Are all automation-resistant skills rewarded? Linguistic skills in the US labour market

Skills that are difficult to automate are expected to increase in demand and reward according to skill-biased technological change advocates, who have identified high rewards for cognitive and social skills. However, such broad skill categories involve numerous essential competencies that can be differentially rewarded or go simply unrewarded. Using US data, this article analyses the demand for and payment of linguistic competency, a cross-cutting kind of skill that is basic for both cognitive and social work in the new economy and is one of the human capacities that is most difficult to automate. While human capital theory predicts an increase in wages as the demand for linguistic skills rises, from cultural/institutional perspectives, it can be theorised that communicative abilities and foreign-language knowledge are socially undervalued because of their association with feminised activities, ethnicity, and low-status service jobs. We analyse the demand and reward for linguistic skills through a two-step analysis of occupational and individual data derived from two sources: the Occupational Information Network and the Current Population Survey. Results show that while 'hard' verbal-reasoning skills are associated with high average salaries, as is predicted by neoclassical theory, the potentially undervalued linguistic skills - interactive and multilingual skills - are unrewarded and even penalised. This evidence requires further political attention, given its implications for large number of workers, especially in feminised, low-status service jobs. JEL Codes: J01, J23, J24, J31, J38