Innovation, market power and the labour share: Evidence from OECD industries

Two issues remain overlooked in empirical investigations of how labour share varies with technological innovation and market power. One is the risk of omitted variable bias that arises from failure to control for both innovation and market power at the same time. The second is the risk of confounding bias that arises from the possibility that innovation and market power affect each other and the labour share at the same time. I address both issues by adopting a simultaneous equations approach and using EU-KLEMS data from 1995 to 2019 on 31 OECD industries and 12 countries. The novel evidence I discover indicates that: (i) innovation always increases with market power, particularly when the latter increase from a high initial level; (ii) market power always increases with innovation, particularly when the latter is extended to include marketing and organisational innovation; (iii) market power is always more detrimental for labour share compared to innovation; and (iv) the combined effect of human capital and labour-market institutions reverses the adverse effect of innovation but it is insufficient to reverse the adverse effect of market power. These findings are robust to a wide range of sensitivity checks and indicate that the major driver of the decline in labour share is not technological innovation per se, but the extent of market power that enables firms to set real wages below the marginal product of labour.