Energy poverty and obesity

Highlights

• We examine the effect of energy poverty on obesity.

• We use 14 waves of nationally representative household panel survey for Australian adult population.

• We find that being energy poor is associated with significant increase in the propensity of being obese.

• We find that amount of sleep, health status and psychological distress are important channel of influence.

• Effects differ by gender and age.

Abstract

Obesity is a major health concern in both developed and developing nations. Yet, evidence on the determinants of obesity is relatively limited. We contribute to the literature on the determinants of obesity by empirically examining the effects of energy poverty on obesity. Using 14 waves of the Household, Income and Labour Dynamics in Australia (HILDA) survey, we find energy poverty is positively associated with obesity. Applying linear probability model, our estimates suggest that being energy poor results between 1.4 and 2.5 percentage points increase in the probability of being obese, depending on how energy poverty is measured. Our results are robust to alternative modelling techniques, inclusion of additional control variables and potential influence of unobservable. We also find that amount of sleep, health status and level of psychological distress are important transmission channels through which energy poverty influences the probability of being obese.

Introduction

Obesity is a major public health challenge that affects adults and children of all ages across both genders. Globally, more than 2.8 million people die each year from several diseases associated with obesity (Hussain et al., 2020), including hypertension, stroke, type 2 diabetes, dementia and myocardial infarction (see, e.g., Blüher, 2019; Casey et al., 2008). Obesity causes increased individual and societal economic burden, given its association with low socio-economic productivity, unemployment and associated social disadvantage (Blüher, 2019). In 2014, the global economic burden of obesity was estimated at approximately US $2 trillion (Tremmel et al., 2017) with over 13% of the adult population diagnosed with obesity (WHO, 2020). Given the current obesity trends, it is projected that by 2025, more than 1 billion adults will be obese (World Obesity, 2020). The existing research largely identifies the role of physical activity (Joslyn and Haider-Markel, 2019; Gray et al., 2018), medications (Wright and Aronne, 2012; Faith et al., 2011), amount of sleep (Hanlon et al., 2019; Sa et al., 2020), genetics (Thaker, 2017; Joslyn and Haider-Markel, 2019), stress and depression (Mannan et al., 2016; Miller and Lumeng, 2018) and poor diet (Sola et al., 2016; Njiru and Letema, 2018) as some of the influences of obesity.

While understanding the factors that influence obesity is an important area of inquiry, relatively little is known about how affordability and access to household energy needs influences obesity. The literature defines households as “energy poor” or in energy poverty if they are unable to access or afford energy (González-Eguino, 2015; Nussbaumer et al., 2012). Studies have shown that energy poverty is associated with a number of outcomes such as health (Oliveras et al., 2020b; Thomson et al., 2017; Awaworyi Churchill and Smyth, 2021), education (Oum, 2019; Zhang et al., 2019) and subjective wellbeing (Awaworyi Churchill et al., 2020b; Thomson et al., 2017; Biermann, 2016). Could there be an association between energy poverty and obesity? To the best of our knowledge, there is no study that has examined the effect of energy poverty on obesity. In this paper, we fill an important gap in the literature by examining how energy poverty influences obesity using panel data for Australia. An implicit motivation for such studies is that by understanding the role of energy poverty in influencing body weight, we can improve our understanding on the determinants of obesity and to allow targeted intervention policies to curb the obesity crisis.

We use 14 waves of the Household, Income and Labour Dynamics in Australia (HILDA) survey, covering the period 2006–2019 to examine the effect of energy poverty on obesity. We measure obesity using a binary variable which equals one if an individual has a Body Mass Index (BMI) score exceeding 30. To measure energy poverty, we use: (1) a subjective indicator that captures households' inability to heat their homes, (2) an objective indicator, which considers high energy costs as well as the low-income status of households and (3) a composite measure of energy poverty. Controlling for individual socio-economic and demographic characteristics, state and time fixed effects along with taking care of endogeneity concerns, our results indicate that energy poverty has positive effect on obesity. These findings are robust to a suite of sensitivity checks. Further, we explore the channels through which energy poverty may transmit to obesity. We find that the amount of sleep, general and mental health status are important channels through which energy poverty influences obesity.

We situate our study in Australia for three reasons. Firstly, Australia has a high obesity prevalence rate in all age groups. Approximately, 31% of Australian adults are obese (AIHW, 2020), making Australia the fifth country with the highest prevalence rates of obesity among OECD countries (James et al., 2020). Secondly, due to high obesity rates, Australia incurs huge burden of disease cost. Specifically, recent estimates suggest that overweight and obesity contribute to about 8.4% of the entire disease burden in Australia (AIHW, 2020), with an annual cost of about $56.6 billion (Colagiuri et al., 2010). This obesity related cost is projected to reach $87.7 billion by 2025 (PwC Australia, 2015). Thirdly, Australia's energy costs have risen. For instance, over the last decade, residential gas prices have increased by about 74% in some states (Department of Industry Innovation and Science, 2016), and electricity prices have almost doubled (see, e.g., Valadkhani et al., 2018; Wood and Blowers, 2017). Given that Australia's real wage growth is low (Commonwealth Treasury, 2017), these energy price increases are claiming a large proportion of income for most households (Hogan and Salt, 2017), pushing many of these households into energy poverty (Awaworyi Churchill and Smyth, 2021). Therefore, understanding the effect of energy poverty on obesity using Australian data would not only help Australian policymaking, but also have important policy implications for many developed countries experiencing high obesity rates.

Our study makes important contributions to at least four bodies of literature. First, we make additions to the literature on the drivers of obesity (see, e.g., Joslyn and Haider-Markel, 2019; Sa et al., 2020; Thaker, 2017; Miller and Lumeng, 2018; Njiru and Letema, 2018). We find that besides the well-established factors that include physical activity, medications, amount of sleep, genetics, stress, depression and poor diet, energy poverty is another important determinant of obesity. Second, we contribute to the existing broad strand of literature that examines the relationship between poverty and obesity (see, e.g., Salmasi and Celidoni, 2017; Levasseur, 2019; Wells et al., 2010; Villar and Quintana-Domeque, 2009; Wen et al., 2010). These studies analysing health effects of income poverty find that income poverty contributes to increase in BMI and the probability of being obese. However, studies in this strand of literature, are at best mixed for men and women (Villar and Quintana-Domeque, 2009). We differ from these studies by focusing specifically on the effects of energy poverty on obesity.

Third, we contribute to the general literature on the impact of energy poverty (see, e.g., Oliveras et al., 2020b; Amin et al., 2020; Awaworyi Churchill et al., 2020b). While this strand of literature has explored the impact of energy poverty on outcomes such as health, economic development, and subjective wellbeing, within the context of health outcomes, the focus has largely been on how energy poverty influences mental health issues (see, e.g., Lin and Okyere, 2020; Thomson et al., 2017; Awaworyi Churchill et al., 2020b). However, we add to this literature by focusing on obesity – a physical health disorder. Fourth, we contribute to the literature by exploring ways to curb the obesity crisis (see, e.g., Hu, 2013; Thangaratinam et al., 2012; Swinburn et al., 2015; Barcellos et al., 2018; Kaur and Briggs, 2019; Seyyed Reza and Mina, 2019). These studies suggest ways such as increasing taxes on foods containing added sugars, promoting physical activities, calorie labelling on food products, strengthening accountability systems in food environments and promoting educational interventions. We show that reduction in energy poverty is an effective way of reducing obesity.

The rest of the paper is organized as follows: the next section discusses the channels through which energy poverty may influence obesity. Section 3 explains the data and variables, while Section 4 outlines the methodology adopted in this study. 5 Results, 6 Robustness checks and other sensitivity checks present the results and robustness checks, respectively. Section 7 concludes.