1. A prize for the science in social science

A prize for the science in social science

Angus Deaton has brought the study of consumption and welfare from a state of hand waving and magic tricks to a rigorous, testable one

By: | Published: October 15, 2015 12:18 AM

Economics, as has been pointed out by commentators and critics alike, is not a science. It deals with the peculiar subject of human beings and their choices, which are inherently unpredictable.

Witness the inability of economists to predict the housing market crash in the US and the subsequent financial crisis. Witness their ongoing inability to prescribe the remedies for the low growth and low inflation paradigm that plagues Western economies today. The mathematical models of economic behaviour, recognised by previous awards of the Nobel Prize to its creators, have failed.

Why, then, do we continue to rely on mathematical models in economics? And why, then, has yet another Nobel Prize been awarded to an economist whose contribution to his field has been even more mathematical models of human behaviour?

Whether you agree or not, the answer lies in the value not of the science, but of the scientific method.

The former applies to fields that study subjects that are regular and predictable. Economics is not one of those fields, like physics is. But there was once a time when the behaviour of light, or atoms, or exothermic reactions, were themselves weird and inexplicable. Now, they have been made commonplace.

The scientific method involves observing phenomena, building a theory around those observations, conducting experiments to test the theory, failing, but repeating until theory and empirics coincide. And it is for bringing that approach to his field that Angus Deaton of Princeton University has been awarded the 2015 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel, for his “analysis of consumption, poverty, and welfare”.

Prof Deaton has brought the study of individual and aggregate consumption and welfare from a state of hand waving and magic tricks to a rigorous, testable one. It might not mean that we fully understand, and it might not mean he is right (as he would be the first to admit). Instead, it means we have an improved set of tools to see where we went wrong, and what we need to fix.

Consider his work on individual and aggregate consumption. Prof Deaton began his career at a time when household data on consumption were becoming widespread, but economic models of aggregate consumption were yet to catch up, relying instead on assumptions of a representative (and rational) consumer.

But the new data were beginning to show that aggregate variables were being contradicted by how individual consumers behaved. Prof Deaton’s contribution was to develop the mechanism by which aggregate consumption could be truly determined as a sum of household consumption, ironing out the contradictions.

As with any scientific method, this involved building a theory that could be tested. The theory consisted of what is called an Almost Ideal Demand System, which describes the relationships between the quantities of different commodities demanded, against prices and total expenditure. The crucial point was that each of those things could now be measured and analysed at a household level.

In practice, this has now allowed us to evaluate the effects of tax and subsidy policies, because the system can more accurately capture the response of expenditure on a particular commodity to changes in income and prices.

Prof Deaton then turned his attention to the question of how much of one’s income or wealth to spend and save at each point in time. This led to a series of insights on macroeconomic phenomena. In the best spirit of the scientific method, he helped convert what was then unknown (or unthought of) to what is now obvious.

The most notable of these is what is now known as the Deaton paradox. Economists used to think that people would vary their consumption based on their income expectations `rationally’. That is, they would save more (and consume) less if they expected a lower future income, and borrow to consume more if they expected to earn more in the future.

Instead, the data showed that aggregate consumption actually varied less than aggregate income did. Nobody could say why.

Prof Deaton noted, in a series of academic articles, that different individuals face different income challenges at the same time, and the aggregate effect on consumption tends to wash out. He also pointed out that individual consumption decisions were affected by liquidity constraints and borrowing costs.  Obvious now, perhaps, but revelatory at the time.

But the real impact of these discoveries was to convince the world that data, and their measurement, are imperfect, and must be carefully considered before use. This is particularly sound advice for those who study data and blindly draw conclusions from them.

Prof Deaton’s other contributions have been in the subject of welfare, particularly in developing countries.

Some of these have to do with the ingenious use of scarce data. For example, repeat surveys of household consumption do not allow a researcher to track the path of a particular household over time, simply because the same household may not appear in each survey. But treating a household as a part of a cohort (by age, income, or other parameters) allows a researcher to use different surveys to track the path of a cohort, and therefore households by implication. This has led to a series of formidable (and well-targeted) poverty reduction initiatives.

Others have to do with the nature of data collected. In the context of measuring poverty, Prof Deaton has shown how to estimate per-capita consumption (and therefore poverty) from household data. This is important because children and adults do not consume the same amount, so a flat per-capita calculation would understate child poverty and overstate adult poverty. His efforts have led to revamping poverty estimates the world over, India included.

Prof Deaton’s lasting legacy, however, will be in terms of the research and methods his work has spawned.

Economics may never reach the status of a science. But with efforts like his, we inch ever closer.

The author has a PhD in quantitative economics from the Indian Statistical Institute, Delhi. He is at present a consultant with the Ministry of Rural Development, Government of India.

  1. No Comments.

Go to Top