Reward-Timing Uncertainty, Languages, and R&D Investment

Linguistic Structures and Economic Outcomes

“…These effects also carry over to behaviour at the corporate level. Recent evidence suggests that firms in weak‐FTR language locations perform better at corporate social responsibility practices (Liang et al ., ), have higher precautionary cash holdings (Chen et al ., ), engage less in earnings management (Fasan et al ., ) and invest more in research and development (Su et al ., ). Interestingly, a study by Liang et al .…”

“…Experiments – natural or laboratory based – are a potentially promising approach to identifying the effect of linguistic structures. Natural experiments may involve, for example, changes from one language to another where the two differ in structure (Chen et al ., ; Su et al ., ), or potentially, structural changes within a single language. These are typically analysed within a difference‐in‐difference framework.…”

“…() provide difference‐in‐differences analyses based on the case of Hong Kong, where a weak‐FTR language (Chinese) relative to a strong‐FTR language (English) became increasingly important after the 1997 transfer of sovereignty from the United Kingdom to China. They use this natural experiment to examine the effect of an increased significance of a weak‐FTR language on cash holdings (Chen et al ., ) and investment in research and development (Su et al ., ) in Hong Kong firms, compared to several control groups. Similar evidence for other contexts would be useful.…”

“…Some studies have additionally considered two alternative continuous measures of language FTR: verb ratio and sentence ratio, defined as the frequency of verbs and sentences respectively that are grammatically future‐marked in weather forecasts (e.g. Chen, ; Paule‐Paludkiewicz et al ., ; Su et al ., ).…”

“…Studies based on cases of single multilingual countries such as Belgium (Su et al ., ), Canada (Chen et al ., ) and Switzerland (Chen et al ., ; Guin, ) have also recently emerged in the literature. By estimating language effects at smaller geographic scope, these approaches potentially allow to control more extensively for unobserved heterogeneity.…”