ABSTRACT
Constructing convincing legitimacy claims is important for securing the stability of authoritarian regimes. However, extant research has struggled to systematically analyse how authoritarians substantiate their right to rule. We analyse a novel data set on authoritarian regimes’ claims to legitimacy that is based on leading country experts’ assessments of 98 states for the period 1991–2010. This analysis provides key new insights into the inner workings and legitimation strategies of current non-democratic regimes. Closed authoritarian regimes predominately rely on identity-based legitimacy claims (foundational myth, ideology and personalism). In contrast, elections fundamentally change how authoritarian rulers relate to society. In their legitimacy claims, electoral authoritarian regimes focus on their ‘adequate’ procedures, thereby mimicking democracies. All regimes also stress their purported success in proving material welfare and security to their citizens.
Acknowledgements
Reverse alphabetical order, both authors contributed equally. We would like to thank André Bank, Alexander Dukalskis, Johannes Gerschewski, Maria Josua and the participants of the 2016 ECPR Joint Sessions workshop ‘Legitimation in Non-Democracies: Concepts, Theories and Empirical Evidence across Regime Subtypes’ in Pisa for their valuable comments.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes on contributors
Christian von Soest is a Lead Research Fellow and head of the Peace and Security Research Programme at GIGA German Institute for Global and Area Studies. His research investigates international sanctions and other foreign policy interventions, conflict processes, and the domestic and international politics of authoritarian regimes.
Julia Grauvogel is a Research Fellow at GIGA German Institute of Global and Area Studies. Her work focuses on domestic opposition movements in countries under sanctions and the legitimation strategies of authoritarian regimes.
Notes
1. We use the terms authoritarian, non-democratic and autocratic interchangeably throughout this paper.
2. For better readability we use both terms interchangeably.
3. Compare Scharpf’s (Citation1997, Citation1999, chap. 1) understanding of input legitimacy as a core function of democratic legitimacy that focuses on institutional arrangements and governing processes. Relatedly, throughput legitimacy signifies the actual integration of people in decision-making processes (Schmidt, Citation2013). We refer to such claims to legitimacy as procedure-based (Beetham, Citation1991).
4. To be sure, Levitsky and Way (Citation2010, p. 16) prefer the concept of competitive authoritarianism to the broader notion of electoral authoritarianism. Schedler (Citation2002, p. 47) himself subsumes ‘competitive’ and ‘hegemonic’ authoritarian regimes under the broad category of electoral authoritarianism.
5. For want of a better term, we follow the terminology used by Diamond (Citation2002), Howard and Roessler (Citation2006) and Freedom House. As Diamond (Citation2002, p. 22) affirms, these electoral regimes ‘fail to meet the substantive test’, particularly when it comes to safeguarding civil liberties.
6. The Wahman et al. (Citation2013) data set, used in this paper to separate authoritarian from liberal democratic regimes, is also based on a broad understanding of authoritarianism and a high threshold for liberal democratic regimes (they use a combined measure of Freedom House and Polity IV).
7. For the identification of country experts based on their prior publication record with respect to the countries and issue in question, we relied on the leading comparative politics journals, including Comparative Politics, the Journal of Democracy, and Democratization. In addition, we also searched the key area studies journals for each region.
8. We very much appreciate the invaluable responses of these experts as well as their extensive and extremely helpful comments.
9. In earlier versions, Hadenius and Teorell (Citation2005, Citation2007) used a 7.5-point threshold.
10. The only exception is Singapore.
11. We work with the original thresholds of 2 and 6; in their 2009 version (Roessler & Howard, Citation2009), Howard and Roessler used thresholds of 3 and 5.
12. In contrast to Howard and Roessler, who examine authoritarian elections, our unit of analysis is the regime spell. For details on all regime spells, see Online Appendix C.
13. The three dimensions are significantly correlated at p = 0.01 as follows: foundational myth–ideology = 0,643380711, foundational myth–personalism = 0,410285827, ideology–personalism = 0,360149604.
14. Only verbatim quotes and major points of information from the survey are explicitly referred to as coming from the RLES. Due to the assessments’ sensitive nature, the list of experts cannot be made publicly available. Many experts only participated because their anonymity was guaranteed.
15. Note again that we use a broad understanding of ideology.