This study analyzes the productive efficiency of the toll motorway sector in Spain over the 2003–2018 period by applying a stochastic frontier analysis with the particularity of including an undesirable output (traffic accidents with victims). Given the multioutput nature of the study and the presence of negative externalities, the method used is based on hyperbolic distance functions. The results obtained reveal that i) the average efficiency in the sector suggests that it is possible to increase the total number of vehicle-kilometers traveled by 29% and simultaneously reduce recorded accidents with victims by 23%, ii) there was technical progress of 16.5% during the study period, iii) reducing by one unit the number of accidents with victims would be equivalent to a monetary value of over 491 thousand euros in terms of the loss of toll revenue paid by light or heavy vehicles, and iv) the shadow price of accidents with victims allows us to determine that the optimal toll rate for heavy vehicles should be 1.62 times that of light vehicles.

本研究通过应用随机前沿分析，分析了 2003 年至 2018 年期间西班牙收费高速公路部门的生产效率，特别是包括了不良产出（造成受害者的交通事故）。考虑到研究的多输出性质以及负外部性的存在，所使用的方法基于双曲距离函数。获得的结果表明，i) 该部门的平均效率表明，可以将车辆行驶总公里数增加 29%，同时将记录的伤亡事故减少 23%，ii) 技术进步 16.5在研究期间，iii) 造成受害者的事故数量减少一个单位，就轻型或重型车辆支付的通行费收入损失而言，相当于超过 491,000 欧元的货币价值，以及 iv) 影子根据伤亡事故的价格，我们可以确定重型车辆的最佳通行费应为轻型车辆的 1.62 倍。