The determinants of household electricity consumption in Taiwan: Evidence from quantile regression

• We analyze the determinants of household electricity use between 1981 and 2011.
• We identify the characteristics of high-electricity-consuming households.
• The effects of predictor variables differ across quantiles and change over time.
• Income variable and household size were significant in all quantiles for each year.
• Strategies for reducing electricity use should focus on high electricity users.

This paper employs quantile regression to analyze the determinants of household electricity consumption in Taiwan over the period 1981–2011. Our results show that the effects of demographic, socioeconomic, and household dwelling characteristics on household electricity consumption may differ across quantiles and may change over time. We found that household income and household size were significant in all quantiles for each year. We identify the characteristics of high-electricity-consuming households. Households with higher income, larger household size, and more elderly members consumed more electricity. In terms of dwelling attributes, larger housing areas, homes with more appliances, and owner-occupied, business-used, and multi-floor houses contributed to higher household electricity consumption. Strategies for reducing electricity consumption should focus on specific groups that tend to exhibit higher electricity use. However, we also found that the low-income and small-size households may have higher electricity consumption on a per capita basis. Thus, as household size decreased, the increase of per capita electricity demand driven by the change of household size should be a matter of concern.