Over the past few decades, several regions within liberal democracies have voted on their independence, as secession referendums are an instrument that can help solve the tension between self-determination and state sovereignty. While unilateral referendums were held in some regions, in other cases the parent state agreed to allow voting. Both groups have points in common, but they are also very heterogeneous and correspond to very different periods in history, as acknowledged by the authors of the study themselves.

So, what are the determinants of agreed referendums on independence in liberal democracies? An academic article recently published in the journal Democratization, involving Marc Sanjaume-Calvet, a professor of Political Theory at the UPF Department of Political and Social Sciences, together with Jean-Baptiste Harguindéguy (Pablo de Olavide University of Seville) and Almudena Sánchez Sánchez (Distance University of Madrid, UDIMA) carries out a systematic analysis of liberal democracies and referendums on independence, both agreed and unilateral, since 1945, and proposes a general model designed to predict the willingness of a liberal democracy to accept them, and under what conditions.

The work, based on a proposal by the political scientist Matt Qvortrup but applied to the decisions of central rather than regional governments, concludes that there are two determining factors when explaining under what conditions the parent state accepts a referendum on independence.

The first is the perception of competition held by the state government (that the parent state can see that it needs to hold the referendum in the face of an electoral or military threat that may cause it to lose power) and/or the proximity of its electorate to the demand for a referendum. Regarding the perception of competition, it is key that the referendum should be perceived as a necessary tool to maintain or consolidate power: “In the case of Spain, for example, the need for parliamentary support from peripheral parties of a minority government counts as a threat to losing power”, Marc Sanjaume asserts.

Marc Sanjaume: “The characteristics, content, consequences and scope of a potential referendum would be decisive. Holding the referendum in Catalonia with a prior agreement on these aspects, or in Spain as a whole, would probably be closer to the preferences of the average Spanish voter”

With regard to the fact that the central government sees that this is a policy that has a certain proximity to its electorate, the UPF professor explains that outside Catalonia, in the rest of the State “this could be the case of Podemos/Sumar and some of the PSOE, but currently, a referendum on independence is clearly low down in the list of preferences of the average Spanish voter”.

And he adds: “However, they are modulable preferences, as shown by the comparative policy. The characteristics, content, consequences and scope of a potential referendum would be decisive. Holding the referendum in Catalonia with a prior agreement on these issues, or in Spain as a whole, as France did in the case of Algeria in a very different context, or as has occurred in the case of constitutional referendums or conflict resolution in several countries, would probably be closer to the preferences of the average Spanish voter”.

For Marc Sanjaume, right now, part, but not all of the model, is fulfilled in Spain. In the case of the Cameron Government in 2012 in the United Kingdom, for example, “the two criteria were partially met, since the Conservatives needed to set themselves apart from Labour, and the proposed referendum in Scotland (and Scotland’s independence) was not entirely unpopular with the rest of the United Kingdom, unlike the case in Spain”, the UPF professor asserts.

Apart from competition and/or proximity to the electorate, a second determinant for a state government to accept holding a referendum depends on the expectation of the low cost of eventual secession, at least in terms of population, geographical area and natural resources. According to competition/proximity and costs analysis, the authors point out four scenarios of probability that arise when combining these determinants.

70 agreed and non-agreed referendums on independence since 1945

Within a broader database containing all referendums on independence held since 1945, the authors focus on the seventy held in liberal democracies after World War II, 58 of which were agreed to and accepted by the parent state, with 33 of the latter leading to the creation of a new state.

Some of the aspects highlighted by the authors are the variations in the use of referendums by secessionist groups over time: the largest number of referendums on independence occurred during the decolonisation process as of the 1950s.

Likewise, most cases are concentrated in just a few countries: the three major powers (the United States, France and the United Kingdom) account for 50 cases. Moreover, several cases correspond to colonial or post-colonial contexts, that is, to territories that either had direct colonial rule from the metropolis and/or formerly depended on it.

Among other conclusions, the authors point out that in the most populated territories, the probability of holding an negotiated referendum is far lower. In fact, since 1945 there have been few cases of territories with more than 10% of the population of the parent state obtaining an agreed referendum.

Outside the colonial realm, there have been only seven cases (Quebec has voted twice), and only one of them led to the formation of a new state (Montenegro). In addition, in only three cases (Quebec, Montenegro and Scotland) did the referendum take place in geographically contiguous territories: the rest corresponded to island or remote territories.

“We observe a trend over time towards a lower acceptance of the demands for a referendum on independence, and we can predict greater territorial conflict around this issue”

On the other hand, interestingly, the authors do not find a direct relationship between democratic quality and the acceptance of the referendum: they claim that even if democracies maintain a growing democratic quality -a controversial assumption given the current trends of democratic backsliding- this does not mean that they become more likely to accept referendums on independence .

“We observe a trend over time towards a lower acceptance of the demands for a referendum on independence, and we can predict greater territorial conflict around this issue. Recent events in the cases of Scotland and Catalonia point in this direction”, the authors conclude.

Reference Work: Sanjaume-Calvet, M., Harguindéguy J.B., Sánchez Sánchez, A. (April 2024) “Self-determination vs. state sovereignty. What are the determinants of agreed-upon independence referendums in liberal democracies?”, Democratization

https://doi.org/10.1080/13510347.2024.2331709