Tools for Measuring and Controlling Energy Subsidies in Middle Income Countries

Abstract

The purpose of this dissertation is the development of econometric tools useful in the design of energy subsidies. The tools are divided between three chapters, a chapter on Energy Demand and the Optimal Level of Subsidies, a chapter on Evaluating Energy Subsidies over Time and a chapter on Targeting and Distribution of Energy Subsidies. Although the primary target for subsidies in this paper are low income households, subsidies designed to target groups by region or by profession are also discussed. The Tools discussed in the Dissertation include: Tool 1: Elasticities with Quantile Regressions - Used for understanding energy demand at the edge of the income distribution Tool 2: Spline Regressions - To show the marginal degree to which energy use increases with income and to study when household versus per capita subsidies are preferable. Tool 3: Engel Curves - To determine the quantity of energy that should be subsidized. Tool 4: Energy Poverty Lines - To help determine who should be eligible for energy subsidies Tool 5: Energy Expenditure Time Series - To help to determine who receives benefits from subsidies and to determine which subsidies are worth supporting. Tool 6: Budget and Pie Charts - To help to determine the cost-benefit of providing a subsidy and to better understand who benefits. Tool 7: Big Data Loops - Used to understand the characteristics of current energy and subsidy users of energy. Tool 8: Geographical Data Visualization - Used to analyze geographical patterns of subsidy use in the country. In addition a justification of why subsidies may continue to be relevant for Middle Income Countries is provided. Four different approaches to providing subsidies are discussed with pros and cons of each given. These are Subsidies that are 1) Targeted , 2) Self-Selecting, 3) Give the Same Quantity to All, and 4) Cash-Transfers. In juxtaposition to much of the current literature, the author makes a case for the first three options. While cash transfers are economically efficient, they are a successful political solution and they require a high degree of information and control that middle income countries not want or be able to provide. While most of this dissertation is generally focused on tools for targeted subsidies, self-selecting subsidies can also be an elegantly simple solution that does not categorize people, and in terms of fairness, giving the same to all is likely the most just solution. Specific suggestions for Thailand subsidies are proposed, including an 80-140 KwH level of free electricity per household and IF NEEDED, a 3-5 kg level of subsidy for LPG or a 20 liter subsidy for transportation fuel per household.