The concluding chapter argues that the scope of refugee protection forms a complex picture of instability. Due to the refugee definition’s double rationale – Convention grounds and persecution – and the consequent conundrum of whether to focus on persecutor or persecuted, the scope of protection is subject to a variety of tensions and ambiguities which overlap and intersect in significant and complex ways. For example, human rights standards can be used as a benchmark when it comes to deciding both, which identities and which actions, fall within or outside of the scope of protection. And although ‘discretion’ reasoning is created at the level of the Convention ground, it moves around and is often debated in other elements of the refugee definition. These complex connections make everything unstable – not only ‘discretion’ itself but also its rejection. ‘Discretion’ functions as a patch for all these instabilities that emerge from the refugee definition. The book shows both the breadth and the depth of the ‘discretion’ logics: it is an unresolvable issue at the core of the refugee concept that surfaces at different layers, in different locations and in different ways – if it is put down in one place and form it will resurface in another place and in another form.

Chapter 8 addresses the competing definitions of the Convention ground ‘particular social group’: the ‘protected characteristics’ approach and the ‘social perception' approach. Whereas both are capable of encompassing sexuality-based claims, they each hold the potential for ‘discretion’ reasoning in different ways. The ‘protected characteristics’ approach is designed to exclude ‘trivial’ claims. If claimants fear harm for what is considered a non-fundamental aspect, they can be returned to be ‘discreet’. The ‘social perception’ test in contrast, which would more appropriately be called the ‘persecutor’s perception’ approach, in principle precludes any a priori exclusion of certain particular social groups. In this approach, it is the persecutor who defines what is persecuted. Yet the chapter shows that even this broader approach is prone to ‘discretion’ logics: the limit that is reverted to is the ‘singling out’ requirement, providing protection only to those who are singled out for persecution whereas those deemed able to pass unnoticed can be returned. As such, in both approaches, the protected group is not the same as the persecuted group, such that those who are persecuted but not protected must remain ‘discreet’.

Chapter 9 argues that a new layer of complication was added with the so-called human rights–based approach to refugee law. The challenge is that refugee protection is lesser in scope than human rights protection. That ‘lesser’ protection is usually located in the persecution element: while reference to human rights has opened up refugee law to a wider range of harms, not every human rights violation reaches the threshold of persecution. The chapter shows that one consequence of the human rights–based approach is the risk of conceptually merging Convention ground and persecution. According to the human rights approach, a violation of the freedom of religion or the freedom of thought can amount to persecution. In other words, prohibitions on the expression of the Convention grounds in themselves can be considered persecutory. The chapter reveals that this has led to the invention of an alternative or additional kind of harm that stands next to the harm that is inflicted in case the claimant does express the Convention ground – or is discovered in another way. This approach builds on and reinforces ‘discretion’ reasoning, because it distinguishes on the basis of the claimant’s future behaviour and requires ‘discretion’ under certain circumstances.