by Scott Duke Kominers
How can the impact of economic policies be measured? One solution is to look for what’s called a “natural experiment,” or a place where a policy’s design — or possibly even nature itself — has created an implicit comparison between affected and unaffected groups.
For their efforts to refine the art of the natural experiment, David Card, Joshua D. Angrist and Guido W. Imbens were awarded this year’s Nobel Memorial Prize in Economic Sciences.
Some of Card’s work — which he undertook with the late Alan Krueger — led to a deeper understanding of the impact of minimum-wage laws. Conventional economic wisdom suggested that raising the minimum wage could cause employment to fall, since it would both raise the cost of hiring workers and increase workers’ competition for jobs. But it was hard to know whether that theory would hold up in practice — and if so, how big the effect would be.
Card and Krueger found a natural experiment thanks to the line separating New Jersey and Pennsylvania. Workers on either side of the border aren’t that different from each other, except that those on the New Jersey side received a minimum wage increase.
Surprisingly, Card and Krueger found no sign that employment in New Jersey decreased relative to Pennsylvania. If anything, it grew. Their scholarship forced a fundamental rethinking of the impact of minimum wage policies, and kicked off a boom in research in labor economics that continues to this day. The work has reshaped policy, bolstering the argument that raising the minimum wage won’t hurt employment.
Angrist, also working with Krueger, used a version of these methods to assess the impact of education on earnings. The two observed that laws mandating compulsory school attendance created a small disparity in the amount of time people who were in the same class year had to stay in school. Those born earlier in the year reached the age at which they could drop out before students born later in the year, which meant that on average those born earlier received a little bit less schooling.
Since a person’s birthday typically isn’t related to his or her earnings, the birthday-driven difference in educational attainment was random relative to earnings. Angrist and Krueger’s analysis based on this insight led to an estimate of the returns of an additional year of schooling: a roughly 9% increase in pay for students who stayed in school.
Looking for implicit randomization that separates otherwise similar groups is incredibly powerful, but it is also limited by human behavior.
People don’t always react to a policy change in the same way. To explain this, the Nobel committee used the example of a policy where employees of a company are given free bicycles. The policy will cause some employees to start biking — and they might see health benefits. But there are other employees who were already biking, and others who won’t take up biking at all — and naturally, we don’t expect to see any change in these employees’ health outcomes. But then how can we learn about the health effects of the policy?
Imbens and Angrist identified a simple and intuitive statistical framework that makes it possible to estimate causal effects in these sorts of settings. They also clarified how we think about the effects themselves, showing that what we’re measuring is the impact of the policy on those who adjust their behavior as a result of the policy change — in this case, those who take up biking.
The Nobel laureates and their collaborators applied these methods to greatly expand the application of natural experiments in economics. In doing so, they ushered in a new era of empirical economics. And they continue to push the frontiers to this day: Card’s recent work has examined youth labor markets and the sources of earnings inequality. Angrist has researched the effect of Uber on the labor market. Imbens, in joint work with Susan Athey, has fundamentally reshaped our understanding of the relationship between econometrics and machine learning. Moreover, all three of the laureates are renowned advisers and teachers and have been tremendous servants to the profession.
Taken together, this makes the trio well worthy of the Nobel — you might even call them “natural” candidates.
Scott Duke Kominers is the MBA Class of 1960 Associate Professor of Business Administration at Harvard Business School and a faculty affiliate of the Harvard Department of Economics.
