Limitarianism: a philosophical dispute

by Ingrid Robeyns on February 21, 2022

In my last post, on the public debate on limitarianism, I responded to Matt’s philosophical doubts about limitarianism by saying that there was a debate on this matter forthcoming in The Journal of Political Philosophy, and that I would post it once published in Early View. So, here it is.

Robert Huseby wrote a critique on limitarianism arguing that we don’t need that idea, given that we have egalitarianism and sufficientarianism. I responded in two ways. First, by saying that there are other ways to judge a philosophical idea than looking at philosophical distinctiveness (in other words, Huseby and I have different views on what we want political philosophy to do, and I argue that this has implications for judging the value of limitarianism). Second, within the meta-theoretical choices made by Huseby, I object to his arguments.

Those interested in this intra-academic-philosophical debate, feel free to chime in. If anyone wants to read the paper by Robert Huseby (since unlike mine, it’s not open access), I’m happy to send it to anyone dropping me an email.

NB: Luigi Caranti and Nunzio Alì published in the Italian Journal Politica e Società a paper with the same title as Huseby’s paper, voicing overlapping criticisms. Email them or me if you want to get hold of a copy.



Murali 02.21.22 at 2:12 pm

Is it me or is limitarianism primarily (motivation-wise) a matter of modus tollens-ing the levelling down objection to certain stricter forms of egalitarianism.


Mike Huben 02.21.22 at 3:59 pm

I think the idea of pluralist accounts of distributive justice is going to be very useful to my thinking. Especially since I tend to argue against greedy reductionist accounts by libertarians.

One major benefit of pluralist accounts is that they can be shortcuts to important factors, even if those could eventually be explained by fewer accounts. Maybe that’s why feminism is so important, for example.


Tim Wilkinson 02.22.22 at 12:24 am

{1} I don’t see how it makes any sense for Huseby to say that SUFF + EGAL entails LIM. if SUFF imposes a floor (or floors?!) and LIM a ceiling, then they are symmetric. Since it is supposed that SUFF can impose a substantial constraint at the bottom end of the distribution rendered by EGAL (ie SUFF is not redundant in ‘SUFF + EGAL’), then surely LIM can do the same at the top.

It would at least be comprehensible to claim that EGAL entails both SUFF and LIM (but that would need to be demonstrated; and wouold be an objection to either only if they are not themselves invoked in the specification or operationalisation of the relevant conception of EGAL).

And that idea might be turned around, and that claim made instead that SUFF + LIM actually provides an adequate specification or operationalisation of EGAL, so that it is really EGAL that is made redundant.

{2} In any case, at a highly abstracted level, egalitarianism does seem to be of a different kind to either SUFF or LIM, whichever side we take of (arguably) the main dichotomy in specifying EGAL. I take that dichotomy to be between {a} procedural or {b} outcome-based approaches.

{a:} a procedural EGAL can in turn be either {i} an outcome-agnostic one like mere equality of opportunity or {ii} an outcome-approximating one like free-market capitalism + (a certain level of) taxation. In either case there is certainly no conceptual equivalence to a ceiling/floor approach, and there no practical guarantee that the outcomes will conform to any particular ceiling or floor.

(b:) An outcome-based EGAL might practically guarantee conformity to some ceiling or floor, but unless it incorporates SUFF or LIM, that doesn’t seem likely except by coincidence – unless (perhaps) we have a very strong form of outcome-EGAL like perfect equality. Of course very few people would want to endorse that, but to stick to a relatively formal approach we can observe that once in operation perfect-outcome-EGAL might end up outputting a level outside (presumably below!) the range specified by a combination of SUFF + LIM. That’s being as charitable as possible to the idea that EGAL might make SUFF or LIM redundant. A more robust approach is to take a reasonably abstract view of what EGAL promises on its face, and to note that increasing equality is a different kind of operation from imposing a floor or ceiling. Reducing the GINI coefficient for example is a matter of the overall distribution and extreme outliers have little impact on (and are certainly compatible with a relatively low value of) the coefficient. If it’s objected that EGAL is about narrowing the gap between the highest and lowest values, then we are getting closer to a position in which EGAL is specified or operationalised by the combination of SUFF + LIM.

{3} It might beuseful to think about income or wealth ceilings as having two different possible consequences, juridically speaking:

{a} disentitlement: anything above a certain level is illegitimate and is to be confiscated
, and the weaker (superset) condition
{b} no-entitlement: anything above the ceiling is not necessarily to be confiscated but is liable to be, if required for some (defined) other purpose.

{a} would make sense where limits of wealth are imposed to prevent individuals from wielding excessive power – not just political power narrowly construed as making electoral donations, but media influence and many other forms of economic power – assuming this theory is going to be implemented in something like current Western society in which just about anything can be bought, often legally (Walzer, Spheres of Justice rings a bell here)

{b} might be a suitably modest (and non-paternalist) way of introducing ideas about use-value, diminishing marginal utility, and even the corrupting influence of wealth. It would have obvious application to a system in which both SUFF and LIM operate – the ceiling (or one of the ceilings) imposed by LIM might be used to create a pool which can be drawn on for redistribution needed to satisfy SUFF.

{4} It’s obviously a more complex and convoluted question than any of the above really acknowledges once a dynamic analysis is introduced with an eye to implementation (assuming agin that we are talking essentially about moiderating market capitalism roughly as we know it). At this point implausibly-shaped graphs of the Laffer curve start being brandished along with all the other arguments about the unintended ill effects of (ex post) ‘re’-distribution. This is potentially quite a welcome kind of objection since it shifts the groung from fundamentalist proprietarian claims to questions of expedience.

There certainly a place here for empirical discussion about what allocations (expected uses) of capital are productive – and perhaps relatedly, which market transfers constitute rents/producers surplus/more-than-minimally-motivating-inducements. Linking these to the concept of ‘unearned’ income might be a useful way to earmark a pool of funds that can be redistributed both legitimately (because unearned) and harmlessly (because unproductive). I think that would actually provide quite a massive pool of funds (in theory, that is – the only remaining problem being how, politically, to prise said funds from the cold dead hands of their current owners.) Whether a quasi-procedural conception like earned v unearned income would marry up particularly well with an outcome-based LIM I’m not sure…


Matt 02.22.22 at 12:26 am

Thanks, Ingrid – I’m very glad to see it’s out. I will hope to read it carefully in a week or two, when I have some urgent work/tasks taken care of. I appreciate you posting it.


John Quiggin 02.22.22 at 11:28 pm

Thanks for this, Ingrid. Both your point about the purposes of political philosophy, and the distinction between the riches line and the limitarian threshold make a lot of sense to me.

I’ve argued (somewhat similarly to your account of Huseby) that utilitarianism yields most of the conclusion you want to derive from limitarianism. But, as you say, that’s not, in itself an argument against limitarianism. In particular, lots of people who object to utilitarianism for reasons like those of Rawls, might accept similar conclusions from a limitarian perspective.

Looking at the riches line, I conceive it as the point the top marginal tax rate be set so as to maximise tax revenue. Implicit in this is the idea that additional income above the riches (but below the limitarian threshold) is neutral in social terms – it doesn’t yield any significant increase in welfare, but it also doesn’t do any harm to the rest of society [I’ll leave aside impacts on natural resources etc, which need a different kind of response]. Maximizing revenue means redistributing as much as possible of the surplus to those below the line.

The limitarian threshold is the level above which wealth is positively harmful because (for example) it gives unreasonable political and social power to a small number of individuals. It’s worth taking steps to limit such concentrations of wealth even if it’s costly to the society as a while to do it.

As I think of things, the riches line would apply somewhere around the income level that puts people into the top 1 per cent, including business owners, senior managers and top professionals. By definition, the top 1 per cent in the US encompasses millions of people.

The limitarian threshold would apply to the kinds of people we mostly associated with the term “rich”, including billionaires and near-billionaires. There might be a few thousand such people.

Comments on this entry are closed.