Working Paper: NBER ID: w25295
Authors: James Harrigan; Ariell Reshef; Farid Toubal
Abstract: We study the impact of firm level choices of ICT, R&D, exporting and importing on the evolution of productivity, its bias towards skilled workers, and the implications for labor demand. We use a novel measure of firm-level technology: firms' employment of workers in occupations related to R&D and ICT adoption, who we call “techies”. We develop a methodology for estimating nested CES production functions at the firm level, which allows us to measure both Hicks-neutral and skill-augmenting technology differences. Using administrative data on French firms we find that techies, exporting and importing raise skill-biased productivity. In contrast, only ICT techies raise Hicks-neutral productivity. On average, higher firm-level skill biased productivity does not affect low-skill employment even as it raises the ratio of skilled to unskilled workers, due to the cost-reducing effect of higher productivity. ICT techies account for large increases in aggregate demand for skill, mostly due to their effect on firm size, less so through within-firm changes. Exporting, importing, and R&D techies have smaller aggregate effects.
Keywords: Productivity; Technology; Labor Demand; Trade; Skill-Biased Technological Change
JEL Codes: D24; F16; O52
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| employing 'techies' (O33) | skill-biased productivity (J24) |
| employing 'techies' (O33) | Hicks-neutral productivity (O49) |
| exporting (F10) | skill-augmenting productivity (J24) |
| importing (F10) | skill-augmenting productivity (J24) |
| skill-biased productivity (J24) | low-skill employment (F66) |
| skill-biased productivity (J24) | ratio of skilled to unskilled workers (J24) |