Carrier Safety Scoring Under Measurement Instability: From CSA Percentile Volatility to Route-Level Risk Allocation

Carrier safety percentile scores generated by the Federal Motor Carrier Safety Administration’s Safety Measurement System inform enforcement prioritization, insurance underwriting, and, increasingly, automated freight-assignment decisions. Several independent strands of research document that these percentiles behave inconsistently for carriers with limited inspection histories. This article connects that documented inconsistency to a specific decision a route-level or lane-level risk-allocation system must make: whether to treat every published percentile with equal statistical confidence. Drawing together evidence on carrier size, inspection frequency, longitudinal score movement, and carrier-level aggregation, the article argues that a below-threshold CSA percentile carries a different statistical status than an above-threshold one, and proposes this distinction as a design requirement for automated routing systems that already handle other forms of scenario uncertainty. The article specifies the data, unit of observation, and modeling approach a future empirical study would need to test this claim, framing the present contribution as the conceptual groundwork for that test.