This article argues that what is seen as a justice cascade may in fact amount to merely an advocacy cascade, which has facilitated justice policies that democratising states would inevitably have pursued (and helped neutralise opposition), but which in causal terms has been far less influential on justice policy than is commonly assumed. Because transitional justice is generally presented in very idealistic terms, scholars in the field have begun to acknowledge that its virtuous effects are more easily presumed than proven (Van der Merwe, 2009, p. 121). Amongst advocates and activists in particular, one sees in the literature an emotional commitment to transitional justice that generally foregoes doubts about its overall efficacy even where isolated shortcomings are accepted. Policy has hitherto proceeded less from analysis to conclusions than from commitments to action. Some argue that ‘the commitment to advocacy has come at the expense of progress in empirical research’ (Vinjamuri and Snyder, 2004, p. 345) – the benefits of certain mechanisms are assumed instead of treated as empirical propositions to be proven rigorously. Because so many of the early debates about transitional justice took the form of partisan advocacy in the dichotomised days of the ‘justice versus peace’ and ‘truth versus justice’, prospective hypotheses about likely outcomes dominated the literature at the expense of retrospective assessments of what generally had or had not worked. For at least a decade, scholars have noted the paucity of studies systematically examining the correlation between transitional justice and social reconstruction. Subsequently, the literature has variously been criticised for its dependency on anecdote and hypothesis (Crocker, 2002, p. 541), analogy (Brahm, 2008, p. 3) and wishful thinking (Olsen, Payne and Reiter, 2010, pp. 25–26). Until recently, scholarship had primarily been based on single or dual mechanism case-studies and comparative qualitative case-studies of a limited number of states, which gave disproportionate emphasis to certain transitions or transition types conducive to study. This in turn made generally applicable policy conclusions difficult to elaborate. It has been argued that human rights research in general, and transitional justice research in particular, are enterprises directed at manufacturing legitimacy for their fields of practice (Gready, 2009, p. 159). Such endeavours must, however, retain at least a threshold plausibility. As a fear developed among practitioners and theorists about the damage to the credibility of transitional justice from wild, unsubstantiated claims, there has emerged in recent years a commendable attempt to clarify the causal relationships (if any) between individual mechanisms and general ends. By employing social science methodologies and hard data, a tentative literature has emerged on how to assess the impact of transitional justice (Thoms, Ron and Paris, 2010). The expectation is that this scholarship can chip away at falsity and overly ambitious claims. This article examines the extent to which two recent works do this. After surveying what is novel about their distinctive methodologies, Part II examines the primary difference between the works, namely the extent to which the work of transnational human rights activists has impacted on the decision of democratising states to pursue criminal accountability for crimes of the past. Part III explores alternative explanations for why states are seemingly more willing to undertake trials and the extent to which the works in question control for these variables. Part IV considers the extent to which the normative change both books note has impacted on the security dilemmas inherent in transitional accountability. The article concludes that transitional justice research has some distance yet to travel in disentangling correlation from cause.