more attention to contextual influences, to departures from this rigid model and try and understand  how and what people actually think. In the private sector, they have been doing this and I think the public sector can learn a little bit from this process. As for role-models, when the president of Brazil was diagnosed with throat cancer, the Google searches for quitting smoking just spiked in Brazil. This shows the power of role models. Leadership, as well as education and entertainment -- movies, soap operas -- can really influence people.

 

 

The evidence suggests that this is an untapped opportunity.

 

 

Well, this is a very important issue. I think that scrutiny of the World Bank is welcome like the scrutiny of any organization. I think we can all improve the work we do. One report can pack in only that much. We didn't focus on the issues you have mentioned, There will be a World Development Report on governance and law, not in 2016 but in 2017, in which we look at, among other things, complaints redressal mechanisms. This is something in my own scholarship and on which I have written about myself, in the past. I'm interested in law, redressal mechanisms and human rights. In this report, we have focussed on psychological and social influences. You have pointed at the staff surveys done at the World Bank. We wanted to make clear that everyone has biases -- not just poor individuals but World Bank staffers as well. We conducted surveys on randomly selected World Bank staffers and found that they are biased in systematic ways. We urge readers to go take a look at that because the results are quite striking.

 

 

Well, fair enough, what I would say is that we tried to stick to the knitting in terms of where the literature is relatively strong. There is a forthcoming very important search agenda on two large macro topics. One is what you mentioned, which is really behavioural finance. We have a chapter on household finance, but institutional investors are also subject to many biases that are not described in this report, of course. Then there is herd behaviour described i the book on 'animal spirits' by John Maynard Keynes. There is a view that what draws the world's largest institutional investors may go beyond mere profit maximization and that they are perhaps subject to psychological and social influences. And that this may, under some circumstances, create bubbles, irrationalities which can have large consequences when they burst. 

 

Now the question arises as to whether those bubbles can be linked directly to the influences described in this report or to something else that is going on, the incentive system, about how investors are paid, in the very short term rather than in the long term, or to psychological influences. To me, the literature is still unclear on these questions. I don't feel comfortable describing these, but I think that you have rightly pointed out that that's an important agenda going forward. The other agenda, the macro agenda, might be called behavioural governance. There is a view that policy-makers themselves are subject to the biases we documented in the World Bank staff. We have to think about what this means when we design processes, regulations, legislation and judicial oversight. We didn't feel that there is enough research to warrant a policy position on some of these issues, at least not as yet.