Comparative education as a field of study in universities (and "comparative education" as practised by nineteenth-century administrators of education in Canada, England, France and the USA) has always addressed the theme of "transfer": that is, the movement of educational ideas, principles and practices, and institutions and policies from one place to another. The first very explicit statement of this way of thinking about "comparative education" was offered in the early nineteenth century in France and was expressed in terms of the expectation that if comparative education used carefully collected data, it would become a science. Clearly--about 200 years later--a large number of systems of testing and ranking, based on the careful measurement of educational processes and product, have provided us with hard data and these data are being used within the expectation that successful transfer (of educational principles and policies and practices from one place to another) can now take place. A transferable technology exists. This article argues that this view--that "we" now have a successful science of transfer--ignores almost all of the complex thinking in the field of "academic comparative education" of the last 100 years; and that it is likely to take another couple of hundred years before it can approximate to being a science of successful social and educational predictions. However, what shapes the article is not this argument per se, but trying to see the ways in which the epistemology of the field of study (academic comparative education) is always embedded in the politics of both domestic educational reform and international political relations -- to the point where research in the field, manifestly increasingly "objective" is also de facto increasingly "political". The article is about the "how" and "why" of that and what has been forgotten and what has not yet been noticed.