Economics/Tutorials

Some footnotes for economists

Economics as a science

Whether economics is a science is a question that would appear to a layman to economics to be esoteric and semantic, but it is of considerable interest to its professionals. It was claimed to be a science by John Stuart Mill, - a claim that was later challenged by Lionel Robbins and critically examined by Paul Samuelson . It has been pointed out that many economic theorems do not satisfy Karl Popper’s criterion that, for a proposition to be considered scientific, it must be falsifiable . The economic theory of utility, for example, yields tools that can be used in economic analysis rather than testable propositions. However, such tools can be used in the scientific analysis of economic activity, just as Euclid’s theorems can be used in the scientific analysis of physical phenomena. To the extent that the analysis yields testable propositions, it can nevertheless be considered scientific. But critics have claimed that the propositions of theoretical economics are seldom put to the test, and that even when the are, academic economists are apt to continue to put forward propositions that fail (The survival of the Hecksher-Ohlin theorem in face of Leontief's Paradox has been cited as an example). Arthur Pigou has been quoted as saying that although abstract arguments cannot demonstrate what is empirically true, they can often reveal what is unproven or wrong.

Economic theory as a language

Milton Friedman considers economic theory to be a mixture of two components, one part being a language and the other a collection of hypotheses. He describes the language component as “a set of tautologies” whose function is to serve as a filing system for organising empirical material and helping to understand it. That part is to be judged by its by its logical consistency and completeness, and by its ability to establish precise categories that have meaningful empirical counterparts. The article in utility provides an example of the tautological nature of some aspects of economic theory.

Normative economics

Normative economics is commonly defined as the study of “what ought to be” in contrast to the study of “what is” . The term was put forward in a book by John Neville Keynes and drawn to the attention of the economics profession by Milton Friedman. In most people’s minds there is a normative background to many branches of economics. The results of research on poverty, inequality and unemployment, for example, are usually presented in the tacit or explicit context of their undesireability. That has led to attempts to classify certain categories of economic theory as normative (welfare economics is a common example) and to divide the body of economic theory into normative and positive categories. But the methodology of economics does not require any normative postulate: it is left to the user to attach one if necessary. The fact that it may sometimes be necessary to attach some normative rules to a theory in order to reach a practical conclusion does not make the theory itself normative. (The need for normative interpretations of economic theory is discussed further on the tutorials subpage of the article on welfare economics.)

Economic theory as a language

Milton Friedman considers economic theory to be a mixture of two components, one part being a language and the other a collection of hypotheses. He describes the language component as “a set of tautologies” whose function is to serve as a filing system for organising empirical material and helping to understand it. That part is to be judged by its by its logical consistency and completeness, and by its ability to establish precise categories that have meaningful empirical counterparts.

Positive economics

Friedman describes the second part as “ a body of substantive hypotheses designed to abstract essential features of complex reality” Controversially, he claims that those hypotheses should be judged, not by the realism of their postulates, but by their ability to produce correct forecasts. He even claims that it will be found that “the more significant the theory, the more unrealistic the assumptions” on the grounds of the importance of being able to exclude factor that turn out to be irrelevant. Testing is difficult, however, not just because of the difficulty of conducting controlled experiments but also for other reasons including the possibility that other hypotheses might also succeed. Friedman comments that the difficulty of testing has fostered a “retreat into purely formal or tautological analysis".