Dave Colander: Where Economics Went Wrong


Milton Friedman once predicted that advances in scientific economics would resolve debates about whether raising the minimum wage is good policy. Decades later, Friedman’s prediction has not come true. In Where Economics Went Wrong, David Colander and Craig Freedman argue that it never will. Why? Because economic policy, when done correctly, is an art and a craft. It is not, and cannot be, a science. The authors explain why classical liberal economists understood this essential difference, why modern economists abandoned it, and why now is the time for the profession to return to its classical liberal roots. Contending that the division between science and prescription needs to be restored, Where Economics Went Wrong makes the case for a more nuanced and self-aware policy analysis by economists.

Where Economics Went Wrong is a somewhat audacious title. Can you briefly tell us what’s wrong with economics?

Why have a firewall? The firewall discourages applied policy economists from trying to be too scientific, and economic scientists from worrying too much about policy implications of their work. The firewall is necessitated by the values inherently applied policy analysis. Scientific methodology isn’t designed to resolve differences in values. If a theorist is thinking about policy, the theory won’t be as creative as it can be. And if applied policy economics is too related to current theory, it won’t be as creative as it can be. Applied policy work requires that scientific methodology be integrated with more open and discursive engineering and philosophical methodologies that are designed to narrow differences in values and sensibilities and arrive at solutions to policy problems.

Our central argument is that scientific work and applied policy work are best done when there is a firewall between science and policy. Classical liberal economists had such a firewall, and we are calling for a return to Classical liberal methodology.

There are a lot of books out there criticizing economics; how does your critique differ?

The biggest difference is that we aren’t criticizing all of economics, but only one aspect of it—how economics relates theory to policy. We see ourselves as friendly critics, critiquing from the inside the economics profession, rather than from outside. In our view most of the outside critiques of economics miss their mark—they don’t convey the way top economists see themselves doing economics, which leads top economists to discount the critiques. Our critique is focused narrowly on economists’ blending of economic science and economic policy methodology.

 How does the subtitle of the book, Chicago’s Abandonment of Classical Liberalism, fit into your story?

Chicago is a useful case study for us because it was the last bastion of Classical liberalism in U.S. economics. It was Classical liberalism’s Alamo. Classical liberalism included both a methodology and a set of policy recommendations. The methodology involved keeping a firewall between economics science and policy for the protection of both science and policy. Classical liberals argued that if scientific researchers had policy views, those policy views would influence their science and their science would be tainted. If economists used scientific justifications for policy, which didn’t make clear that policy had to have a value component, policy would be tainted. It was a broad tent, not a narrow tent, methodology, and it reached its high point with the work of John Stuart Mill.

In the 1930s that changed; Classical liberalism was abandoned and was replaced with a new semi-scientific Pigovian welfare economics that blended science and policy into one field. Solutions to policy problems were to be found in better science, not in reasoned discourse.

The applied policy revolution started outside Chicago—at schools such as MIT and Harvard,and was quite pro-government interventionist. It seemed as if economic science was directing government to intervene in the economy. Chicago economists, led by Frank Knight, objected to both the change in methodology and the interventionist nature of the policy recommendations.

With the advent of the Chicago school of economics, the intellectual leadership of Chicago economics moved from Knight to Milton Friedman and George Stigler. They gave up Knight’s methodological fight, and concentrated on objecting to the interventionist nature of the new policy approach. They developed a pro-market scientific economic theory based on the Coase Theorem that led to the policy results they wanted. They presented it as a scientific alternative to the newly developed government interventionist scientific economics theory. In doing so they abandoned Classical liberal methodology, which held that science did not lead to policy recommendations. So the Chicago case study nicely highlights where economics went wrong.

What’s your solution to what’s wrong with economics?

Our solution is to bring back the firewall between science and policy. Using the Classical liberal approach, economic science includes only those aspects of economic reasoning and thinking that all economists agree can be scientifically determined. By design, there should be almost no debate about scientific economic theory. If there is serious debate about the theory, then the theory hasn’t reached the level of scientific theory; it is simply an hypothesis that needs further empirical study. Policy analysis uses economic science, but it also uses any other insights and analysis that the policy economist finds useful to arrive at policy conclusions.

The approach we are advocating for applied policy has much in common with engineering methodology. It is much looser and more open than scientific methodology. Engineering methodology is designed to solve problems, not to find truth. For an applied policy economist a scientific theory is simply a useful heuristics, to be used when useful. Engineering methodology specifically allows for the integration of values and does not present itself as infallible. It invites challenges and discursive exploration. Using an engineering methodology will make values in economics more transparent, and more subject to philosophical debate that can clear up some of the value and sensibility differences.

Can you be more explicit about how an engineering methodology differs from a scientific methodology?

Adopting an engineering methodology involves a change in how economists think about theory and policy. For an applied policy economist, theory becomes simply a useful heuristic.Debates about science are reduced enormously because the domain of economic science is reduced. In policy analysis a much broader pluralistic methodology is used. Scientific methodology is designed to discover truth, which means it must be very precise. Engineering methodology is designed to solve problems in the least cost fashion. It is far less precise because precision is costly.

How do you see such a change coming about?

Slowly, but surely. We see it more as an evolutionary change than revolutionary change. The change is already occurring. Many top economists are already following the Classical liberal methodology we advocate—they just don’t call it that. So one of the goals of the book is   to highlight their work and encourage young economists to use it as a role model. In the last chapter of the book we consider the work of six top economists who do quite different types of economics—they include theorists,empirical economists, and applied policy economists—who are all currently following what we call a classical liberal methodology. We show how that methodology influences the work they do and the interpretation they give to their work.

Our advice to other economists is to follow their lead. That means that:

  • in policy work, economists should be far less worried about carefully following scientific methodological guidelines; they should replace those scientific guidelines with educated common sense engineering guidelines designed to answer the type policy questions they are dealing with.
  • in theoretical work economists should stop worrying about relating theory to policy and let their imagination roam without concern about policy. They should go where few economists have gone before.
  • in blended theoretical and empirical work, economists should be more creative and less concerned about dotting i’s and crossing t’s. Leave that for the theoretical clean-up crew.
  • in econometric work, economists should use all the evidence that sheds light on the issue, not just the limited evidence that meets the profession’s current version of scientific rigor.

Our advice is for economists to free themselves from historically determined methodological scientific conventions and replace those conventions with pragmatic state-of-the-art conventions that take advantage of technological computational and analytic advances.

David Colander is Distinguished College Professor at Middlebury College. His many books include The Making of an Economist, Redux and Complexity and the Art of Public Policy (both Princeton). Craig Freedman is the author of Chicago Fundamentalism and In Search of the Two-Handed Economist.