Economics Is Not A Science

What is economics? To experts like Harvard Economics professor Raj Chetty ’00, economics is a science: Economists form hypotheses, gather data to test those hypotheses, and draw conclusions. If economics is a science, though, why do economists disagree? According to Chetty, disagreement in the field can be traced to methodological shortcomings. The lack of lab-like settings to run randomized controlled trials forces economists to resort to quasi-experimental methods and econometric wizardry. Though this imperfect form of experimentation causes disagreement, as the field progresses, methodologies will improve and data will become more readily available. Given time, Chetty argues, the field of economics will be as scientific as medicine.

This conception of economics is steadily gaining momentum at Harvard. Chetty teaches a course titled “Economics 50: Using Big Data to solve Economic and Social Problems,” which prides itself on introducing Harvard undergraduates to the world of economic research. For example, with the help of Harvard Economics professor Nathaniel Hendren and University of California, Berkeley Economics professors Emmanuel Saez and Patrick Kline, Chetty analyzed U.S. Census Bureau data to illustrate the causal effect of neighborhoods on upward mobility. Their policy recommendations include making investments in low-opportunity areas and helping families move to other locations. With the proper econometric training, students will be able to replicate their findings and perhaps craft policy proposals on their own.

While advances in data analysis have given researchers, students, and policymakers great insight into the surrounding world, Chetty and other econometric magicians are trying to turn economics into something it cannot and should not be. The direct step from statistical research to policy making dangerously masks the normative realities of crafting legislation.

The methodology of medicine is roughly as follows: Utilizing randomized control trials, researchers find treatments to cure the ailments of a population. Chetty likens the economist to a type of doctor. The ailments are of a different nature (unemployment, poverty, inequality, etc.), and so are the treatments (taxes, regulation, the lack thereof, etc.), but according to him, the paradigm is fundamentally the same.

Economics and medicine, however, are very different. One defining characteristic of policy making is its medium: the state. Passing legislation necessitates reflection on normative positions regarding the function of government. Some examples of such considerations include: What is the proper role of government? What defines a government that acts justly? Should morality be the impetus for policy creation? Who should the government prioritize if a policy has winners and losers?

Take the debate surrounding unemployment insurance. A simple policy question would be: Should the government grant a 10-week extension in unemployment benefits to those out of work?

According to Chetty, consensus has been reached:“[S]tudies have uniformly found that a 10-week extension in unemployment benefits raises the average amount of time people spend out of work by at most one week. This simple, unassailable finding implies that policy makers can extend unemployment benefits to provide assistance to those out of work without substantially increasing unemployment rates.”

The certainty is extraordinary. Let us assume, as he does, that this finding is incontrovertible (which is a generous assumption, given that there is at least one study concluding otherwise). Does it immediately follow that 10-week extension policies are desirable, given such remarkable evidence? No, it does not, and for good reason.

This policy, like all government policy, has winners and losers. The winners seem obvious — those who receive the unemployment extension. The losers are less obvious, but it would be a mistake to assume they don’t exist. One potential ramification of the policy is an increase in government spending and subsequent increase in the national debt, which could lead to disastrous long-term consequences for the entire population. Policymakers need to ask themselves if the debt accumulation is worth it or not. The question, like all of welfare economics is centered on interpersonal utility comparison: To what extent should we sacrifice overall welfare in the future for the welfare of the disadvantaged today? Another area of concern is justice. Is it just to redistribute money from those who have jobs to those who do not? Again, this question gets at interpersonal utility comparison, as well as a whole host of other normative judgements.

Positive certainties do not lead to policy certainties. As elementary as normative economics seem, economists today tend to shy away from such philosophical discussions. Moral claims and debates about justice are often vague and subjective, and “big data” offers little help. The experts of the field, however, should not so readily abandon normative considerations, at least if they wish to continue making direct or indirect policy recommendations. On the contrary, policymakers need to be honest about which value sets they take on and why they choose to take them. This isn’t to say econometrics should be abandoned. Positive and normative deliberation working in tandem is essential to good economics. Incorporating both into policy discussions will lead not only to better policies, but also to a more genuine public discourse.

Dody T. Eid ‘20 is an Economics concentrator in Leverett House.