If You’re a Paternalist, Why Aren’t You an Epistocrat Then?
Paternalism is often thought of in the context of consumption behavior in the marketplace. Because people are misinformed and biased, they make choices that are detrimental to their own welfare. For some, this justifies that the state takes coercive or milder measures (taxes, nudges) to steer people toward welfare-improving choices. If you’re among those proponents of paternalist interferences, then consistency entails that you should also be a “political paternalist” and therefore endorse some form or another of epistocracy in politics. This is in substance the claim made by Jason Brennan and Christopher Freiman in a recent paper just published in the Journal of Ethics and Social Philosophy.
Brennan and Freiman’s argument in support of this claim is straightforward. It builds on postulated symmetry between consumption behavior on the market and voting behavior in politics. A now impressive number of studies in many disciplines document a long list of biases that affect people’s decisions: framing effects, present bias, availability bias, … They affect consumer’s decisions in a way that has long been recognized to justify openly paternalistic laws such as bans on some products and/or their promotion. More recently, philosophers and social scientists have defended “soft” forms of paternalism which, though not coercive, are nonetheless though effective at influencing individuals’ behavior for their own good. Now, studies also indicate that the same biases operate in the political domain. Actually, their effects may even be more dramatic here:
“[T]he assumption of voter competence is even more doubtful than the assumption of consumer competence. A priori, we would expect that every flaw in consumers to be worse in voters because the expected cost of an uniformed and biased consumption choice is higher than an uninformed and biased voting choice. A consumer bears most of the cost of their decision to smoke. But unlike consumers, voters never have unilateral decision-making power. Their votes are thrown in with everyone else’s.” (p. 337)
It follows therefore that if one makes the symmetry, one should agree that there are at least as strong reasons to be a political paternalist than a consumption paternalist. And political paternalism probably entails epistocracy. Paternalists should for instance support epistocratic institutions such as “government by simulated oracle” consisting of inferring people’s “true” political preferences, i.e., the political preferences they would have revealed through their votes have they been perfectly informed and unbiased, controlling for demographics and other factors. This highlights that epistocracy can be defended not only based on the adverse “external” effects that political incompetence produces on everyone, but also on the fact that voters tend to “self-harm”.
The authors consider several objections to their claim. All of them dispute one way or another the consumption/vote symmetry:
Objection #1: Individual consumer choices are efficacious, not individual votes.
Objection #2: In the political domain, individual incompetence is overcome by the “wisdom of crowds”.
Objection #3: A mechanism like epistocratic vetoes would prevent benefits for some, thereby harming a few to benefit the many.
Objection #4: The right to vote is a basic political liberty that must be protected, as long as it does not conflict with other basic political liberties.
Objection #5: Democracy has intrinsic value.
Objection #6: Voters’ incompetence can be partially solved by better education and better institutions disseminating the relevant information.
Objection #7: Paternalist regulation of vote is more vulnerable to rent-seeking and other forms of self-interest captures.
Objections 1, 4, and 6 are probably the more sensible. Regarding the first, Brennan and Freiman rightly reply that if we have reasons to believe that a group of individuals will concomitantly make decisions that, put together, will have adverse effects on each’s welfare, then someone who endorses economic paternalism would properly agree that we have reason to interfere. They also convincingly respond to objection #4 framed in a Rawlsian way. Basically, there is no reason to consider that the right to vote is needed for people for developing and exercising their moral powers. This is related to what I would call the “vote-based reductionist misconception of politics”, that is the mythical idea that political expression and participation finds its ultimate form in the right to vote and the act of voting. If we agree that political participation and expression have value (an assumption that might be disputed), then it should be realized that there are many ways to realize this value beyond the act of vote. Actually, these ways (e.g., militantism, education, production and diffusion of ideas, deliberation in public fora, making “ethical” consumption choices…) are probably far more effective than casting a ballot in influencing behavior in the society. Objection 6 is important and indeed an epistocrat should also endorse the need for better education and information for citizens. But this objection is as a rule not available to paternalists, who are generally skeptical of the effectiveness of this approach to remedying consumers’ rationality failures.
I think the authors have overlooked one more objection that paternalists could endorse. Paternalists generally argue that their objective is “to make people better-off as judged by themselves”. In other words, paternalist interferences are directed at the means that individuals are using to satisfy their ends, and do not aim at changing these ends. It can be argued on the other hand that politics, and voting behavior in particular, is necessarily enmeshed with values: you do not vote for a candidate or a political program as means to realize your values; you’re expressing your values through the act of vote, notably because in a democracy politics is foremost the fact of making collective choices about values. In this sense, paternalists are more technocrats than epistocrats because they endorse the means/ends and the related facts/values dichotomies. By the way, some critical commentators have pointed out the key role played by behavioral sciences in modern “technocratic democracies”. Political paternalism would be quite the contrary a paternalism about ends and values that is antinomic to the democratic ideal. Going further, it could be argued that while it is possible to assess the truth value of “technical” judgments in the domain of means, truth has no bearing in the political domain of value judgments, a claim that Rawlsians among others would make.
As it happens, I think that while distinguishing means and ends (as well as facts and values) is useful as a heuristic, it does not have the moral and political significance that is often postulated. In practice, paternalists’ claims are often based on an implicit conception of the good and a normative theory of prudence. Moreover, the dichotomy assumes that people have underlying “authentic” and “inner” preferences that ought to be realized, an assumption that is problematic[1] and may have dire normative implications.[2] The fact that ends and means are not strictly separated goes both ways: facts are “value-ladened”, even when they are studied based on scientific methods; but scientific methods and other devices of knowledge production help to form “enlightened” value judgments. Paternalism about consumption and paternalism about vote are thus closer than one would expect based on the means/ends dichotomy.
This leads me to a last remark. I’m not certain that “political paternalism” is the best characterization of epistocracy. It is also not its best defense, just because there are many objections to paternalism. But the analogy hints at an important point: the defense of paternalism cannot obviously be made on purely “matter-of-fact” grounds. It calls for both teleological and deontological forms of justification. The former because it is impossible to justify paternalistic interferences without some conception of the good; the latter because considerations of individual rights cannot be sidestepped, at least in liberal societies. The same is true for epistocracy. In particular, the case for epistocracy must be made based on a certain conception of the good that, I conjecture, includes a perfectionist component. It must also engage with a thorough consideration of what the ideal of political equality means in a liberal society.
[1] Incidentally, that affects the normative relevance of the epistocratic mechanism of “government by simulated oracle” mentioned above.
[2] Isaiah Berlin’s critical take on the concept of positive liberty in his essay “Two Concepts of Liberty” for instance largely builds on the myth of the “authentic self”.