TY - JOUR
AU1 - Bohr,, Jeremiah
AU2 - McCreery, Anna, C
AB - Abstract For many households, energy consumption represents a non-discretionary portion of their budget and directly relates to quality of life. As researchers continue to study the environmental impacts of energy behavior, it is important to explore how energy consumption relates to socio-economic wellbeing. This paper examines the economic impacts of being energy-burdened in the United States, defined as spending at least 10% of household income on heating and electricity services; energy burdens are partially, but not entirely, driven by income, since energy needs and costs can vary substantially due to housing characteristics, utility rates, and other factors. Using panel data of US household income and energy expenditures during 1999–2017, this analysis demonstrates that energy-burdened households were at about 150%–200% greater risk of transitioning into or extending the duration of economic poverty over a two-year timeframe relative to non-burdened households. This analysis indicates that dedicating inordinate amounts of income to energy services can threaten a household’s economic well-being over time, possibly by preventing a household from engaging in other economic activities or compounding existing economic hardship. These results emphasize the importance of energy assistance and energy efficiency for low-income households, drawing attention to how structures of energy consumption, the welfare state, and social stratification intertwine. Introduction Energy consumption in the United States is often viewed in terms of excess. But energy scarcity—that is, access to affordable energy—will become increasingly relevant if policies internalizing the environmental costs of energy production lead to increased prices and leave some households vulnerable to energy burden (Price, Brazier, and Wang 2012). Non-discretionary energy consumption can impose financial hardships for many US households despite rationing and efficient behavior intended to minimize spending (Hernández and Bird 2010; Evens 2015). Rather than behavioral overconsumption, the financial burdens from energy use for these households primarily result from inefficient housing and insufficient capital to invest in improvements (Boardman 2012). Although scholarly attention to energy burdens in the industrial and post-industrial world is growing, fewer studies have examined energy burdens in the United States. Given the relatively weak US welfare state, where benefits often flow toward more affluent households (subsidizing for energy efficient appliances, e.g.), we should expect rising energy costs to push vulnerable populations to remain or fall into economic insecurity. In this paper, we examine household energy costs in the United States, focusing on households that spend 10% or more of their total income on heating and electricity. Using longitudinal data, this analysis investigates the relationship between household energy burden and vulnerability to future economic poverty. While households vulnerable to economic poverty are by definition vulnerable to energy burdens, we assume that since energy budgets are at least partly beyond the agency of individual households, they present barriers to activities that help households avoid or transition out of economic poverty. In other words, by demonstrating an association between energy burden and future economic poverty, this analysis suggests that energy burdens serve as an obstacle to preventing or transitioning out of poverty. We conclude with future lines of research considering these results, making connections between energy burden, social stratification, and environmental social science. Energy Burden and Poverty Transitions into poverty and economic insecurity constitute a vast and important area of research in sociology and related disciplines. Without reviewing this scholarship in full, it is fair to state that researchers often focus—with good reason—on changes to earnings and employment status as the predominant predictors of poverty entry (Bane and Ellwood 1986; Ruggles and Williams 1987; Jenkins 2000; McKernan and Ratcliffe 2005; Brady, Finnigan, and Hübgen 2017). In addition to these major factors, researchers have examined “trigger events” that put households at greater risk of poverty entry or bankruptcy, such as family dissolution, injury, or ill health (DiPrete and McManus 2000; Smith 2004; Himmelstein et al. 2009; Maroto 2015; Tach and Eads 2015). Despite the rich body of work in this area, the complexity of unraveling the causal pathways to poverty entry leaves room for additional research connecting events to economic insecurity (Western et al. 2012). Energy burdens may constitute a risk factor that has received less attention in the existing research. Research on energy insecurity lends insight into how this experience may lead to future poverty spells. Hernández (2016) provided a conceptual framework, outlining the economic components of energy insecurity and the various burdens to which it exposes households. In addition to the health consequences of living in cold, dark, and damp conditions, which can lead to debt accumulation or lost wages, Hernández (2016) found that many energy insecure residents prioritize housing payments and food costs over utility bills when forced to choose (a point reinforced by Finnigan and Meagher 2019). The resulting utility debt can lead to an economic cycle of “playing catch up.” Such debt can limit the ability to secure new housing as arrears can prevent the transfer of utility services to new addresses, potentially locking households into structurally inefficient homes, and demanding resources that households could otherwise invest in avoiding or transitioning out of poverty (work-related capital, schooling, or training, etc.). Utility disconnection is also linked with increased risk of eviction (Desmond 2016; Finnigan and Meagher 2019), food insecurity, telephone disconnection, unmet medical needs, and other forms of material hardship (Heflin 2006). In the case of eviction, Desmond’s (2016) qualitative research in Milwaukee, Wisconsin finds that evictions spike in the summer and early fall when low-income households prioritize catching up on their utility bills to avoid disconnection before the winter shut-off moratorium begins, and evictions decline over the winter when households can stop paying their utility bill without risking a disconnection. Interviews conducted by Hernández (2016) of low-income families revealed utility bills cumulating to thousands of dollars, restricting discretionary income, and putting households into long-term debt. In their ethnographic research, Harrison and Popke (2011) shed light on the challenges of energy-burdened households deciding between paying utility bills and covering their medical and nutritional needs. At a meso-level, the experience of energy burden can prevent families from moving out of low-resource neighborhoods and establishing social connections that facilitate economic opportunity. While social networks can provide private safety nets that help mitigate the consequences of triggering events (Heflin, London, and Scott 2011), a large majority of Americans experiencing missed utility payments report receiving no outside help; of those who do, about half receive support from friends and family (Finnigan and Meagher 2019). Researchers have linked inadequate access to affordable energy with household food insecurity (Cook et al. 2008; Bhattacharya et al. 2003), health and child development risks (Cook et al. 2008; Hernández 2013; Hernández 2016), social network disruption (Harrison and Popke 2011), and mortality during heat waves (Klinenberg 2002). The trade-off between paying energy bills and sacrificing basic needs is documented across several pieces of research. A US Energy Information Administration report indicated that about a third of households faced challenges in paying their energy bill in 2015, and about a fifth of households reduced or forwent basic necessities for at least a month to pay their energy bill (U.S. EIA 2015). A whitepaper produced for the Senate Committee on Energy and Natural Resources explored how energy costs can push families to poverty entry or cause food insecurity (Murkowski and Scott 2014). A statewide report in Colorado found the inability to pay utility bills as one of the leading causes of homelessness behind job loss and family breakup (Center for Education Policy Analysis 2006). Utility shutoffs can exacerbate or cause illness (Hernández and Bird 2010), and landlords may use disconnection as grounds for eviction in some circumstances (Verclas and Hsieh 2018). Eviction, food insecurity, and other material hardships often associated with energy burdens could initiate or prolong a cycle of poverty. The inability to access affordable energy services can prevent participation in certain economic activities and serve as a barrier to upward mobility (Sovacool, Sidortsov, and Jones 2014). We suggest that energy burdens may be one of several triggering events preceding poverty entry (Bane and Ellwood 1986; DiPrete and McManus 2000). The concept of energy burden provides a structure for theoretical and empirical analysis of energy access and its implications for human welfare. We can conceptualize energy burden in terms of access to affordable, safe, and sufficient energy services or as a disproportionate (and non-discretionary) share of household resources dedicated to energy needs. A key part of this concept is that, to a certain extent, energy consumption is non-discretionary. Just as some basic amount of food consumption is essential for human welfare, some basic amount of energy consumption is essential for maintaining well-being, such as for heating and cooling, cooking, and lighting. Households are constrained in how much energy use can be reduced without compromising health and well-being, while energy efficiency improvements can both reduce energy needs and improve health (Wilson et al. 2016; Büchs et al. 2018). Various researchers have defined burdens in terms of actual expenditures using 6%, 10%, or 20% thresholds, estimated energy costs associated with heating a home to a particular temperature, or combinations of high costs with low incomes (Boardman 1991; APPRISE 2005; Moore 2012). While energy burdens are clearly linked with income and thus economic poverty, the two are not perfectly aligned. Moderate income households may be susceptible to energy burdens just as households living in economic poverty are not necessarily dedicating a large share of their income to energy needs. Furthermore, focusing on expenditures fails to capture energy rationing among lower income households; energy expenditures are thus distinct from “affordable warmth” discussed in the fuel poverty literature (Boardman 2010), which takes into account necessary costs to reach a particular temperature. The amount of total income resources dedicated to energy varies across class boundaries, with middle- and upper-income households typically spending 5% or less while poor households may spend 10%, 20%, or even greater proportions of their income on energy services (Evens 2015; Drehobl and Ross 2016). Income dedicated to energy services limits the resources available for other actions, so fluctuating energy costs can greatly affect resources available to satisfy other basic needs, including food (Bhattacharya et al. 2003). At the same time, energy consumption does not increase in a simple linear fashion with income, and some low-income households use more energy per square foot of housing space due to less efficient appliances and temperature control systems (Evens 2015, using electricity data for Chicago-area households). Qualitative research in the United States indicates that low-income households experiencing high energy burdens are not profligate in their behavior, displaying high levels of rationing energy consumption in everyday life. Some low-income households restrict energy use to the point where they are unable to maintain World Health Organization recommended minimum home temperatures (Hernández and Bird 2010; Middlemiss and Gillard 2015). The relationship between energy burden and poverty is not simply a matter of income, but also a function of home characteristics and related factors that influence energy needs (the Methodology section provides further details on the correlation between energy burdens and poverty). Research in the United Kingdom has identified constraints on controlling energy costs driven by: housing quality; tenancy relations and the “split incentive” problem (see below for details); utility rates (which vary substantially across the United States) as well as fees and debt; and non-negotiable energy needs of household members (Middlemiss and Gillard 2015). Poverty clearly increases the likelihood of energy burden simply due to low income, but may also exacerbate energy burden on the cost side through these factors. Recent research has found that energy use intensity, or the consumption of heating or cooling energy per square foot of living space, is higher in lower income neighborhoods in Detroit, Michigan (Bednar, Reames, and Keoleian 2017) and Kansas City, Missouri (Reames 2016). On the other hand, experiencing energy burden may place a household at greater risk of poverty, since reduced discretionary funds can limit a household’s ability to invest in education, job training or economic capital, or make contributions to savings. Energy-burdened households may become vulnerable to utility debt, further adding to these constraints. Parallel research on housing affordability has found that families facing housing cost burdens (commonly defined as costs greater than 30% of household income) face a variety of challenges: increased odds of poor self-rated health and various medical conditions (Pollack, Griffin, and Lynch 2010); less ability to invest in child education (Newman and Holupka 2014); and less money set aside in savings (Mendenhall, Kramer, and Akresh 2014). Like housing costs, energy costs are at least partially fixed and essential to household wellbeing. Although energy costs are usually substantially smaller than direct housing costs (i.e., rent, mortgage payments), higher energy costs nonetheless reduce discretionary funds and likely impact low-income families in similar ways to housing cost burdens. Additionally, housing is a productive asset that can support investments in social or human capital, or generate income (Moser 1998). Applying the asset vulnerability framework, energy inefficient housing may less effectively guard against poverty spells. In general, housing is a critical factor on both the energy and poverty sides and has a significant impact on household energy needs. Tenants in rental housing face additional hurdles to managing their energy costs, since they are typically unable to invest in energy efficiency upgrades for their homes. This situation compounds behavioral and other barriers to energy efficiency that poor households face, since most households within 150% of the federal poverty level reside in rental housing and about 75% of renters pay their own utility bills (U.S. EIA 2015). From an ownership standpoint, landlords have minimal incentive to improve the energy efficiency of the units they oversee. This falls under a “split incentive problem,” where landlords do not accrue direct benefits from their investments in energy efficiency, while occupants responsible for paying energy bills do not have the opportunity to make structural changes to their dwelling (Gillingham, Newell, and Palmer 2009; Hernández and Bird 2010). Thus, the rental homes that many poor households occupy are often less efficient and subject to higher energy costs for heating and cooling, which may contribute to higher levels of energy rationing and less healthy home temperatures (Gillingham, Harding, and Rapson 2012; Middlemiss and Gillard 2015). Energy Assistance and the Welfare State The US has had a federally coordinated energy assistance policy since 1981, when the Low Income Home Energy Assistance Program (LIHEAP) was established to assist low-income households in meeting their home energy needs. The most common LIHEAP qualification threshold for household income is 150% of federal poverty guidelines, although many states use other thresholds. LIHEAP typically covers only a small portion of total energy bills and has historically reached only a fraction of eligible households, between 10% and 20% (APPRISE 2005; Kaiser and Pulsipher 2006). There are various other federal, state, local, and utility programs that encourage energy efficiency and renewable energy or provide assistance to low-income households. However, assistance programs generally do not have sufficient funds to serve all applicants, while energy efficiency programs often require partial payments from customers that present a barrier to low-income households (Evens 2015). Since Lutzenhiser and Hackett’s (1993) call for sociologists to connect environmental problems and energy with stratification and social welfare, relatively few scholars have answered. Welfare state research includes diverse contributions on tax systems (Prasad 2011, 2012), health care reform (Hacker 2004), and other topics, but little consideration to date regarding energy burdens. A comprehensive review of the welfare state literature lies beyond the scope of this paper, but it is worth highlighting the theoretical framework developed by Weir, Orloff, and Skocpol (1988), who explain the relatively weak US welfare state in terms of institutional structure: unlike the centralized bureaucratic states of Western Europe and their comparatively generous welfare expenditures, regional patronage networks and piecemeal bureaucratic-professional reforms characterize US history, preventing a robust welfare state from developing. Consistent with this, Hacker (2002) and Mettler (2011) both emphasize the extent to which US policy operates via indirect benefits (a tax break rather than direct provision, for example), which can easily fail to reach vulnerable populations in favor of the relatively affluent. These perspectives are also consistent with the various energy efficiency and clean energy tax credits and deductions, and the fact that these benefits primarily go to wealthy households (Borenstein and Davis 2016). Given such trends, we should expect that many households negatively impacted by energy burdens find insufficient welfare state resources available to help, and the patchwork of federal, state, local, and utility programs are unlikely to fill the gap (Evens 2015). We can logically assume that at the level of households, transitioning into economic poverty and the lower income it entails will increase the likelihood of experiencing energy burden. If the transition into economic poverty is preceded by a loss of household income, then—following the logic of Engel’s law of food elasticity (Engel 1857)—we can assume that the share of income dedicated to energy consumption will rise above operational definitions of energy burden. Yet the reverse direction has not been investigated, leading to the key research question here: does energy burden increase the likelihood of extending or entering economic poverty? In this analysis, we test the hypothesis that it does. Given the relatively non-discretionary nature of energy consumption, the cycles of debt energy burdens can initiate, and the social and health consequences of energy insecurity, we expect to find connections between energy burdens and future economic poverty status. Data Data for this project come from the Panel Study of Income Dynamics (PSID), a longitudinal survey of households that began in 1968 with an initial sample of nearly 5,000 US families. Data for this analysis come from household energy consumption gathered between 1999 and 2017 (waves 31–40). Three measures of poverty serve as dependent variables, using income thresholds for defining poverty according to whether a household lives within 100%, 150%, or 200% of per capita household poverty guidelines contained in the PSID. These thresholds are based on detailed poverty tables from the Census Bureau that account for family size and number of related children.1,2 (In 2017, families of four with two children and a total income of $24,944 or less fell within 100% of poverty guidelines.) Most households living within 150% of the poverty level qualify for LIHEAP subsidies, although different states use different thresholds.3 Additionally, it is common practice when using PSID data in particular to operationalize poverty at greater than 100% of federal guidelines (Cellini, McKernan, and Ratcliffe 2008) or when studying US poverty in general to analyze it according to multiple indicators (Ruggles 1992). Respondents were asked how much they spend on heating or gas fuel and electricity per month.4 Although energy costs are expected to fluctuate across seasons, the PSID survey does not capture this seasonality but instead asks about typical monthly costs.5 To test the effect of energy burden, this amount was calculated as a share of total household monthly income from the previous year. A dummy variable was created with 1 = households that spend at least 10% of their monthly income on heat and electricity combined, conceptualized as households that face energy burdens. Respondents whose utility bills are included in their rent were dropped from analysis (accounting for about a fifth of all renters). This metric uses energy expenditures, and does not account for the level of energy use necessary for a household to maintain healthy home temperatures, such as the “low income/high cost” measurement modeled by Hills (2011). The PSID data likely suffer from a “rounding strategy” used by respondents to self-report consumption expenditures (Pudney 2008), although the binary measure of energy burden should minimize this concern. Additionally, the PSID does not instruct respondents in how to handle utility subsidies when estimating monthly expenditures. Other independent variables are included to control for common predictors of poverty status. Since loss of income is an obvious potential cause for entering economic poverty, a dummy variable was created where 1 = a loss of at least 10% of household income from the previous wave. Ill health status was measured according to whether household heads assessed their health as fair or poor, with “good health” (those self-assessing as excellent, very good, or good) as the reference category. Both concurrent and lagged ill health were analyzed to control for the impacts of household heads’ ill health on economic wellbeing over time. Several dummy variables were created to account for social structural location. These include whether household heads are female (which is time-varying in cases where female respondents become household heads for the first time), young (under 30), senior (65 or older), college graduates, and own their home. Dummy variables were also created for the type of health insurance a respondent had: private health insurance, other health insurance (provided for military veterans, e.g.), Medicare and/or Medi-Gap, or Medicaid, with uninsured as the reference category. Since economic poverty is measured on a per capita basis, new additions to the family unit could qualify a household as living in poverty even if their income remains stable. To account for this, a dummy variable was created, where 1 = households that reported new family members since the previous wave. Marital status is also controlled for, with concurrent dummy variables accounting for whether the household head is married, divorced, separated, or widowed, with single as the reference category. Lagged variables for marital status are not included because of likely collinearity with female head of household: since the PSID defined the head of household as male for all opposite-sex married or long-term cohabitating couples across all but the most recent wave included in this analysis, a household that transitions from a male head in a previous wave to female head concurrently would also experience a related change in marital status. Table 1 shows descriptive statistics of variables used in this analysis. Table 1. Descriptive statistics Variable Proportion (200% poverty ratet-1 = 1) Proportion (200% poverty ratet-1 = 0) Energy-burdenedt-1 0.46 0.09 10% income loss 0.23 0.45 Ill health 0.23 0.14 Private Insurance 0.24 0.44 Other Insurance 0.03 0.03 Medicaid 0.20 0.06 Medicare 0.04 0.05 New family member 0.19 0.18 Married 0.53 0.68 Divorced 0.14 0.11 Separated 0.07 0.05 Widowed 0.05 0.04 Homeowner 0.49 0.68 Female HH 0.39 0.23 Young HH 0.19 0.17 Senior HH 0.07 0.08 College grad 0.11 0.26 Variable Proportion (200% poverty ratet-1 = 1) Proportion (200% poverty ratet-1 = 0) Energy-burdenedt-1 0.46 0.09 10% income loss 0.23 0.45 Ill health 0.23 0.14 Private Insurance 0.24 0.44 Other Insurance 0.03 0.03 Medicaid 0.20 0.06 Medicare 0.04 0.05 New family member 0.19 0.18 Married 0.53 0.68 Divorced 0.14 0.11 Separated 0.07 0.05 Widowed 0.05 0.04 Homeowner 0.49 0.68 Female HH 0.39 0.23 Young HH 0.19 0.17 Senior HH 0.07 0.08 College grad 0.11 0.26 Open in new tab Table 1. Descriptive statistics Variable Proportion (200% poverty ratet-1 = 1) Proportion (200% poverty ratet-1 = 0) Energy-burdenedt-1 0.46 0.09 10% income loss 0.23 0.45 Ill health 0.23 0.14 Private Insurance 0.24 0.44 Other Insurance 0.03 0.03 Medicaid 0.20 0.06 Medicare 0.04 0.05 New family member 0.19 0.18 Married 0.53 0.68 Divorced 0.14 0.11 Separated 0.07 0.05 Widowed 0.05 0.04 Homeowner 0.49 0.68 Female HH 0.39 0.23 Young HH 0.19 0.17 Senior HH 0.07 0.08 College grad 0.11 0.26 Variable Proportion (200% poverty ratet-1 = 1) Proportion (200% poverty ratet-1 = 0) Energy-burdenedt-1 0.46 0.09 10% income loss 0.23 0.45 Ill health 0.23 0.14 Private Insurance 0.24 0.44 Other Insurance 0.03 0.03 Medicaid 0.20 0.06 Medicare 0.04 0.05 New family member 0.19 0.18 Married 0.53 0.68 Divorced 0.14 0.11 Separated 0.07 0.05 Widowed 0.05 0.04 Homeowner 0.49 0.68 Female HH 0.39 0.23 Young HH 0.19 0.17 Senior HH 0.07 0.08 College grad 0.11 0.26 Open in new tab Methodology A challenge of this research area is the role of income: it determines whether a household falls below the poverty thresholds, and by definition, energy burden is a function of income. Income is therefore part of the outcome (i.e., poverty), and included among the predictors as a component of energy burden. To address this mechanical correlation, we restrict inclusion of energy burden to its lagged state (two years prior), so that income is not used to calculate both concurrent poverty status and state of energy burden. When inspecting the relationship between lagged energy burden and corresponding poverty status after each model, all correlations were below 0.11. Furthermore, by themselves, these poverty thresholds explain only a small proportion of the variation in energy burdens: we ran a series of logistic regressions predicting energy burden from different levels of poverty, predicting energy burden from total family income, and predicting each level of poverty from energy burden, resulting in a pseudo r2 between 0.2 and 0.3. In other words, we cannot characterize the relationship between poverty and energy burden as a simple mechanical or spurious correlation due to low income. While our measure of energy burden is operationally tied to income, as is poverty, these data support our arguments that energy burden and poverty are neither conceptually nor empirically reducible to each other, and thus it is appropriate to use them in the same statistical model.6 This characterization is consistent with research on material hardship; although income loss is obviously and consistently related to medical, food, or housing hardship, such experiences are not reducible to income alone and can also be impacted by factors such as changes in health status, caring for family members with a disability, or drug use (Heflin and Butler 2012; Heflin 2016). Given the binary dependent variables and panel structure of the PSID data, fixed effects logistic regression is utilized in this analysis (Allison 2009). Fixed effects models offer the ability to measure within-unit change in the outcome over time, taking the following form: $$ \log \left(\frac{p_{it}}{1-{p}_{it}}\right)=\kern0.5em {\mu}_t+\beta{\mathbf{X}}_{it}+\gamma{\mathbf{Z}}_i+{\alpha}_i,t=1,2\dots, T $$ where p refers to the probability of household i at time t of displaying a positive outcome, μ is the intercept that may vary over time, X represents a vector of time-varying predictors, Z represents a vector of time-invariant predictors, and α represents the total effects of unobserved variables that remain constant over time. One consequence of restricting analysis to within-unit change is the inability to estimate time-constant variables such as race. Additionally, any unit that displays no variation in the outcome over the course of data collection is dropped from analysis. However, if researchers can partial out unobserved and time-constant heterogeneity while focusing on within-unit change in the outcome, they can make stronger claims to causal inference not available with cross-sectional survey analysis. While it is difficult to make direct claims about causality with observational data, by establishing a statistically significant relationship between dependent and independent variables, accounting for omitted variable bias, and establishing a temporal ordering of events, fixed effects panel designs offer more rigorous analysis than cross-sectional approaches (Finkel 1995; Frees 2004). An additional challenge regards distinguishing between poverty entry and poverty extension. Prior and current states of poverty are correlated. To understand the relationship between energy burden and poverty status, it is important to know whether prior experience with energy burdens increases the likelihood of new entry into poverty, extends the duration of poverty, or both. With continuous outcomes, researchers often address this by estimating dynamic panel models that include lagged states of the dependent variable as predictors, using Arellano-Bond estimators to deal with resulting bias (Arellano and Bond 1991). However, there is less consensus regarding dynamic logit models. Instead of including lagged dependent variables as predictors, and taking inspiration from Stewart and Swaffield’s (1999) study of low pay transitions in Britain, we restrict samples to whether a household was below a poverty threshold at the previous wave (models predicting the endurance of poverty), or above a poverty threshold at the previous wave (models predicting poverty entry). Sociologists working with repeated binary outcomes characterized by state dependence should consider adopting this stratifying approach, which is not commonly employed in the discipline. Temporal ordering is established through the inclusion of lagged independent variables, including energy burden. For all models, unstandardized regression coefficients are presented in tables, while substantive discussion reports odds ratios. Finally, we ran sensitivity tests with several alternative model specifications: The addition of geographic controls, specifically fixed effects for state of residence. Operationalizing energy burden at 6% rather than 10% of household income, and as a continuous variable, to account for the depth of energy burden. The addition of housing type (single family homes, mobile homes, and multiunit structures). The addition of subsidies for rental housing and heat (as a continuous variable; about 5% of the sample received a government-funded heat subsidy, with an average annual payment of $390).7 A subset model: predicting 100% poverty within the population of families with income below 200% of the poverty threshold. This approach was used as an additional check regarding the role of income. Results for these alternative model specifications were consistent with the results presented below. Stata code for the full analysis is available online. Results Table 2 presents results from fixed effects logistic regression estimating whether households fall within various poverty thresholds, restricted to households below the corresponding poverty level at the previous wave. These models represent whether a household endures a poverty status across concurrent waves. For context, most households in the positive outcome for Models 1 and 2 qualify for federal heating assistance. Table 2. Unstandardized coefficients (and robust standard errors) for time-varying predictors from fixed effects logistic regression predicting different levels of poverty among households below the corresponding poverty level at the previous survey wave. Models control for year (not displayed) Model 1: 100% poverty ratet-1 = 1 Model 2: 150% poverty ratet-1 = 1 Model 3: 200% poverty ratet-1 = 1 Coef. Robust St. Err. Coef. Robust St. Err. Coef. Robust St. Err. Energy-burdenedt-1 1.019*** 0.090 0.674*** 0.06 0.761*** 0.048 10% income loss 3.286*** 0.132 3.219*** 0.099 3.139*** 0.093 Ill healtht-1 0.210* 0.097 0.064 0.065 0.035 0.057 Ill health 0.080 0.083 0.223*** 0.064 0.039 0.054 Private Insurancet-1 −0.593*** 0.175 −0.232* 0.099 0.108 0.076 Private Insurance −0.682*** 0.157 −0.338*** 0.086 −0.431*** 0.068 Other Insurancet-1 1.148*** 0.218 0.581*** 0.139 0.516*** 0.129 Other Insurance 0.189 0.200 0.100 0.15 0.202 0.133 Medicaidt-1 0.247* 0.106 0.108 0.076 0.466*** 0.07 Medicaid 0.487*** 0.100 0.692*** 0.074 0.686*** 0.071 Medicaret-1 0.412 0.220 0.366* 0.158 0.328* 0.15 Medicare 0.198 0.220 0.576*** 0.155 0.373** 0.137 New family member 0.194* 0.092 0.045 0.059 0.195*** 0.05 Married 0.192 0.198 0.907*** 0.148 0.266* 0.115 Divorced −0.716*** 0.201 −0.716*** 0.14 −0.400*** 0.107 Separated −0.079 0.180 −0.072 0.157 −0.282* 0.127 Widowed −0.942*** 0.266 −0.571** 0.203 −0.201 0.177 Homeownert-1 −0.485*** 0.122 −0.100 0.086 −0.310*** 0.068 Homeowner −0.269* 0.111 −0.233** 0.079 −0.443*** 0.067 Female HH 1.286*** 0.205 1.842*** 0.153 1.081*** 0.11 Young HH −0.326** 0.101 −0.286*** 0.074 −0.207** 0.065 Senior HH −0.060 0.279 0.159 0.168 0.813*** 0.145 College grad −1.034*** 0.254 −1.073*** 0.134 −0.772*** 0.112 N 7,971 13,982 19,589 Model 1: 100% poverty ratet-1 = 1 Model 2: 150% poverty ratet-1 = 1 Model 3: 200% poverty ratet-1 = 1 Coef. Robust St. Err. Coef. Robust St. Err. Coef. Robust St. Err. Energy-burdenedt-1 1.019*** 0.090 0.674*** 0.06 0.761*** 0.048 10% income loss 3.286*** 0.132 3.219*** 0.099 3.139*** 0.093 Ill healtht-1 0.210* 0.097 0.064 0.065 0.035 0.057 Ill health 0.080 0.083 0.223*** 0.064 0.039 0.054 Private Insurancet-1 −0.593*** 0.175 −0.232* 0.099 0.108 0.076 Private Insurance −0.682*** 0.157 −0.338*** 0.086 −0.431*** 0.068 Other Insurancet-1 1.148*** 0.218 0.581*** 0.139 0.516*** 0.129 Other Insurance 0.189 0.200 0.100 0.15 0.202 0.133 Medicaidt-1 0.247* 0.106 0.108 0.076 0.466*** 0.07 Medicaid 0.487*** 0.100 0.692*** 0.074 0.686*** 0.071 Medicaret-1 0.412 0.220 0.366* 0.158 0.328* 0.15 Medicare 0.198 0.220 0.576*** 0.155 0.373** 0.137 New family member 0.194* 0.092 0.045 0.059 0.195*** 0.05 Married 0.192 0.198 0.907*** 0.148 0.266* 0.115 Divorced −0.716*** 0.201 −0.716*** 0.14 −0.400*** 0.107 Separated −0.079 0.180 −0.072 0.157 −0.282* 0.127 Widowed −0.942*** 0.266 −0.571** 0.203 −0.201 0.177 Homeownert-1 −0.485*** 0.122 −0.100 0.086 −0.310*** 0.068 Homeowner −0.269* 0.111 −0.233** 0.079 −0.443*** 0.067 Female HH 1.286*** 0.205 1.842*** 0.153 1.081*** 0.11 Young HH −0.326** 0.101 −0.286*** 0.074 −0.207** 0.065 Senior HH −0.060 0.279 0.159 0.168 0.813*** 0.145 College grad −1.034*** 0.254 −1.073*** 0.134 −0.772*** 0.112 N 7,971 13,982 19,589 * p < 0.05, ** P < 0.01, *** P < 0.001 Open in new tab Table 2. Unstandardized coefficients (and robust standard errors) for time-varying predictors from fixed effects logistic regression predicting different levels of poverty among households below the corresponding poverty level at the previous survey wave. Models control for year (not displayed) Model 1: 100% poverty ratet-1 = 1 Model 2: 150% poverty ratet-1 = 1 Model 3: 200% poverty ratet-1 = 1 Coef. Robust St. Err. Coef. Robust St. Err. Coef. Robust St. Err. Energy-burdenedt-1 1.019*** 0.090 0.674*** 0.06 0.761*** 0.048 10% income loss 3.286*** 0.132 3.219*** 0.099 3.139*** 0.093 Ill healtht-1 0.210* 0.097 0.064 0.065 0.035 0.057 Ill health 0.080 0.083 0.223*** 0.064 0.039 0.054 Private Insurancet-1 −0.593*** 0.175 −0.232* 0.099 0.108 0.076 Private Insurance −0.682*** 0.157 −0.338*** 0.086 −0.431*** 0.068 Other Insurancet-1 1.148*** 0.218 0.581*** 0.139 0.516*** 0.129 Other Insurance 0.189 0.200 0.100 0.15 0.202 0.133 Medicaidt-1 0.247* 0.106 0.108 0.076 0.466*** 0.07 Medicaid 0.487*** 0.100 0.692*** 0.074 0.686*** 0.071 Medicaret-1 0.412 0.220 0.366* 0.158 0.328* 0.15 Medicare 0.198 0.220 0.576*** 0.155 0.373** 0.137 New family member 0.194* 0.092 0.045 0.059 0.195*** 0.05 Married 0.192 0.198 0.907*** 0.148 0.266* 0.115 Divorced −0.716*** 0.201 −0.716*** 0.14 −0.400*** 0.107 Separated −0.079 0.180 −0.072 0.157 −0.282* 0.127 Widowed −0.942*** 0.266 −0.571** 0.203 −0.201 0.177 Homeownert-1 −0.485*** 0.122 −0.100 0.086 −0.310*** 0.068 Homeowner −0.269* 0.111 −0.233** 0.079 −0.443*** 0.067 Female HH 1.286*** 0.205 1.842*** 0.153 1.081*** 0.11 Young HH −0.326** 0.101 −0.286*** 0.074 −0.207** 0.065 Senior HH −0.060 0.279 0.159 0.168 0.813*** 0.145 College grad −1.034*** 0.254 −1.073*** 0.134 −0.772*** 0.112 N 7,971 13,982 19,589 Model 1: 100% poverty ratet-1 = 1 Model 2: 150% poverty ratet-1 = 1 Model 3: 200% poverty ratet-1 = 1 Coef. Robust St. Err. Coef. Robust St. Err. Coef. Robust St. Err. Energy-burdenedt-1 1.019*** 0.090 0.674*** 0.06 0.761*** 0.048 10% income loss 3.286*** 0.132 3.219*** 0.099 3.139*** 0.093 Ill healtht-1 0.210* 0.097 0.064 0.065 0.035 0.057 Ill health 0.080 0.083 0.223*** 0.064 0.039 0.054 Private Insurancet-1 −0.593*** 0.175 −0.232* 0.099 0.108 0.076 Private Insurance −0.682*** 0.157 −0.338*** 0.086 −0.431*** 0.068 Other Insurancet-1 1.148*** 0.218 0.581*** 0.139 0.516*** 0.129 Other Insurance 0.189 0.200 0.100 0.15 0.202 0.133 Medicaidt-1 0.247* 0.106 0.108 0.076 0.466*** 0.07 Medicaid 0.487*** 0.100 0.692*** 0.074 0.686*** 0.071 Medicaret-1 0.412 0.220 0.366* 0.158 0.328* 0.15 Medicare 0.198 0.220 0.576*** 0.155 0.373** 0.137 New family member 0.194* 0.092 0.045 0.059 0.195*** 0.05 Married 0.192 0.198 0.907*** 0.148 0.266* 0.115 Divorced −0.716*** 0.201 −0.716*** 0.14 −0.400*** 0.107 Separated −0.079 0.180 −0.072 0.157 −0.282* 0.127 Widowed −0.942*** 0.266 −0.571** 0.203 −0.201 0.177 Homeownert-1 −0.485*** 0.122 −0.100 0.086 −0.310*** 0.068 Homeowner −0.269* 0.111 −0.233** 0.079 −0.443*** 0.067 Female HH 1.286*** 0.205 1.842*** 0.153 1.081*** 0.11 Young HH −0.326** 0.101 −0.286*** 0.074 −0.207** 0.065 Senior HH −0.060 0.279 0.159 0.168 0.813*** 0.145 College grad −1.034*** 0.254 −1.073*** 0.134 −0.772*** 0.112 N 7,971 13,982 19,589 * p < 0.05, ** P < 0.01, *** P < 0.001 Open in new tab Table 3. Unstandardized coefficients (and robust standard errors) for time-varying predictors from fixed effects logistic regression predicting different levels of poverty among households above the corresponding poverty level at the previous survey wave. Models control for year (not displayed) Model 4: 100% poverty ratet-1 = 0 Model 5: 150% poverty ratet-1 = 0 Model 6: 200% poverty ratet-1 = 0 Coef. Robust St. Err. Coef. Robust St. Err. Coef. Robust St. Err. Energy-burdenedt-1 1.109*** 0.068 0.921*** 0.072 0.453*** 0.076 10% income loss 4.050*** 0.087 3.598*** 0.067 3.091*** 0.054 Ill healtht-1 0.243*** 0.07 0.028 0.066 0.144* 0.069 Ill health 0.157* 0.066 0.248*** 0.064 0.184** 0.064 Private Insurancet-1 −0.072 0.085 0.015 0.076 0.093 0.072 Private Insurance −0.546*** 0.083 −0.448*** 0.071 −0.329*** 0.064 Other Insurancet-1 0.103 0.186 0.440** 0.162 0.501** 0.152 Other Insurance 0.540*** 0.146 0.459** 0.145 0.636*** 0.140 Medicaidt-1 0.098 0.093 0.619*** 0.098 0.744*** 0.113 Medicaid 0.642*** 0.084 0.576*** 0.083 0.814*** 0.092 Medicaret-1 0.060 0.171 0.316* 0.155 0.182 0.133 Medicare 0.105 0.157 0.194 0.139 0.131 0.123 New family member 0.499*** 0.067 0.702*** 0.060 0.714*** 0.056 Married −0.743*** 0.112 −0.596*** 0.095 −0.657*** 0.098 Divorced −0.407*** 0.117 −0.692*** 0.113 −0.345** 0.117 Separated −0.729*** 0.125 −0.477*** 0.122 −0.043 0.138 Widowed −0.369* 0.187 −0.032 0.170 0.038 0.180 Homeownert-1 −0.152* 0.071 −0.205** 0.067 −0.055 0.067 Homeowner −0.702*** 0.072 −0.903*** 0.067 −0.859*** 0.061 Female HH 0.319** 0.101 0.695*** 0.098 0.272** 0.094 Young HH 0.379*** 0.074 0.401*** 0.071 0.440*** 0.069 Senior HH −0.039 0.156 0.044 0.136 0.071 0.126 College grad −0.722*** 0.111 −0.659*** 0.095 −0.499*** 0.086 N 23,531 28,517 30,406 Model 4: 100% poverty ratet-1 = 0 Model 5: 150% poverty ratet-1 = 0 Model 6: 200% poverty ratet-1 = 0 Coef. Robust St. Err. Coef. Robust St. Err. Coef. Robust St. Err. Energy-burdenedt-1 1.109*** 0.068 0.921*** 0.072 0.453*** 0.076 10% income loss 4.050*** 0.087 3.598*** 0.067 3.091*** 0.054 Ill healtht-1 0.243*** 0.07 0.028 0.066 0.144* 0.069 Ill health 0.157* 0.066 0.248*** 0.064 0.184** 0.064 Private Insurancet-1 −0.072 0.085 0.015 0.076 0.093 0.072 Private Insurance −0.546*** 0.083 −0.448*** 0.071 −0.329*** 0.064 Other Insurancet-1 0.103 0.186 0.440** 0.162 0.501** 0.152 Other Insurance 0.540*** 0.146 0.459** 0.145 0.636*** 0.140 Medicaidt-1 0.098 0.093 0.619*** 0.098 0.744*** 0.113 Medicaid 0.642*** 0.084 0.576*** 0.083 0.814*** 0.092 Medicaret-1 0.060 0.171 0.316* 0.155 0.182 0.133 Medicare 0.105 0.157 0.194 0.139 0.131 0.123 New family member 0.499*** 0.067 0.702*** 0.060 0.714*** 0.056 Married −0.743*** 0.112 −0.596*** 0.095 −0.657*** 0.098 Divorced −0.407*** 0.117 −0.692*** 0.113 −0.345** 0.117 Separated −0.729*** 0.125 −0.477*** 0.122 −0.043 0.138 Widowed −0.369* 0.187 −0.032 0.170 0.038 0.180 Homeownert-1 −0.152* 0.071 −0.205** 0.067 −0.055 0.067 Homeowner −0.702*** 0.072 −0.903*** 0.067 −0.859*** 0.061 Female HH 0.319** 0.101 0.695*** 0.098 0.272** 0.094 Young HH 0.379*** 0.074 0.401*** 0.071 0.440*** 0.069 Senior HH −0.039 0.156 0.044 0.136 0.071 0.126 College grad −0.722*** 0.111 −0.659*** 0.095 −0.499*** 0.086 N 23,531 28,517 30,406 * p < 0.05, ** P < 0.01, *** P < 0.001 Open in new tab Table 3. Unstandardized coefficients (and robust standard errors) for time-varying predictors from fixed effects logistic regression predicting different levels of poverty among households above the corresponding poverty level at the previous survey wave. Models control for year (not displayed) Model 4: 100% poverty ratet-1 = 0 Model 5: 150% poverty ratet-1 = 0 Model 6: 200% poverty ratet-1 = 0 Coef. Robust St. Err. Coef. Robust St. Err. Coef. Robust St. Err. Energy-burdenedt-1 1.109*** 0.068 0.921*** 0.072 0.453*** 0.076 10% income loss 4.050*** 0.087 3.598*** 0.067 3.091*** 0.054 Ill healtht-1 0.243*** 0.07 0.028 0.066 0.144* 0.069 Ill health 0.157* 0.066 0.248*** 0.064 0.184** 0.064 Private Insurancet-1 −0.072 0.085 0.015 0.076 0.093 0.072 Private Insurance −0.546*** 0.083 −0.448*** 0.071 −0.329*** 0.064 Other Insurancet-1 0.103 0.186 0.440** 0.162 0.501** 0.152 Other Insurance 0.540*** 0.146 0.459** 0.145 0.636*** 0.140 Medicaidt-1 0.098 0.093 0.619*** 0.098 0.744*** 0.113 Medicaid 0.642*** 0.084 0.576*** 0.083 0.814*** 0.092 Medicaret-1 0.060 0.171 0.316* 0.155 0.182 0.133 Medicare 0.105 0.157 0.194 0.139 0.131 0.123 New family member 0.499*** 0.067 0.702*** 0.060 0.714*** 0.056 Married −0.743*** 0.112 −0.596*** 0.095 −0.657*** 0.098 Divorced −0.407*** 0.117 −0.692*** 0.113 −0.345** 0.117 Separated −0.729*** 0.125 −0.477*** 0.122 −0.043 0.138 Widowed −0.369* 0.187 −0.032 0.170 0.038 0.180 Homeownert-1 −0.152* 0.071 −0.205** 0.067 −0.055 0.067 Homeowner −0.702*** 0.072 −0.903*** 0.067 −0.859*** 0.061 Female HH 0.319** 0.101 0.695*** 0.098 0.272** 0.094 Young HH 0.379*** 0.074 0.401*** 0.071 0.440*** 0.069 Senior HH −0.039 0.156 0.044 0.136 0.071 0.126 College grad −0.722*** 0.111 −0.659*** 0.095 −0.499*** 0.086 N 23,531 28,517 30,406 Model 4: 100% poverty ratet-1 = 0 Model 5: 150% poverty ratet-1 = 0 Model 6: 200% poverty ratet-1 = 0 Coef. Robust St. Err. Coef. Robust St. Err. Coef. Robust St. Err. Energy-burdenedt-1 1.109*** 0.068 0.921*** 0.072 0.453*** 0.076 10% income loss 4.050*** 0.087 3.598*** 0.067 3.091*** 0.054 Ill healtht-1 0.243*** 0.07 0.028 0.066 0.144* 0.069 Ill health 0.157* 0.066 0.248*** 0.064 0.184** 0.064 Private Insurancet-1 −0.072 0.085 0.015 0.076 0.093 0.072 Private Insurance −0.546*** 0.083 −0.448*** 0.071 −0.329*** 0.064 Other Insurancet-1 0.103 0.186 0.440** 0.162 0.501** 0.152 Other Insurance 0.540*** 0.146 0.459** 0.145 0.636*** 0.140 Medicaidt-1 0.098 0.093 0.619*** 0.098 0.744*** 0.113 Medicaid 0.642*** 0.084 0.576*** 0.083 0.814*** 0.092 Medicaret-1 0.060 0.171 0.316* 0.155 0.182 0.133 Medicare 0.105 0.157 0.194 0.139 0.131 0.123 New family member 0.499*** 0.067 0.702*** 0.060 0.714*** 0.056 Married −0.743*** 0.112 −0.596*** 0.095 −0.657*** 0.098 Divorced −0.407*** 0.117 −0.692*** 0.113 −0.345** 0.117 Separated −0.729*** 0.125 −0.477*** 0.122 −0.043 0.138 Widowed −0.369* 0.187 −0.032 0.170 0.038 0.180 Homeownert-1 −0.152* 0.071 −0.205** 0.067 −0.055 0.067 Homeowner −0.702*** 0.072 −0.903*** 0.067 −0.859*** 0.061 Female HH 0.319** 0.101 0.695*** 0.098 0.272** 0.094 Young HH 0.379*** 0.074 0.401*** 0.071 0.440*** 0.069 Senior HH −0.039 0.156 0.044 0.136 0.071 0.126 College grad −0.722*** 0.111 −0.659*** 0.095 −0.499*** 0.086 N 23,531 28,517 30,406 * p < 0.05, ** P < 0.01, *** P < 0.001 Open in new tab The key variable of interest—whether a household was energy-burdened in the prior wave—is statistically significant in all models. Net of other factors, we see in Model 1 that energy-burdened households had about 175% greater odds of remaining in poverty two years later relative to their non-burdened peers. In Model 2 among households previously within 150% of the poverty threshold and net of other factors, those with prior energy burdens were about twice as likely to endure this economic status compared with households without prior energy burdens. In Model 3, we again see the impact of prior energy burdens resulting in over 100% greater odds of enduring a status below 200% of the poverty threshold. As noted above, this energy burden metric does not account for rationing or the level of energy use needed to maintain healthy home temperatures. Given the energy rationing described in qualitative studies, if data were available to capture the higher cost of “necessary” energy use, we would likely see an even more deleterious impact on economic status. Several household demographics historically associated with economic poverty are controlled in these models. Unsurprisingly, adding new family members increased the odds of staying in poverty. Women who became new household heads had about triple the odds compared with male household heads of staying within 100% of the poverty threshold. Focusing on transitions into poverty, table 3 presents results from fixed effects logistic regression, restricted to households above the corresponding poverty level at the previous wave. Across all three models, lagged energy burden had a statistically significant relationship to concurrent poverty status. Relative to non-burdened households, those with prior energy burdens had about double the odds of transitioning into poverty (Model 4), 150% greater odds of transitioning within 150% of the poverty threshold (Model 5), and about 60% greater odds of transitioning within 200% of the poverty threshold (Model 6). The impact of prior energy burdens on concurrent poverty status again accounts for whether a household experienced a substantial loss in income, which is obviously the single largest driver of the outcome across all models. As with Models 1–3, Models 4–6 control for a variety of other variables commonly associated with transitions into economic poverty. Both married and divorced heads of household were less likely to transition into poverty relative to single household heads. Likewise, owning a home or graduating from college were positive predictors of economic wellbeing. Although a full discussion of the role of health is beyond the scope of this paper, it is worthy to note the impact of health on economic wellbeing. In Model 4, for example, household heads rating their subjective health as fair or poor had about 20% greater odds of entering poverty two years later, compared to their counterparts rating their health as good, very good, or excellent. Concurrently having private health insurance predicts economic wellbeing. This is consistent with other research suggesting that access to health insurance through the Affordable Care Act marketplace may lower home mortgage delinquency rates (Gallagher et al. 2017), and is aligned with research linking maternal health with entrance into material hardship in the forms of food insufficiency, utility disconnections, unmet medical needs, and housing problems (Heflin 2013). Discussion These results clearly point to the economic disadvantage faced by energy-burdened households. As policymakers and activists continue to address energy consumption as part of climate change mitigation, it is important to recognize how increased energy prices could affect economically vulnerable households. Impacts from increased energy costs can be immediate, for instance if a household reduces expenditure on other necessities in order to heat the home (Bhattacharya et al. 2003), or when a household falls behind on their utility bills and consequently harms their credit rating (see Fourcade and Healy 2013 on the stratifying effect of credit ratings in the United States). Activists and policymakers should keep in mind that too many households lack a proper safety net to secure them from policies that increase energy costs. For researchers and theorists, this analysis highlights how energy burdens may act as triggering events for poverty entry and draws attention to the inadequacy of welfare state programs addressing this. We hope this analysis points to how energy consumption, the welfare state, and social stratification are intertwined. Environmental activists, program implementers, and policymakers should keep these energy equity issues in mind when confronting potential transitions to cleaner energy. For example, programs offering energy efficiency or clean energy services for low-income households can simultaneously work toward economic equity (by ensuring access to affordable energy) and environmental mitigation (by ensuring cleaner and/or more efficient energy consumption), social concerns historically brought up by Lutzenhiser and Hackett (1993). Ideally, renewable energy and energy efficiency programs would be accessible to low-income households, with funding assistance if necessary, since utility hardships are linked with other forms of hardship and assistance may reduce risks in other areas (Finnigan and Meagher 2019). In reality, activists and policymakers need to consider energy burdens in three policy areas: energy assistance, energy efficiency programs, and clean or renewable energy programs. These results demonstrate that the benefits of energy efficiency and weatherization could extend beyond energy use to include reduced risk of economic poverty. For example, home insulation can last 80 years or more, effectively serving the duration of residency for a given household. Direct energy assistance may offer similar benefits in the short term, although future research should confirm that direct and weatherization assistance can mitigate the results found in this analysis. In terms of program design, high-quality and well-funded low-income energy programs may help reduce the risk of economic poverty while supporting environmental, climate, and other goals. We suspect, however, that current funding levels for LIHEAP and weatherization programs insufficiently mitigate inadequate energy access, at least at the federal level. Historically, LIHEAP and federal weatherization assistance funding has served only a fraction of eligible households (Higgins and Lutzenhiser 1995), although it can be a critical form of assistance for those who receive it (Murray and Mills 2014), and federal weatherization assistance programs have reduced energy bills by 12% on average (U.S. DOE 2015). Nonetheless, policy or activist efforts to address climate change mitigation could consider energy consumption as a site of intersection that simultaneously addresses social and environmental concerns (Hernández 2015). To maximize the benefits of energy efficiency programs, these programs must be accessible to low-income households. This potentially requires greater funding for direct assistance or to address building health and safety issues that must be fixed before a retrofit can safely proceed (Scheu et al. 2018). Furthermore, to the extent that low-income households engage in energy rationing, low-income efficiency programs may see smaller energy savings but greater benefits in terms of healthy home temperatures and other non-energy benefits. For energy efficiency programs to be accessible to renters, these programs need to target multifamily building owners. Landlords remain key to the reform of energy consumption. Bird and Hernández (2012) suggest incentivizing landlords to take on energy efficiency upgrades for their units by covering the cost of loan financing in exchange for committing their property to low-income renting for the period of the loan. Future research could help to encourage multifamily energy efficiency by evaluating the benefits for landlords who do not pay utility bills, for instance more comfortable housing units that reduce tenant turnover and increase net operating income for the building (Philbrick, Scheu, and Brand 2016). Energy efficiency programs targeting multifamily buildings can offer services to a significant share of low-income renters. For example, in Chicago, 59% of multifamily housing units are in low-cost, low-income neighborhoods, or are subsidized (Corso, Garascia, and Scheu 2017), and programs targeting these buildings may reduce energy burdens among low-income tenants. Research in the UK, however, indicates that income poverty spells typically last longer than fuel poverty (consistent with data used here), although rates of long-term fuel poverty exit vary across rural/urban settings, indicating that policies targeting housing stock may see inconsistent results (Roberts, Vera-Toscano, and Phimister 2015). Nonetheless, improving the efficiency of low-income housing will likely produce environmental benefits while potentially reducing the risk of energy burdens triggering poverty entry. In addition to implications for activists, policymakers, and program implementers, these results also point to several fruitful directions for further research. We suspect the relationship between energy burden and economic poverty reflects multiple points of vulnerability and resource constraints relating to physical housing structure, neighborhood context, and consequences for social networks, physical health, or mental health. Energy burdens are under-studied and need much more attention from social researchers to refine the mechanisms that connect this experience with poverty, and to understand the myriad coping strategies households employ. Comparative research could illuminate the effectiveness of policies and programs in mitigating these problems. Other research areas could benefit from considering energy burdens as possible mechanisms, moderators, or complicating factors for other relationships. As one example, these results could hold implications for research on environmental risk perceptions and support for clean energy. Overall, lower-income individuals perceive greater long-term environmental risks than higher-income individuals (Trumbo et al. 2011; Lo 2014). Researchers could inquire how energy insecurity affects support for clean energy or climate change mitigation policy, particularly when those policies may result in slightly higher energy prices or inequitable impacts on consumers. Although renewable energy costs have fallen dramatically and are near parity with the cost of other fuels (Lazard 2016), carbon taxes or related instruments may nonetheless contribute to higher energy costs for some consumers, or the perception of higher costs. Additionally, tax credits and rebates for clean energy technologies can exacerbate inequity because upfront costs or credit requirements make them accessible primarily to the rich: 60% of US clean energy tax credits have gone to the top income quintile (Borenstein and Davis 2016). It is thus important to understand how perceptions of energy burdens or inequities affect support for social and environmental change. Energy justice could be crucial for working toward the goals held by mainstream environmentalists and clean energy movements. Along this line of thought, several topics are proposed for future research to extend insights into energy burdens in the United States: Does the experience of energy burden erode support for clean energy (assuming clean energy options are perceived as more expensive than conventional energy options)? To what extent does energy consumption constitute a barrier to social mobility if basic needs exceed energy burden thresholds? How does energy burden compare to equivalent concepts of food and housing cost burden in terms of volatility and temporal patterns, and interrelationships? What are the relative impacts on poverty and social mobility of these and other independent causes of poverty? In what ways does experiencing an energy burden contribute to poverty, and do the mechanisms vary in meaningful ways? What are the roles of housing (physical structure as well as affordability), health, financial constraints, cognitive capacity (Mani et al. 2013), utility rates and disconnection policies, and other factors? What is the role of utility assistance, low-income weatherization assistance, LIHEAP, or other energy programs in moderating the relationship between energy burden and poverty? Do the programs that have the largest and longest-lasting impact on reducing energy use (e.g., weatherization) also have a large impact on reducing energy burden? The first question brings a concern from social stratification (financial hardship) into the realm of environmental sociology and related social science by suggesting an analysis of support for clean energy development. The second question brings a concern from environmental sociology (energy consumption) into the realm of social stratification and material hardship and ties energy policy to broader poverty alleviation efforts, while the third question delves deeper into the processes that link energy and financial hardships. The fourth question addresses a potential moderating factor in that relationship and may have implications for program design and our theoretical understanding of the welfare state. When analyzing energy consumption in the United States, sociologists typically focus on excessive use. But household energy scarcity could constitute an important topic for environmental sociology and social science more broadly, particularly in contexts where natural resource consumption to meet basic needs results in financial hardship. Sociological research in general would benefit from greater attention to energy burden and energy insecurity, in terms of financial hardships as well as variation in utility rates8, the impact of utility disconnection, the intersection of housing and energy, and other topics. This presents an opportunity for social scientists to engage in a larger discourse about social inequality that will be crucial to future conversations about energy and climate policy. About the Authors Jeremiah Bohr is an assistant professor of sociology at the University of Wisconsin Oshkosh. His research interests include climate change denial, environmental sociology and politics, social stratification, and computational social science. His work has appeared in journals such as Environmental Politics, Environmental Sociology, and Climatic Change, and has been featured in media outlets such as The Guardian and Time. Anna C. McCreery is a research manager at Elevate Energy, a non-profit dedicated to smarter energy use for all. Her research areas include household energy insecurity, “big data” analysis of residential hourly electricity usage, interconnections between housing, energy, and health, and environmental social movements. Her work is published in Organization and Environment, Environment and Planning A, and the International Journal of Sustainability in Higher Education, and also appears in industry white papers and policy factsheets. Footnotes 1 The models predicting income of 150% and 200% of poverty thresholds are included to address concerns in the literature that the 100% threshold is too low and reflects outdated modes of consumption. A full discussion of this issue is beyond the scope of this paper, but see Ruggles (1990, 1992) and Blank (2008) for examples. Additionally, since many states use 150% and 200% thresholds to determine qualification for energy assistance, it makes sense to extend analysis to these levels. 2 Although the supplemental poverty measure could account for differences in cost of living, it was not published by the US Census until 2011 and thus not available for many of the PSID waves analyzed here. 3 Another common threshold is 60% of the state median income, which is often higher than 150% of the federal poverty guideline, and some states extend the threshold to 200% of the poverty level. The LIHEAP Clearinghouse provides eligibility details for each state: https://liheapch.acf.hhs.gov/snapshots.htm. 4 Before 2011, this question asked about “gas or other types of heating fuel,” and was revised in 2011 to “gas or other types of fuel.” There are other small changes in PSID question wording over time, which we do not believe are likely to influence results. Models restricted to 1999–2009 produce results consistent with the models presented here. 5 There still may be a seasonal aspect to these data because respondents provide different estimates of typical monthly costs depending on when you ask them (e.g., average gas costs among respondents interviewed in March–May is $164, versus $157 in June–August). We ran basic one-way ANOVA tests for the influence of interview season, defining winter as December through February and summer as June through August, which is aligned with heating and cooling months for most of the United States. These models found a statistically significant effect of interview season on heating costs and electricity costs (both p < .001), and a marginally significant effect of interview season on combined heating and electricity costs (p < .1). We therefore ran additional sensitivity checks controlling for interview season, and the results are consistent with the models presented here. 6 We also explored the reverse relationship—different thresholds of economic poverty leading to energy burdens in a fixed effects logistic regression model. Results are available upon request, and indicate that lagged economic poverty is positively correlated with future energy burden status (controlling for factors described in the Data section). 7 Heating subsidies are not included in the main model due to potential double-counting (if respondents track subsidies as reduced expenditure, then they are incorporated into the energy burden metric), and a potential spurious correlation due to eligibility criteria. LIHEAP “crisis” assistance is only available to households at risk of disconnection (i.e., who are behind on their utility bills), and many states use higher income eligibility thresholds and/or provide larger subsidies for crisis assistance. These characteristics would select for households at risk of poverty entry or extension. Historically, LIHEAP participation was strongly predicted by households receiving other forms of social welfare (SSI, e.g., Higgins and Lutzenhiser 1995). 8 State-level average residential prices for electricity in the contiguous United States in May 2018 ranged from 9.53 cents/kWh in Louisiana to 21.69 cents/kWh in Massachusetts (U.S. EIA 2018). Rates and rate structures also vary within states, for example, flat rates, time-of-use rates, inclining or declining block rates, flat customer charges versus volumetric charges, etc. We would like to thank Cindy Buckley, Anne Evens, Margaret Garascia, Tim Liao, Adam Mayer, Deborah Philbrick, Rachel Scheu, and the anonymous reviewers at Social Forces for their constructive feedback on earlier versions of this research. All shortcomings remain our own. Direct correspondence to Jeremiah Bohr, University of Wisconsin Oshkosh, Department of Sociology, 800 Algoma Blvd., Oshkosh, WI 54901, USA References Allison , Paul 2009 . Fixed Effects Regression Models . Thousand Oaks, CA : SAGE Publications . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Applied Public Policy Research Institute for Study and Evaluation (APPRISE) . 2005 . LIHEAP Energy Burden Evaluation Study . Accessed August 2018 . https://www.energy.gov/sites/prod/files/2018/03/f49/HHS_LIHEAP%20study_july_2005.pdf Arellano , Manuel , and Stephen , Bond . 1991 . “ Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations .” Review of Economic Studies 58 : 277 – 97 . Google Scholar Crossref Search ADS WorldCat Bane , Mary Jo , and David T. Ellwood . 1986 . “ Slipping Into and Out of Poverty: The Dynamics of Spells .” The Journal of Human Resources 21 ( 1 ): 1 – 23 . Google Scholar Crossref Search ADS WorldCat Bednar , Dominic J. , Tony , Gerard Reames , and Gregory , A. Keoleian . 2017 . “ The Intersection of Energy and Justice: Modeling the Spatial, Racial/Ethnic and Socioeconomic Patterns of Urban Residential Heating Consumption and Efficiency in Detroit, Michigan .” Energy and Buildings 143 : 25 – 34 . Google Scholar Crossref Search ADS WorldCat Bhattacharya , Jayanta , Thomas DeLeire , Steven Haider , and Janet Currie . 2003 . “ Heat or Eat? Cold-Weather Shocks and Nutrition in Poor American Families .” American Journal of Public Health 93 ( 7 ): 1149 – 54 . Google Scholar Crossref Search ADS PubMed WorldCat Bird , Stephen , and Hernández , Diana . 2012 . “ Policy Options for the Split Incentive: Increasing Energy Efficiency for Low-Income Renters .” Energy Policy 48 : 506 – 514 . Google Scholar Crossref Search ADS PubMed WorldCat Blank , Rebecca M . 2008 . “ How to Improve Poverty Measurement in the United States .” Journal of Policy Analysis and Management 27 ( 2 ): 233 – 254 . Google Scholar Crossref Search ADS WorldCat Boardman , Brenda . 1991 . Fuel Poverty: From Cold Homes to Affordable Warmth . London : Belhaven Press . Google Preview WorldCat COPAC ———. 2010 . Fixing Fuel Poverty: Challenges and Solutions . London : Earthscan . WorldCat COPAC ———. 2012 . “ Fuel Poverty Synthesis: Lessons Learnt, Actions Needed .” Energy Policy 49 : 143 – 148 . Crossref Search ADS WorldCat Borenstein , Severin , and Davis , Lucas W. . 2016 . “ The Distributional Effects of U.S. Clean Energy Tax Credits .” Tax Policy and the Economy 30 : 191 – 234 . Google Scholar Crossref Search ADS WorldCat Brady , David , Ryan M. Finnigan , and Sabine Hübgen . 2017 . “ Rethinking the Risks of Poverty: A Framework for Analyzing Prevalences and Penalties .” American Journal of Sociology 123 ( 3 ): 740 – 786 . Google Scholar Crossref Search ADS WorldCat Büchs , Milena , AbuBakr Bahaj , Luke Blunden , Leonidas Bourikas , Jane Falkingham , Patrick James , Mamusu Kamanda , and Yue Wu . 2018 . `` Sick and Stuck at Home–How Poor Health Increases Electricity Consumption and Reduces Opportunities for Environmentally-Friendly Travel in the United Kingdom .'' Energy Research & Social Science 44 ( 2018 ): 250 – 9 . Google Scholar Crossref Search ADS WorldCat Cellini , Stephanie Riegg , Signe-Mary McKernan , and Caroline Ratcliffe . 2008 . “ The Dynamics of Poverty in the United States: A Review of Data, Methods, and Findings .” Journal of Policy Analysis and Management 27 ( 3 ): 577 – 605 . Google Scholar Crossref Search ADS WorldCat Center for Education Policy Analysis 2006 . Colorado Statewide Homeless Count . Denver, CO : Graduate School of Public Affairs, University of Colorado Denver . Google Preview WorldCat COPAC Cook , John T. , Deborah A. Frank , Patrick H. Casey , Ruth Rose-Jacobs , Maureen M. Black , Mariana Chilton , Stephanie Ettinger de Cuba et al. 2008 . “ A Brief Indicator of Household Energy Security: Associations with Food Security, Child Health, and Child Development in US Infants and Toddlers .” Pediatrics 122 ( 4 ): e867 − e875 . Google Scholar Crossref Search ADS PubMed WorldCat Corso , Abigail , Garascia , Margaret , and Scheu , Rachel . 2017 . Segmenting Chicago Multifamily Housing to Improve Energy Efficiency Programs . Chicago, IL : Elevate Energy . Google Preview WorldCat COPAC Desmond , Matthew . 2016 . Evicted: Poverty and Profit in the American City . New York, NY : Penguin Random House . Google Preview WorldCat COPAC DiPrete , Thomas A. , and Patricia , A. McManus . 2000 . “ Family Change, Employment Transitions, and the Welfare State: Household Income Dynamics in the United States and Germany .” American Sociological Review 65 : 343 – 370 . Google Scholar Crossref Search ADS WorldCat Drehobl , Ariel , and Lauren Ross . 2016 . Lifting the High Energy Burden in America’s Largest Cities: How Energy Efficiency Can Improve Low Income and Underserved Communities . Washington, DC : ACEEE . Accessed January 2017. http://aceee.org/research-report/u1602. Google Preview WorldCat COPAC Engel , Ernst . 1857 . “ Die produktions-und konsumptionsverhältnisse des königreichs Sachsen .” Zeitschrift des Statistischen Bureaus des Königlich Sächsischen Ministeriums des Innern 8 : 1 – 54 . WorldCat Evens , Anne . 2015 . “ The Utility of the Future and Low-Income Households .” The Electricity Journal 28 ( 10 ): 43 – 52 . Google Scholar Crossref Search ADS WorldCat Finkel , Steven E. . 1995 . Causal Analysis with Panel Data . Thousand Oaks, CA : SAGE Publications . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Finnigan , Ryan , and Meagher , Kelsey D. . 2019 . “ Past Due: Combinations of Utility and Housing Hardship in the United States .” Sociological Perspectives 62 ( 1 ): 96 – 119 . Google Scholar Crossref Search ADS WorldCat Fourcade , Marion , and Healy , Kieran . 2013 . “ Classification Situations: Life-Chances in the Neoliberal Era .” Accounting, Organizations and Society 38 : 559 – 572 . Google Scholar Crossref Search ADS WorldCat Frees , Edward W. . 2004 . Longitudinal and Panel Data: Analysis and Applications in the Social Sciences . Cambridge, UK : Cambridge University Press . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Gallagher , Emily A. , Gopalan , Radhakrishnan , Grinstein-Weiss , Michal , Roll , Stephen P. , and Davison , Genevieve . 2017 . Home Delinquency Rates are Lower Among ACA Marketplace Households: Evidence from a Natural Experiment . St. Louis, MO : Washington University in St. Louis Center for Social Development, CSD Research Brief , pp. 17 – 01 . Google Preview WorldCat COPAC Gillingham , Kenneth , Richard G. Newell , and Karen Palmer . 2009 . `` Energy Efficiency Economics and Policy .'' Annual Review of Resource Economics 1 ( 1 ): 597 – 620 . Google Scholar Crossref Search ADS WorldCat ———, Matthew Harding , and David Rapson . 2012 . “ Split Incentives in Residential Energy Consumption .” The Energy Journal 33 ( 2 ): 37 – 62 . WorldCat Hacker , Jacob S. . 2002 . The Divided Welfare State: The Battle Over Public and Private Social Benefits in the United States . Cambridge, UK : Cambridge University Press . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC ———. 2004 . “ Privatizing Risk Without Privatizing the Welfare State: The Hidden Politics of Social Policy Retrenchment in the United States .” American Political Science Review 98 ( 2 ): 243 – 260 . Crossref Search ADS WorldCat Harrison , Conor , and Jeff Popke . 2011 . “ ‘Because You Got to Have Heat’: The Networked Assemblage of Energy Poverty in Eastern North Carolina .” Annals of the Association of American Geographers 101 ( 4 ): 949 – 961 . Google Scholar Crossref Search ADS WorldCat Heflin , Colleen . 2006 . “ Dynamics of Material Hardship in the Women’s Employment Study .” Social Service Review 80 ( 3 ): 377 – 397 . Google Scholar Crossref Search ADS WorldCat ———. 2013 . “ Why do Women Enter and Exit from Material Hardship? ” Journal of Family Issues 34 ( 5 ): 631 – 660 . Crossref Search ADS WorldCat ———. 2016 . “ Family Instability and Material Hardship: Results from the 2008 Survey of Income and Program Participation .” Journal of Family and Economic Issues 37 ( 3 ): 359 – 372 . Crossref Search ADS WorldCat Heflin , Colleen. , and J. S. Butler . 2012 . “ Why do Women Enter and Exit from Material Hardship? ” Journal of Family Issues 34 ( 5 ): 631 – 660 . Google Scholar Crossref Search ADS WorldCat Heflin , Colleen , Andrew S. London , and Ellen K. Scott . 2011 . “ Mitigating Material Hardship: The Strategies Low-Income Families Employ to Reduce the Consequences of Poverty .” Sociological Inquiry 81 ( 2 ): 223 – 246 . Google Scholar Crossref Search ADS WorldCat Hernández , Diana . 2013 . “ Energy Insecurity: A Framework for Understanding Energy, the Built Environment, and Health among Vulnerable Populations in the Context of Climate Change .” American Journal of Public Health 103 ( 4 ): e32 − e34 . Google Scholar Crossref Search ADS PubMed WorldCat ———. 2015 . “ Sacrifice Along the Energy Continuum: A Call for Energy Justice .” Environmental Justice 8 ( 4 ): 151 – 156 . Crossref Search ADS PubMed WorldCat ———. 2016 . “ Understanding ‘Energy Insecurity’ and Why it Matters to Health .” Social Science & Medicine 167 : 1 – 10 . Crossref Search ADS WorldCat Hernández , Diana , and Stephen Bird . 2010 . “ Energy Burden and the Need for Integrated Low-Income Housing and Energy Policy .” Poverty & Public Policy 2 ( 4 ): 5 – 25 . Google Scholar Crossref Search ADS PubMed WorldCat Higgins , Lorie , and Loren Lutzenhiser . 1995 . “ Ceremonial Equity: Low-Income Energy Assistance and the Failure of Socio-Environmental Policy .” Social Problems 42 ( 4 ): 468 – 492 . Google Scholar Crossref Search ADS WorldCat Hills , John . 2011 . “Fuel Poverty—The Problem and its Measurement.” In Interim Report of the Fuel Poverty Review. CASE . London : LSE . Google Preview WorldCat COPAC Himmelstein , David U. , Deborah Thorne , Elizabeth Warren , and Steffie Woolhandler . 2009 . “ Medical Bankruptcy in the United States, 2007: Results of a National Study .” The American Journal of Medicine 122 ( 8 ): 741 – 746 . Google Scholar Crossref Search ADS PubMed WorldCat Jenkins , Stephen P. . 2000 . “ Modelling Household Income Dynamics .” Journal of Population Economics 13 : 529 – 567 . Google Scholar Crossref Search ADS WorldCat Kaiser , Mark J. , and Allan G. Pulsipher . 2006 . “ Concerns Over the Allocation Methods Employed in the US Low-Income Home Energy Assistance Program .” Interfaces 36 ( 4 ): 344 – 358 . Google Scholar Crossref Search ADS WorldCat Klinenberg , Eric . 2002 . Heat Wave: A Social Autopsy of Disaster in Chicago . Chicago, IL : University of Chicago Press . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Lazard . 2016 . Lazard’s Levelized Cost of Energy Analysis – Version 10.0 . Accessed November 2017 . https://www.lazard.com/media/438038/levelized-cost-of-energy-v100.pdf. Lo , Alex Y. . 2014 . “ Negative Income Effect on Perception of Long-Term Environmental Risk .” Ecological Economics 107 : 51 – 58 . Google Scholar Crossref Search ADS WorldCat Lutzenhiser , Loren , and Bruce Hackett . 1993 . “ Social Stratification and Environmental Degradation: Understanding Household CO2 Production .” Social Problems 40 ( 1 ): 50 – 73 Google Scholar Crossref Search ADS WorldCat Mani , Anandi , Sendhil Mullainathan , Eldar Shafir , and Jiaying Zhao . 2013 . “ Poverty Impedes Cognitive Function .” Science 341 ( 6149 ): 976 – 980 . Google Scholar Crossref Search ADS PubMed WorldCat Maroto , Michelle Lee . 2015 . “ Pathways to Bankruptcy: Accumulating Disadvantage and the Consequences of Adverse Life Events .” Sociological Inquiry 85 ( 2 ): 183 – 216 . Google Scholar Crossref Search ADS WorldCat McKernan , Signe-Mary , and Ratcliffe , Caroline . 2005 . “ Events that Trigger Poverty Entries and Exits .” Social Science Quarterly 86 : 1146 – 1169 . Google Scholar Crossref Search ADS WorldCat Mendenhall , Ruby , Karen Z. Kramer , and Illana R. Akresh . 2014 . “ Asset Accumulation and Housing Cost Burden: Pathways to (Not) Saving .” Housing Policy Debate 24 ( 2 ): 387 – 414 . Google Scholar Crossref Search ADS WorldCat Mettler , Suzanne . 2011 . The Submerged State: How Invisible Government Policies Undermine American Democracy . Chicago, IL : University of Chicago Press . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Middlemiss , Lucie , and Gillard , Ross . 2015 . “ Fuel Poverty from the Bottom-Up: Characterising Household Energy Vulnerability Through the Lived Experience of the Fuel Poor .” Energy Research and Social Science 6 : 146 – 154 . Google Scholar Crossref Search ADS WorldCat Moore , Richard . 2012 . “ Definitions of Fuel Poverty: Implications for Policy .” Energy Policy 49 : 19 – 26 . Google Scholar Crossref Search ADS WorldCat Moser , Caroline O.N. . 1998 . “ The Asset Vulnerability Framework: Reassessing Urban Poverty Reduction Strategies .” World Development 26 ( 1 ): 1 – 19 . Google Scholar Crossref Search ADS WorldCat Murkowski , Lisa , and Tim Scott . 2014 . “ Plenty at Stake: Indicators of American Energy Insecurity .” Energy 20/20 White Paper 113th United States Congress . Accessed February 2019 . https://www.energy.senate.gov/public/index.cfm/files/serve?File_id&#x003D;075f393e-3789-4ffe-ab76-025976ef4954. Murray , Anthony G. , and Bradford F. Mills . 2014 . “ The Impact of Low-Income Home Energy Assistance Program Participation on Household Energy Insecurity .” Contemporary Economic Policy 32 ( 4 ): 811 – 825 . Google Scholar Crossref Search ADS WorldCat Newman , Sandra J. , and Holupka , C. Scott . 2014 . “ Housing Affordability and Investments in Children .” Journal of Housing Economics 24 : 89 – 100 . Google Scholar Crossref Search ADS WorldCat Philbrick , D. , Scheu , R. , and Brand , L. . 2016 . Quantifying the Financial Benefits of Multifamily Retrofits . Golden, CO : National Renewable Energy Laboratory (NREL) . Contract No. DE-AC36-08GO28308 . Google Preview WorldCat COPAC Pollack , Craig Evan , Beth Ann Griffin , and Julia Lynch . 2010 . “ Housing Affordability and Health Among Homeowners and Renters .” American Journal of Preventive Medicine 39 ( 6 ): 515 – 521 . Google Scholar Crossref Search ADS PubMed WorldCat Prasad , Monica . 2011 . “ Tax ‘Expenditures’ and Welfare States: A Critique .” Journal of Policy History 23 ( 2 ): 251 – 266 . Google Scholar Crossref Search ADS WorldCat ———. 2012 . The Land of Too Much: American Abundance and the Paradox of Poverty . Cambridge, MA : Harvard University Press . WorldCat COPAC Price , Catherine Waddams , Brazier , Karl , and Wang , Wenjia . 2012 . “ Objective and Subjective Measures of Fuel Poverty .” Energy Policy 49 : 33 – 39 . Google Scholar Crossref Search ADS WorldCat Pudney , Stephen . 2008 . “ Institute for Social and Economic Research, University of Essex .” Heaping and Leaping: Survey Response Behaviour and the Dynamics of Self-Reported Consumption Expenditure. No. 2008–09. ISER Working Paper Series . Reames , Tony Gerard . 2016 . “ Targeting Energy Justice: Exploring Spatial, Racial/Ethnic and Socioeconomic Disparities in Urban Residential Heating Energy Efficiency .” Energy Policy 97 : 549 – 558 . Google Scholar Crossref Search ADS WorldCat Roberts , Deborah , Vera-Toscano , Esperanza , and Phimister , Euan . 2015 . “ Fuel Poverty in the UK: Is There a Difference Between Rural and Urban Areas? ” Energy Policy 87 : 216 – 223 . Google Scholar Crossref Search ADS WorldCat Ruggles , Patricia , and Williams , Roberton C. . 1987 . Transitions In and Out of Poverty: New Data from the Survey of Income and Program Participation . Washington, DC : Census Bureau . Google Preview WorldCat COPAC Ruggles , Patricia . 1990 . Drawing the Line: Alternative Poverty Measures and Their Implications for Public Policy . Washington, DC : Urban Institute Press . Google Preview WorldCat COPAC ———. 1992 . “ Measuring Poverty .” Focus 14 ( 1 ): 1 – 9 . WorldCat Scheu , Rachel , Parham Azimi , Mary Ellen Guest , Amanda Gramigna , and Brent Stephens . 2018 . “ Why Equity Matters: Energy Use and Health Disparities by Neighborhood: Stories (and Data) from Families Living in Chicago’s Bungalow Belt .” Proceedings from the ACEEE Summer Study on Energy Efficiency in Buildings . Pacific Grove, CA : ACEEE . Smith , James P. . 2004 . “ Unraveling the SES—Health Connection .” Population and Development Review 30 : 108 – 132 . Google Scholar Crossref Search ADS WorldCat Sovacool , Benjamin K. , Sidortsov , Roman V. , and Jones , Benjamin R. . 2014 . Energy Security, Equality, and Justice . New York : Routledge . Google Preview WorldCat COPAC Stewart , Mark B. , and Joanna K. Swaffield . 1999 . “ Low Pay Dynamics and Transition Probabilities .” Economica 66 ( 261 ): 23 – 42 . Google Scholar Crossref Search ADS WorldCat Tach , Laura M. , and Eads , Alicia . 2015 . “ Trends in the Economic Consequences of Marital and Cohabitation Dissolution in the United States .” Demography 52 : 401 – 432 . Google Scholar Crossref Search ADS PubMed WorldCat Trumbo , Craig , Michelle Lueck , Holly Marlatt , and Lori Peek . 2011 . “ The Effect of Proximity to Hurricanes Katrina and Rita on Subsequent Hurricane Outlook and Optimistic Bias .” Risk Analysis 31 ( 12 ): 1907 – 1918 . Google Scholar Crossref Search ADS PubMed WorldCat U.S. Department of Energy (DOE) . 2015 . Weatherization Assistance Program National Evaluations: Summary of Results . Accessed June 2019 . https://weatherization.ornl.gov/wp-content/uploads/2018/06/WAPNationalEvaluationWxWorksv14blue8515.pdf. U.S. Energy Information Administration (EIA) . 2015a . Residential Energy Consumption Survey, “One in Three U.S. Households Faced Challenges in Paying Energy Bills in 2015 .” Accessed February 2019. https://www.eia.gov/consumption/residential/reports/2015/energybills/. ———. 2015b . “ Residential Energy Consumption Survey, Household Demographics Table HC9.2 .” Accessed February 2019. https://www.eia.gov/consumption/residential/data/2015/. ———. 2018 . “ Electric Power Monthly with Data for May 2018, Table 5.6.A, Average Price of Electricity to Ultimate Customers by End-Use Sector, by State, May 2018 and 2017 (Cents per Kilowatthour) .” Accessed August 2018. https://www.eia.gov/electricity/monthly/current_month/epm.pdf. Verclas , Kirsten , and Eric Hsieh . 2018 . “ From Utility Disconnection to Universal Access .” The Electricity Journal 31 ( 6 ): 1 – 8 . Google Scholar Crossref Search ADS WorldCat Weir , Margaret , Orloff , Ann Shola , and Skocpol , Theda (eds.) 1988 . In The Politics of Social Policy in the United States . Princeton, NJ : Princeton University Press . Google Preview WorldCat COPAC Western , Bruce , Bloome , Deirdre , Sosnaud , Benjamin , and Tach , Laura . 2012 . “ Economic Insecurity and Social Stratification .” Annual Review of Sociology 38 : 341 – 359 . Google Scholar Crossref Search ADS WorldCat Wilson , Jonathan , Jacobs , David , Reddy , Amanda , Tohn , Ellen , Cohen , Jonathan , and Jacobsohn , Ely . 2016 . Home Rx:The Health Benefits of Home Performance – A Review of the Current Evidence . Washington, DC : U.S. Department of Energy . Google Preview WorldCat COPAC © The Author(s) 2019. Published by Oxford University Press on behalf of the University of North Carolina at Chapel Hill. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Do Energy Burdens Contribute to Economic Poverty in the United States? A Panel Analysis
JF - Social Forces
DO - 10.1093/sf/soz131
DA - 2006-01-01
UR - https://www.deepdyve.com/lp/oxford-university-press/do-energy-burdens-contribute-to-economic-poverty-in-the-united-states-LFXZZeBfzu
SP - 1
VL - Advance Article
IS - 
DP - DeepDyve
ER -