Bryan Caplan over at EconLog is asking how normative economics should be. For those of you who may not know, normative means "pertaining to a norm" and is opposed to positive. In other words, normative statements tell you what should be which positive statements tell you what is. (I am flashing back to first day High School Econ class)
I for one believe that economics should never venture into normative statements because it then ceases to be economics. Strong opinion, I know, but hear me out: The social scientific method relies upon building testable, falsifiable theories which allows us to gain predictive power. How do normative theories measure to that standard?
A normative theory is effectively a positive theory with something added at the beginning and the end. Take for example the theory of comparative advantage. A very simplistic version states that unrestricted trade through specialization leads to greater prosperity. At the beginning, add "We should do things that lead to prosperity." and at the end add "Therefore we should do unrestricted trade." At best those statements are irrelevant, at worst, they make the theory untestable and therefore unscientific. If the actual conclusion is: "We should do unrestricted trade," the theory is untestable. You cannot take facts, compare them to that statement and determine whether those facts infirm or confirm the statement without bringing in the original statement that "We should do things that lead to prosperity." But, if you bring in that assumption, you are redefining the word "should" such that the actual conclusion is: "Unrestricted trade leads to prosperity" And at that point, you have not added anything with the normative statements.
Furthermore, when you bring in an assumption, you are opening up the field for its discussion. However, how can you discuss "We should do things that lead to prosperity" as an economist? The answer is, you can't. You are doing philosophy, politics, religion which are noble pursuits (or at least can be) but you are not doing economics.
The only valid answer as an economist when asked what should be done is to say: "Tell me what you want, and I will help you get it." But to pretend that economists know what should be done is simply preposterous.
32 minutes ago