Cost efficiency of Tunisian water utility districts: Does heterogeneity matter?

We examine the cost inefficiencies in the Tunisian water industry by controlling the heterogeneity bias. Two different cost-stochastic frontier models are tested to determine how observed and unobserved heterogeneity affect water utility districts' performance rankings and efficiency scores. The correction for heterogeneity significantly affected rankings by efficiency measures, and the two models yielded different results. The true fixed-effects model is more reliable than the pooled stochastic model. Then district's utility set still exhibits latent heterogeneity and significant cost inefficiencies are observed. The model estimates demonstrate the existence of economies of scale and density with respect to the size of the operation. Small district utilities exhibit economies of scale, while large district utilities exhibit diseconomies of scale. • Latent heterogeneity affects Tunisian water utilities' cost efficiency and ranking. • True fixed effects model incorporating exogenous factors produces accurate results. • The Tunisian water industry should be restructured to reduce costs. [ABSTRACT FROM AUTHOR]