Work

Applicability of quantitatively predictive logical models at subnational level of governance: Testing the Seat Product Model on Polish, Czech, Hungarian and Slovak regional elections

Year: 2025
Type: article
Abstract: This paper extends Taagepera's predictive logic (quantitively predictive logical models) to test whether a Seat Product Model derived from national level of governance and based on institutional variables alone – the country's average district magnitude ( M ) and assembly size ( S ) – has the same predictive capacity on subnational levels of governance. To address the gap in subnational-level research of quantitively predictive logical models, we test the Seat Product Model on an original new cross-sectional dataset from elections to regional legislatures in four Central European countries (Poland, Czechia, Hungary, Slovakia) between 1994 and 2024 (i.e. all elections since the establishment of their respective self-governing regions). We analyse 403 regional elections overall to verify the logical model's capacity to predict the number of seat-winning parties, the seat shares of the largest parties, and the effective number of parties. We find that when confronted with the empirical data, the Seat Product did not reach the same level of accuracy in regional elections as in national elections. However, as the results in some cases were not too far from theoretical expectations (especially after incorporating the temporal dimension, which suggested that electoral system outcomes only tend to align with institutional expectations after several electoral cycles), further research should aim to expand the Seat Product to include the multi-level perspective to control for any effects of the institutional arrangements that exist on the higher (national) level of governance. Indeed, regional party systems are embedded systems and can only be fully understood when broader institutional context is taken into account.
Language: English
Open Access status: hybrid
Estimated APC paid: $2,840
OpenAlex ID: W4417060437
FWCI: 11.78
Cites: 29
Cited by: 3
Related to: