Crises of forced migration are, unfortunately, nothing new. At the time of the writing of this paper, at least two such crises were in full swing – mass movements from the Middle East and parts of Africa to the E.U., and major movements from Central America to the Southern U.S. border, including movements by large numbers of families and unaccompanied minors. These movements are complex, with multiple causes, and it is always risky to attempt to craft either general policy or philosophical positions in response to salient crises. However, both of these instances do bring to the foreground important questions about the proper purpose and extent of refugee protection as a means of dealing with crises of forced migration. In particular, both of these instances force us to consider what role persecution on the basis of a protected ground – race, religion, nationality, membership in a particular social group, or political opinion – ought to play in granting refugee protection, and whether our response to those in danger should change if the agents of persecution are non-state actors. This paper is not primarily about the problems arising from Syria or Central America. In a way similar to how hard cases make bad law, I contend that a too central focus on salient crises tends to lead to bad theory and often to bad general policy. However, if I am successful in my goal of clarifying the place of persecution and non-state action in refugee protection, then we may in turn be better able to think clearly about our current crisis situations. In this paper I will first draw on my previous work on the normative logic of the refugee convention to argue that, while persecution should play an important, and even central, role in our thinking about refugees, this importance is shallow and pragmatic rather than deep and fundamental. Next, I will show how this conclusion supports the claim that harms amounting to persecution by non-state actors may ground an asylum claim, at least in some cases, both when the state is unwilling and when it is unable to protect its members. I consider two cases: first, instances where the authority and power of the state has been usurped by another power, and second, when the state has (implicitly or explicitly) delegated its power or authority to non-state actors. I will show how this leads to extending asylum to a broader range of people than traditional accounts would
