Marshall begins Chapter 3 (economic generalizations or laws) with the basic steps of science, i.e., collecting data, exploring interdependencies, and using both deductive and inductive methods to “discover the relations between cause and effect” [p 24]. His reference to inductive (looking for patterns in real data) caught my eye, since mathematical economists use deductive methods (deriving relations and testable hypotheses from basic axioms), and referees dislike my inductive methods. (We gave up on that paper rather then going out a third time for data.) Marshall then says different methods complement each other, just as new facts complement ongoing analysis of existing facts.
§2. Marshall explains that “science” in economics is not due to precision but an ongoing effort to test hypotheses, reject those that fail, and classify those that persist and predict as “laws” that anyone can use.
§3. Marshall contrasts exact laws (e.g., gravitation) with probabilistic laws (e.g., tides) to put economic laws in context [pp 26-27]:
The laws of economics are to be compared with the laws of the tides, rather than with the simple and exact law of gravitation. For the actions of men are so various and uncertain, that the best statement of tendencies, which we can make in a science of human conduct, must needs be inexact and faulty. This might be urged as a reason against making any statements at all on the subject; but that would be almost to abandon life. Life is human conduct, and the thoughts and emotions that grow up around it… And since we must form to ourselves some notions of the tendencies of human action, our choice is between forming those notions carelessly and forming them carefully.
I agree that we want to understand regularities in human behavior without dismissing or undervaluing the lessons of exceptions.
§4. Putting laws on an “exactness” continuum, Marshall compares social laws (“a certain course of action may be expected under certain conditions from the members of a social group”) to a narrower set of economic laws (“conduct in which the strength of the motives chiefly concerned can be measured by a money price”). Marshall says they are not always easy to separate but that prices can mean economic laws are more precise [p 27].
Marshall clarifies his use of “law” in the sense of regularity rather than legal exactness, adding that these regularities are called “norms” in terms of frequency rather than morality. Thus, it might be “normal” for an egg to cost one penny for most of the year but 3p (“thruppence“) when hens are tired. Likewise, “normal” may include competition, cooperation, or some mix of the two. The goal is a decent set of predictions — not rules for the good life nor a rationalization for “normal” suffering.
§5. Economic laws depend on simplifying assumptions (e.g., “holding all else equal” or “for the average man”). Abstract laws have more assumptions whereas applied cases depend on specifics to be more exact.
I recommend the ideas in this chapter to anyone introducing economic ideas to lay people. Economic insights can be useful, but they are not always right.