In a tribute to Angus Deaton, recipient of this year’s Nobel Prize in Economics, Maitreesh Ghatak, Professor of Economics at the London School of Economics, outlines Deaton’s contribution to economic and policy analysis, and to bridging the gap between theory and empirics. He also highlights the strong connection to India in his work.
There is a cultural divide in Economics - theory is theory, and empirics are empirics, and seldom do the twain meet. Every now and then someone comes along who can straddle both worlds, and the recipient of this year´s Nobel Prize in Economics, Angus Deaton, is one of them.
Deaton’s key contribution has been to look at the behaviour of individuals that lies behind economic aggregates used in macroeconomic or policy analysis, like the saving rate or the extent of poverty, building them up from household-level data on consumer behaviour. He has helped develop methods that allow us to see how millions of individual decisions regarding saving and spending, like little dots in a granular picture, aggregate up to the economy-wide level. The ultimate goal is to find out how the welfare of individuals changes when some of these economic aggregates move. The biological analogy would be looking at the behaviour of individual cells in the human body to be able to make inferences about overall health, and, conversely, using reports on overall health to infer what is going on in individual cells.
One of Deaton’s key contributions to studies of household consumption behaviour is to demonstrate and explain why consumption varies more than income at the individual level but not at the aggregate level. Deaton showed that, since current changes in income are often indicative of future changes in income, individuals tend to consume some of this expected income increase right away. But at the economy-wide level these individual income fluctuations often average out, since income goes up for some and goes down for others; so average consumption varies much less when average income varies.
Deaton’s work does not follow the tradition of developing a grand theoretical framework of a Paul Samuelson or a Kenneth Arrow or subsequently a Gary Becker or a Joseph Stiglitz. Nor has he come up with a striking theoretical or empirical tool or a concept that changes how we do economic analysis – such as game theoretic methods pioneered by John Nash, rational expectations in macroeconomics by Robert Lucas, or methods of correcting for selection bias in empirical analysis by James Heckman.
The value of Deaton’s work lies in honing the tools and techniques that go behind construction of data series that are then used for everyday economic or policy analysis. It may appear less glamorous than coming up with a novel theory or an empirical technique but ultimately, from the point of view of the practical usefulness of economics as a subject, is no less important and no less demanding of brilliance and ingenuity. One of his major contributions, for instance, has been to help construct appropriate price indices to enable comparison of living standards across different places and over time, which are key building blocks in the World Bank’s cross country dataset on poverty. In this respect, Deaton follows the empirical tradition of Richard Stone who pioneered national income accounting and happens to be one of his advisors, and Simon Kuznets, who pioneered construction of historical time series on national income.
Even within such august company, Angus Deaton stands apart with his close combination of theory, measurement, and careful empirical analysis to answer policy questions of great interest. As in a detective story, theory tells us what the possibilities are behind a phenomenon, and tells us where to go to look for evidence and what kind of evidence to collect. Once the evidence is collected, empirical analysis helps identify which theory explains the facts the best. For example, a lot of research in development economics these days is about discrimination against girl children within families. Deaton came up with a really clever idea to test for it. When a child is born, a family’s per capita income falls by definition and so it has to curb some of its spending. Deaton’s conjecture was that, in the presence of gender discrimination, some “adult” consumption goods like tobacco and alcohol would be cut more following the birth of a male child. Subsequent work has indeed found evidence of gender discrimination using this method.
This example shows that theory and empirics are close complements, not rival camps, in making sense of real world phenomena. Deaton has always cautioned against theory and empirics not moving in tandem, and against fetishising method over substance, and recently has been a strong critic of the excessive focus on the method of randomised control trials in development economics.
There is a strong connection to India in Deaton’s work, and those who know him, vouch for the fact that this bond is not just academic. Deaton begins his book Analysis of Household Surveys with a fact that is not as well-known as it should be: one of the first large-scale scientific sample surveys in the world was carried out in India by Prashanta Mahalanobis, the Founder-Director of the Indian Statistical Institute (ISI) in Calcutta. Not just that, he credits Mahalanobis to be a pioneer of modern household surveys in the form of the National Sample Survey (NSS) that started collecting data on household consumption (among other things) regularly from 1950 and continues to date. Incidentally, Deaton was instrumental in enabling public access to NSS micro-data, which has happened fairly recently.
In his own writings he jokingly attributes his fascination with India to growing up in the cold and grey climate of Edinburgh and to his resultant longing for the warmth and colours of the tropics. However, his choice of topics, from the measurement of poverty to health, reflects a moral dimension to the challenge of finding solutions to the great deprivation that exists in the world amidst plenty. The fact that more than one-fourth of the world’s poor live in India is certainly one of the reasons for his abiding interest in the Indian economy. His contributions on this topic (often working jointly with Jean Drèze) include attempts to gauge whether growth has reduced poverty and the impact of this on food intake and nutrition.
As a personality, Angus Deaton is larger than life, well known for being outspoken and his distinctive brand of humour. There is a story about a seminar speaker Deaton once hosted at Princeton. The speaker apparently said that economists need to understand the criteria bureaucrats use to evaluate policies and take those as the right benchmark, as opposed to how they ought to be evaluated based on economic principles. A sceptical Deaton felt that economists should stick to their own judgement criteria. The legend goes that at some point he quipped that some of my best friends like to sing duets with their pet donkeys but that does not mean the rest of us should be doing that too (knowing him, I suspect this is a sanitised version of what he actually said). There is no point trying to read too much into a joke, but this story does reflect Deaton’s overall vision as an applied economist: economics is a real-world subject, the real world is a messy place, and to interpret and evaluate it we need to stick to core economic principles.
Reprinted from THE INDIAN EXPRESS with the permission of The Indian Express Limited © 2014. The link to the original piece is here.