An accidental economist: a brief history
A series of accidents, mostly fortunate, made me the human being and economist that I am. It started with the accident of my birth, to a couple who were both economists. This, in my generation, was already a minor miracle – in the mid 1950s my mother somehow managed to persuade her parents, who were relatively conventional Maharashtrians of the time, to take the very unusual step of sending her to the London School of Economics. This is where she met my dad. My dad was really not meant to be there; he was a college drop-out who, bored with his clerical job in Kolkata, made his way to London with a goal of studying more. He took night classes at the Kensington Polytechnic while working various jobs as a manual laborer. With all that he did so well in his part 1 BSc exam that one of the examiners, Richard Lipsey, a professor at London School of Economics, went out of his way to seek him out and eventually secured for him a scholarship to go to LSE for his final year. But for this chance encounter with such extraordinary generosity, it is likely that he would have never gone to LSE and my parents would have never met.
As a child of economists, I knew that economics was one field I must avoid. My father was a famously charismatic teacher, who adored and was much adored by his many students. He would often talk about just how brilliant some of them were, and it was clear to me that I had nothing to gain and much to lose by inviting comparisons with them. In my deeply anti-intellectual high school, it was made very clear that we should all aspire to study engineering or medicine because they led to good jobs (the lure of jobs in finance came many years later). They made an occasional exception, in the case of an unusually brilliant student, for studying physics. I had no desire to be an engineer or a doctor and prepping for physics required consorting with our physics teacher, a man who seem to take genuine pleasure in inflicting pain. What else could I do? I loved literature and history, philosophy and math; my parents were against the first three. Their stated grounds were that I could always go from math to those at a later stage but not the reverse, though my guess now is that they were not sure that I was good enough to make a living in the humanities, given the shape of the labor market. In any case their argument for math appealed to my instinct for trying to postpone all hard choices. Math it was going to be.
I applied and got into the undergraduate program at the Indian Statistical Institute in Kolkata, reputed to be far and away the best place to study math (or mathematical statistics) in India. ISI, as it is popularly known, is – to my knowledge – the only Indian academic institution that was a true world leader in its field; many of the early leaders of the field of statistics were professors at ISI, along with many of prominent scholars across the discipline of pure and applied mathematics. Many of those teaching us in our entering classes were well-known scholars and the overall quality of teaching was very good.
But almost as soon as I got there it was clear to me that I needed to get out. It might have been my attempts to crack the occasional risqué joke that typically met with looks of disapproval or bafflement from my classmates. Or the fact that classes started early in the morning (for me, anything before noon was kind of early) and continued till dusk, and then there were a few dozen math problems to solve for class the next morning. Or perhaps the more general sense, seemingly shared by my teachers and classmates alike, that this was a monastery and we were novices waiting to be inducted into its mysteries. I liked and enjoyed math, but I was both too young and too undisciplined to accept to make it my entire life (or at least that is what it felt like). For those who know the classic musical Sound of Music, I was Maria – not meant to be a “credit to the Abbey”.
What then? The Indian system in those days was not forgiving of deviants. Deadlines for many of the standard options were long gone. I was back to trying to decide between literature, history or philosophy, when my parents tentatively sprung the idea of doing economics. I was aghast – my father was the head of the economics department at Presidency College in Kolkata, widely recognized as the best place to study economics in the country. I would have to take classes with him – what would he call me? What would I call him? And suppose I did badly …?
It turned out however that several of my close friends were applying. It seemed like a cool group to be with. I covertly picked up an economics textbook by Lipsey, the same man who was my father’s savior many years ago. On a quick read, introductory economics seemed almost entirely disconnected from what I imagined economics would be about, the many fraught social issues that my parents discussed at breakfast and dinner. In particular, the word poverty never seemed to come up.
On the other hand, it seemed easy enough. I asked my friends about job prospects, wary of showing too many of my cards to my parents. One can do an MBA, I was told, which sounded dire. Or work for the media, which sounded more plausible – I kind of fancied myself as a writer of expansive op-eds about beauty and justice. My parents planted the thought that it is easy to move from economics to other fields. Satyajit Ray studied economics. So did another wonderful filmmaker, Shyam Benegal. Amartya Sen himself was almost as much of philosopher as he was an economist. Perhaps I could do it too. At the very least it postponed making a choice. I was sold.
I hugely enjoyed my three years at Presidency. The teaching was very good; I found microeconomics rather bland, but my father, who taught it, was extremely entertaining. Mihir Rakshit taught us macro and managed to make the ideas crystal clear without trivializing them. One left with the impression that there was something more to learn. My favorite was economic history, beautifully taught by the chronically self-effacing Nabendu Sen.
But what made it truly special were my fellow students. There was always someone who knew much more than me about literature, history, philosophy, politics, art, music, theater or film – in other words everything that I really cared about. And they were passionate and keen to share; it was the most incredible learning opportunity. I read voraciously, attended every show and showing that was on offer, and argued with everyone, in part to test my growing understanding. And I discovered just how much I love getting to know new people – their passions and their interests, their anxieties and their hesitations, what makes them smile or frown. I was always gregarious, a result, I suspect, of the way we grew up. My parents were warm and hospitable people, and even when money was tight, welcomed the large and diverse community of their friends into our house. As a result, my brother and I grew up to be very comfortable around adults, even adults who had much too much to drink. That openness to people quite naturally extended to my own cohort in the very congenial environment of Presidency College. I now have more close friends than almost anyone I know; I have close friends whom I have known since I was five, and close friends I met a year ago, close friends who are twenty years older than me and close friends that are much, much younger. It is one of the unadulterated joys of my life.
Three years passed quickly and then there was another decision to be taken. What next? I still didn’t absolutely love economics, but I was doing well and in the spirit of why not postpone any decision that can be postponed, I figured that a Masters degree would be handy. It also seemed exciting to move to a different city and live in a dormitory. The question was where – Delhi School of Economics was known as the best place to study economics in India and my father, in his gentle and always carefully balanced way, was nudging me to go there. The alternative was Jawaharlal Nehru University (JNU) in New Delhi, famous for its lively political life and Marxist economics. I was slightly leaning towards the latter – I had decided that I needed to learn about politics and given that I found conventional economics somewhat uninspiring, I had hopes that Marxism would deliver the punch I was looking for. However, what actually decided me was a recruitment visit to the Delhi School, where a brash young faculty member told me that DSE was the royal road to PhD programs in the US. That made up my mind – I was looking to be inspired, not a prep school for some future step that I had no burning desire to take. I was on my way to JNU. I wonder where my life would have gone had he not taken on to convince me.
I have written elsewhere at some length about my JNU experience. That blinding moment of insight that I was waiting for never arrived. Marxist economics, the way we were taught, was much like all the other economics I had seen, just with some assumptions tweaked (and a bunch of self-righteousness thrown in for good measure). Since I had no particular stake in the assumptions of standard economics, I found it easy to believe that one could make others, but I found it unsatisfying to be told that these assumptions were better because they were in the Marxist tradition. I had tried to read Marx as an undergraduate, and found him, in equal measure, illuminating and confusing – I saw no reason to take his words as a say so. The fact that the market system was not the paragon of all virtues that the standard micro textbook implied, seemed obvious to me – the question that troubled me was how do we get to a convincing diagnosis of where the problem lies? Is it all to be blamed on the capitalist class? Or are there other problems that might be with us even after revolution? How would one check? It was only some thirty years later that I caught, in the writings of my friends Thomas Piketty, Daron Acemoglu and Jim Robinson (and their many collaborators), a glimpse of where some genuinely creative Marxist thinking could take us.
The teaching at JNU, both from the Marxists and the more orthodox, was generally very good. But I was not particularly challenged, thanks to the excellent grounding in basic economic logic that I came with from Presidency. This combined with my growing disinterest in economics, meant that I learnt very little of the field in those two years. There were exceptions – Anjan Mukherji and Satish Jain, two of the most generous teachers I ever had, gave me a glimpse of what made high theory exciting to them and the wonderful Krishna Bharadwaj was a very convincing advocate for the History of Economic Thought – but on the whole I learnt much more from my many historian friends (including by occasionally sneaking into their classes with them).
Perhaps the most important thing I got from JNU was a sense of what politics in the Indian context meant. I learnt activism from the activists, pamphleteering from the master pamphleteers, slogans from the demonstrators. I listened to the Marxists talk about class, the Gandhians about caste and religion, the feminists about gender. The world was suddenly a more interesting and dangerous place.
Unfortunately, the two wonderful years were soon about to run out. My grades were good. My teachers and my parents wanted me to apply for a PhD. I dithered, but eventually decided that I might as well take the GRE and see if I was good enough to get in. I did well, which made it that much harder to not apply. I decided to go for it, but gamble on getting into one of the top places and if not try something else. Since applying was a huge pain – the forms were endless – I went for just a handful of places. To avoid being pressured to apply to too many places, I did not consult anybody; this is why I missed out on applying to Princeton and MIT, which were probably the two best places for someone like me. Instead, I applied to Harvard, which I knew from Love Story, and to Berkeley and Stanford, because I had heard of Gerard Debreu and Kenneth Arrow, which was a bit like converting to Catholicism with the goal of making friends with the pope.
I guess I will never know why Harvard admitted me. My application was hand-written in my rather scratchy hand, and the research proposal was border-line nonsensical. I was a good student, even a very good one, but at a university which had never sent anyone to Harvard.
In any case they did, and I felt that not going would be rude to my presiding deity who was clearly looking out for me. In any case, it was not as if I had plan B. Nonetheless I must confess that I was slightly relieved when a student demonstration landed me in prison, and many people confidently told me that the Americans would not give me a visa. I applied for jobs at a few NGOs and shocked many well-wishers by being entirely nonchalant about the whole affair.
Bizarrely enough I did get the visa, despite the fact that we were all (quite unfairly) accused of various rather heinous crimes. My first days at Harvard were hard; I was homesick, pining for my girlfriend in India and unused to working hard. The teaching was at an entirely higher level than I was used to and two years at JNU had ruined my work-habits, never great to start with. Many of my classmates, I could tell from their questions and comments during lectures, were way ahead of me. I was a part of a great class: Tyler Cowen, Mathias Dewatripont, Steve Kaplan, Miles Kimball, the late lamented Alan Krueger, Kiminori Matsuyama, John Nachbar, Nouriel Roubini and Rob Stavins were to leave their imprint on the profession in one form or the other, and other like Allen Sangines could have easily done the same but opted not to. Many of them came much better prepared than me and they all worked very hard and talked about it; I discovered that in the US, unlike in India, it was cool to work hard. I think this is what saved me; I decided that I need to try harder. Those work habits stayed with me; I still work long hours, interrupted by regular doses of distraction every hour or two – I play with the kids, chat with friends, cook or play a game of ping-pong.
It still took me a long time to properly catch up. I have always been obsessed by how every bit of what I know fits together with the rest, which means I get very confused whenever I meet something entirely new. Classes at Harvard were a mixed bag. I loved listening to Andreu Mas-Colell, who tried to get us to appreciate the simple, often geometric, logic behind a lot of the theorems in microeconomics and not to get bogged down by the notation and the bombast. He would eventually become one of my PhD advisors. David Kreps gave us an early and brilliant introduction to both Game Theory and Behavioral Economics, but I was too unsophisticated to fully appreciate what he did. Larry Summers introduced us to empirical economics circa 1983 in his own unique way, a blizzard of intriguing insights and confusing words and math. Econometrics was boring and similar to what I had already studied at JNU, so I decided to mostly skip it. It remains one of my abiding regrets – the result was that the only econometrics class that I (partially) attended was Time Series, the one part of econometrics that has since never crossed my life. Macro was deeply confusing – the application of “rational” expectations to various dynamic models generated a certain black magic that I learnt, painfully, to reproduce, but found entirely implausible. Nobu Kiyotaki, a macroeconomist who I admire greatly, who was then our kindly Teaching Assistant, and therefore the target of my many queries, once told me “you have many puzzles, don’t you”.
Predictably perhaps, I had trouble deciding what field to study. My original plan was to study economic history, but it turned out that economic history at this time was moving away from telling illuminating stories, which I enjoyed (and still enjoy), towards more conventional applied micro. Development economics was in throes of computable general equilibrium models – the view seemed to be that we knew how the world worked in general, all we lacked was the computationaI capacity to figure out what happens in specific cases. A couple of lectures on computational methods decided me to look elsewhere.
Trade was another way into development, at least in those days, and I enjoyed the first class I took with Kala Krishna and Sue Collins and ended up being hired to work as a research assistant to Jeff Sachs, who was then flying around fixing up macro economies in Latin America. But Jeff quickly decided that I neither had the computational skills to crunch his data nor the personality to bang the desks in finance ministries, so, being a gentleman, gave me something that was mostly make work. I got the message. Development was not for me.
In the spirit of not committing to anything as long as possible, I studied macro and theory as my main fields, telling myself that if I had a strong enough base I could decide later. It was clear to me however that the classes I most enjoyed were the theory classes, Game Theory with Dilip Abreu, General Equilibrium with Andreu and a Contract Theory class that Oliver Hart taught at MIT. Andreu in particular, despite the formidable formalism of his own research, was very good at making abstract ideas seem quite concrete and connecting them to the world, and Oliver’s lessons offered an object lesson in how to apply theoretical ideas to illuminate very practical issues.
The challenge for me was to reconcile my urge to do something useful with my love of elegant arguments; my mother, who spent her life reacting, often with extreme vehemence, to the unfairness and wanton waste that she saw everywhere in the world, was speaking in my right ear, while my father, who loved elegant arguments and clever mathematics, was speaking in my left. On the strength of what I heard in Oliver’s class and encouragement from Andreu, I decided that I could be useful as one of these applied theorists who tells stories about the world, in part to undermine the many unfounded sureties that economists often carry with themselves. I obviously did not know how I would get there.
This is when Eric Maskin arrived from MIT. Eric, not knowing quite what he was getting into, encouraged me to formalize the ideas I had and show them to him. I, blissfully unaware of Eric’s reputation as one of the most intimidatingly brilliant minds of his generation, did exactly that. Eric saw many silly ideas, and patiently took them down. This is when I got lucky. One of the ideas I was playing with turned out to be more interesting than it had any right to be, and literally within days I had a paper on the “Economics of Rumors” to go on the job market with.
This was also when I was beginning to really enjoy economics. It helped that I had two wonderful playmates, Mike Spagat and Andy Newman, who were, like me, left-leaning theorists who wanted to use models to tell stories that usually do not get told. We spent hours together, part crafting models and part shooting the breeze. The work with Andy Newman on the foundations of poverty traps, which started in those years, was to eventually to change my life. I was married by then and waiting for my wife to finish her degree before looking for a job. Given that I had my job market paper in hand, this gave me a year and half with lots of time in hand. I filled it with coming up with an idea a day, usually worthless. I was having a great time.
I went on the job market in the winter of 1987. Everybody told me it would be stressful to be surrounded by so many others who are all looking for the same jobs, so I decided to distance myself from it. I parked myself at Mike Spagat’s house, some 20 miles from downtown Chicago where the meeting was, ignoring all advice to stay as close as possible. As luck would have it, there was a massive snowstorm on one of the days when I was supposed to be interviewed for a job, so I had to get up in the dark, walk a couple of miles in knee-high snow, and managed to find some public transportation that eventually got me there. But probably as a result of all that anxiety, I left my jacket in a taxi while going from one site to the other and ended up being interviewed in my overcoat (I was cold).
To compound my misjudgment, I had decided to avoid the days right after the job market when famously, job candidates spend their time anxiously trying to figure out how everyone else’s market was going. I went to the UK, to visit Eric who was on sabbatical there, and work on our joint project. This was in the days before cellphones, and as I quickly discovered, it was a terrible idea. I did have an answering machine but calling into it involved feeding innumerable coins into the aging phones in the bright-red British phone booths, and often ended with half the message being inaudible. I almost missed my invitation to give a job talk at Stanford.
Overall, my job market performance was solid rather than spectacular. All the “top” schools, except Stanford, had decided to pass. When I reported that at Harvard, two of the kindest economists I know, Jerry Green and Kala Krishna, decided to go out and bat for me. This is how I ended up with a job at Princeton.
My wife Tuli was starting a PhD in French Literature at Columbia. We lived in Manhattan and I commuted to Princeton. It is while waiting for trains to Princeton that I noticed that there were days when a long queue would form on the platform at the wrong place – the door would open somewhere else and those not in the queue would get in and grab the seats. In trying to think of why these systematic and correlated errors occur. Partly inspired by my work on rumors, I wrote a paper called “A Simple Model of Herd Behavior”, which in many ways built my career. It somehow struck a chord and suddenly I was getting feelers from here and there, and eventually an offer to return to Harvard. This was great for me – it gave me exposure and made it possible for me to move to MIT the year after, in 1993. It was probably not good for Tuli, who was writing her dissertation and very generously agreed to move with me. The resulting relative isolation from the academic network at Columbia might have been one reason why despite her great charm and obvious brilliance, she had a very hard time finding an academic position. She eventually turned this around and is now one of the heads of MIT’s fund-raising team, but she suffered for many years, always with a brave face to the world, and I fear I was not always as supportive as I should have been.
The move to Harvard made me a development economist. I remember Andrei Shleifer asking me what advanced PhD class I would want to teach, since the theory classes, which is what I taught at Princeton, were already covered. On an impulse, I suggested development. It is a tribute to the generosity of the existing development faculty at Harvard – leaders of the field like Dwight Perkins and Jeff Williamson, as well the younger Jonathan Morduch – that they allowed me to take over a PhD class, despite the fact that I had never studied development and my main claim to research in the field was a theory paper with Andy Newman with development in the title. As I said, I moved to MIT the next year, and continued to be allowed to teach PhD development. I had chanced upon on an opportunity of a lifetime.
Looking back, I think it was actually an advantage being an utter novice. Development at the time was a mine field, with many strong personalities and powerful minds who did empirical research and did not get along. I knew nothing of these battles and as a result, taught a course that was very different from what was standard, both more theory driven and more skeptical of what passed for empirical research. In particular a lot of the empirical work seemed too obsessed with showing that markets (or non-market institutions) delivered efficient outcomes, which I found hard to credit. I started to think about how to get more credible evidence, and my one insight was that we needed fresh data, collected with the relevant theory in mind. Kaivan Munshi and Maitreesh Ghatak, two of my earliest PhD students, were my accomplices in this enterprise, which Chris Udry at Northwestern and Rob Townsend at Chicago were also pursuing.
Around this time, there were two more critical accidents that entirely reshaped my life. First, there was another opening in development teaching and I somehow persuaded Michael Kremer, who had just finished his PhD and joined us at MIT as a brilliant young macroeconomist, to take it on. That was quite a coup, since macro was a serious field and development, quite plainly, was not, as one of my senior macro colleagues kindly pointed out. The result was that I was exposed to Michael’s rather unique perspective on the world; in particular I remember a conversation where he wondered aloud why we economists don’t do randomized controlled trials (RCTs). It turned he was doing one. This was I think in 1995. I went home and thought hard about it and was entirely persuaded. RCTs offered so many possibilities that one could never get to otherwise. Emboldened by the fact that I had some confidence in our ability to collect data, I started organizing my own first RCT – and persuaded Michael join me. This was in 1996.
The second accident was that Esther Duflo applied to our PhD program. With the sort of perfect irony that one would never dare to make up, I rejected her application, but thankfully Thomas Piketty rescued it. In the Fall of 1995, I taught development to Esther, Eliana La Ferrara, Jishnu Das, Asim Khwaja, Stuti Khemani, and Chris Spohr, all whom have made major contributions to the field. Development was on its way. In a little while Esther asked me if she could work for me as a research assistant, which is how I really got to know her and learnt to appreciate the way she, better than anyone I know, combines deep understanding of theory and empirical techniques with intense practicality, pitch-perfect common sense, infinite energy and boundless generosity. She might be my student, but she taught me a big chunk of what I know about empirical research.
As my work with Esther and RCTs in general flourished, we had another lucky break. A white envelope arrived from the MIT administration. Esther, Sendhil Mullainathan and I, were offered 100,000 dollars each to promote our research. The unspoken goal was to keep us at MIT. We decided to pool the money and build an institution around RCTs, and somehow that idea caught the fancy of Susan Hockfield, who was MIT’s new president. This brought in Mohammad Jameel as a donor and a supporter – I believe that he saw more potential in us then, than we ourselves did. That was also true of Bengt Holmström, our brilliant department chair and the wonderful Rachel Glennerster, Michael Kremer’s wife and a friend, who I persuaded to give up her lucrative IMF job and join us – convincing her is the one part of the success of the Abdul Latif Jameel Poverty Action Lab (J-PAL) that I take credit for. The rest is history.
Amidst all the lucky breaks that shaped my life, exactly as the J-PAL story unfolded, a tragedy was brewing. Tuli’s and my son, Sasha, was going from a wonderfully precocious boy to a troubled adolescent. That, among other things, contributed to our breakup. Sasha eventually moved in with me. We had an extraordinarily close relationship – we would talk five times a day or more, but I, and Tuli, who also tried her best, and his community of devoted friends failed to persuade him that life was worth fighting for. Sasha died on the 10th of March 2016.
I have learnt to live with his absence, though there are days when I still refuse to believe it. That makes me all the more grateful for the other piece of unbelievable good fortune I have enjoyed, my extraordinary and wonderful wife Esther and our two intensely adorable children, Noémie and Milan. Touch wood, or as we say in Hindi, chasme baddoor.
This autobiography/biography was written at the time of the award and later published in the book series Les Prix Nobel/ Nobel Lectures/The Nobel Prizes. The information is sometimes updated with an addendum submitted by the Laureate.
See the full schedule