The Netherlands provided the perfect setting for Chu’s research. Unlike the United States’ two-party system, where voters typically have established party preferences and primarily need mobilisation, Dutch elections involve dozens of parties competing for visibility, creating a dynamic environment in which to study persuasion.

To understand voters' exposure and responses to political targeting in real-life settings, Chu tracked their browsing history on Facebook and requested their political ad exposure data across various social media platforms in the lead-up to the 2021 national elections in the Netherlands.

Repeated exposure

By tracking people’s real-life exposure to political ads on Facebook, Chu found that their true power lies in repetition. A single ad rarely changes a person’s mind, but repeated exposure does increase the likelihood of a voter supporting the party behind the message. These effects may seem small in isolation, yet in close elections, even slight shifts can prove decisive.

Chu’s study also highlights that some citizens are more susceptible than others. People with less political knowledge or weaker understanding of online privacy were found to be more easily swayed by repeated exposure to advertising. This raises important questions about equity in the democratic process, as those who are least equipped to critically evaluate digital messages may be seeing them the most.

Personal relevance and political identity

Relevance emerged as another crucial factor. Ads that spoke directly to voters’ own concerns – whether climate change, healthcare, or the economy – were consistently judged more persuasive than generic appeals. When issue-based appeals were combined with references to a voter’s favoured party, the impact was strongest, suggesting that targeting both personal relevance and political identity makes a message particularly convincing.

The reality of political targeting

Interestingly, despite the fear that political microtargeting relies on highly invasive use of personal data, the study found that Dutch parties typically employed much simpler methods and less accurate targeting. Strict European privacy laws and limited campaign budgets meant that most parties relied on broad categories rather than detailed psychometric profiles.

The future of political targeting

Taken together, the research paints a picture of political targeting as a double-edged sword. On one hand, it can enhance democracy by helping voters engage with issues that matter to them and by giving smaller or newer parties a cost-effective way to reach the public. On the other, it can also magnify inequalities, since larger parties with more resources are able to maintain greater visibility, and it may contribute to voter manipulation, especially among less informed groups.

Meta recently announced its plan to halt all political, electoral and social issue advertising on Facebook and Instagram within the European Union. Will that make political targeting a thing of the past? Chu argues that in the new era, political targeting will reemerge in a new form. When it does, transparency, voter education and responsible regulation will be key to ensuring that online political ads strengthen rather than undermine democracy.

Chu urges both caution and perspective. Political advertising is not the all-powerful force that some fear, but neither is it harmless background noise. Its influence is small but cumulative, building gradually through repetition and working most effectively when messages feel personally relevant. ‘Understanding these subtler effects is crucial,’ Chu concludes. ‘Even minor shifts in voter behaviour can change election outcomes.’

Defence details

Xiaotong Chu will defend the dissertation 'The persuasiveness of political targeting'. Supervisors are Prof. S. Kruikemeier and Prof. S.K. Lecheler. The co-supervisors are Dr L.P. Otto and Prof. C.H. de Vreese.

Time and location

Thursday, 2 October, 14:00 in the Aula.