21 July 2010

Darwinian liberalism

I'm enjoying the conversation at CATO Unbound about Larry Arnhart's ideas about the connection between Darwinism and liberalism. It seems to me that Arnhart is on to something, but that he is not presenting the argument very well, and thus the other contributors (PZ Myers, Lionel Tiger, and Herbert Gintis) have failed to appreciate it properly. Here is how I would reframe it (and perhaps change it’s conclusion a bit?):

Given the relatively unchanging human nature, various types of social order (i.e. different sets of formal+informal institutions) are not equally conducive of a particular social goal X; i.e. one cannot manufacture the “ideal human nature” for advancing social goal X, but instead has to take human nature as it is and search for institutional reforms that lead to X. Whether or not social goal X is better achieved by institutions A or B is an objective fact (although sometimes, due to the complexities involved, it is difficult to demonstrate A or B). This objectivity stems from human nature’s lack of malleability, as Arnhart argues and the others seem to agree.

The problem now is that there are many possible social goals that people may consider, and there's no consensus about which are “really” more important than others. This is where Arnhart’s weak point in the argument seems to be. He seems to be saying simply that, given human nature as it is, certain worthy social goals (e.g. welfare) are best achieved via classical liberal institutions. However, this isn’t enough to conclude that liberalism is a necessary consequence of human nature because other people may rather focus on different social goals (e.g. more equality, respecting tradition etc.). Various social goals are in conflict and there’s no objective way in which we can decide how to make the trade-offs between them. All we can hope for is a negotiation process among various people with different views, i.e. politics.

There is thus a problem if one wants to derive a normative position from the Darwinian theory of human nature. However, one doesn’t need to push for a normative position, and the above relativist argument actually fails to push Darwinism to its natural limits. There is a particular social “goal” that is objective, in the sense that it is derived from an application of the weak anthropic principle to the evolution of institutions: present institutions are here because they have survived (so far) in the competition with other institutions. In other words, any social goal X that one might favor has to obey a higher Darwinian “goal”: the institutions conducive of goal X have to be able to survive in the competition with other institutions.

Institutions survive and spread due to three possible reasons:
1) they further the average interests of individuals (classical sociological individualist functionalism and economics, e.g. see Mises for a rather extreme idealization of this position);
2) they further the power of the group (classical sociological group functionalism, and, notably, Hayek, David Sloan Wilson);
3) they simply spread, regardless of whether they actually bring any individual or group benefits, because people find them intuitive due to their evolved social instincts (Cosmides & Tooby, Sperber, Richerson & Boyd, Boyer, Atran).

To put it differently, institutions spread either because they are imitated from one society to another (and they’re imitated because of 1 or 3) or because a society conquers another and imposes new institutions to it (i.e. 2, a task somewhat difficult but not unseen).

Thus, from a purely descriptive point of view, asking what follows from the existence of a universal and inflexible human nature means asking: what ensemble of institutions furthers average individual interest better (the average being hypothetically computed by weighting the interest of each individual according to its power and influence within the group), makes society more powerful relative to other societies (such that it resists being conquered or it succeeds in conquering them), and is relatively in-sync with our evolved social and moral intuitions?

So, how does classical liberalism fare this test? Is it really the kind of institutional system we would expect to see spreading and conquering all others? First of all, it has a real intuition problem (especially because it conflicts with our intuitions about equality, bad luck, and strangers). Arnhart acknowledges this, which is why he sides with a liberal-conservative fusion, rather than with pure libertarianism. This intuition problem is compensated to some extent by the fact that a) liberalism increases individual welfare faster than any other system, b) it diminishes conflicts with other societies due to free trade, albeit c) greater wealth also translates into greater military power. Thus, at least prima facie, it seems that what we should expect to see spreading is basically the US-EU institutional model – i.e. a somewhat protectionist social-market economy with a strong military. And of course, this, rather than classical liberalism, is precisely the system that’s spreading (to China etc.).*

* my own liberal/libertarian normative beliefs are not exactly along the lines of this conclusion, but, heck, what can I do?