Classifying Political Regimes, 1900-2019 a New Dataset Based on V-Dem

• In Event: Regime Type

Abstract

This paper contributes to the ongoing debate about how to classify political regimes and how to detect change from one regime type to another. We use the Varieties of De-mocracy (V-Dem) dataset in a novel way to improve on previous classifications. We define democratic, semi-democratic, electoral authoritarian, and closed authoritarian regimes. Based on 19 expert survey questions from V-Dem, we then classify regimes into these four categories. Countries must meet a series of minimal thresholds on these 19 questions to be coded as democratic, semi-democratic, and electoral authoritarian. Countries that do not satisfy these necessary conditions are coded as closed authoritarian.
Our classification of regimes has four advantages. First, it is based on data from the V-Dem project, which has advantages over previous measures of democracy for classifying regimes. V-Dem is methodologically more sophisticated than Freedom House and Polity, and it allows for a more nuanced measure of regime characteristics than previous indicators (Coppedge et al., 2018; Pemstein et al., 2018). V-Dem uses 3200 coders, and their case knowledge is useful for develop-ing valid categorical classifications. The large number of knowledgeable coders should reduce measurement error and enhance validity compared to previous attempts to measure qualitative as-pects of democracy. Because V-Dem is updated annually, our measure can be regularly updated by scholars who want to use it after the publication of our paper. In this respect, it has advantages over the excellent works of Boix et al. (2013), Cheibub et al. (2010), and Geddes et al. (2014).
Second, our classification and dataset have advantages over Lührmann et al.’s (2018) Re-gimes of the World (RoW) dataset, which is the most prominent categorical classification that is based on V-Dem. Lührmann et al. begin with Dahl’s (1971) definition of democracy, but their op-erationalization of electoral democracy leaves aside important dimensions of his definition. For two of their three necessary conditions, they use cut points that are insufficient to characterize a regime as democratic. Their primary means of distinguishing between authoritarian and democratic re-gimes is requiring that democracies meet a value of 0.50 on V-Dem’s electoral democracy score. This threshold, however, has no substantive meaning. It does not reveal a great deal about how a regime functions since, by construction, lower values on some dimensions of democracy may be compensated by higher values in other dimensions. Thus there is a gap between their concept and their measurement strategy. We resolve that gap by beginning with a definition of the four regime types and using thresholds based on 19 V-Dem questions to operationalize them.
Third, hybrid regimes have long been an important reality, and their number has increased since the end of the Cold War. Our four-fold classification takes into consideration these hybrids in a way that a simple democratic/authoritarian dichotomy does not. Some of the most important scholarship on political regimes in the last fifteen years has focused on ways in which electoral au-thoritarian regimes differ from closed authoritarian regimes. Many scholars have emphasized the importance of hybrid regimes since the end of the Cold War, so it is useful to go beyond a democ-racy/autocracy duality.
Fourth, our measure is based on a broad procedural definition of democracy that goes be-yond electoralism. We define democracy as a regime in which power is allocated via free, fair, broad, and decisive elections, and exercised within limits. These limits are of two orders. On one hand, power is limited by a separation of powers and by oversight mechanisms. On the other hand, democracies must respect basic individual liberal rights of expression, movement, and organization.
A definition and operationalization that go beyond electoralism is useful in an era of wide-spread electoral authoritarian regimes that meet the criteria of electoralist definitions of democracy. Minimal electoralist definitions do not capture the most important distinctions between electoral authoritarian regimes and democracies. Likewise, for many scholarly purposes, it is useful to dis-tinguish between democracies and borderline cases that we call semi-democracies.
The paper is organized as follows. Section 1 defines democracy. Section 2 operationalizes democracy and defines and operationalizes semi-democracy, electoral authoritarianism, and closed authoritarianism. The operationalization of the four regime types is based on 19 variables taken from V-Dem or constructed on the basis of V-Dem. Section 3 shows the distribution of regime types over time. Sections 4 to 7 discuss the advantages of our classification. Section 8 compares it empirically to other classifications. Section 9 runs a replication exercise on the relationship between the level of development and democracy. Section 10 concludes.

Authors

• Scott Mainwaring, University of Notre Dame

• Fernando Bizzarro Neto, Harvard University

• Natan Skigin, University of Notre Dame