Why Do Firms Train? Theory and Evidence
Abstract
This paper offers a theory of training whereby workers do not pay for the general training they receive. The superior information of the current employer regarding its employees' abilities relative to other firms creates ex post monopsony power, and encourages this employer to provide and pay for training, even if these skills are general. The model can lead to multiple equlibria. In one equilibrium quits are endogenously high, and as a result employers have limited monopsony power and provide little training, while in another equilibrium quits are low and training is high. Using microdata on German apprentices, we show that the predictions of our model receive some support from the data.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Labor economists have documented that some training investment occurs because labor market imperfections such as wage compression, information asymmetries, and monopsony power can make general training investment privately rational under certain conditions.
“Labor economists have documented that despite this market failure, some training investment occurs because labor market imperfections (wage compression, information asymmetries, monopsony power) can make general training investment privately rational under certain conditions (Acemoglu & Pischke, 1998;Stevens, 1994).”
From The Tragedy of the Cognitive Commons: How AI Could Disrupt the Regeneration of Professional Expertise (2026) · cites this paper as Acemoglu & Pischke, 1998
Cited by
Sorted by citations
Lovett, N. (2026) · Human Resource Development Review · doi:10.1177/15344843261470602 · Cited by 7 · see citing works
Revisiting lesson-drawing: methods to assess policy’s institutional transferability
Yamada, M. (2026) · Policy Design and Practice · doi:10.1080/25741292.2025.2574112 · Cited by 1 · see citing works
- Authors
- Acemoglu, D., & Pischke, J.-S.
- Affiliations
- Massachusetts Institute of Technology (US)
- ORCID
- Pischke, J.-S.
- Venue
- The Quarterly Journal of Economics
- Publisher
- Oxford University Press (OUP)
- Published
- 1998-02-01
- Volume
- 113
- Issue
- 1
- Pages
- 79-119
- ISSN
- 0033-5533 (ISSN-L)
- 1531-4650
- OpenAlex
- W2100316976
- Cited by
- 992
- References
- 0
- Citations by year
- 2026: 35 · 2025: 33 · 2024: 36 · 2023: 48 · 2022: 37 · 2021: 33 · 2020: 50 · 2019: 39 · 2018: 44 · 2017: 38 · 2016: 44 · 2015: 50 · 2014: 40 · 2013: 27 · 2012: 40
- FWCI
- 42.23
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- The Quarterly Journal of Economics (publishedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- HighWire Press Open Archive (submittedVersion)
- Topics
- Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Labor market dynamics and wage inequality
- Social Sciences › Economics, Econometrics and Finance › General Economics, Econometrics and Finance › Global trade and economics
- Social Sciences › Political Science and International Relations › Education and Labor Relations
- Keywords
- employee training, monopsony power, transferable skills, multiple equilibria, employee turnover, apprenticeship, human capital
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review