In this article, we examine the methodological presuppositions underlying comparative research into higher education. We begin by criticizing the familiar, external approaches to comparing nationally differentiated higher education systems, on the basis of structural-functional or system theory. We then argue in favour of the power-source analysis developed mainly by Michael Mann (1986), and from this basis offer a modified general model of social dynamics, with the aim of achieving a better insight into the methodologically critical areas of comparative educational research. For comparative social research, it is inadequate merely to model educational systems; we also need to locate historically and explicate the key events, the moments of “switching” (to borrow Max Weber's metaphor of the “switchmen of history”) when higher education, for instance, took its subsequent direction. What we propose as methods appropriate for a comparative, historically aware sociological analysis of higher education, drawn from the armoury of qualitative research, are research strategies under the “casing” label; but there is no sense in an exclusive polarization with quantitative research. The aim of the research approach outlined here is to generate theoretically articulated, but above all meaningful and useful descriptions of the reality of higher education. What we are particularly interested in is the theory of social power, from a methodological standpoint based on the principle of sufficient reason.