"Let us say that there exists a social state of nature when the coordination of behavior for common purposes is not yet established, when compliance with any agreed-upon terms is not generally assured, or when participants are unable to adjust their organization over time, whether by centralized control or by informally modifying the many particular choices that sustain the practice, through individual reform and joint efforts of moral argument and social sanction. Let us also say that there exists a political state of nature when, but only when, one lacks assurance that others will not bargain over the terms of a practice purely out of self-interest."
If you have a situation where both social and political states of nature prevail you have no hope of raising questions of distributive justice with regard to trade he suggests. If, however, there is only a political state of nature but not a social one then there are grounds for thinking of trade as susceptible to pressures of distributive justice. Does this distinction between two types of states of nature in itself show that there are reasons for thinking of trade being open to pressures of distributive justice independently of any hopes or otherwise that one might have for superseding the political state of nature?
The reason for thinking that trade is susceptible to such pressure is that we can identify a distinct level of concern separate from that applicable to specific transactions, a level that describes the background conditions of transactions in general. These background conditions are the subject of such regulation as exists (WTO etc) but could they not (and are they not?) sites of contestation in which global distribution of resources could be altered?