Libertarianism without inequality (2)

by Chris Bertram on October 25, 2003

This is the second installment in a series of postings to accompany a reading group around Michael Otsuka’s “Libertarianism Without Inequality”: (first installment “here”: ). I’ll put the meat of the posting below the fold. Comments are again welcome from others who are reading or who have read the book.

Otsuka’s second chapter asks what can be done about those persons who can only be supported via a distribution of resources which requires encroaching on the robust right of self-ownership he defended in chapter one. That’s to say, assume there are some people whose basic needs can only be met if some (able-bodied) people are forced to work for their benefit – how are we to square this with libertarian commitments? He puts forward the unusual suggestion that libertarians and liberal egalitarians can both agree that it would be permissible to raise the funds required by taxing those who have been convicted of crimes. (The conviction has to be sound, and the crime has to be something that is justifiably criminal.)

I have to say that this chapter has a slightly odd feel to it. I wasn’t sure why this apparatus was being introduced at this stage in the book and I can only presume that it is because Otsuka wants to make use of this conclusion later on. Often I found that Otsuka’s intuitions about cases (such as conscript armies or the comparative duties of siblings towards elderly parents) just didn’t mesh closely enough with my own. But perhaps I need to wait and see.

Otsuka, still pursuing his goal of reconciling libertarians and egalitarian liberals puts different arguments to each camp at this point in the book. He recognizes the attachment of egalitarians to universal taxation, but invites us to compare various alternatives to it including universal giving, non-universal giving and taxation of the unjust as means to raise funds to support a welfare state. Otsuka’s strategy here is to start from universal giving as the best way of providing for the disabled on the grounds that it is both egalitarian and non-coercive. But if such a method were to fail — as well it might — then how are we to compare the merits of the fallback options of non-universal giving
and universal taxation. One is less coercive than the other, the second is more egalitarian than the first.

Here the argument is, to my mind, somewhat weak. Otsuka proceeds by analogy with liberal egalitarian reactions to another case: military defence. Here, faced with a choice between universal conscription and having a volunteer army, liberals seem happy with the latter. In other words, or so he suggests, they prefer non-coercion over universal burden-sharing. But the case of voluntary donors and that of volunteer soldiers seems significantly disanalogous. Volunteer soldiers aren’t best thought of as altruistic givers: their motives for volunteering may include a hankering for the romance of military life and overseas travel. And, also, they get paid for their efforts.

Having muddied the waters a little to suggest that the egalitarian might hesitate between universal taxation and non-universal giving, Otsuka then proceeds to argue that taxation of the unjust is rather like non-universal giving. Following Hart, he suggests that the criminal who is taxed in this way (or otherwise punished) is merely paying the previously announced price for their course of behaviour.

I have to say, I find this model of what punishment is rather unconvincing. In a passage Otsuka cites from Hart, Hart even suggests that this way of looking at things, where criminals elect to act in the light of the tarrif for the crime they are contemplating, thereby maximizes individual freedom. But many of us will feel uncomfortable with this. Surely, when a criminal commits a crime after contemplating that the punishment is a price worth paying, that very calculating attitude to the authority of the law suggest to us that _additional_ punishment is merited.

The case Otsuka makes for libertariana taxing the unjust is rather more straightforward. Since the libertarian already accepts that those who transgress others rights should be punished and should have no special objection to that punishment taking the form of a fine or forced labour (and we can again wheel in the Hart tarrif model to suggest free choice of the penalty by the convict), the only real worry the libertarian needs have concerns the justifiable extent of punishment. Here Otsuka seems almost alarmingly cavalier in his willingness to contemplate (almost) however harsh a sanction is necessary on the unjust to generate the necessary surplus to provide for the disabled. I found my concerns about proportionality of punishment to crime had not gone away after reading his account.

A slightly elusive and unsatisfying chapter, then. But perhaps it won’t seem that way in the light of the ones to come.



Jacob T. Levy 10.25.03 at 11:13 pm

My coblogger Randy can speak for himself, but he’s probably got the best-developed distinctively libertarian theory of punishment and restitution– and I’m pretty sure that on his account this would be necessarily unjust– a punishment that is neither aimed at making the victim whole nor at deterrence nor at restraint. The portion of the punishment that exceeds what’s needed for those three purposes is unjust; if it’s being done for the sake of someone else’s benefit, then the criminal is being unjustly used as a means.

(It’s been a while since I ready Randy’s stuff on this– I may be conflating his theory with bits of my own intuition, here. But there’s nothing intrinsically libertarian about saying : the criminal has violated someone else’s rights, so let’s extract every last penny we can out of him.)


Jay Conner 10.26.03 at 3:55 pm

As a retired lawyer with some criminal justice experience, I can only shake my head in wonder at the towering ivory-towerism of this idea. Obviously to me, these thinkers have never made contact with any real criminals, most of whom are the inept of society, barely able to earn enough to take care of themselves, let alone anyone else.


self 10.26.03 at 4:54 pm

otsuka seems to propose an alternative vehicle for collections intended for a group dependent on some form of social insurance. Yet the only advantage is that the money is collected from less sympathetic and politically disenfranchised subjects.
Two problems:
1)taxes will be collected not just for social insurance but for use of infrastructure. So it is only the additional increment of each taxpayer’s contribution to social insurance (over lifetime minus received benefits) that creates distortions or injustice. I’m guessing this is not substantial enough to claim that taxes as a whole is an unjust system.
2) Insuring for an outcome that occurs involuntarily from birth or calamity is markedly different from constructing rules of behavior with penalties for actions that harm others. A system of general revenue collection would seem to target the potential population of disabling outcomes much better than extracting additional tariffs from the incarcerated for an unrelated social outlay. I’m not convinced that Otsuka’s solution is more just, rather it’s a suggestion that probably leads to greater distortions.
Also missing is some discussion of the implicit assumption of equal detection and prosecuion (and incarceration) of offenders…i don’t think anyone is buying that.
Let’s review: taxes aren’t a form of punishment and law enforcement is not equally distributed across any reasonable metric that predicts criminal behavior. Back to the drawing board (or ivory tower).


Michael Otsuka 10.26.03 at 5:49 pm

I guess Jay Conner doesn’t think white-collar criminals are ‘real criminals’.

“The most common white-collar offenses include: antitrust violations, computer/internet fraud, credit card fraud, phone/telemarketing fraud, bankruptcy fraud, healthcare fraud, environmental law violations, insurance fraud, mail fraud, government fraud, tax evasion, financial fraud, securities fraud, insider trading, bribery, kickbacks, counterfeiting, public corruption, money laundering, embezzlement, economic espionage, and trade secret theft…. According to the Federal Bureau of Investigation, white-collar crime is estimated to cost the United States more than $300 billion annually.” (

These people seem to manage to take care of themselves. Those who have studied or taught at ivory tower institutions will have made contact with a number of students who would go on to commit tax evasion, insurance fraud, and other white collar crimes.


Pete Guither 10.26.03 at 7:38 pm

What disturbs me about this idea is that society becomes dependent on taxing criminals in order to fund a somewhat unrelated activity.

If past experience is any indication (think asset forfeiture), this is likely to lead to a continual increase in crimes determined eligible for the tax, and ultimately informal quotas in arrests (think small town speed traps) in order to provide sufficient funding.


Michael Otsuka 10.26.03 at 11:25 pm

Chris writes: ‘…the case of voluntary donors and that of volunteer soldiers seems significantly disanalogous. Volunteer soldiers aren’t best thought of as altruistic givers: their motives for volunteering may include a hankering for the romance of military life and overseas travel…’

This particular disanalogy works in my favour, since it shows that liberals prefer a volunteer army to universal conscription even when the motives of the volunteers are not entirely noble, thus highlighting the importance which liberals place on avoiding coercion as opposed to promoting other values.


Troy 10.27.03 at 12:27 am

Henry George added a lot to this discussion over a 100 years ago.

The central argument is that Unimproved Land, being the product of no one, belongs in common stewardship of the “community”, and those who wish to monopolize the profits of Land by excluding others owe compensation to this community being excluded.

Real-world implementation of this simple, self-consistent libertarian philosophy is taxing the unimproved value of land as close to 100% as possible.

Geolibertarian FAQ


Micha Ghertner 10.27.03 at 2:20 am

Troy, the problem with Georgism is that it is next to impossible to determine the “unimproved value of land” because the value of land is directly related to the value of the surrounding areas, which may or may not have been improved upon.

That being said, a property tax based on some other standard other than improvements is probably the least offensive tax, even for libertarians.


David W. 10.27.03 at 9:42 am

If I were a libertarian, I think I’d say the least offensive tax would be the estate tax, as it doesn’t take anything away from the people who did the work. Tax on umimproved land, if recently purchased with money gained from labor, could still be taking away from someone’s actual labor.

(I’m still waiting for my copy of the book to arrive, so I shouldn’t even be commenting–but I must say I’m curious to see how an estate tax fits in. My hunch is that a hefty estate tax should be central to any reconciliation of egalitarian and libertarian commitments)


david w. 10.27.03 at 9:51 am

And to be clear, even when I’m not dressing up like a libertarian, I see the estate tax as the best imaginable tax. In fact, I wouldn’t say “least offensive,” since I don’t find it offensive at all.


Chris Bertram 10.27.03 at 12:24 pm

Thanks to Mike Otsuka for his responses.

I guess I wasn’t clear enough about why I don’t think the volunteer army case doesn’t do the work that Mike wants it to do.

The case arises in the context of a discussion where some X cannot be provided by universal voluntary donation. So, pressed for a second-best solution, should we abandon universality or voluntariness. Mike wants to say that the reaction of most liberal egalitarians to volunteer armies (as opposed to universal conscription) shows that they value non-coercion above universality. But my thought is that the volunteer army case doesn’t really establish this.

If we are a bit sophisticated about what universality requires here, then we ought to allow that it has not been breached in the volunteer army case because the individuals who volunteer (non-altruistically) get fully compensated for their contribution. It remains the case that the net cost of providing for national defence (viz, zero) remains equally distributed across the population.

There’s room for quite a bit if argument and redescription of cases here, but it seems to me that whilst moving from universal to non-universal giving (of money) is clearly a breach of universality because the donors receive nothing in compensation, a volunteer army is at the very least, not a clear breach, because of the compensations the volunteers get.


Michael Otsuka 10.28.03 at 9:27 am

A reply to Chris’s latest post:

(i) Military service: In the US, poor members of racial minorities without a university education bear a disproportionate share of the burden of military service, since the option of enlisting in the military which many middle class, university-educated whites would regard as radically inferior to their other job opportunities is the best available choice for many enlistees. In spite of this well-known fact, few liberals in the US advocate universal conscription. And the stated reason is the illiberal, coercive nature of conscription (though the fact that liberals tend to be middle class, university-educated whites is surely also relevant).

(ii) Non-universal giving: I don’t think the case for non-universal giving over universal taxation depends on the altruism of the givers. For even if these givers gave purely out of a self-interested desire to gain a reputation for altruism, this would still be a case of non-universal provision, and I would still maintain the following: ‘If, for example, liberal egalitarians knew that enough money fully to provide for the disabled had been voluntarily raised and donated to the state by a small portion of the population, I doubt that many would insist that the state go ahead and coercively tax the non-[donating] majority of the population and give the [donating] minority partial rebates to ensure that each citizen make an equal sacrifice.’ (Libertarianism without Inequality, pp. 44-45, with the word ‘generous’ replaced by ‘donating’.)


Mike Huben 10.28.03 at 6:34 pm


Unimproved value of the land need only be approximated to give more equitable results with a Georgist tax than the firstcomer owns it outright model.

Yes, it can be difficult to estimate: yet that is a task done routinely by real estate appraisers. And it can be coroborated in cases where improvements are destroyed by catastrophes such as fires. It can also be bounded by cases where purchasers remove previous improvements to make new ones.

Single tax based on unimproved land values might be fine for rural areas, but it doesn’t make sense in urban areas where a great deal of property value is due to community infrastructure and the value of density and proximity. Then, taxation on full value would seem like a better proxy.

Comments on this entry are closed.