Compounded Housing Vulnerabilities and Energy Burden Among Low-Income Households: A Latent Classification for the U.S.

Energy burden (EB) is an urgent yet underdiscussed issue in social work. In the U.S., 16% of households experience energy poverty, which inequitably affects populations by race, income, gender, and other factors. Specifically, low-income households face a double burden. They are not only constrained by limited financial resources but are also more likely to reside in energy-inefficient housing, which further amplifies their vulnerability to EB. However, existing research has paid limited attention to how compounded housing vulnerabilities shape the EB, focusing solely on socioeconomic factors. To address this gap, this study proposes the following research questions: (1) How are housing vulnerabilities among low-income households compounded and classified into distinct types? (2) How do these types differently influence EB?

We used data from the 2023 American Housing Survey, focusing on low-income households below 100% of the poverty threshold (N=6,759). A Latent Class Model (LCM) was used to classify housing vulnerability types based on observed patterns across multiple housing condition indicators, and a Logistic Regression Model was used to examine the relationship between the identified latent classes and EB. Housing vulnerability variables related to energy efficiency, such as tenure type, year of construction, number of rooms, unit square footage, and primary heating fuel, were included as indicators. EB was defined as the ratio of household energy expenditures to residual household income, with a threshold of greater than 6%. Socioeconomic characteristics, including gender, age, race/ethnicity, educational attainment, marital status, disability, household size, and income, were included as control variables.

The LCM identified three distinct latent classes of housing vulnerability. Class 1 was characterized by renter occupancy, electric heating, and a high proportion of smaller housing units (n=3,351; 49%). Class 2 was distinguished by owner occupancy, heating with electricity and gas, and a high proportion of older housing units (n=2,754; 41%), whereas Class 3 was marked by owner occupancy, heating with gas, and a high proportion of relatively newer and larger housing units (n=654; 10%). Regarding the impact of compounded housing vulnerabilities on EB, households in Class 3 (OR=10.39; 95% CI: 5.62-19.21) and Class 2 (OR=7.57; 95% CI: 5.81-9.87) had a significantly higher EB than those in Class 1.

The findings highlight the importance of understanding housing conditions that are fundamentally related to energy efficiency, as a compounding factor shaping EB among low-income households, beyond individual socioeconomic characteristics. Specifically, larger housing units were associated with higher energy expenses or mortgage payments, thereby serving as a stronger predictor of higher EB than other housing vulnerabilities among low-income households. The study suggests that future policies aimed at mitigating the EB on low-income households should incorporate tailored, housing-focused strategies that consider energy efficiency. As climate impacts and energy crises intensify, this research may contribute to achieving environmental justice by addressing the association between the latent housing vulnerabilities and EB in low-income households.