COMMENT: Evidence-based policy mistakes

It is important to recognise that data alone are not enough, common sense matters

COMMENT | KAUSHIK BASU | After years of stressing the importance of evidence-based policymaking, economists have clearly had some influence on politicians. What economists now need to do is to impress upon those same politicians that citing any evidence before adopting any policy is not evidence-based policymaking at all.

Turkish President Recep Tayyip Erdoğan has thrown around numbers to defend his decision to flood the Turkish economy with state-guaranteed credit. But the truth is that the policy was a politically motivated effort to win public support by engineering short-term growth (at the cost of driving inflation to a nine-year high of 12%).

Likewise, U.S. President Donald Trump cites simplistic trade-deficit figures to justify protectionist policies that win him support among a certain segment of the U.S. population. In reality, the evidence suggests that such policies will hurt the very people Trump claims to be protecting.

Now, the chair of Trump’s Council of Economic Advisers, Kevin Hassett, is attempting to defend Congressional Republicans’ effort to slash corporate taxes by claiming that, when developed countries have done so in the past, workers gained “well north of” $4,000 per year. Yet there is ample evidence that the benefits of such tax cuts accrue disproportionately to the rich, largely via companies buying back stock and shareholders earning higher dividends.

It is not clear whence Hassett is getting his data. But chances are that, at the very least, he is misinterpreting it. And he is far from alone in failing to reach accurate conclusions when assessing a given set of data.

Consider the oft-repeated refrain that, because there is evidence that virtually all jobs over the last decade were created by the private sector, the private sector must be the most effective job creator. At first glance, the logic might seem sound. But, on closer examination, the statement begs the question. Imagine a Soviet economist claiming that, because the government created virtually all jobs in the Soviet Union, the government must be the most effective job creator. To find the truth, one would need, at a minimum, data on who else tried to create jobs, and how.

Moreover, it is important to recognise that data alone are not enough to determine future expectations or policies. While there is certainly value in collecting data (via, for example, randomised control trials), there is also a need for deductive and inductive reasoning, guided by common sense – and not just on the part of experts. By dismissing the views and opinions of ordinary people, economists may miss out on crucial insights.

People’s everyday experiences provide huge amounts of potentially useful information. While a common-sense approach based on individual experience is not the most “scientific,” it should not be dismissed out of hand. A meteorologist might detect a coming storm by plugging data from myriad sources – atmospheric sensors, weather balloons, radar, and satellites – into complex computer models. But that doesn’t mean that the sight of gathering clouds in the sky is not also a legitimate sign that one might need an umbrella – even if the weather forecast promises sunshine.

Intuition and common sense have been critical to our evolution. After all, had humans not been able to draw reasonably accurate conclusions about the world through experience or observation, we wouldn’t have survived as a species.

The development of more systematic approaches to scientific inquiry has not diminished the need for such intuitive reasoning. In fact, there are important and not obvious truths that are best deduced using pure reason.

Consider the Pythagorean Theorem, which establishes the relation among the three sides of a right triangle. If all conclusions had to be reached by combing through large data sets, Pythagoras, who is believed to have devised the theorem’s first proof, would have had to measure a huge number of right triangles. In any case, critics would likely argue that he had looked at a biased sample, because all of the triangles examined were collected from the Mediterranean region.

Inductive reasoning, too, is vital to reach certain kinds of knowledge. We “know” that an apple will not remain suspended in mid-air, because we have seen so many objects fall. But such reasoning is not foolproof. As Bertrand Russell pointed out, “The man who has fed the chicken every day throughout its life at last wrings its neck instead, showing that more refined views as to the uniformity of nature would have been useful to the chicken.”

Of course, many policymakers – not just the likes of Erdoğan and Trump – make bad decisions not because of a misunderstanding of the evidence, but because they prefer to pursue politically expedient measures that benefit their benefactors or themselves. In such cases, exposing the inappropriateness of their supposed evidence may be the only option.

But, for the rest, the imperative must be to advocate for a more comprehensive approach, in which leaders use “reasoned intuition” to draw effective conclusions based on hard data. Only then will the age of effective evidence-based policymaking really begin.

*****

Kaushik Basu, former Chief Economist of the World Bank, is Professor of Economics at Cornell University and Nonresident Senior Fellow at the Brookings Institution.

Copyright: Project Syndicate, 2017.