Do Voting Advice Applications Affect Political Behaviour? Experimental Evidence from the Czech Republic

Voting advice applications (VAAs) are used by millions of voters across Europe, making it crucial to understand how their use influences political decision-making. Prior to the 2025 Czech Chamber of Deputies election, we conducted a pre-registered experiment on a representative sample of the adult population (n = 1209) to examine the effects of VAAs on political attitudes and behaviour. Participants completed a modified version of a voting advice application developed by KohoVolit.eu, and we experimentally manipulated the information available to them when measuring outcome variables. We find that participants incorporate information provided by the VAA by updating their estimates of agreement with political parties relative to their prior beliefs. However, participants' propensity to vote for a particular party remains unaffected in general, as VAA use influences vote choice only when users update their perceived agreement with political parties. Although participants whose policy preferences do not align with those of any relevant party are less likely to intend to vote, exposure to VAA results, which provides them with such information, does not affect their turnout intentions. Finally, we find no evidence supporting our hypothesis that monetary contributions to the VAA provider are positively associated with the extent of belief updating, defined as the change in voters' perceived policy overlap with a given party before and after completing the VAA. Monetary contributions are not explicable by any other characteristic such as gender, age, education, or political interest.