McConnell, Brue, and Flynn's textbook, Macroeconomics, 19th edition, makes the claim, on p. 5, that economics is a science which uses the scientific method. This is often claimed and is part of the way economists legitimize themselves.
I dissent, however.
Some economic "laws," such as the Law of Demand (that people will demand less of a good when the price goes up, ceteris paribus), could stand up to pretty rigorous testing. Other key concepts, such as the concept of market equilibrium and the market clearing assumption, probably could not be formulated in such a way as to be vindicated through rigorous testing. There are even key concepts, such as the assumption that "markets clear," which are in a sense directly refuted by the facts: the supermarket shelves are never "clear," there are always a lot of goods waiting to be sold. Yet economists are justified in continuing to use the assumption for many purposes (although they may sometimes overuse and misinterpret it). Why?
Economics-- or so I would claim-- is not a science in the same sense that the natural sciences are. It is at a disadvantage in one way, but has a secret weapon to compensate. It is at a disadvantage because human beings are not deterministic: they have free will, they make choices, they cannot be predicted, at least not comprehensively and reliably. If you are philosophically committed to determinism, (I think you are wrong but) you can arrive at a similar insight by another way, namely that since we are human beings, we can never predict ourselves, because the information that we use as the basis for predicting our own behavior will also affect our behavior so as to falsify the prediction. If the world's greatest genius in predicting the stockmarket deduces that the Dow will rise by 500 points tomorrow, he may be right; but if he says so, he will be wrong, because his words will change what happens to the market. I cannot predict the invention of the wheel, because if I am able articulate what a wheel is, I have already invented it. Another way of putting the point is that human beings are too complex for science to predict them.
So much for the disadvantage. The secret weapon is that, since we are human beings, we know some things by introspection. If some utterly alien race were to observe the Earth, I do not see how it would be possible to know that human beings are purposive, or that resources are insufficient to satisfy their wants. Certainly, economists do not begin by setting aside (somehow) their introspective knowledge of rationality and scarcity, and then attempting to deduce it from the purely external kind of observation that natural scientists must rely on. They establish rationality and scarcity by appealing to introspection, and proceed from there. Theoretical models in economics have the character of tautologies: they make a set of assumptions, and deduce things from the assumptions. Sometimes this is a fault, in that the models are too abstract to be applied, but not always. It is possible to derive useful conclusions from mere logic starting with mere introspection, bringing in the real world only in the form of the loosest stylized facts, and making predictions which are precise only in the model and which when applied to the real world are too vague ever to be really tested in a rigorous way, letting mere experience leavened with common sense suffice to confirm the (rough) truth of the theory. Thus the Econ 101 supply and demand model, for example.