To start off: Why should libertarians be bothered about, specifically, rights to natural resources?[1. This paper is a slightly revised version of a paper first published in Public Reason 1 (1) 2009: 1-8] Why single out these entitlements, rather than looking at property rights in general and asking how any of them can be justified, if they can be justified at all? After all, the domain of distributive justice – of moral rights – includes much more than merely rights to natural resources. Natural resource rights are only a subset of moral property rights. And it has to be acknowledged that, for most people interested in problems of distributive justice, questions about the nature and location of rights specifically to natural resources are still not seen as being of any singular importance.

So I’m going to devote the first part of this paper to saying why I think they are of singular importance. That is, I’ll try to set out the conditions under which their special importance emerges more clearly. I won’t, however, spend a lot of time justifying or explaining these conditions and shall, instead, supply references to where I’ve done this elsewhere at some length.

One obvious condition for according singular importance to natural resource rights is that, like Locke, we see them as having a special generative or foundational relation to other moral property rights. Property rights to other things are, in some sense, derived from natural resource rights and their justifiability is therefore seen as at least partly predicated on the justifiability of natural resource rights.

Another, logically anterior condition is that we see property rights, in general, as libertarians standardly see them: that is, as being parametric for any other rights and liberties people can have. Here I’m alluding to the idea that any coherent set of rights and liberties needs to satisfy the requirement of compossibility – a requirement that the various correlative duties entailed by any such set of rights must all be jointly performable and none can be mutually obstructive. This implies that these duties must be the correlative entailments of mutually differentiatable claims: that their respective compliances must take place within non-intersecting portions of action-space. And there are good reasons for holding that only action-spaces whose descriptions are, or are reducible to, references to things can be mutually differentiated in the requisite way. One implication of this is that, in a compossible set of rights, all rights are funded: that is, the sets of items respectively required for compliance with each of their entailed duties are specifiably distinct from one another. Rights to actions (performances or forbearances) which can be described only in irreducibly intensional terms – in terms of their purposes or intended consequences – lack the requisite mutual differentiatability. They are ones which can be incompossible and, hence, the principle implying them needs to be modified in order to eliminate the duty-conflicts and contradictions they would otherwise generate.[2. Incompossibility is what underlies the complaints of many libertarians (and others), when they deplore the ‘rights explosion’ implicit in many policy proposals and the theories offered in justification of them; cf. Robert Nozick, Anarchy, State and Utopia, (New York: Basic Books, 1974), p. 238; also L. W. Sumner, The Moral Foundations of Rights, (Oxford: Oxford University Press, 1987), pp. 1-8. On the nature and conditions of rights-compossibility, see my An Essay on Rights, (Oxford & Cambridge MA: Blackwell, 1994), ch. 3(C, D), and my ‘Working Rights’, in Matthew Kramer, Nigel Simmonds & Hillel Steiner, A Debate Over Rights: Philosophical Enquiries, (Oxford: Oxford University Press, 1998), pp. 262-274.]

Finally, a still more anterior condition is that we understand the concept of rights along the lines proposed by the Will or Choice Theory of rights. That is, that rights are things whose correlative duties are controllable – permissibly waivable or enforcible – solely at the discretion of their holders. On this account, a necessary and sufficient condition of being the holder of a Hohfeldian claim-right (or immunity) is being vested with the powers to waive and, alternatively, demand/enforce compliance with that claim’s correlative duty (or disability). Since property rights are standardly like this anyway, I won’t here try to mount any general defence of the Will or Choice Theory.[3. An elaboration and defence of the Will or Choice Theory is to be found in An Essay on Rights, ch. 3(A), and ‘Working Rights’, pp. 233-301. The rival Interest or Benefit Theory, in regarding possession of such duty-control as neither necessary nor sufficient for having a right, is incompatible with libertarianism inasmuch as it thereby underwrites the possibility of right-holders’ rights being exercised paternalistically (i.e. by others) on their behalf.]

Granted these three conditions, we pretty much have the basis of the case for the salience of natural resource rights. In fact, the last two conditions strongly point to the first. That is, if coherent sets of rights just are (or are reducible to) sets of property rights, and if their correlative duties are controllable by rights-holders, then, since all non-natural (i.e. made) things are immediately or ultimately derived from natural resources, the validity of any rights to those made things inescapably depends on the validity of the rights to their natural antecedents – since those made things can only have come about precisely through various permissible or impermissible uses of those natural resources and of the things successively created by those uses.

Justified titles to made things therefore have pedigrees exhibiting two key features: (i) they consist in a series of previous justified titles to those things or their component factors; and (ii) they thereby originate in justified titles to natural resources. Or to put it only slightly more concisely, nothing gets made from nothing. All made things have natural resources as ancestors. And hence rights to those made things can be no more valid than the titles to each of their ancestors, in roughly the same sense that Elizabeth’s title to the throne of England depends on those of William the Conqueror and his predecessors.

But, of course, natural resources can’t be the only ancestors of made things which, ipso facto, must also include various bits of labour among their ancestors. Labour is the stuff that does the making. Since the justified ownership of made things depends on pedigrees, it depends on the justified ownership of that labour as well as of natural resources. So, to whom does that labour justifiably belong? I think there are good reasons for holding, and libertarians do hold, that justice vests all persons with the titles to any labour which they haven’t contracted away to others. And it does this on the basis that each person is what has come to be called a ‘self-owner’, the owner-occupier of his/her body. This premiss seems to be the clearest basis – and perhaps the only one – for explaining our fairly fixed conviction that the titles to things made from the labour of slaves are not justifiably vested in slave-owners. And, by a complex extension of this argument, things made from exploited labour don’t morally belong to exploiters.[4. Cf. Hillel Steiner, ‘A Liberal Theory of Exploitation‘, Ethics, 94 (1984): 225-241; ‘Exploitation:  A Liberal Theory Amended, Defended and Extended’, in Modern Theories of Exploitation, (ed.) A. Reeve, (London & Los Angeles: Sage, 1987); An Essay on Rights, ch. 5(D); ‘Exploitation Takes Time’, in Economic Theory and Economic Thought: Essays in honour of Ian Steedman, (eds.) J. Vint, J.S. Metcalfe, H.D. Kurz, N. Salvadori & P.A. Samuelson, (London & New York: Routledge, 2010).]

At this point, we need to take a detour into what has seemed to many to be a serious problem besetting this idea of self-ownership and, hence, the labour ownership that’s said to derive from it. The problem is worth the detour, I think, because its appropriate resolution has important implications for our understanding not only of natural resource rights, but also of several other seemingly unrelated issues lying very much at the heart of arguments about distributive justice. This problem arises from the fact that persons themselves are clearly products of other persons’ labour. Regardless of the circumstances of our conception and gestation – whether by conventional means or in some clinical test-tube – other persons were evidently hard at work in operating these processes. It is, I take it, a conceptual truth that owners cannot be owned. How, then, can the ownership of our selves – as made things – be permissibly vested in us and not in our makers?[5. I address this ‘paradox of universal self-ownership’ in An Essay on Rights, ch. 7(B).]

Students of the history of political thought will know that the answer given by Locke’s contemporary adversary, the royalist Sir Robert Filmer, is, simply and boldly, that it can’t. On the basis of the very libertarian principle we’ve been exploring – that made things belong to the owners of the labour that makes them, or to whomever they choose to transfer that ownership – Filmer argued that, as the Bible suggests, God the maker transferred human species ownership to Adam, from whom that title legitimately descends, primogenitally and patriarchally, to some current person who, as the only self-owner in town, is a rightful absolute monarch.

Now the difficulty here is that, even if we set aside Filmer’s historical, theological and sexist premisses, we appear still to be left with the question of how we – as made things – can own ourselves and our labour. Why aren’t we owned by our makers or by whoever owns them? My proposed solution to this problem relies on the claim that, notwithstanding the fact that labour – usually considerable labour – is involved in making us, it’s not the only kind of production factor involved.

Again, nothing can be made from nothing. And in this particular case, the labour required has to be applied to – or in Locke’s terminology, ‘mixed with’ – a lot of other stuff, including germ-line genetic information. That appropriated genetic information, I want to suggest, is a natural resource because, as Darwin and Dawkins tell us, it’s been transmitted from creatures who were neither persons nor made things.[6. An adaptation of this argument, in the light of the possibility of synthesizing germ-line genetic information, is developed in my ‘Silver Spoons and Golden Genes: Talent Differentials and Distributive Justice’, in Justine Burley (ed.), The Genetic Revolution and Human Rights: The Oxford Amnesty Lectures 1998, (Oxford: Oxford University Press, 1999); see also ‘Universal Self-Ownership and the Fruits of One’s Labour: A Reply to Curchin’, Journal of Political Philosophy, 16 (2008), 350-355.] So although Adam and Eve, as primordial moral agents, might truthfully claim that they made their children, they have to acknowledge that one of the factors used in that manufacturing process was made neither by them nor by any other person. Accordingly, their rights over those children cannot be derived exclusively from rights to self-ownership and to labour. Their rights over those children must partly depend upon – and can be restricted by – whatever rule applies to the ownership of natural resources.[7. Compare the claims, to their respective herds of livestock, of those owners who are breeders with those who are not. Ceteris paribus, natural factors account for more – and human labour for less – of the latter’s herds than the former’s.] And cutting a long story short, I suggest that one element of that restriction can be the standard limitation on the duration of parental rights over their children: namely, that those rights expire upon their children’s attainment of adulthood or moral agency, which is a necessary condition of being an owner (including a self-owner).

Now my guess is that many libertarians (and others) will think that this solution to the universal self-ownership problem looks a bit like using a sledgehammer to crack a nut. Why not simply lay it down and declare by moral fiat, so to speak, that everyone just is a self-owner? What I’d like to suggest, however, is (i) that this is an unsatisfactory way of proceeding within the terms of the larger argument we’re considering and, further, (ii) that the genetic information solution implicitly supplies a unifying foundation for many of the otherwise conflicting intuitions we all have about the vexed issue of how justly to distribute the results of persons’ differential endowments of abilities and disabilities.

Let’s take the first of these arguments first. Simply declaring each person morally to be a self-owner is unsatisfactory because it’s going to leave open a question that libertarian and many other conceptions of justice and unjust exploitation want to close: namely, the question of who is morally entitled to his/her labour and to the products of that labour. For if the labour of conception, gestation and post-natal nurturing doesn’t in some way entitle the labourer, then it’s entirely open to others to advance the unwelcome suggestion that neither do other types of labour. Much better, for a variety of reasons including congruence with a large array of our own intuitive judgements in these matters, is a strategy that can consistently reconcile what we all recognise to be the special claims of parents – a strategy that can sustain some entitling effect of their labour – with an affirmation of their offspring’s self-ownership. The fiat strategy of simply declaring all persons to be self-owners not only has the theoretical disadvantage of simply overriding the libertarian principle that labour entitles, but also, and counter-intuitively, it leaves entirely indeterminate the location of liability for the injuries and damage that pre-adult children can cause to others. For we obviously cannot impose that liability on those children themselves. It’s surely myself, and not my three year old son, who should bear the liability for his injuring your child or damaging your property. Our understanding of distributive justice, and of where (in whom) it locates rights and duties, needs to be carefully contoured so as to take account of such considerations. The fiat strategy simply can’t do that.

Why does this way of resolving the self-ownership paradox give us a unifying handle on many diverse intuitions we have about differential abilities and the fair distribution of their results? It’s a common thought – and one by no means confined to libertarians – that, ceteris paribus, people are entitled to the fruits of their abilities. That is, we think this is especially true if those abilities were themselves mainly acquired through their possessors’ own efforts. When it comes to abilities that are primarily the products of others’ efforts, we hesitate a bit. And when those abilities are largely attributable to favourable genetic endowments, we hesitate a lot.

Why these graduated hesitations? Why does it seem to make a morally relevant difference whether the wonderful state of Pavarotti’s vocal chords was chiefly the result of disciplines he imposed on himself, or the result of childhood training secured by his parents, or the result of sheer genetic good fortune? And something more or less symmetrical with this can be said about disabilities and the suffering they engender. Self-inflicted injuries entitle least, brute misfortune entitles most, and harms inflicted by others may come somewhere in between. No doubt, in the real world, all three sorts of factor combine in the production of many instances of ability and disability. And countless research projects and judicial proceedings are devoted to sorting out the relative contribution of each such factor to these production processes.

Now I want to suggest that what these graduated hesitations reflect is a wider distributive intuition we have about what I’ll call choice and circumstance. Gains and losses are most acceptably shifted when they’re primarily the results of circumstance, and least acceptably shifted when they’re principally the products of choices made by those who incur them. And what counts as circumstance, I suggest, is pretty adequately captured by what we would include under the heading of ‘nature’. ‘Nature’ covers a lot: there are places where it rains all the time and places where it never rains; places with oil deposits and places with serious geological faults; crowded and less crowded cyberspace locations; and genes that code for Kentucky blue grass, poison ivy, viruses, koala bears, cystic fibrosis, schizophrenia, Pavarotti-type vocal chords, some elements of human intelligence, and so forth.

Rights to natural resources – to nature, compendiously construed – are rights to bits of all these various, and variously valued, things. So if we follow Locke and a number of other thinkers in that tradition, if we hold that anyone claiming ownership over some bits of nature must leave ‘enough and as good for others’, we’re led by a series of plausible steps to the conclusion that, in a fully appropriated world, each person is entitled to an equal portion of the value of these bits of nature. That is, all owners of natural resources must pool the value of what they own in a fund – ultimately a global fund – to an equal portion of which everyone everywhere has a moral right.

In that sense, our just rights to natural resources entitle each of us to what has come to be called an ‘unconditional basic income’ or, in its non-paternalistic form, an unconditional initial capital grant. And what’s especially important for libertarians to note in this regard is that we’re owed this grant not as a basic positive right – for on this sort of theory, there are no positive rights which are basic, but only negative ones, with all positive rights being derived solely from antecedent contractual understandings or rights-violations. Rather, we’re owed it as a matter of redress by those who do not forbear from acquiring or retaining more than ‘enough and as good’ natural resources – a negative duty which they have by virtue of our ultimately foundational right to equal freedom.

When you appropriate an unowned natural resource, you create an enforcible right to exclude me and everyone else from it. You thereby deny to each of us a freedom that we previously had and otherwise would continue to have. Unless there are equivalent other resources to which each of us has unexcludable access or, if not, unless you compensate us for that loss of freedom, you have engrossed more freedom than we each possess. It’s this fundamental right to equal freedom that gives us both our rights to self-ownership and our rights to natural resources. And all our other just rights are created solely by exercises of these two rights and of the rights successively derived from those exercises.

Before concluding, however, I think I need to say a bit more about abilities and disabilities. As was suggested previously, our distributive intuitions about choice and circumstance tend to allow self-chosen gains and losses to stay where they are and to require circumstance-caused gains and losses to be shifted. And I argued, in essence, that circumstance-caused gains and losses are ones due to nature: they are, if you like, ‘nature-chosen’ ones. As such, they’re required to be pooled and divided equally.

Now the question that needs to be addressed here is this: Is the set of abilities and disabilities that we’re equipped with from childhood a product of choice or circumstance? Rawls and many others seem to take the view that these come entirely under the heading of ‘circumstance’ and, hence, are eligible for pooling. But there are at least two reasons for rejecting this view: one moral and the other empirical.

The moral reason, which is one internal to libertarianism and to many other theories as well, is simply that a pooling of abilities and disabilities – that is, enforced compensation of the disabled by the enabled – is, in itself, an incursion on self-ownership. It implies an enforcible duty in the enabled to deploy their abilities in ways sufficient to generate the amount of compensation they’re each assessed as owing. Doubtless, most of us firmly believe that the enabled should make transfers to the disabled. And we would be absolutely right to criticise – and even stigmatise – those who don’t. But many of us also believe that such transfers must be voluntary and, in that sense, cannot be a requirement of rights and justice.

The empirical reason for rejecting Rawls’ classification, of children’s abilities and disabilities as products of circumstance, is simply that it’s false. And you certainly don’t need to have raised a child yourself to know that it must be false. What’s broadly true, of course, is that children’s abilities are not self-chosen. But the fact that they’re not self-chosen doesn’t even remotely imply that they’re unchosen. What are people doing, if not engaging in just such choices, when they spend long hours in ante- and post-natal clinics, in teaching at home and in schools, in working to pay for kids’ music lessons, holidays and baseball equipment, and so on and so forth? If children’s abilities were typically construed as products of circumstance and not of choice, it would be pretty difficult to know what conceivably could count as a product of choice. Or to put it in Dworkinian terms, a distribution that claims to be ambition-sensitive and endowment-insensitive can hardly afford to ignore the fact that children’s abilities reflect some of the most deeply felt ambitions that adults standardly have.

Well then, what’s the kernel of truth that lies at the core of that otherwise entirely mistaken view? It is, surely, the undeniable fact that, along with the many hard-earned, labour-embodying inputs used to construct children’s abilities, there’s another essential factor employed in these processes that is a deliverance of nature: namely, those children’s genetic endowments. It’s this production factor of those abilities that properly falls under the heading of ‘circumstance’ and that is therefore eligible for pooling.

So, cutting another long story short, the inference seems to be that, under the general rule for rights to natural resources, we should tax parents on the value of their children’s genetic endowments. Or more precisely, we should tax them on the value of the germ-line genetic information they jointly appropriate in conceiving an offspring.[8. See An Essay on Rights, pp. 246-9, where I discuss how this germ-line genetic information, despite being located within the precincts of a self-owned body, can be construed as an unowned (until appropriated) natural resource.] And this tax, like taxes on people holding rights to other natural resources, goes into the global fund, on which everyone has an equal claim. What this tax does is to effect a net transfer from those who have genetically well-endowed children to those who don’t. And those with poorly-endowed children are thereby supplied with extra resources to develop their children’s abilities and, thus, to offset their genetically-predisposed disabilities.[9. Can parents be said to be subject to a correlative duty to apply these resources to that development? The problem here is to identify the holders of the corresponding right – given that minors, lacking self-ownership, lack the conditions for qualifying as rights-bearers. Perhaps one solution lies in the possibility that parents who fail so to apply these extra resources, and who thereby impose on their child a lesser degree of ability development that endures into his/her adulthood (self-ownership), would then be held responsible for that injury and accordingly be then liable to him/her for compensation.]

Let me conclude, then, not by further elaborating the details of this natural-resource-based account of just redistribution, but rather by briefly listing some of the pre-theoretical intuitions it succeeds in sustaining. First, in entitling persons to the fruits of their labour, it rules out exploitation. Second, in generating an unconditional initial grant as a basic right, it gives everyone some minimum material entitlement, some initial portion of action-space. Third, in extending this entitlement globally, it reflects the view that basic rights are universal: that is, that they are human rights. Fourth, in differentially taxing children’s genes, it simultaneously corrects for the unequal advantages these can deliver, but also avoids relieving adults of responsibility for their own procreative and nurturing choices. And finally, in refusing to mandate transfers from the enabled to the disabled, it precludes what Dworkin and others have called the ‘slavery of the talented’ and thereby allows unencumbered occupational choice: brilliant brain-surgeons can abandon their lucrative jobs to become mediocre poets, if they want to.

By locating the line between choice and circumstance in the right place – by isolating all of what counts as nature, and then distributing its value equally – libertarians can more easily do what they want, philosophically, to do: which is to pass, coherently, through the eye of the needle formed by many of our diverse and conflicting moral intuitions. Of course, such intuitions are not – and can never be – the final arbiters of what’s right. We’d have to be very peculiar people indeed to pass through this needle’s eye with all of our intuitions still intact. And anyway, the continuing market for jobs in moral and political philosophy strongly suggests that there’s no immediate danger of this happening.[10. Many of the arguments advanced here have benefited greatly from Jerry Cohen’s pressing criticisms. For a more general account of the family of left libertarian theories, see my entry, ‘Left Libertarianism’, in Gerald Gaus & Fred D’Agostino (eds.), Routledge Companion to Social and Political Philosophy, (London & New York: Routledge, forthcoming); also Peter Vallentyne & Hillel Steiner (eds.), The Origins of Left-Libertarianism: An Anthology of Historical Writings, (London & New York: Palgrave,  2000), and Peter Vallentyne & Hillel Steiner (eds.), Left-Libertarianism and Its Critics: The Contemporary Debate, (London & New York: Palgrave, 2000).]