The cost of quality of service improvement in electricity distribution in Burundi: a parametric shadow price approach.

The quality of service (QoS) in the sector is of paramount importance to the industry, to customers and to the regulator. It is measured by the duration and frequency of electricity interruptions and their consequences. QoS is used in the comparison of decision-making units that use the same inputs and outputs. Improving the QoS in the electricity sector cannot be achieved without costs. In this paper, we estimate the cost of preventing losses due to electricity interruption at REGIDESO, a vertically and horizontally integrated company in Burundi.We use a parametric output distance function approach, assuming that the utility's objective is to increase the number of customers and the amount of electricity delivered. Electricity losses enter the production process as inputs, imperfect substitutes for capital, with the causes of electricity outages being related to, for example, low generation capacity. We identify the sources of inefficiency and the trade-off between poor QoS and capital. We use a dataset of 19 districts located in 5 regions under the responsibility of REGIDESO, for the period 2014 to 2017. The estimation of the multi-input and multi-output translog model shows that the cost of preventing one MWh of losses due to power interruption is US$55.37. The marginal cost of quality tends to be higher in Bujumbura than in other regions.