The Nature of Economic Theories

An economic theory is an explanation of how the economy, or some part of the economy, works. Such explanation can be produced in two ways: by induction and by deduction.

Induction

Think of induction as the business of "going and seeing". It is the attempt to work out what is happening by carrying out very careful observations. An example might help to illustrate this.
Suppose you are trying to produce a theory of consumer behaviour. Your aim is to explain why consumers behave in the ways they do. After much careful study you might decide that, in general and perhaps with some exceptions, consumers try to get the best value they can for the money they spend. This would lead to certain ideas about how they would behave in particular circumstances. You could then examine some of those particular circumstances were borne out. After a while you might come up with a fully worked out theory (or explanation) of consumer behaviour.
The problem with this inductive method is that it is difficult to be sure that the general conclusion you have reached is really correct. How can you make a general statement, which is supposed to apply to everything, when all you have done is to examine a number of particular cases? Strictly speaking, of course, you cannot. It could be that, for instance, you have selected non-representative groups. Or for some reason the groups selected might have been behaving slightly oddly at the times when you were observing them. For reasons such as this, you can never be 100 per cent certain that what you have concluded is wholly correct.
This is sometimes referred to as "the problem of induction". All the available evidence may support a general theory but it can never be said to prove it beyond all possible doubt. Inductive theories must always be subject to some element of uncertainly. Nevertheless, people do derive conclusions in that way, and act on them, all the time. Economists are no exception. They make the best observations they can and try to draw the most sensible conclusions possible. From time to time, as happens with other disciplines, it may be found that have made a mistake. In that case, the theory must be revised and a new and better one produced. But that theory in turn must always be at least a little open to doubt. It is always possible that, some day, someone may come up with some additional piece of evidence or some new situation in which the theory does not apply. Inductive theories must always be tentative. They can only ever be the best explanation we have - so far.

Deduction

The alternative method of deduction is best thought of as "sitting and thinking" rahter than "going and seing". What you do is to assume certain things to be the case. Then you try to work out what would follow from being so.
Suppose for instance that you wanted to work out a theory of producer behaviour. You might, perhaps, start by supposing that producers were in business to maximize their profits. On that masis, you could work out what they would do in certain specified conditions. After quite a lot of this kind of thinking, a theory of producer behaviour could be worked out in detail.
Again there is a problem. yo can never be sure that your original assumption; but the assumption would have been wrong in the first place.
Before you leap to the conclusion that the solution to this difficulty is to make correct assumptions in the first place, just think of how you would have to do that. You would have to find out which assumption would be the correct one. In other words you would have to use induction; you would have to go and see. You have already seen the difficulties that can involve.