This volume is an impressive addition to the small but growing body of literature on global justice that tries to find a midpoint between cosmopolitanism and statism or nationalism (other books in this category include Kok-Chor Tan's Justice without Borders and Gillian Brock's Global Justice: A Cosmopolitan Account). The aim is to make room for the idea that we owe special duties to our fellow citizens by virtue of some feature of our relationship, while at the same time to show that these duties are only defensible if we also acknowledge certain cosmopolitan responsibilities. Vernon follows an original path to this conclusion. The argument is quite intricate, and I cannot engage with all of it in the space of a short review, so I will focus on what I take to be the book's central claims.

Vernon rejects two popular accounts of the special duties we owe to fellow citizens. One appeals to the idea of reciprocity, and portrays states as mutual-benefit societies: special duties are owed as a matter of fairness for the benefits we receive, through the state and in other ways, from our fellow citizens. Another appeals to national identity. It claims that special duties are simply part and parcel of the strong sense of solidarity we feel with our compatriots. Finding both of these flawed, Vernon proposes a third account that has not, to my knowledge, been put forward before. He points out that when people associate together under the aegis of the state, they impose on each other a series of risks that would not otherwise obtain. He refers here both to risks that the state imposes directly, such as the risk that it may oppress or even kill its subjects, and to risks that arise indirectly as a result of state provision, such as the risk of being injured in a car accident on a public road. Because citizens are complicit in imposing these risks, he claims, they owe one another forms of protection that they do not owe to outsiders—hence, the justification of special duties.

Cosmopolitan responsibilities arise, meanwhile, from what Vernon calls the Iteration Proviso, which (as initially stated) holds that “a group of people can legitimately set out to confer special advantages upon each other if others, outside that group, are free to do the same in their own case” (p. 104). So for advantage-conferring states to be legitimate, they must take steps to ensure that outsiders are not prevented from establishing viable states of their own. Some of the practical implications are spelled out in the last three chapters of the book, which discuss, respectively, humanitarian intervention, international criminal law, and a “global harm principle” that would apply to, for example, the trade and tax policies of rich states. Like the rest of the book, these chapters are finely argued, but since the conclusions reached are somewhat uncontroversial (who in the literature on global justice has ever failed at some point to condemn the agricultural subsidies offered by the governments of wealthy states, for instance?), I will not comment further on them here.

What are we to make of the risk-imposing argument? If we say that people associated in political communities impose risks on each other, we must be assuming some baseline at which the risks in question were not present. What might this be? Given the broadly Lockean provenance of Vernon's argument, it must be something like a pre-political state of nature. Here, for example, there would be no risk of being falsely imprisoned by the state. But there would certainly be other risks, such as the risk of attack or theft. Now, it is normally taken for granted that when a state is legitimate, it must impose fewer, or less severe, risks than would occur in its absence. Ex ante, therefore, joining the state is a good bargain. You might be unlucky, of course, if you are one of those picked on by state officials for maltreatment, but in general the risks that the state imposes are more than compensated for by the risks that it averts, compared to the baseline. So it is not clear why special duties (over and above the duty to maintain a legitimate state) could arise on this basis. If everyone benefits, who owes what to whom?

Similar considerations apply to Vernon's automobile traffic example. It is clearly true that when we drive we expose others to the risk of death or injury, and they likewise expose us, but we judge that the benefit outweighs the risk—or else we do not drive (if the odds do not appeal, we can take out insurance). The special duty argument could only get a foothold in the case of people for whom the very existence of the traffic system imposed an unavoidable net burden.

Turning to the iteration argument for cosmopolitan responsibilities, it seems at first as though leaving outsiders free to set up their own viable states could only entail negative duties, such as the duty not to invade their territory or to dump nuclear waste upon it. But as Vernon proceeds, the proviso becomes more demanding. By p. 110, for example, it has turned into a “global equal-opportunity principle”—the opportunity in question being “the possibility of enjoying the political and economic conditions that rich countries standardly seek.” To achieve this possibility would presumably require economic intervention and social engineering on a massive scale. Of course, it depends on how “opportunity” and “possibility” are understood. Take state failure, one of the cases in which “cosmopolitan regard” is said to require rich states to act. Suppose a state fails because rival ethnic groups within it cannot achieve a political accommodation and civil war breaks out. Does this mean that the opportunity to create a viable state did not in fact exist, or was the opportunity simply squandered? Despite showing some sympathy for the position expressed in J. S. Mill's famous essay on nonintervention, Vernon appears to end up taking the view that intervention in such cases may be required. But it is surely paradoxical that the legitimacy of the British or any other developed state should be put in question by its failure to intervene in, say, Somalia, which is what the demanding interpretation of the proviso appears to imply.

This reveals the problem in trying to cantilever outward from domestic duties to global ones, Vernon's ingenuity notwithstanding. To reach cosmopolitan conclusions, you may need (strong) cosmopolitan premises. Nevertheless, this is an engaging and well-argued book, strongly recommended to those seeking to give the idea of citizenship a cosmopolitan dimension.