Yesterday we kicked off the fall 2019 Tax Policy Colloquium, our 25th "season," with my colleague and co-convenor Lily Batchelder's paper Optimal Tax Theory as a Theory of Distributive Justice.
This is an illuminating paper that reviews and elucidates relevant literature, while also convincingly supporting a perhaps counter-intuitive conclusion, to the effect that, despite fervent denials by both Ronald Dworkin, as a proponent of "resource egalitarianism," and prominent welfarists (such as Louis Kaplow) that Dworkin's theory has much in common with egalitarian welfarist theories, in fact they substantially converge in their conclusions, even if reaching them by different means and based on different concerns.
The paper also discusses the longstanding tax literature sub-theme (which I may have helped launch in legal academic circles) regarding endowment taxation, the conundrum regarding "slavery of the talented" and the hypothetical beachcomber who could have been an investment banker, etc. Its conclusions in this area are certainly compatible with mine, although more thoroughly worked out.
Dworkin's theory, which I have always found more of an impressive effort than actually persuasive, has the following main premises (quoting Lily's paper): It aims to "treat all members of society with equal concern and respect, as determined from a position of equality. this has two components. First, everyone should begin their life with equal transferrable resources, as measured by the value of those resources to others. Second, the state should provide insurance against inequalities arising from non-transferrable resources, such as one's talents and health. But it should only insure against the effects [of] brute luck - not option luck, which is the result of deliberate and calculated gambles."
Egalitarian welfarism is the same as utilitarianism, except that it assigns greater social welfare weight to the utility of worse-off individuals. Both it and utilitarianism are sometimes viewed as compatible in theory with imposing an endowment or ability tax, such that the above beachcomber who could have been an investment banker is viewed as no worse-off (and hence, if you add a string of unrealistic assumptions, as rightly owing just as much tax) as her peer who actually went the investment banker route. But in the welfare framework what matters is utility - endowment only matters instrumentally since in a particular reasoning chain one might think (all else equal) that, the higher your potential earnings, the higher your total and marginal utility. We're far from the world in which such assumptions would hold, in multiple dimensions.
There are a few main reasons why the two sets of theories converge to a degree. Both may support addressing brute luck, since as defined this isn't the individual's fault or responsibility (Dworkin) and in the absence of choice won't have adverse incentive effects (welfare egalitarianism). In both one has to be leery of addressing option luck, in Dworkin as a matter of principle and in welfare egalitarianism merely insofar as incentive effects might lead to adverse efficiency consequences. And once Dworkin started pursuing in earnest his analysis of insurance against inequalities pertaining to non-transferrable resources, he inevitably found himself in territory where welfare-type considerations were going to matter (e.g., moral hazard et al would now affect the character of the insurance scheme people would hypothetically want).
Here are two quick thoughts I'd add regarding the underlying theories that the paper discusses:
1) I have always found the case for egalitarian welfarism to be a bit askew. That is, I'm sympathetic to the motivation, but find it peculiarly expressed. It's a very odd set-up, and inconsistent with how one would think about distribution issues behind Harsanyi's version of the veil of ignorance, in which one would want to maximize expected utility. In effect, the social planner who is making a social welfare assessment has declining marginal utility with regard to people's utility - a very abstract set-up that lies in a land far beyond where our moral intuitions can reasonably take us, one way or the other.
What I see as motivating it is the judgment, which I share, that focusing just on the declining marginal utility of consumption and leisure to a particular individual falls far short of explaining fully why - to some of us, at least - inequality appears to be a bad thing. As I discussed in my recent paper, The Mapmaker's Dilemma in Evaluating High-End Inequality, the standard economic assumption that people's utility is solely a product of their consumption of market goods plus leisure, may work well enough in price theory, but not here.
To quote that paper:“We are an intensely social species, and often a rivalrous one, prone to
measuring ourselves in terms of others, and often directly against others.
People thus “have deep-seated psychological responses to inequality and social
hierarchy,’ creating the potential for extreme wealth differences to “invoke
feelings of superiority and inferiority, dominance and subordination” that
powerfully “affect the way we relate to and treat each other."
Case in point, the Wilkinson-Pickett data suggesting that inequality “does
not just reduce happiness for all groups—the rich as well as the poor—but even
has measurable adverse effects on social trust, economic mobility, life
expectancy, infant mortality, children’s educational performance, teenage
births, homicides and other violence, imprisonment rates, mental illness, drug
and alcohol addiction, and obesity."
Anyway, back to the weighted welfarists. My thought is that perhaps their recognizing intuitively the inadequacy of the standard story in which it's all DMU and nothing else (based on the radically impoverished utility function that was used for reasons of tractability), and that inequality was also bad for reasons not included in the model, led to the ad hoc solution of weighting towards the bottom, based on the observer's arbitrarily posited inequality aversion. But however one thinks otherwise about utilitarianism's adequacy as a moral theory (given all the standard challenges and conundra it raises), I would say that is not really the best response to this particular problem. Rather, one should broaden the relevant inputs to posited utility, however indeterminate that makes the analysis, keeping in mind that determinacy within the boundaries of a little toy model doesn't really do or mean much anyway.
2) Turning to Dworkin, and recognizing that there's a rich literature here that I don't know well, I've always been troubled by his use of the brute luck versus option luck distinction, despite its being related to issues of definite interest. The distinction between bad outcomes that don't versus do reflect choice matters, even purely in a welfare framework, because (a) the individual's choice may shed light on her underlying preferences or utility function, and (b) the less chosen, the less we have to worry about incentive issues. But I don't join Dworkin in wanting to moralize the distinction to the degree that he does, because I've long found beneficence to be a more attractive principle than desert, "you've made your bed now lie in it," etcetera. I also wonder about such problems as (a) the brute luck of being a bad chooser with respect to option luck, and (b) the brute luck aspect of having been handed particular choices on which to exercise option luck. E.g., assuming a counter-party I can bet as I like, and as high or low as I like, on tonight's US Open matches. But when I choose a profession, activity, etc., I didn't choose the payoff set that I have. Maybe I'd have preferred to have had different and better choices and payoffs. And I suspect this causes difficulties for the distinction that others must have explored.