In Friday's edition of NY Times, David Brooks wrote a very interesting column (link) discussing the state of economics. Although subtle and rigorous in its assertions I doubt that the field of economics needs a fundamental change in the philosophical origins of economic methodology. I agree with the author that economists often ignore the notion of moral philosophy and history in economic analysis but that doesn't mean that old textbooks on classical microeconomics need to be disposed. The author points out the role of economic forecasters who often appear on the TV, essentially trying to forecast economy's future path. Econometrics, which constitutes time-series forecasts on which most economic projections are based, is a real discipline to which numerous new scientific articles are devoted, published mostly in The Econometrics Journal and Econometrica. In his famous 1983 article (link) Let's take the con out of econometrics, Ed Leamer wrote how econometric modelling can be misused if a researchers ditch the true role and significance of assumptions. A prominent example is the study of death penalty on crime deterrence. As you can read in more detail in Leamer's article, one study found out that each capital punishment deters more than 9 murders while another one found out that each additional capital punishment causes more murders.
The contradicting evidence doesn't imply the falsification of the scientific method in economics. It merely reveals the hidden difference in how economists set assumptions regarding the behavior of individuals, firms and countries. David Brooks pointed out that economists failed to predict the recent financial crisis. However, many economists (myself included) pointed out the true dangers of an over-leveraged economy and monetary easing which led to subprime mortgage crisis and the consequential aftermath.
Many of us have had clear evidence, models and studies that showed how an over-leveraged financial sector can induce a significant economic downturn. However, many policymakers ignored the evidence of the behavior of the financial system which could be easily compared with chaos theory in mathematics. In my recent paper on Iceland's financial crisis I showed that the depository banks' overall leverage and indebtedness in the small country was growing exponentially, beyond the limit of capital adequacy.
As Paul Krugman recently noted, lessons from the Great Depression were not learned because people forgot it too quickly. Although mainstream economics, as every field within economic science, needs some major cures, I disagree with author's assertion that economics is not a real science but a moral philosophy. True, economics is not exact science because human behavior is not as experimental as particle analysis in physics but economics tries to resemble the scientific methodology through models, data, statistical inference and evidence. In fact, new interdisciplinary fields within economics are emerging such as behavioral economics, neuroeconomics and transport economics in which economic analysis is combined with other disciplines such as law, psychology, sociology and neuroscience.
The limits of mathematical recourse in economics were already discussed by economic thinkers such as John Maynard Keynes. And agreeably, mathematical reasoning is bounded by economic reasoning. However, from the real point of view, we need models to address human behavior mostly because, as F.A Hayek noted, it's too complex to capture its essence in a scientific entirety. If the neccesity of assumptions, models and evidence were not the case, hardly any lessons would be learned from the episodes of crises, booms, busts and periods of economic advancement in human history.