Split into two articles, the Empirical Paper Series on Secession and Constitutionalism explores the relationship between constitutionalism and secession. The first paper of this series examined the relationship between, on the one hand, the constitutional prohibition and legalisation of secession – ie indivisibility and secession clauses – and, on the other hand, the prevalence of secessionist movements. This second article addresses the very heart of the doctrinal debate between the Indivisibilist and Secessionist schools, namely the relationship between a constitution’s framework towards secession and the actual occurrence of secession. Between 1900 and 2018, 24 subnational units seceded from a total of 16 central states. Of these 24 cases of secession, 25 % occurred in central states where secession was prohibited, 13 % in states whose constitution was silent on the issue, and 63 % in states whose constitution contained a right to secession, which indicates a very strong statistical relationship between secession clauses and the actual occurrence of secession. Yet, a closer look into the history of these provisions suggests that the causal relationship between secession clauses and secession itself – if any – goes in a direction opposite to what one would assume. Indeed, not a single one of the 13 secession clauses identified in this paper series ever gave rise to a pro-independence movement. In the vast majority of cases, it is actually the pre-existence of secessionist groups that forced the constitutionalising of a right to secession, either downstream, to pacify a violent secessionist conflict, or upstream, to accommodate a region that set the constitutionalising of such a right as a precondition of its joining a new country that would otherwise not have existed in the first place. Liechtenstein is the only recorded example of a state that constitutionalised a right to secede without being faced with pre-existing secessionist tensions, and none has arisen since secession became legal in Liechtenstein.