Thick or Thin?: An Empirical Intervention

Abstract

This note provides evidence about the relationship between state legitimacy and liberal rights in 72 states, and compares the strength of that relationship to other plausible legitimacy sources. It concludes that liberal rights have an equal but non‐superordinate status to good governance and material development. The ‘thick’ versus ‘thin’ argument about an appropriate global theory of state legitimacy may be misplaced. In its place, we should adopt a partial and plural approach.

Key Words:

• Legitimacy
• governance
• liberal rights
• consultation hierarchy
• urgent rights
• development
• tradeoffs

Notes

1. Given the high level of inter‐relatedness between many plausible variables, as well as the probably weak but nonetheless significant inverse causality from legitimacy to those variables, simple correlations offer the best starting point for such an investigation. In order to meet the demanding assumptions of multivariate analysis, one would need to develop instrumental variables to cope with endogeneity, or else develop longitudinal data to ensure temporal ordering.

2. I define the bottom two categories of the seven‐point scale (defined as ‘unlimited authority’ or an intermediate category above it) as failing the decent consultation hierarchy test. Above that, there are ‘slight to moderate limitations on executive authority’ and the decent consultation hierarchy test is met.

3. ‘Uzbekistan: Lifting the siege on the truth about Andizhan: Summary’, Amnesty International, 20 September 2005. The official death toll was 187.

4. Using either multiple regression or a combined variable of governance, development, and liberal rights gives a squared correlation coefficient (r²) of approximately 0.53, meaning that universalistic sources can explain roughly half of legitimacy variations across the 72 states.

5. One obviously concerns the conceptualization and measurement of legitimacy and of the three main variables themselves. In addition, while micro‐level evidence can be cited in support of the view that these correlations reflect causal relationships, an outstanding question is whether the linear model is the best approximation of the strength of those relationships. In addition, the large amount of ‘noise’ that characterizes any explanation of legitimacy outside of the more predictable West raises the question of how accurate this picture is in most countries and whether it would change across time.

6. An obvious list of uncertainties about the moral reasoning behind this theory of legitimacy would include: whether individuals or peoples are the appropriate subjects; the definition of procedural fairness through which tradeoffs are decided; whether liberal rights properly defined would include good governance itself; and the role of international society in second‐guessing domestic legitimacy evaluations.