Enforcement activity or administrative visibility? Decomposing two decades of labour inspection fines in Peru, 2002–2023

Background Administrative fines are the most widely available empirical trace of labour inspection, yet counts of fines are jointly determined by the incidence of violations, the intensity of inspection, the capacity of the enforcing agency and the coverage of the recording system itself. Aim To characterise the distribution of labour inspection fines imposed on employers in Peru between 2002 and 2023, and to assess whether the observed national trend reflects enforcement activity or the expansion of administrative visibility. Methods Nationwide retrospective analysis of 60,315 sanctioning records retrieved in April 2025 from Peru's publicly accessible labour inspection resolution registry. Penalties were expressed in tax units (Unidad Impositiva Tributaria, UIT) of the year of issuance. Segmented Poisson regression with department fixed effects and cluster-robust standard errors was fitted to annual counts for 2013–2021; earlier years were excluded as a period of incomplete registry coverage and 2022–2023 as right-censored. Results Nationally, fines grew by 26.0% per year between 2013 and 2019 (incidence rate ratio [IRR] 1.26, 95% CI 1.07–1.48). In Metropolitan Lima, the only jurisdiction with complete coverage throughout, growth was 11.7% per year (IRR 1.12, 95% CI 1.04–1.19), with no discontinuity at 2020. Across the remaining 25 departments it reached 72.0% per year (IRR 1.72, 95% CI 1.47–2.02), accelerating to 145.4% after 2020 (IRR 2.45, 95% CI 1.97–3.06). Departments recording ten or more fines annually rose from 2 to 26 and territorial concentration fell from a Herfindahl–Hirschman index of 0.95 to 0.14, while sectoral concentration was unchanged. The median penalty was 2.25 UIT and the median number of workers covered per fine was one in all twelve sectors. Conclusions The nearly sevenfold national increase in sanctioning records reflected the progressive incorporation of regional inspection offices into the public registry rather than intensified enforcement. Administrative fines index where and when labour regulation becomes administratively visible, and should not be read as a measure of enforcement intensity, detected non-compliance or worker protection until registry coverage has been explicitly modelled.