Visible or Covert? The Causal Effect of Inspector Visibility on Fare Evasion Detection: A Causal Machine Learning and Policy Learning Approach

Fare evasion generates substantial revenue losses for public transport operators and is typically combated through fare inspections, yet little is known about how the mode of inspection-uniformed versus plainclothes-affects detection efficiency. Using a unique dataset of 178,686 individual inspection records from PostAuto, the largest regional bus operator in Switzerland, we apply causal machine learning to estimate the causal effect of inspector visibility on inspection efficiency, defined as detected fare evaders per inspection hour. Our results indicate that wearing a uniform reduces efficiency by 0.156 incidents per hour, or approximately 26% of the sample average. This can also be interpreted as the immediate deterrent effect of visibility: passengers who would otherwise be detected may instead choose to comply. Heterogeneity analyses find that this effect is larger in areas with a high share of travelcard holders. When applying optimal policy learning to optimally target subgroups by one or the other treatment, plainclothes inspections are recommended for the large majority of contexts (91.8%). Uniformed inspections are only suggested for lines with a low share of foreign residents but a relatively large population, and for lines with a high share of foreign residents and many total inspection hours in the previous month.