Dozens of countries have established Sovereign Wealth Funds (SWFs) in the last decade or so, in the majority of cases employing those funds to manage the large revenues gained from selling resources such as oil and gas on a tide of rapidly rising commodity prices. These funds have raised a series of ethical questions, including just how the money contained in such funds should eventually be spent. Chris Armstrong's article in the winter issue (EIA 27.4) engages with that question, and specifically seeks to connect debates on SWFs with debates on global justice. Just how good are national claims to the great wealth contained in SWFs in the first place? Using the example of Norway's very large SWF – derived from selling North-Sea petroleum – he shows that national claims are at least sometimes very weak, with the implication that the wealth in many such funds is ripe for redistribution in the interests of global justice.
To read the article in full for a limited time, click here.
We invited four experts to respond to Professor Armstrong's argument. Read their responses below and join the conversation in the comments section.
A Response by Oliviero Angeli
In recent years criticism of sovereign wealth funds has been confined mainly to their misuses and lack of transparency. Few have seriously questioned their basic legitimacy. In this respect, Chris Armstrong’s article is surely distinctive. His criticism, however, is not directed against sovereign wealth funds as intergenerational transfer mechanisms. Chris challenges their legal and moral premises. Do states have a right to exploit and use those natural resources whose revenues finance the sovereign wealth funds? The question is indeed worth asking. Suppose all natural resource-rich states start setting aside a certain share of revenues obtained from selling their natural assets to continue benefiting current and future generations of their citizens. Most (non-right libertarian) political theorists might welcome this policy as both sustainable and fair on future generations. These theorists are opposed to privatizing natural resources, yet see no objection to using them primarily for national purposes. But what if these natural resource-rich states are benefiting from resources that are not their own? In the article, Chris assesses the merits of three reasons for granting exclusive (or at least preferential) rights over natural resources to states in whose territory these resources are located: improvement, attachment, and self-determination. In his view, none of these is fully adequate. While I agree with the general tenor of his criticism, it seems to me that his way of approaching the subject could lead to confusion. Let me explain.
Consider the attachment argument. Chris concedes that people can be deeply attached to their land and therefore extremely reluctant to lose control over it. He even grants the argument that communities can be attached to resources with little, if any, symbolic meaning—like petroleum. But petroleum is not what sovereign wealth funds are made up of. These funds are derived from revenues generated from selling petroleum. In Chris’s view, Norwegians cannot therefore claim attachment to petroleum as a ground for income rights over natural resources. If Norwegians were truly attached to their petroleum, he contends, they would not sell it. One problem with this line of reasoning, however, is that it makes the attachment argument appear weaker than it really is. As I see it, attachment does not extend only to resources themselves, but equally (if not more) to the ability to control these resources. To put it bluntly, people can become deeply committed to the ability to control and change patterns of use within a certain area, even though they have no particular attachment to the things within that area. Depriving them of this ability can be disruptive to their dignity and self-esteem as citizens who can take responsibility for the environment. Within a system of territorially bounded states the natural expectation is that resources such as petroleum are tied to the choices of those citizens, in whose territory they lie. It is therefore reasonable for citizens to think of themselves as having a say on whether and how to use such resources.
Now, Chris may reply that this line of thought begs the question about whether the expectation to rule over territories is especially worthy of protection, considering the impact that revenues generated from natural resources can have on global poverty. This is true. The state’s rights over natural resources cannot be absolute. They must be limited (especially with respect to the right to income from natural resources). But this is quite a different thing than arguing (as Chris does) that since the distribution of natural resources is arbitrary from a moral point of view, territorial states should provide reasons for controlling natural resources within their territories in the first place. For to argue in this way would amount to ignoring that politics in democratic states is organized upon the expectation that citizens have a say on the development of their environment and that this expectation is worth being protected (or at least not frustrated without reason). My disagreement with Chris is therefore not so much on the conclusions he draws as on his starting point. Natural resources do not come from nowhere, out of nothing. They occupy places in territories within which citizens exercise their democratic right to be active members of their society. This alone means that natural resources are goods to which particular communities already have special claims. These claims may not ground a property-like entitlement over natural resources. But they provide a suitable starting point in dealing with controversies over profits derived from using and selling natural resources.
A Response by Andreas Follesdal
(The author was a member of the Council on Ethics of the Norwegian Pension Fund for eight years.)
I welcome Armstrong’s attempt to bring standards of global justice to bear on the Norwegian Sovereign Wealth Fund (SWF), when he asks if Norway’s claim to sovereign control over its oil resources is such that this claim might override calls for global redistribution. The revenues from the fund amounted to a net gain to the Norwegian government of 344 billion NOK (approximately $56.7 billion) in 2013.
I happen to disagree with Armstrong concerning the substantive distributive norms of global justice. He believes that minimizing global equalities is a defensible standard of global justice. Others (such as Brian Barry , Thomas Scanlon, and Charles Beitz ), however, have defended a more modest global difference principle. And yet still others, including John Rawls, Thomas Nagel, Samuel Freeman, and Michael Blake, have defended principles that allow larger global economic inequalities, even if they agree on the need for some international obligations, including social and economic human rights. The following observations do not hinge on such disagreements. They stem from the vantage point of domestic debates about the Norwegian SWF, which, rarely for a SWF, focus not only on questions of internal distribution but also on issues of global justice. Indeed, global justice has been addressed both in public debates in Norway—yielding a commitment to development-oriented investments by the present government (see here, especially p. 25)—and in scholarly contributions, including by myself and several other members of the Council on Ethics of the Norwegian Pension Fund.
Armstrong argues convincingly that Norway should share some of the money stemming from the sale of oil that currently accumulates in its SWF by paying a global tax on that income. This conclusion is, on the one hand, not exhaustive of the requirements of global justice and, on the other, somewhat hasty.
The requirements of global justice would also seem to include such norms as to avoid complicity with evil; help prevent the wrongdoing of others; contribute to the generation of wealth in poor countries: contribute to a just distribution of such wealth; and possibly distribute other benefits more fairly as well. I submit that the ethical constraints on investment together with the active ownership strategies of the Norwegian SWF contribute to secure compliance with several such norms. The Norwegian Parliament has unilaterally taken on such obligations, arguably contributing modestly to make the “global basic structure” somewhat more just, even in the absence of functioning legitimate institutions of global governance. It should be noted that the economic performance of the fund, contra Armstrong’s claims (p. 3), has not suffered identifiable losses due to its divestment from corporations that would otherwise make the fund complicit in severe human rights violations, gross environmental damage, production of particularly problematic weapons, and so on. Such a decisive loss is certainly a hypothetical possibility.To my knowledge, however, the Norges Bank Investment Management has not reported any systematic negative effects on performance, though there are transaction costs inflicted by the divestment and reinvestment, as with any such transactions.
As to Armstrong’s suggestion that a global tax is required by global distributive justice, note first that there are currently no credible institutions that can gather and distribute such a tax with sufficient efficiency. Second, I suggest that we may (or must) consider other institutions of the “global basic structure” than a global tax authority to engender a more just global distribution of wealth. Norway currently contributes in at least three ways, albeit insufficiently. First, the foreign investments of the SWF are arguably to the benefit not only of Norway and the corporation but also to the workers and the host community where capital is being invested – some of which is in poor states. Second, as active shareholders, the fund managers promote OECD guidelines and other components of global justice. Third, the Norwegian government maintains a commitment to spend 1 percent of GNP on development cooperation. In 2013 this was equal to 30 billion NOK ($4.86 billion). If this development assistance is regarded as a global tax on the net income from oil, it amounts to close to 9 percent. I do not claim that these contributions fully meet the obligations of global justice, but they should be included in these discussions—discussions that Armstrong laudably advances.
A Response by Angela Cummine
Chris Armstrong’s article is a welcome addition to the literature on sovereign wealth funds (SWFs), particularly the small but growing body of work on the funds’ ethical implications.[1. See Angela Cummine (2013) "A Citizen’s Stake in Sovereign Wealth Funds: The Management, Investment and Distribution of Sovereign Wealth," Unpublished Doctoral Thesis, Department of Politics and International Relations, University of Oxford, esp. Introduction.]
Armstrong breaks new ground by focusing on a relatively neglected set of ethical issues raised by SWFs—those of global justice. In questioning whether the domestic character of sovereign funds is legitimate in light of potential global claims to their underlying assets—typically, the natural resources of an establishing state—Armstrong offers a novel critique on the moral status of SWFs. In advancing this critique, Armstrong considers three reasons for limiting the benefit of natural resource returns held in sovereign funds to domestic citizens, present and future, of a sponsoring state. These include the claims that a community may enjoy a priority right to territorial resources where (1) they have improved those resources, increasing their value; (2) certain resources have a special significance for a community; and (3) control over natural resources is essential for collective self-determination. If persuasive, such arguments provide a justification for the establishment and operation of sovereign funds as domestic entities. Armstrong finds all three arguments wanting, testing each through the prism of Norway’s sovereign fund, one of the largest, most transparent, and ethically responsible SWFs. Consequently, Armstrong claims we have good reason to think that the Norwegian fund and other sovereign funds of similarly affluent nations may violate the demands of global justice, and that “theorists . . . ought therefore to pay attention to [SWFs’] potential to fund progress toward global justice” (p.12). Three points need raising in relation to this critique.
Are Sovereign Funds Natural Resource Funds?
Much of Armstrong’s critique rests on a characterization of sovereign funds as commodity funds or natural resource funds. This is true for Armstrong’s primary case-study of the Norwegian fund, established with proceeds from the country’s North Sea oil and gas, and in many other countries with SWFs. It is also consistent with the focus of much academic literature on the use of sovereign funds as mechanisms for managing commodity wealth in resource-rich nations.[2. See for instance Adam Dixon and Ashby. H.B. Monk (2011) The Design and Governance of Sovereign Wealth Funds: Principles and Practices for Resource Revenue Management, 30 October 2011 at www.ssrn.com/abstract=1951573; Jeffrey Frankel 2010) ‘The Natural Resource Curse: a Survey’ Harvard Kennedy School Faculty Research Working Paper Series, February 2010 RWP10-005J.] While historically commodity management has been the dominant reason for establishing such funds [3. Cornelia Hammer, Peter Kunzel and Iva Petrova (2008a) Sovereign Wealth Funds: Current Institutional and Operational Practices (Washington DC: IMF) at www.imf.org/external/pubs/ft/wp/2008/wp08254.pdf], today, non-commodity funds account for around 40 percent of total assets under management in SWFs.[4. Yael Selfin et al (2011) Sovereign Wealth Funds—the Key to Economic Success, October 2011 (London: PwC), p. 3] Indeed, recent research has explored how a state does not have to be “resource-rich” to establish a sovereign fund. The taxable asset base for a state extends far beyond natural resources, covering a wide array of public assets that offer potential revenue for states to establish community funds.[5. See Gary Flomenhoft (2012), "Applying the Alaska Model in a Resource-Poor State: The Example of Vermont,"in Karl Widerquist and Michael Howard, eds., Exporting the Alaska Model: Adapting the Permanent Fund Dividend for Reform Around the World, (New York: Palgrave Macmillan) pp.85-108.]
If resource wealth is not the only type of underlying asset for a SWF, then much of Armstrong’s global justice critique loses critical purchase, since that critique targeted arguments advocating territorial rights to natural resources. This begs the question: what is the morally relevant asset when it comes to analyzing the legitimacy of rights to sovereign wealth? I suggest two responses. One is to treat the original underlying assets as the morally relevant object and restrict Armstrong’s critique of SWFs to commodity-based funds while developing a separate critique of funds established with other financial assets (from, say, trade surpluses or nationalization proceeds). An alternative response might be that the relevant underlying asset of an SWF is the government wealth in these funds, irrespective of its origin. The analysis then would focus less on the origin of the initial seed capital and more on whether the citizens of a country have a particular claim to the surplus wealth of their nation over non-nationals.
Special Claim from Attachment to Sovereign Funds?
If one adopts this broader view of the underlying assets of sovereign funds, a further question arises about whether the “special significance” argument—what Armstrong terms the “special claim from attachment”—becomes plausible in justifying a statist view of rights to sovereign fund assets. Since this argument is based on the idea that “some particular resources come to be crucial props for communal identification and come to have great symbolic value to them” (p. 8), then surely this rationale may apply to sovereign funds themselves, many of which assume an important role in the public policy apparatus and identity of national communities.
Consider that multiple sovereign funds have been established to fund projected shortfalls in government pension commitments to citizens.[6. For instance, the New Zealand Sovereign Fund which was established to help the government fund its universal retirement scheme for over 65’s, or the Australian Future Fund set up to fund shortfalls in government pension liabilities from 2020.] Similarly, in the Gulf States, where sovereign funds form a crucial part of the fiscal architecture that supports large public spending commitments, these SWFs are surely crucial to those citizens and their life plans. Such funds are also some of the largest in the world and a source of national pride.
One might respond, as Armstrong does, that the fungible nature of sovereign wealth means that citizens cannot be attached to the particular money in their SWFs over and above any other public wealth. As long as their needs are met by the state, there can be no special attachment to their SWF as opposed to their national wealth. But in some cases, public attachment exists not just to SWF assets but to the very fund itself. Consider that the sheer scale of the Norwegian fund has been blamed for a psychic effect on Norwegians’ declining propensity to save, with some arguing that the enormous per capita wealth of the fund has bred complacency in the small nation regarding personal savings. Equally, in Alaska, which boasts an impressive $45 billion sovereign fund that pays out an annual dividend share of earnings to every Alaskan—of which there are only 600,000—it would be “political suicide” to cease the dividend or to abolish the fund.[7. Scott Goldsmith (2005), "The Alaska Permanent Fund Dividend: An Experiment in Wealth Distribution" in G. Standing, ed., Promoting Income Security as a Right (London: Anthem Press), p. 563.] In a recent survey, which asked Alaskans if they would prefer to abolish their fund and take their per capita share of its assets, citizens showed overwhelming support for the continued existence of the fund as an expression of the communities’ commitment to the rights of future generations to this wealth.[8. See Cummine (2013), Chapter 8.]
These examples suggest that the very establishment of a domestic sovereign fund with an important nation-building or financing mission may create a particular resource to which a population enjoys a legitimate attachment, giving those domestic citizens some claim to their SWF assets over non-nationals on the grounds of attachment.
Special Claim from Improvement to Sovereign Funds?
In a similar vein, one might also wonder if Armstrong was too quick to dismiss the “special claim from improvement” argument to sovereign funds. Again, if the morally relevant unit of analysis is the fund rather than its original underlying asset, could the very creation of a sovereign fund be considered an “improvement” to resources, whether natural or otherwise? Consider that by definition, a SWF seeks to remove surplus wealth from a spendable government budget and quarantine it in an investment vehicle dedicated to asset appreciation. In the case of both commodity and non-commodity funds, there is an improvement. A community that establishes a fund for its commodity windfalls could be seen as improving those natural resources through its conversion of finite assets in the environment to an infinite financial asset that will generate financial returns for citizens across time. Alternatively, a community that builds up a fund from non-commodity proceeds has still committed to preserving one-off windfalls and augmenting those for both current and future citizens.
If persuasive, would the idea that a sovereign fund improves the value of its seed capital (assuming successful market investment) not give the people of the sponsoring state a special claim to those resources and thus blunt the global justice claim to domestic sovereign wealth? In sum, then, the arbitrary assumption about the restricted nature and scope of justice implied by SWFs as domestic entities may be more defensible than this article argues.
A Response by Paul Segal
Norway may not be morally entitled to all its oil revenue. This is Armstrong’s argument, and to an egalitarian cosmopolitan it is surely plausible. What is less clear is the extent to which this argument may help to reduce global injustice. In my view it is incumbent on a cosmopolitan philosopher to consider the real-world constraints that face the furthering of cosmopolitan ideals. While Armstrong asks if “the proceeds of these [sovereign wealth] funds [could] not be better used to attack global poverty, or to reduce global inequalities,” he is silent on how this might be achieved.
On the cosmopolitan view, every individual is worthy of equal regard, independently of where they live or what nationality they are. This implies that any cosmopolitan egalitarian redistribution between countries must be aimed at the individuals within those countries who comprise the global poor. The essential, practical step omitted by Armstrong is how to ensure that global tax revenues help the global poor.
The problem with focusing on international redistribution is that there is little reason to believe, and much reason to doubt, that countries are effective at redistributing resource revenues to minimize poverty within their own borders.[1. Paul Segal, “How to Spend It: Resource Wealth and the Distribution of Resource Rents”, Energy Policy, Vol. 51, (2012), pp. 340–348.] Armstrong acknowledges this point himself, but does not draw the obvious conclusion: that it is therefore doubtful that simply transferring oil rents from the government of Norway to the government of a poor country would have the desired effect on the poorest people within that country. The robust debate on the effectiveness of aid extends this doubt to expenditures by aid agencies and international organizations. Armstrong raises the normative question of “how the proceeds can best be spent” without considering the practical question of how they will be spent.
Consider Nigeria. In 2011 its oil rents were worth PPP$742 per person, yet 54 percent of its population of 164 million people lived below PPP$1.25 per day—which is substantially less than those per capita rents.[2. Data from World Development Indicators and PovcalNet. The use of PPP$ takes account of price differences across countries and are used for cross-country comparisons of well-being; here they are given at 2005 prices.] Even the easiest-to-administer pro-poor revenue distribution policy, namely an untargeted, unconditional universal transfer or basic income funded by those rents, would easily eliminate this poverty.[3. Paul Segal (2011), “Resource Rents, Redistribution, and Halving Global Poverty: The Resource Dividend,” World Development, Vol. 39, No. 4, (2011), pp. 475-489.]
Thus Nigerian oil revenues are evidently not being used to reduce poverty as effectively as they might. Why assume poor countries in general will do better, if they receive oil rents from Norway?
There is therefore a certain inconsistency in Armstrong’s approach: he wants to justify the taxation of oil rents on egalitarian cosmopolitan grounds, without considering the likelihood of the revenues being spent in a manner consistent with egalitarian cosmopolitanism. This is not to say that Armstrong is wrong about Norway's sovereign wealth fund. But in my view two questions are prior. First, why do countries that contain very poor people help those poor people so little? Second, how could they help them more? Moreover, these questions are prior in two senses: because progress in answering them would already help to reduce global poverty, directly furthering egalitarian cosmopolitan aims; and because without progress in answering them, there is little reason to believe that more international redistribution will further egalitarian cosmopolitan aims. The most important constraints on global pro-poor spending are not global, but national.
It is also necessary to clarify the division of labor in the hydrocarbon sector, and the nature of resource rents, which in my view are inadequately explored in Armstrong’s article. In the case of natural resources, rents are simply payment over and above the cost of extraction (where costs should be competitive, and should ideally include social and environmental externalities).[4. See Segal (2011) for discussion.] If Iraqi oil cost $2 a barrel to extract but the Iraqi government can sell it for $100 per barrel, then the rents are $98 a barrel. Nice work if you can get it.
Why does oil generate such rents? First, the enormous share of rents in Iraq’s oil is due to the fact that its extraction cost is so much lower than the marginal barrels of oil that get extracted globally. These include unconventional oil such as tar sands, which might cost $100 per barrel to extract. Thus these rents are partly Ricardian differential rents—equal to the difference in costs between Iraqi oil and the most expensive oil. Second, at times there has been a deliberate policy by the major oil exporters that comprise OPEC to restrict the supply of oil in order to bid up its price, raising the scarcity rents to all oil.
It is a common misperception that it is international oil companies (IOCs) who benefit primarily from these rents. True, IOCs make enormous profits. But this is because they are enormous companies investing enormous amounts of capital: their rate of return on capital is not unusually high. Thus they do not, in most cases, receive much in the way of rents. This is because today—and unlike 50 years ago—there is a lot of competition between IOCs to extract oil for the governments that own it. Indeed, the $2-a-barrel extraction cost mentioned above was the extraction fee per marginal barrel in the winning bid for an Iraqi oil field, made by a consortium that included the IOCs BP and China’s CNPC.
Armstrong comments that “In practice, markets may be uncompetitive in the sense that a few sellers artificially restrict supply, driving up commodity prices and thereby gaining what economists call ‘scarcity rents’ in the process. Perhaps crude oil costs $10, total costs are $5, but sellers hold out for a price of $25, thereby achieving ‘rents’ of $10.” But in this scenario Armstrong’s initial “cost” of $10 is a chimera: he has inserted middle-men—“a few sellers”—who do not exist. The more accurate model is that there is a $25 oil price and a $5 cost of extraction, leaving $20 of rents that should accrue to whoever owns the oil—usually a government. If extraction is performed by a particularly voracious IOC, then they may succeed in appropriating some of those rents and not passing them on to the government. But that does not change the logic.
While this process is simpler than Armstrong envisages, the next step is more complex. This is because most oil-importing countries charge tax on petroleum products such as petrol (gasoline) and diesel. At present the U.K. government charges a fuel duty of 57.95 pence per liter in addition to VAT, which together comprise more than half the price faced by the consumer; many other rich countries have similar taxes. These taxes are probably sufficient to affect global demand, dampening the oil price, from which it follows that governments of fuel-importing countries are expropriating some of oil producers’ rents. (Indeed, when importing countries expressed dismay on behalf of their consumers at the rise in oil prices of the 1970s, some oil producers suggested they simply cut fuel taxes.) A good estimate of the global distribution of resource rents would have to take these general equilibrium effects into account.