Evidence makes a comeback

January 19, 2021

People gather at the Washington Monument before marching on Constitution Avenue in the March for Science on Earth Day in 2017.Hilary Swift for The New York Times

By Paul Krugman

Opinion Columnist

“What do we want? Evidence-based policy. When do we want it? After peer review.” That was one of the chants at the March for Science, an international set of demonstrations in April 2017. The marchers were, of course, reacting in part to the election of Donald Trump and the rise of populism elsewhere; they were concerned that governments would make bad policy decisions that ignored scientific evidence.

And in the United States and the United Kingdom, where the demonstrations received the most attention, the marchers’ fears proved prescient. Both governments responded to the coronavirus pandemic by dismissing the views of epidemiologists, and the two countries have had far more Covid-19 deaths per million people than other major nations.

As an economist watching this disaster, I was mainly horrified. But I couldn’t help also thinking, “Welcome to my world.” My sense is that many medical experts are still shocked to see what should be matters of science utterly politicized. But in the social sciences it has been that way all along.

In fact, many people — and not just on the right — seem to assume that all economic analysis must be political. Today’s column was in part about the case for a higher minimum wage, a topic on which economists’ views have shifted dramatically over the past 25 years. When I describe that shift to lay people, I find them assuming that it must have been politically driven — a reaction to things like the Fight for 15, the movement by fast-food workers to demand higher wages.

Continue reading the main story

But that’s not what happened. The sea change in economic opinion on minimum wages was driven by, dare I say it, science: New evidence came in, and it refuted old conventional wisdom.

The change started with a remarkable paper by the labor economists David Card and Alan Krueger, who had the bright idea of surveying fast-food restaurants near the Delaware River before and after New Jersey raised its minimum wage, while Pennsylvania did not. As far as I can tell, they expected to see employment declines in the former relative to the latter. But they didn’t.

This result — no noticeable employment decline after an increase in the minimum wage — has since been replicated many, many times. The evidence is now overwhelming that minimum wage hikes don’t have major negative effects on employment, while they do raise workers’ incomes and reduce poverty. This isn’t a conclusion driven by politics, although at least some of the economists who still refuse to accept this evidence are being political.

While the minimum wage literature is a really striking example of economists being scientific, it’s not unique. Another example, close to my heart, has been a shift of the profession toward the Keynesian view that deficit spending is good for a depressed economy. This shift was driven in large part by what happened between 2010 and roughly 2013, when some but not all advanced countries were forced into harsh austerity policies — and experienced severe economic contractions, just as Keynes would have predicted.

Continue reading the main story

If these examples may make it seem as if taking the evidence into account always pushes economists to the left, that’s misleading. It may be true on average, if only because discourse in general has been dominated by a right-leaning orthodoxy, so that new evidence usually pushes us left. But it’s not hard to find contrary examples.

For example, many people would like to believe that universal health care saves money, because people get more preventive care and have less reason to visit emergency rooms. Unfortunately, that doesn’t seem to be true: Better coverage means that people get more health care, which costs money.

Which is not to say that we shouldn’t guarantee health care for all! For evidence can’t actually tell us what our policies should be — that is, in the end, a political decision that must reflect values as well as science.

But evidence can help us make policy. And I, for one, am thrilled that 23½ hours after this newsletter goes out we’ll have an administration that understands that.

Continue reading the main story

Quick Hits

These days, Arindrajit Dube is the go-to guy for minimum wage research.

David Card, who helped start this work, says that it “cost me a lot of friends.”

It’s not just minimum wages: economics has gone through an empirical revolution.

Coming next: the good things that happen when we help poor children.

Feedback
If you’re enjoying what you’re reading, please consider recommending it to friends. They can sign up here. If you want to share your thoughts on an item in this week’s newsletter or on the newsletter in general, please email me at krugman-newsletter@nytimes.com.