Working Paper: CEPR ID: DP17389
Authors: Leonardo Ridolfi; Carla Salvo; Jacob Weisdorf
Abstract: Is mechanisation labour displacing? We use two all-inclusive industrial censuses from 19th-century France to examine the effect on wages and employment of one of the greatest waves of mechanisation in history: the diffusion of steam power. Many and costly workers were positively associated with the later adoption of steam. This suggests that mechanisation intended to save on the labour bill. But after controlling for selection effects, our diff-in-diff analysis shows that steam-adopting industries ended up employing both more and better-paid workers on average than their non-steam-adopting counterparts. This disputes the widespread view that past mechanisations entailed technical unemployment and falling labour compensation.
Keywords: employment; health; human capital; industrialisation; innovation; labour; motive power; wages
JEL Codes: I15; J42; J31; L92; O14; O33
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Adoption of steam power (L94) | Increase in employment (J23) |
| Adoption of steam power (L94) | Increase in wages (J31) |
| Adoption of steam power (L94) | Higher wages and employment (J39) |
| Adoption of steam power (L94) | Increase in employment without raising wages (J39) |
| Adoption of steam power (L94) | Labour-augmenting rather than labour-displacing (J24) |