In the context of two-stage data envelopment analysis (DEA) for efficiency correction, we shift the focus from the common central tendency orientation in its second stage to an individually oriented procedure. We propose to evaluate the influence of contextual variables on each unit's performance relative to the other operating units. This results in an alternative approach in which the second stage inherits the principal property of DEA in the first stage of putting each individual unit in its best possible light. We demonstrate the applicability of our approach using data from the energy sector in the domain of incentive regulation, where operators are natural monopolies. In such systems of incentives, ensuring fair performance evaluations is crucial, given the influence of contextual variables beyond management control. Our approach contributes to efficiency correction procedures under these circumstances. The results not only encourage operators to economize costs and improve service quality but also motivate them to, for example, minimize environmental impact of operations, adopt eco-friendly technologies, and invest in renewable energy sources.

中文翻译：

---

数据包络分析中的个性化第二阶段修正


在用于效率校正的两阶段数据包络分析（DEA）的背景下，我们将焦点从第二阶段的共同集中趋势导向转移到单独导向的过程。我们建议评估背景变量对每个单位相对于其他运营单位的绩效的影响。这导致了一种替代方法，其中第二阶段继承了第一阶段中 DEA 的主要特性，即让每个单独的单元处于最佳状态。我们使用来自能源部门的激励监管领域的数据来证明我们的方法的适用性，其中运营商是自然垄断的。在这样的激励体系中，考虑到超出管理控制范围的背景变量的影响，确保公平的绩效评估至关重要。我们的方法有助于在这些情况下提高效率修正程序。结果不仅鼓励运营商节约成本、提高服务质量，还激励他们尽量减少运营对环境的影响、采用环保技术、投资可再生能源等。