Language segmentation of the online sports labor market: The case of chess coaches

Language simultaneously enhances communication and segments labour markets. We study how language shapes outcomes in a global online sports labour market using novel data on chess coaches from the popular Lichess platform. Combining 1,314 public coach advertisements with official International Chess Federation (FIDE) records, we construct a coach–language panel that links hourly lesson prices to coaches’ linguistic repertoires, native speaker status, currency of transaction, country characteristics, and language-specific audience sizes. We estimate regressions of log hourly fees on audience size, its interaction with GDP per capita, and indicators of multilingualism, native language, and currency choice. Prices increase with the size of the potential audience only when the relevant language is anchored in high-income countries; in lower-income linguistic markets larger audiences are associated with lower prices, consistent with stronger competition and lower purchasing power. Multilingual ability per se is not rewarded once observable characteristics are controlled for, while native-speaker status commands at most a small premium confined to the lower and middle parts of the price distribution. Listing prices in a foreign currency is robustly associated with higher fees, particularly in the upper market segment. Overall, our findings show that digitalisation does not eliminate linguistic segmentation but reshapes it.