Over at EconLog, Bryan Caplan asks, “How normative should economics be?”
Since it is an impossibility to separate economics from normative judgement, I would have to say that economics must seek to be carefully normative about both individual behavior and government policy — but use special care in informing individual behavior. To the extent that economics can inform the individual about how to maximize expected utility, economics as a science must keep in mind individual preferences (which are not always utility-maximizing) and liberty.
For many outside of economics, there can be a confusion between scientific analysis and normative judgment. Indeed, Amazon.com is inundated by books that claim the pillar of scientific analysis, which are simply using stylized numbers to make brash normative statements. This problem arises from the fact that economics is unable to perform controlled experiments, and thus must rely on often noisy real-world data. This does not discount methodology, nor the importance of robust models that provide accurate depictions of the real world — it just makes the task much more difficult. Unfortunately, economics never be can never divorced from normative judgement. In this sense, economics must use care in presenting data in a way that makes the correct (implied) normative statement about the given subject.
- The Scope and Method of Political Economy, by John Neville Keynes (Particularly chapter 2)
- The Methodology of Positive Economics, by Milton Friedman
Although, I do not agree with Friedman’s conclusion that what matters the accuracy of prediction of economic models, regardless of the assumptions made. For instance, one could have a model of water freezing at 32 degrees Fahrenheit, with the causal assumption that at 32 degrees, nano-elves come and replace water with ice. The implications would be borne out in the real world — water does freeze at 32 degrees — but the assumption is, of course, outlandish.