Correlation is not effectiveness. There is an important truth behind this clich. January of this year, for example, had the most severe lockdown in the UK and had the highest death rate from Covid. There were no deaths and some restrictions in New Zealand. Still, no matter what your favorite YouTube conspirator says, the lockdown doesn’t create waves of cowardice. The reason for the lockdown in Coved.
But while “reciprocity is not effectiveness” is an important warning, when policymakers come to ask questions, it doesn’t respond very well.
For example: Why are more educated people leaning towards higher incomes? Is education the cause of high income, or are smart, enterprising people thriving in both school and the workplace?
Why are there so many workers of foreign descent in prosperous places? Is it because immigrants increase income or where people go for money?
Lots of kids in stark places. Is it because Stark gives birth to a baby or is there room for both the big races?
The example of Stork and Kids is part of a cautionary tale, as I explained in my book, How to add up the world. In 1965, Darrell Huff, a well-known statistics communicator, told a U.S. Senate that the correlation between smoking and cancer was exactly between Stork and children. This is a deadly example of how health doubts can easily prevent slander.
All of this explains why I got excited after the announcement of the Nobel Memorial Prize in Economics on Monday. The award winners, David Card, Joshua Angrist and Guido Imbes, led the charge in what became known as the “Credibility Revolution” in the economy.
Faced with chaotic real-world information, economists ask, “Does education increase income?” It is tempting to turn away from such an important question. And “Do Immigrants Increase Productivity?” Card, Angrist and Imbens professions have shown that we can be more ambitious.
In 1992, New Jersey raised its minimum wage from 4.25 to $ 5.05 per hour. Can some fast-food workers be too expensive to hire? Card and Alan Krueger saw a natural test: East Pennsylvania sat next to New Jersey with the same economy, but Pennsylvania did not change its minimum wage. Card and Kruger compared employment in New Jersey and East Pennsylvania, and found no signs of a fast-food job leaving while raising the minimum wage in New Jersey.
It was a highly influential search, but perhaps the most important part of it was not the results, but the demonstration that economists can find information to answer serious policy questions.
Angrist and Kruger tackled the question of education income by observing the whims of the education system in the United States. Consider two children, one born in late December and the other a few weeks later in early January. The child in December started school a full year ago. However, both children can legally drop out of school on their 16th birthday, a few weeks apart. The difference may seem trivial, but in 1991 Angrist and Krueger showed that January children spend very little time in school and earn very little.
Of course some children drop out of school at the age of 16; Most do not. This is the case with natural tests: instead of randomly allocating drugs and placebo, natural tests randomly allocate something vaguely, such as the chance to leave school soon.
It’s a statistical headache, but Imbens, along with Angrist, created a toolkit to help researchers identify genuine causal relationships from obscure natural experiments. Economics has become a field full of clever empirical results, and most of them stand on Angrist-Imbens foundations.
This year’s Nobel Prize is bitter and sweet. It commemorates Alan Kruger’s suicide in 2019-2019.
This is an example of the gap between political rhetoric and the best information intelligence work. For example, one of the most influential papers on the card is the most talked about topic in British politics today: Can you raise wages by restricting immigration? Prime Minister Boris Johnson has said he can and he will.
The data suggest a different story. Card Mariel Botlift studied the deportation of 125,000 people from Cuba to the United States in 1250. Most of those people came and stayed in Miami and most were relatively incompetent.
While Miami’s unskilled workers increased nearly 20 percent over the months, the card found no sign of frustration with unskilled wages in Miami. Instead of using labor arrivals to reduce wages, Miami businesses have found ways to hire these new workers.
This is not just a study, but the work of the card prompted economists to reconsider simple models of immigration. The balance of evidence now suggests that immigrants are more likely to increase productivity than suppress it.
The world is full of interesting information, but it is not full of rigorously controlled experiments. It is very easy to cherish treacherous statistics to argue that lockdown is the cause of covid. But dismissing the evidence altogether is not very good, reassuring people that cigarettes are probably safe because intercourse is not the cause.
We can do better. As Kruger once said: “The idea of turning economics into a truly experimental science, where the basic theories can be rejected, is a big, revolutionary idea.”
That’s right. It is really possible to turn statistics into insights. And we have to try.
Tim Harford’s “The next fifty things that modern economics has created”Now out of paperback
Follow FTMag Learn about our latest stories first on Twitter