عملکرد و بهره وری سیستم توزیع قدرت در کلمبیا : اثرات اصلاحات سال 1994

We assess evolution in performance, efficiency and productivity of Colombia's power distribution utilities before and after the 1994 regulatory reform that introduced electricity market activities for the power sector in 12 distribution companies from 1985 to 2001. Performance is evaluated contrasting changes in mean and median by Wilcoxon Rank Sum and Pearson tests on financial and other performance indicators. Technical efficiency is measured by means of Data Envelopment Analysis (DEA). The nature of the dataset allows the estimation of Malmquist productivity index and its evolution in time. Results show a recovery after the reform in the main performance indicators of profitability, partial input productivity, and output. Plant efficiency and productivity increased after the reform, mainly in the largest utilities used as benchmarks in the DEA efficiency scores measures. Meanwhile, the less efficient power distribution companies did not improved after the reform and were not able to undertake plant restructuring to catch-up in plant efficiency with respect to the Pareto efficient input allocations. Econometric results on DEA efficiency scores suggest a positive effect of policy reform.

The 1994 regulatory reform of the Colombian power sector was one of the first reforms in Latin America to introduce a market system for wholesale electricity transactions, and the first to implement a bidding system for its pool electricity market in the region. In this sense, the reform took a step forward from the Chilean and Argentinean experiences, where wholesale electricity prices were based on declared costs rather than on marginal supply prices by 1994.1 The reform introduced competition, established a new industry structure and a new independent regulatory agency, set up the basis for expansion and diversification of power generation sources, and improved both the sector's efficiency and reliability. The reform focused on offering incentives for utility efficiency and productivity levels through the introduction of market competition, independent grid access, and markup price regulation for power distribution. Inspired by the British reform, the regulatory reform split the traditional vertical monopoly structure of the power sector into four different activities: generation, transmission, distribution and commercialization of electricity. Power distribution by a domiciliary public service provider faces two types of regulation. The first one is price regulation. The regulatory commission, Comisión de Regulación de Energía y Gas (CREG), currently sets the markup formula for distributors and the design of the pass through component in the final user's tariffs. In particular, CREG determines: i) direct purchase costs such as the pool sale price and transportation charges, ii) capacity charges, and iii) costs of the reserve provisions to stabilize the system and prevent bottlenecks in the transmission system.2 The second type of regulation concerns quality control, companies are subject to sanctions if their service fails to meet minimum quality standards. The reforms and regulations led power holdings to undertake a generalized divestiture process across electricity holdings in order to fully separate power generation, transmission, distribution, and the setting up of new commercialization activities. Thus, privatization arose as one instrument for promoting market competition and industry restructuring, and became a complementary policy within a broad deregulatory context. The aim of this paper is to provide new empirical evidence on the effects of the regulatory reform in power distribution in Colombia. To the best of our knowledge there is no micro-study assessing the effects of a regulatory reform for power distribution. Nonetheless, several studies on the Colombian electricity sector after a decade of the regulatory reform have been published recently. Pombo and Ramírez (2005) test performance in privatized power holdings and measured plant efficiency for Colombia's thermal stations showing a generalized increase in productive efficiency due to market entry, introduction of cost-saving technologies, and a positive effect of the new regulation that implied the setting up of a non-regulated market of large clients that boosted transactions of forward electricity contracts. García and Arbelaez (2002) evaluate the likelihood of merging among power generators acting in the wholesale electricity market. Larsen et al. (2004) present a set of aggregate statistics of Colombia's power sector to highlight the lessons derived from the implementation of market deregulation policies in Colombia for network industries since 1994. Despite the above, studies on several dimensions of the electricity market are still pending, i.e., price collusion in the pool market, consumers' welfare effects, quality regulation and regulatory capture, among others topics. This paper provides empirical evidence on the effects induced by the regulatory reform, focusing on the efficiency and productivity effects upon power distribution. The paper presents an ex post performance analysis for regional power distribution companies based on four elements: i) direct measures of productive efficiency scores and Malmquist productivity indices through data envelope analysis programming (DEA), ii) changes in means and medians of firm performance variables of profitability, operating efficiency, labor, investment and sales, iii) an econometric analysis regarding the determinants and micro-fundamentals of firm efficiency scores by the regional power distributing companies, and iv) an evaluation of the policy's effectiveness on plant efficiency using a two step DEA decomposition procedure of changes among policy regimes. The paper is organized in four additional sections. Section 2 describes the data set and the methodology employed to assess utility efficiency in power distribution. Section 3 analyses the results of the performance indicators through the changes in means and medians and the measurement outcomes of plant efficiency and productivity through DEA efficiency scores and the construction of Malmquist indices. Section 4 reports an estimation of the impacts of the industry reform on plant efficiency and an econometric analysis of the determinants of utility efficiency scores. Section 5 concludes the paper and summarizes its results.

This paper has conducted an in-depth study of utility performance and technical efficiency trends for a time span long enough to test performance, efficiency and productivity benefits in power distribution of the 1994 reform in Colombia. Based on a sample of 12 utilities covering the largest cities of the country and a time span of 16 years, the findings suggest that urban power distribution has improved during the post-reform years. The performance analysis shows a recovery in profitability rates. This is partly explained by gains in labor and capital productivity across companies, but higher residential tariffs also unbound utility financial constraints. This result contradicts in some sense the public version of a structural financial crisis within power utilities. Nonetheless, the companies that are facing financial problems respond to their own particular challenges.35 The efficiency analysis yields important insights. On one hand, the results of efficiency trends across power distributors did not boost efficiency within the inefficient power distributors, as they became less efficient after 1995. However, the efficient distributors, which are the larger utilities, remain on the best practice frontier. This trend is confirmed with the Malmquist productivity indices. Efficiency gains relied on boundary shifts rather than efficiency catch-up across utilities below the frontier. This implies that those plants are away from the benchmark's Pareto efficient inputs allocation. The benefits of the reform are also observed by disentangling managerial from policy efficiency. The inefficiency in the post-reform period, after eliminating any managerial inefficiency in each policy group, is dramatically lower than in the pre-reform period. Finally the econometric exercise provides evidence regarding the determinants of efficiency scores. They are influenced by firm characteristics such as plant size and factor intensities. Market density has a positive effect on utility productive efficiency. Regulatory policy has had positive effects on power distribution efficiency according to the regression equations. This regulation dummy turned out to be a robust regressor that on average implied an efficiency gain of 5% after the reform. This finding is consistent and goes in the same direction with previous results of efficiency analysis in power generation for the case of Colombia reported in Pombo and Ramírez (2005). The policy implications of the above results lie on the fact that ownership does not itself guarantee efficiency improvements. The ownership dummy turned out to be a non-significant explanatory variable. This finding is consistent with results found in other case studies where there is no evidence of a positive impact of private ownership in plant efficiency in power generation and distribution, such as those reported in Pollit (1995).