This chapter concludes by discussing the scholarly and policy implications of the book’s findings. First, it discusses implications for data on conflict onset, arguing that the book has shown that omissions in existing data prevent current work from learning about the start of conflict, and suggests avenues for generating better data. Then, it argues that these omissions have created major inferential issues in prior work about the role of ethnicity in conflict onset. It uses evidence from Uganda presented in prior chapters to show how this problem plagues data used in prominent study that found a strong relationship between ethnic exclusion and rebellion onset. Third, it argues that this book has shown the importance of civil intelligence institutions to state building and notes the thorny ethical issues associated with this finding. Finally, the chapter considers how the early stages of rebel group formation would differ in stronger state contexts, arguing that in such contexts barriers to new rebel entry are higher and thus we should expect rebellion to be more rare and more explosive.

This chapter introduces the puzzles at the core of the book: How does rebel group formation initially start, and why do many rebellions fail before becoming viable challengers to a government? Few prior studies have systematically answered this question because rebel groups typically form in remote areas of weak states and are clandestine. These features of nascent rebellion have thus limited knowledge about conflict onset; most recent conflict onset studies rely on datasets that do not capture the start of violence. This chapter presents the book’s core argument; describes its contributions to literatures on conflict, ethnicity, and state building; describes its methodological approach of retracing all incipient rebellions in Uganda since 1986, and probes the relevance of findings from Uganda in other Eastern and Central African states.