Approaches to Researching Territorial Autonomy: Reflections on Methodology and Data

Felix Schulte

The European Centre for Minority Issues

Research on territorial autonomy is a dynamic and growing field. Both Google Scholar and the Social Science Citation Index show significant growth rates since the end of the Cold War. Yet, it is essential for us, as an academic community, to recognize that quantity does not equate to quality. In our pursuit of knowledge, self-reflection becomes paramount. We must conscientiously examine our scientific efforts and acknowledge our progress whilst also identifying existing gaps and weaknesses. This introspective process should extend beyond reflections on theoretical approaches to include the methodologies and research designs used, as well as the data collected. I will focus on the latter aspects in this post. As we navigate this reflective journey, we become more aware of what we truly understand and where our understanding falls short. This awareness is critical, not least because our research holds practical relevance, considering, for example, the significant number of peace agreements that include provisions for territorial autonomy. Condensing my reflections into four observations, I shall now delve into each of them in the following sections.

Conceptual confusion

There is an unnecessarily large variety of conceptualizations of “territorial autonomy”. Much like other contested concepts, it is subject to diverse understandings, with definitions ranging from minimalist to broader, to operational, and more abstract ones. However, the level of conceptual confusion surrounding territorial autonomy is overwhelming and complicates the research on territorial autonomy and its effects. Too often, the lines between different functional logics of power-sharing and power-dividing become blurred, mistakenly treating territorial autonomy as just another form of power-sharing (“territorial power-sharing”), irrespective of its sui generis character. This conceptual ambiguity is exacerbated further by the use of overly broad terminology (“asymmetrical federalism”) or even misnomers. The phrase “non-territorial autonomy”, frequently used to distinguish minority rights from territorial autonomy, is one example of the latter with authors ignoring the fact that most cases labelled as such neither constitute instances of political rule nor have a ‘non-territorial’ character. The direct empirical implications of this ongoing conceptual confusion are evident: despite being designed to measure the same concept, some of the most frequently used autonomy indices exhibit (relatively) weak cross-correlations. Consequently, the choice of an index may yield different outcomes. Conceptualization and operationalization matter. If scholars employ diverse terminologies to describe the same phenomenon or, worse still, use identical terms to refer to distinct phenomena, the process of scientific communication transforms into an unproductive exchange.

Empirical coverage

The second issue that frustrates is that there is (still) a considerable emphasis on studying well-known autonomy cases as well as ‘relevant’ autonomous groups. Qualitative research sometimes gives the impression of being mere Catalonia, Scotland, or South Tyrol research with a one-sided focus on these selected cases. The geographical scope tends to be limited to Europe, post-Soviet Eurasia, and the Balkans, while sub-state regions in other parts of the world receive much less attention. This Eurocentric focus is concerning as it does not accurately reflect the actual spread of territorial autonomy arrangements. Despite remarkable advances in recent years, the existing landscape of large-n data projects in the field is still paper-based and highly fragmented, limiting comprehensive insights. Currently, only two widely used datasets have become the go-to resources: the Regional Authority Index (RAI) at the entity-level, and the Ethnic Power Relations Data (EPR) at the group-year level. Despite their popularity, these datasets come with their own set of limitations. The EPR dataset, for instance, only considers ‘politically relevant’ autonomous groups, and the RAI dataset is constrained by its inability to explore territorial autonomy as a sui generis type.

So far, autonomy researchers have focused mostly on formal institutions and de jure autonomous provisions. In contrast, de facto rules and practices, attitudes, and elite behavior receive far less attention in empirical studies. Survey data that could be used to examine citizens' attitudes across autonomous regions and over time is scarce. However, for a comprehensive test of our theories, including the basic assumption of consociationalism (i.e., that institutional reforms promote moderate behavior), it is essential to improve the availability of such data. Moreover, to enhance our understanding, increase our confidence in our theories, and validate the reliability of our empirical findings, I believe it is crucial for us to prioritize conducting comprehensive meta-studies and replications that are currently missing in the field. By addressing this gap, we can gain a deeper insight into the core findings so that we truly comprehend and identify the central questions that are (not) strongly supported by empirical evidence.

The quali-quanti gap

This symposium is titled “The Many Faces of Territorial Autonomy.” It is not only autonomy that has many faces, but autonomy research. The field is relatively fragmented into several isolated academic islands. One contributing factor to this fragmentation lies in the diverse methodological approaches adopted, often falling at opposite ends of the quantitative-qualitative spectrum. Quantitative works tend to overlook case-study research and vice versa. While this statement may sound drastic, there is certainly a lack of middle ground; few studies adopt mixed-methods approaches to bridge this gap. There is a scarcity of research that goes beyond either case-based analyses or large-n regression-based studies. We must adopt a broader range of methodological approaches to build further confidence in our findings and grasp the complexity of individual cases while, at the same time, being able to identify general patterns, measure effects, and generalize beyond the individual case.

Causal inference

In our discussions, we often take for granted that autonomy reforms produce various normatively desirable outcomes. However, exploring causal relationships in our field poses significant challenges. First, we encounter the complexity of dealing with ambiguous and contested concepts (autonomy, peace, social cohesion, etc.). Second, the causal relationships are rarely straightforward. For instance, determining whether autonomy leads to peace or vice versa presents a complex challenge. Additionally, autonomy reforms may serve as a necessary but not sufficient condition for mitigating conflict, or may be just one of several potential paths. While we may have well-founded theories and empirical evidence from our process tracing or regression analysis, the comprehensive investigation of alternative explanations and reverse causality often remains inadequate. Nonetheless, this goes beyond abstract epistemological discourse. It has serious implications. For instance, we observe higher conflict levels in regions with greater autonomy. Is there conflict despite high autonomy levels, or is the conflict caused by autonomy? Though we have substantial evidence supporting the former, it remains a challenging question to test. Ignoring this can easily lead us to either overstate or underestimate the impact of territorial autonomy, as seen in many early studies on the autonomy-conflict nexus. It is thus crucial for us to delve even deeper into reflections on these issues, both in terms of theory development, research design, and the methods we employ in our investigations. Accommodative institutions have contributed to the decline in ethnic conflicts. This is not least a merit of the research that has been undertaken. The emergence of new and unresolved conflicts underscores the pressing need for our expertise.

This blog post is a result of a workshop of the network of scholars of territorial autonomy initiated by the Åland Islands Peace Institute. The network first met and discussed ‘The Many Faces of Territorial Autonomy’ at a workshop in Berlin, convened in cooperation with the Finland Institute in Berlin. The blog symposium is hosted by the IACL Research Group on Constitutionalism and Societal Pluralism: Diversity Governance Compared.

Felix Schulte is Senior Researcher and Head of Cluster “Conflict & Security” at the European Centre for Minority Issues (ECMI) in Flensburg, Germany

Suggested Citation: Felix Schulte, ‘Approaches to Researching Territorial Autonomy: Reflections on Methodology and Data’ IACL-AIDC Blog (26 September 2023) https://blog-iacl-aidc.org/territorial-autonomy/2023/9/26/approaches-to-researching-territorial-autonomy-reflections-on-methodology-and-datanbsp.