Read on Twitter
Labor folks! Anna Sokolova and Todd Sorensen's fantastic meta-analysis of monopsony is now out in @ILRReview.
Want a summary of what the empirical lit says about the elasticity of labor supply to the firm? They collect *1,320* estimates from 53 studies https://journals.sagepub.com/doi/10.1177/0019793920965562
Want a summary of what the empirical lit says about the elasticity of labor supply to the firm? They collect *1,320* estimates from 53 studies https://journals.sagepub.com/doi/10.1177/0019793920965562
Takeaway (1): A big spread of estimates!
Looking at "direct" estimates, there's a median of approx 1.4. In a simple model this would imply a wage markdown of 58%.
The median *inverse* elasticity corresponds to a firm LS elasticity of 14, implying a much lower markdown of 7%
Looking at "direct" estimates, there's a median of approx 1.4. In a simple model this would imply a wage markdown of 58%.
The median *inverse* elasticity corresponds to a firm LS elasticity of 14, implying a much lower markdown of 7%
Takeaway (2): Why do estimates differ so much? A few things stand out:
(1) Inverse elasticities higher than directly estimated (as in prev. tweet)
(2) Recruitment elasticities higher than separation elasticities
(3) Elasticities for teachers, nurses, lower than other workers
(1) Inverse elasticities higher than directly estimated (as in prev. tweet)
(2) Recruitment elasticities higher than separation elasticities
(3) Elasticities for teachers, nurses, lower than other workers
Takeaway (3): How to reconcile all these together? Sokolova and Sorensen construct "estimates associated with best practices in the literature" from the separation elasticity estimates, obtaining an estimate of 7.1, which under a simple model would imply a wage markdown of 12%.
Important post-script: as per Berger, Herkenhoff, and @Simon_Mongey, even well-identified estimates of reduced-form elasticities =/= the *structural* elasticity that firms face! They show how one might map reduced-form estimates -> structural elasticities
https://www.nber.org/papers/w25719
https://www.nber.org/papers/w25719
