Abstract:
Accounting for network operators' heterogeneity is of crucial importance for regulators. In contrast to observed heterogeneity, the consideration of unobserved differences is far more challenging. Most estimation models try to account for unobserved factors that impact the network operators' costs and performance but neglect the possibility of heterogeneous technologies. Assuming a common technology represented by a joint cost function across network operators implies, e.g., identical marginal costs and economies of scale for all network operators. As it is questionable that this assumption holds in practice, efficiency estimates may be biased as technological heterogeneity is misunderstood as inefficiency. To overcome this misspecification, a latent class model is applied to a comprehensive database of German electricity distribution network operators between the years 2011 and 2017, explicitly accounting for technological differences across network operators. The results indicate that German distribution network operators can be unambiguously classified into three statistically different classes that share a common cost function. The findings show significant differences in the size, capacity of distributed generation from renewable energy sources and identify distributed generation capacity as an important driver of the network operators' technology.