2023-11-23-纽约联储-入职员工的拥挤和粘性研发_74页_1mb

Summary

Congestion in Onboarding Workers and Sticky R&D
Justin Bloesch, Jacob P. Weber (Federal Reserve Bank of New York Staff Report No. 1075, November 2023)

Key Findings

The paper introduces congestion in onboarding as a microfounded source of convex adjustment costs in R&D investment.

1. Theoretical Mechanism

• Congestion occurs when the probability of converting new hires (junior workers) into productive R&D workers (senior workers) decreases with the number of new hires.
• Firms optimally hire slowly during positive shocks (due to diminishing returns from rushed onboarding) and hoard workers during negative shocks (to avoid adjustment costs).
• This mechanism generates delayed, hump-shaped responses in R&D investment, similar to standard models with ad hoc convex adjustment costs.

2. Empirical Evidence

• Using GitHub data on software developers, the authors estimate that the onboarding probability (transition from junior to senior workers) is a decreasing function of the ratio of juniors to seniors (ρ(x) < 0).
• Approval times for code contributions decrease significantly during the first six months of project experience, supporting the model's assumption of "team-specific capital" acquisition.
• Congestion is quantitatively significant, as it aligns with general equilibrium models, producing realistic hump-shaped responses to shocks.

3. Quantitative Model

• Incorporating the estimated ρ(x) into a New Keynesian model with heterogenous firms confirms that congestion in onboarding fully explains the stickiness of R&D investment.
• The model's impulse responses to monetary policy and TFP shocks match empirical patterns in R&D investment volatility.

4. Conclusion

• Congestion in onboarding provides a microfoundation for the observed stickiness in R&D investment.
• This result suggests that the sluggish adjustment of intangible investment is not invariant to policies but reflects inherent structural lags in human capital transfer.
• The findings support the view that team-specific human capital, lost upon worker departure, creates implicit adjustment costs in R&D production.