Are the conclusions true? Before I address this issue, I want to observe that it is not clear that they are always intended to be true, that is, to correspond to the actual world. Rather, they sometimes represent a form of science fiction—an analysis of the action and interaction of ideally rational agents, who have never existed and never will. The analysis of ever-more-refined forms of strategic equilibria, for instance, is hardly motivated by a desire to explain or predict the behaviour of actual individuals. Rather, the motivation seems to be an aesthetic one. Two of the most accomplished equilibria theorists, Reinhart Selten and Ariel Rubinstein, have made it quite clear that they do not believe their models have anything to say about the real world. When addressing the workings of the latter, they use some variety of behavioural economics or bounded rationality. To cite another example, social choice theory—the axiomatic study of voting mechanisms—became at one point so mathematically convoluted and so obviously irrelevant to the study of actual politics that one of the most prominent journals in economics, Econometrica, imposed a moratorium on articles in this area.
An interesting question in the psychology and sociology of science is how many secret practitioners there are of economic science fiction—hiding either from themselves or from others the fact that this is indeed what they are practicing. Inventing ingenious mathematical models is a well-paid activity, but except for the likes of Selten and Rubinstein payment will be forthcoming only if the activity can also be claimed to be relevant; hence the incentive for either self-deception or deception. To raise this question might seem out of bounds for academic discourse, but I do not see why it should be. Beyond a certain point, academic norms of politeness ought to be discarded.
Jon Elster, Explaining Social Behavior: More Nuts and Bolts for the Social Sciences, Cambridge, 2007, p. 461