The Economics Nobel rewards “natural experiments”

On 11 October 2021 the Royal Swedish Academy of Sciences awarded the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel, known as the Nobel economics prize, to three economists who work in the United States. Half goes to David Card, born in Canada and a professor at the University of California, Berkeley, for his empirical contributions to labour economics. The other half is shared by Joshua Angrist, of the Massachusetts Institute of Technology (MIT), and Guido Imbens, of Stanford University, for their methodological contributions to the analysis of causal relationships. The prize is worth 10 million Swedish kronor, about $1.1 million.

What a natural experiment is

In medicine, to find out whether a drug works, patients are split at random between those who receive it and those who do not. In economics that is almost never possible: you cannot raise the minimum wage for only half of a country’s firms to see what happens. A natural experiment takes advantage of the fact that, sometimes, reality makes that split by itself: a law that applies in one state and not in the next, a lottery, a cut-off date. If the two groups are comparable, the difference between them comes close to what a controlled experiment would show.

Card applied that idea in the early 1990s. In his best-known study, with the economist Alan Krueger, he compared employment in fast-food restaurants in New Jersey, which had raised its minimum wage in 1992, with that in neighbouring Pennsylvania. The result ran against what the standard theory predicted: the rise did not reduce employment and may even have increased it. That work led many researchers to revisit the idea that raising the minimum wage always destroys jobs.

Another of his studies looked at the arrival of Cuban refugees in Miami in 1980, the Mariel boatlift. According to the study, that inflow of workers had very little effect on employment and wages in the local labour market. These are results from specific cases: they need not repeat themselves in another country, sector or period, which is why economists keep testing them against new data.

What Angrist and Imbens contributed

A natural experiment has its pitfalls: the groups being compared are not always identical, and not everyone affected by a rule changes their behaviour because of it. Since the 1990s, Angrist and Imbens have developed the statistical framework that shows what can be concluded in those conditions. Their best-known concept, the local average treatment effect (LATE), measures the effect of a measure on the group of people whose behaviour really changes because of it, rather than on the whole population. According to the Academy, that framework has radically changed how researchers approach empirical questions.

For a business owner or a family, the prize changes no rule, but it helps in reading economic information. When a measure (a minimum wage rise, a change in subsidies, a new regulation) is announced as having a given effect, it is worth asking who it was compared with, in which country and at what time.

What to bear in mind in Spain

In Spain, the national minimum wage is set every year by the government, after consulting the most representative trade unions and employers’ associations. It takes into account consumer prices, average national productivity, labour’s share of national income and the general economic situation (article 27 of the Workers’ Statute). Card’s studies do not say what should be done here, but they are a reminder that the effect of a rise depends on the sector, the area and the state of the labour market.

That is why it is worth distinguishing between the average result of a study and the case of each company: a minimum wage rise has more impact where more workers are close to that level. Ask your employment adviser how it applies to your case.

Natural experiments teach that, in economics, what seems obvious is worth checking against data.