A new Washington University in St. Louis study takes on the Herculean task of studying the linguistic choices of more than 800 political parties from 87 democracies around the world, including the United States.
The authors analyzed 4 million Facebook posts made by political parties between 2016 and 2022 to develop the first classification of monolingual and multilingual parties around the world. The cross-national dataset provides the most comprehensive picture of parties’ multilingualism in contemporary democracies, revealing how and when political parties communicate with citizens in multiple languages.
The article “Speaking their language?: Multilingualism in party communication across democracies” — by WashU Arts & Sciences political scientists Christopher Lucas, Jacob Montgomery and Margit Tavits and recent graduates Dahjin Kim (PhD ’25) and Taishi Muraoka (PhD ’19) — was published in the April 2026 issue of the American Journal of Political Science.
In multilingual democracies, such as the United States, a political party’s linguistic choices shape political inclusion, mobilization and representation, according to the authors. When political parties and their candidates recognize and use minority languages, they can effectively dampen language-based divisions, which could in theory decrease the level of political conflict and instability, the authors write.
The authors focused on social media posts, rather than other sources of communication such as campaign speeches or party manifestos, because they provided the most direct, unmediated, fine-grained data for observing how parties talk to voters day-to-day, not just before or after an election.
Perhaps unsurprisingly, their research found that parties are more likely to adopt multilingual communication strategies in countries with greater linguistic diversity. According to the authors, this fits a basic “seat-maximizing” logic: as linguistic diversity rises, parties have stronger incentives to appeal to multiple language communities by posting in more than one language.
This relationship is not uniform, though. Parties operating in a majoritarian electoral system — where winning broad support is crucial — were more likely to adopt multilingualism than parties in proportional representation systems, a finding that was somewhat counterintuitive, said Tavits, the Dr. William Taussig Professor in Arts & Sciences.
“Conventional wisdom holds that proportional representation systems are better for minority inclusion. Our paper reverses this expectation for language representation: majoritarian systems increase parties’ incentives to cross language lines as diversity grows, because minority voters become pivotal to building a winning coalition,” she said. “Under proportional representation, minority electoral strength scales proportionally and predictably, so there is less urgency to court these voters with multilingual outreach.”
‘A party’s language strategy is therefore not just a communication decision; it is also a choice about what kind of democracy the party wants to help build.’
Margit Tavits
The authors found that political ideology — specifically left-leaning ideology — was a strong predictor of multilingualism. They also found a strong link between party-level and candidate-level behavior. Candidates nominated by multilingual parties tend to mirror their parties’ strategies, posting in multiple languages.
“Our research shows party-level language choices have downstream consequences beyond the party itself,” Tavits said. “Multilingual parties tend to produce multilingual candidates. In other words, party organizational culture around language shapes the entire supply side of representation in linguistically diverse democracies: who runs for office, how they campaign and ultimately which communities feel represented.
“A party’s language strategy is therefore not just a communication decision; it is also a choice about what kind of democracy the party wants to help build.”
Multilingualism in American politics
According to Tavits, the paper’s framework predicts exactly what we observe in the U.S.: left-leaning parties in linguistically diverse districts are more likely to communicate in multiple languages, especially Spanish. The paper also helps explain why Republicans are less likely to adopt a similar strategy.
“The ideological constraints we identify generate concerns about backlash among English-speaking monolingual Americans,” Tavits said.
Additionally, the finding about majoritarian electoral systems is highly relevant to the United States.
“Single-member districts create precisely the pivotal-minority dynamics we theorize,” Tavits said. “In competitive districts with substantial Spanish-speaking populations, minority-language voters can be decisive, which should push both parties toward multilingual appeals. Whether this actually occurs is an empirical question, but our framework generates clear, testable predictions at the district level that future research could examine using U.S. data.”
While this paper did not focus on how party communication and language use change before and after elections or in response to key political or social events, Tavits said that is one possible avenue for future research.
“The data are very rich, and there are many possibilities for further research. We have only scratched the surface,” she said.
