Tatiana Kostadinova's critical remarks and questions are spot on. Some of them had been on my mind while researching the book, and they all deserve to be raised. I would argue that the possible ways by which my case study could be viewed as relevant to more generalizing approaches to corruption are twofold.

First, my case study of Lebanon identifies a plausible causal variable, that of postwar political settlements. Although the precise qualities of such settlements vary, it indeed is striking that since the 1990s, templates for postwar conflict management, as propagated by the agenda of the “liberal peace” in multiple contexts, are nearly always informing complex, highly inclusionary politics that more often than not produce institutional stalemate, corruption, and a fragile peace at best. True, a case study by itself cannot generate a theory. But neither can large-n quantitative studies persuasively identify causal mechanisms without unpacking that messy black box between cause and effect. In this sense our methodologies are complementary.

Second, “corruption” as a dependent variable is, in my view, of such an unwieldy nature that uncritically lumping together highly varied phenomena and forms of behavior at different times and in different places will render sound analysis ineffective. Before arriving at the level of generalization (and indeed regression analyses), even within single-country studies, one needs to demonstrate that practices deemed corrupt sufficiently cohere to warrant causal explanation. In this sense, my case study is intended to remind generalists that corruption remains an aggregate term for a wide variety of practices. Consequently, it cannot be assumed that there are clear “causes” for it; this first has to be demonstrated. Other driving factors indeed compete with my own explanation grounded in postwar politics, including the ones suggested by Kostadinova. I am not arguing that these do not matter. For instance, a weak “civil society” may have contributed to one corruption scandal or the other. Yet I hope to have demonstrated that none of such alternative factors explain all corruption cases that I studied. I argue that Lebanon's postwar political settlement appears to cover them better. More specifically, tolerance levels of corruption, constructed as one such alternative independent variable, are extremely hard to measure. Whose tolerance matters? And what if people accuse each other of such high tolerance levels but are quick to denounce corruption themselves, as many opinion surveys in Lebanon and the Arab world at large indicate?

Finally, there certainly is a dilemma in designing inclusionary postwar political settlements that cause flawed institutions and corruption and yet may be imperative for bringing about some sort of truce or relative stability. Yet only by acknowledging the narrow parameters that such settlements allow for “administrative reform” and building sound state institutions can one assess their costs and risks, and hope to address them. In Lebanon, this would have to entail addressing its highly inclusionist and dispersed distribution of power, for instance by introducing office rotating mechanisms and by gradual prosecution of former warlords-turned-ministers; empowering parliament by way of electoral laws fostering ties between members and their constituencies; and perhaps most importantly, building a national consensus whereby seeking support or patronage from outside powers becomes a political liability and not, as is the case now, a valuable resource in a crowded, gridlocked, and potentially violent political game. Admittedly, these are no easy tasks, certainly not with a civil war in neighboring Syria knocking at Lebanon's door.