The idea originates in Milton Friedman’s essay “The Methodology of Positive Economics” (1953), the opening chapter of his book Essays in Positive Economics (University of Chicago Press, 1953, pp. 3–43).
Basically,
- Friedman’s central idea is, a theory should be judged by the accuracy of its predictions within its intended domain — not by whether its assumptions are realistic or descriptively accurate. And further — “the more significant the theory, the more unrealistic the assumptions.”
- So we should use the “as if” method. Model agents as if the assumptions held. Three examples: the expert billiard player (plays as if solving the physics), leaves on a tree (arranged as if maximizing sunlight), the profit-maximizing firm (behaves as if optimizing; competition weeds out those that don’t).
- Tie-breakers. When theories predict equally well, prefer greater simplicity and fruitfulness (scope, precision, new research).
Friedman’s original essay is ambiguous, contested but never settled. I printed it as part of my Price Theory I readings and it now lies with my Bach sonata sheets…
“Problems of Methodology — Discussion,” American Economic Review 53(2), 1963, pp. 231–36.
Samuelson’s F-Twist basically refuted Friedman’s idea. The “twist” is a logical slide. Formally, a theory’s assumptions (A) and consequences (C) are logically equivalent, so you can’t say the truth of predictions matters while the truth of assumptions doesn’t — false assumptions imply false conclusions.