The project aims to provide a statistically sound forecast of the 2026 Hungarian parliamentary elections and to further develop and validate methodological innovations that improve the accuracy of survey-based estimates in real-world settings. The research builds on the research group’s previous results, which introduced a novel post-stratification procedure based on political interest to reduce bias in survey estimates. A detailed description of the related, earlier EKA-funded project is available here.

During the project, large-scale data collection will be carried out in cooperation with a professional survey agency, complemented by advanced statistical modelling and the application of the “poll of polls” method, which aggregates results from multiple opinion polls. This approach makes it possible to produce election forecasts that are more stable and reliable than conclusions drawn from a single survey.

In addition to methodological development, the project seeks to achieve a better understanding of structural changes in voter behaviour and to strengthen professional dialogue in Hungarian election research. As part of this effort, the research group will organise a national scientific workshop at ELTE prior to the 2026 elections, where the forecasts and the applied methodology will be presented.

As a result of the project, a validated and further developed estimation model will be produced, scientific publications will be prepared, and the research will contribute to the methodological renewal of election forecasting in Hungary.

The project is funded by the University Excellence Fund (EKA) of ELTE.