In his generous review of Political Order and Inequality, Francis Fukuyama writes that the book’s limitations come from its commitment to a rational choice framework that models human beings as self-interest individuals and that forgets the independent role of ideas. Fukuyama affirms, instead, that the human mind, “never the blank slate imagined by John Locke or contemporary neoclassical economists,” is both “filled with evolved structures that promote social cooperation … through faculties rooted in the emotions” and “hard-wired” for “kin selection and reciprocal altruism.”

As a matter of fact, the book acknowledges that we have always lived in families and bands and that “affective or emotional reasons, such as the ties that bind parents and children together” (p. 32) play an important role in human cooperation. It recognizes, too, the role of reciprocal altruism, particularly in informal social interactions. Indeed, altruism as well as other moral senses (such as fairness or duty) seem to be universal dispositions in humans (cf. J. Q. Wilson’s The Moral Sense [2003]). The problem, however, is that this universality in moral predispositions clashes with the fact that actual social cooperation is rather limited and fragile (particularly without formal institutions). My answer is that even if we may be inclined to cooperate, the extent to which we do is fundamentally shaped by the economic and political structures explored in Political Order and Inequality.

An alternative way to explain varying levels of human cooperation consists in asserting, first, that the extent of altruistic sentiments varies across human communities (Jonathan Haidt and Selin Kesebir, “Morality,” in Susan T. Fiske et al., Handbook of Social Psychology, 2010) and, second, that less developed societies have or had a weaker predisposition to altruism than modern ones (cf. Edward Banfield’s work on amoral familism and more recent work by economist Ernst Fehr and coauthors). In that alternative account, some critical historical moments or shocks would have arguably amplified the altruistic orientation of humans. Those shocks would have been of two kinds: material shocks (e.g., the agrarian revolution) that would have reshaped the predisposition to cooperate for the optimization reasons defended by functionalism and that I criticize in my book; and ideational shocks of the kind examined by Max Weber in his analysis of the economic consequences of Reformation and called for by Fukuyama.

Mankind has indeed experienced important ideational shifts. In the domain of religious ideas, for example, there have been momentous changes both in the object of worship (evolving from animism through polytheism to Karl Jaspers’s axial religions) and in the relationship between spiritual and temporal powers. But to claim that religious (or any other kind of) ideas matter, we need to establish two things: that they are exogenous and that they have a causal impact on our outcome of interest. Looking here (for space limitations) at two cases (religion in the transition from bands to tribes or lineages, and the emergence of axial religions) should make us skeptical about the causal role of ideas (as worldviews or principled beliefs).

Both bands and tribes worshipped ancestors and did not differ systematically in their beliefs about the existence of gods. Still, lineages probably worshipped a single ancestor (common to the lineage) while bands believed in many. Yet the exogeneity and the causality conditions are not met: The difference correlates with other fundamental material changes; it is unclear how the biogeographical parameters triggering those changes could lead to the religious shift directly; and such a minor religious transformation cannot account for their differences in social stratification. In turn, the creation of axial religions (between 700 and 300 b.c.) seems exogenous to nonideational factors, but it cannot be matched to any subsequent key political or economic transformation.