Different languages encode gender nouns in differently. In some cases, such as Spanish, differences are stark. Spanish has only two genders (masculine and feminine), the distinction is based on biological sex, they assign gender also to inanimate objects (like tables or chairs), and they even have a male and a female ``we''. So, do these differences lead to more separate gender roles and higher labor market disparities?

Previous research based on cross country evidence suggest that indeed this is the case. Yet, the research has had a hard time separating language from other country specific institutions and characteristics. We intend to build on this research by studying the behavior of migrants in several countries. By analyzing migrants in the same institutional context, be it in the USA, or Brazil, we isolate the effect of cultural baggage (embodied in the language) from other characteristics of the home country. Moreover, since we look at several countries, the effects are not driven by specific combinations, that is migrants from country a during well in b because of cultural proximity.

The results from this ongoing research suggest that language matters. Migrants from countries with more stark gender distinctions present greater wage inequality. Moreover, there seems to be also a destination country effect. Moving to a country where gender distinctions are less marked is connected to a lower penalty. Moreover, the point estimates suggest a strong effect, up to 19 pp difference. However, they are also noisy, which reduces the precision. Enlarging the sample, including more destination countries and languages, could help increase precision and make results even more representative.