Mises on Ricardo Assumptions
To add to Reisman, more than fifty years ago, Ludwig von Mises recognized and carefully spelled out the limited applicability of Ricardo's law of comparative cost or advantage. In those conditions of free international movement of capital and labor in which the law does not apply (e.g., in the later nineteenth century and currently), Mises reached precisely the opposite conclusion of Paul Craig Roberts. Mises believed that monetary calculation in the form of differing rates of return on investment would direct capital to its most value productive uses extending and intensifying the global division of labor and raising living standards for all participating nations, just as the the unhampered movement of capital within the U.S. has enormously increased the living standards over the past two centuries for the citizens residing in all of its states. This is how Mises described Ricardo's theory in Human Action (Scholar's Edition, pp. 161-63):
The theorem of comparative cost...does not deal with value or with prices. It is an analytic judgment; the conclusion is implied in the two propositions that the technically movable factors of production differ with regard to their productivity in various places and are institutionally restricted in their mobility. The theorem...can disregard problems of valuation because it is free to resort to a set of simple assumptions. These are: that only two products are to be produced; that these are freely movable; that for the production of each of them two factors are required; that one of these factors (it may be either labor or capital goods) is identical in the production of both, while the other factor (a specific property of the soil) is different for each of the two processes; that the greater scarcity of the factor common to both processes determines the extent of the exploitation of the different factor. In the frame of these assumptions, which make it possible to establish substitution ratios between the expenditure of the common factor and the output, the theorem answers the question raised....With the law of comparative costs we compare the output of two different products. Such a comparison is feasible because we assume that for the production of each of them, apart from one specific factor [i.e., land], only nonspecific factors of the same kind [i.e. labor] are required.If we do not want to deal wit the law of comparative cost under the simplified assumptions applied by Ricardo, we must openly employ money calculation. We must not fall prey to the illusion that a comparison between the expenditure of factors of production of various kinds and of the output of products of various kinds can be achieved without the aid of money calculation....It has been asserted that Ricardo's law was valid only for his age and is of no avail for our time which offers other conditions. Ricardo saw the difference between domestic trade and foreign trade in differences in the mobility of capital and labor. If one assumes that capital, labor and products are movable, then there exists a difference between regional and interregional trade only as far as costs of transportation come into play. Then it is superfluous to develop a theory of international trade as distinguished from national trade...Ricardo, however, starts from the assumption that there is mobility of capital and labor only within each country, and not between various countries. He raises the question what the consequences of the free mobility of products must be under such conditions.... The theory of comparative cost answers this question. Now, Ricardo's assumptions by and large held good for his age. Later, in the course of the nineteenth century, conditions changed. The immobility of capital and labor gave way; international transfer of labor and capital became more and more common. Then came a reaction. Today  capital and labor are again restricted in their mobility. Reality again corresponds to Ricardian assumptions.