Institutions vs. divisions: A comparative assessment of conflict levels in unitary and decentralized states (1990-2020)

Governing multinational states is not a walk in the park. Facing ethno-territorial debates, secessionist pressures, and regional sensitivities, many of these countries are burdened by political conflicts, gridlocks, and instability. As such, their internal diversity is often said to come with a ‘heterogeneity cost’ (Alesina and Spolaore, 1997). Additionally, due to their institutional and historical features, some multinational states are argued to be even more conflict-prone than others. In this respect, many scholars have stressed the importance of -amongst others- the level of decentralization, the number of substates, and the centrifugal or centripetal nature of these states (‘coming together’ or ‘holding together’). This paper seeks to address two questions: (1) Are multinational states indeed more conflict-prone than other states – as is often assumed? (2) Are certain systems indeed more conflict-prone than others? To date, empirical studies on these questions are scarce and mainly preoccupied with indicators of violent conflict (civil war, terrorism) or the survival of states as such (e.g. Brancati, 2009; Cordell and Wolff, 2016; Horowitz, 2000). Meanwhile, very little attention has gone to mundane conflicts between the political parties and actors that govern divided states. Such everyday political clashes deserve our attention too. It is well-known within the field and stressed by consociational theory (Lijphart, 2002, 1969) that the everyday behavior of politicians is key to the stability of divided states. Assessing such behavior is especially important in a context in which rising levels of electoral volatility and party fragmentation put political cooperation under pressure (Siaroff, 2019). This study aims to address this gap. The paper presents a novel, unprecedented large-N dataset on three decades of everyday political conflicts in over 100 national cabinets in 10 countries with varying levels of internal heterogeneity and different institutional systems: Belgium, the Netherlands, Canada, Denmark, France, Germany, Italy, Spain, Switzerland, and the UK (1990-2020). Using a recently developed approach to measuring cabinet conflicts (Vandenberghe, 2022), conflicts at the central level are tracked by hand-coding the Political Data Yearbooks of the authoritative European Journal of Political Research (EJPR). Using this data allows us to compare countries and discern five of their features: the degree of heterogeneity (multinationalism), the federal/unitary nature, their level of decentralization, the number of substates (bipolar/multipolar), and their centrifugal/centripetal nature. Doing so allows us to add both empirically and theoretically to ongoing debates on the so-called paradox of federalism (Erk and Anderson, 2010) and on conflict and conflict-management in multinational states more generally (Coakley, 2009; Wolff and Yakinthou, 2012; Cordell and Wolff, 2016; Keil and Anderson, 2018). Furthermore, the dataset presented here offers great potential for different fields, most notably that on coalition conflicts (Bergman et al., 2021).