Working Paper: CEPR ID: DP5098
Authors: Ori Haimanko; Michel Le Breton; Shlomo Weber
Abstract: In this paper we introduce the notion of stability threshold that quantifies the minimal returns to size sufficient to prevent credible secession threats by regions of the country. Severity of internal tension has been linked to degree of polarization of citizens' preferences and characteristics. We show that the increasing degree of polarization does not, in general, raise the stability threshold, even though this hypothesis holds in some asymptotic sense. Thus, somewhat counter-intuitively, the relation between polarization and the stability threshold is ambiguous. We also examine the question of the number of smaller countries to be created if the unity of the large country is not sustainable, and investigate the link between this number and the degree of the country polarization. We find that the stable number of countries also behaves non-monotonically with respect to polarization indices. However, monotonicity does emerge when the stable number is large, and the stable number decreases when polarization rises.
Keywords: clusters; polarization; secession; stability threshold
JEL Codes: D70; D73; H20
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| fixed-clusters polarization effect (F12) | stability threshold (C62) |
| variable-clusters polarization effect (C38) | stability threshold (C62) |
| number of clusters (large and heterogeneous) (C55) | stability threshold (C62) |
| centrally located cluster (C38) | stability threshold (C62) |
| polarization indices (C46) | stable number of countries (F55) |