世界银行-收入和消费数据的统计匹配：罗马尼亚能源和收入贫困评估（英）-2024.9-57页_1mb

Statistically Matching Income and Consumption Data: An Evaluation of Energy and Income Poverty in Romania

Summary
This study investigates the complex relationship between energy poverty and monetary poverty in Romania by combining data from two separate surveys: the EU Statistics on Income and Living Conditions (EU-SILC) and the Household Budget Survey (HBS). The EU-SILC focuses on income and welfare indicators, while the HBS collects detailed expenditure data, including energy spending. Since no single survey provides both income-based welfare measures and energy consumption details, statistical matching techniques are employed to integrate the datasets.

The research proposes using predictive mean matching (PMM), an imputation method, to fill in missing energy expenditure information in the EU-SILC using variables from the HBS. PMM was found to be the most effective method based on multiple performance criteria, including comparison of distribution shapes and subgroup analyses. The integrated dataset reveals that nearly all households experiencing monetary poverty (at-risk-of-poverty) also face energy poverty, but a significant additional portion of the population is identified as energy-poor. Energy spending shares are highest among lower-income households, exacerbating their burden during energy crises like the one triggered by the Ukraine invasion.

The study highlights challenges, such as systematic differences in sampling and harmonization across surveys, which may affect the accuracy of the imputation. Future work can improve these methods by enhancing data harmonization, incorporating survey weights, and integrating expenditure data into the EU-SILC. These findings inform policy recommendations, emphasizing the need for targeted energy assistance programs and energy efficiency measures, particularly for vulnerable groups.