Evaluating the Impact of Energy Poverty in a Multidimensional Setting.

AuthorDelugas, Erica

    Even in wealthy countries, there may be a portion of the population that is unable to purchase a basic set of goods and services based on energy use. According to the Building Performance Institute Europe (Atanasiu et al., 2014), in 2012, about 10.8% of the European population was unable to maintain adequate warmth in their homes or were living in energy poverty (henceforth EP). The size of the problem has been increasing over the last 15 years. People exposed to EP not only usually spend a high share of their income on electricity, oil, and gas; they also live in inefficient and unhealthy dwellings; and are exposed to severe consequences concerning health, social exclusion, and overall household welfare. When looking to EP from a macroeconomic perspective, access to modern, clean and affordable energy facilities is considered key to reduce poverty and foster economic development in lower-income countries. The correlation is less clear for countries with high levels of income (Karekezi et al., 2012). In this case, in spite of potential complete access to energy, some households could experience an unsatisfactory availability of energy for economic reasons or lack of infrastructure facilities. In fact, three main economic mechanisms display a concurrent role in shaping actual energy utilization: technical availability of modern services, their affordability in terms of price, and reliability in terms of being usable for most of the time (IEA, 2017).

    In developed economies, the earliest policies to support vulnerable citizens took place in the United Kingdom in the early 1990s. In more recent times, other European countries have begun to recognize EP as a distinct phenomenon vis-a-vis income poverty and to implement specific supporting programs. Since 2006, the European Union has pushed for spreading policies supporting the energy poor across all European countries. (1) According to the latest projects (e.g. the European Energy Poverty Observatory) run by the European Commission, EP should be officially considered a distinct phenomenon from income poverty that should be separately analyzed. (2) This view embraces similar considerations made in several studies that reported EP as a complex phenomenon, where a concurrent role is played by factors such as high energy prices, the inefficiency of buildings and appliances, the low income of households and their specific energy needs and practices: Thomson et al., 2017) with its peculiarities (such as the higher absolute cost for the poor to have an adequately warm dwelling, the additional detrimental effect that the lack of adequate access to energy has among income poor individuals in terms of illness, mental health and social exclusion: Hills, 2012). Considering EP as a distinct phenomenon from income poverty entails that the identification and measurement of energy poor people should not be (exclusively) based on monetary indicators derived from variables such as energy prices and expenditures. In operative terms, the measurement of EP can be obtained starting from an information set that comprises a few deprivation indicators made available at the individual and/or household level in household surveys.

    Most of the existing literature points to a set of objective welfare EP indicators (e.g. Boardman, 1991; Hills, 2011; Moore, 2012; Legendre and Ricci, 2015). However, the scope for including subjective measures in the economic analysis is nowadays embedded in the economic debate on welfare evaluation, where the use of subjective well-being (henceforth SWB) approaches has become common practice. (3) SWB approaches have been applied to different fields, e.g., health care (Ferrer-i Carbonell and van Praag, 2002), social science (Frey and Stutzer, 2002), evaluation of public goods (Luechinger, 2009) and energy provision mix (Welsch and Biermann, 2014a,b). Accordingly, even in the analysis of EP, subjective indicators have been recently considered in a few studies. This is the case in the recent works by Welsch and Biermann (2017), who investigate the effects on life satisfaction of electricity, oil, and gas prices (standard objective measures) in different European countries; and by Biermann (2016), who finds that fuel poverty measures related to households' expenditure on energy are always associated with a significant negative effect on SWB that adds to that of income poverty. Other studies have adopted an SWB perspective by trying to define a subjective measure of EP (Papada and Kaliampakos, 2016; Rehdanz et al., 2015; Lawson et al., 2015; Waddams Price et al., 2012).

    To the best of our knowledge, what is apparently missing in the extant literature is an analysis of individuals' well-being where the combined information from objective and subjective measures of EP, considered within a multidimensional approach, is exploited to econometrically assess the relationship between EP and SWB. With the aim to widening the set of the methodological tools that can be used in this field of economic analysis, we first show how to subsume a set of available indicators (pointing to both subjective and objective dimensions of households'energy deprivation), in a single multidimensional energy poverty index (henceforth MEPI) that provides information on EP even at the individual level. This is done by adapting to EP analysis (and the data at hand) the methodology that Alkire and Foster (2011) have proposed for standard multidimensional poverty measurement. Considering subjective indicators of EP makes these kinds of indices trivially endogenous in their relationship with SWB. Coupled with its ordinal nature, at least in our application, this endogeneity issue impacts on the detection of an appropriate econometric modeling strategy. We suggest estimating the individual-level relationship between SWB and the MEPI by employing a bivariate ordered probit model with exclusion restrictions. This allows us to account for the correlation between the two variables. Moreover, provided that an opportune set of instruments is available, this solution is adequate to face a general set of endogeneity problems related to unobservable factors. This approach is valid even in a cross-sectional environment and could be potentially applied when using alternative multidimensional indices partially based on subjective measures. (4)

    We develop our MEPI and carry out the empirical analyzes by using the Italian version of the European Union Survey on Income and Living Conditions (henceforth ITSILC). As for the information on SWB, we exploit a question about the degree of life satisfaction included in a specific module on social exclusion, which is asked to be evaluated on an 11-point scale. (5)

    We first provide an exploratory analysis that shows the potential from using the MEPI to identify EP while pointing at the same time to the discrepancies between multidimensional indices and traditional monetary indicators of fuel poverty. Subsequently, we econometrically assess the relationship between SWB and the MEPI by identifying the causal relationship between EP and life satisfaction using exclusion restrictions referred to the year of construction of the dwellings. The results not only confirm theoretical predictions, by detecting a significant negative relationship between subjective well-being and energy poverty intensity, but also point to the capability of multidimensional subjective indicators in explaining the impact of EP on SWB compared with traditional expenditure-related measures.

    The paper is structured as follows. In Section 2, we sketch a background of the relevant literature. Section 3 describes the construction of multidimensional poverty indices and their application to the data at hand. Section 4 illustrates the conceptual model in which the empirical analysis is framed. Section 5 illustrates the results of the econometric analysis, and Section 6 contains a few concluding remarks.


    The two main topics in which our work is framed are the EP measurement methods and the relationship between SWB and EP.

    2.1 Energy Poverty Measurement

    Approaches to the analysis of EP measurement can be broadly categorized as either affordability or energy deprivation. The former is inherently unidimensional, being based on reference monetary thresholds that define the maximum level of income or expenditure share spent on energy (the term fuel is often used) that can be considered affordable by individuals or households. Boardman (1991) provides a starting point for this approach by simply stating that EP occurs when any household needs to spend more than 10% of its disposable income on total fuel use (the so-called 10%Rule). Variations of this elementary approach are the so-called 2M indicators, double mean, or double median, which count as energy poor those individuals whose energy expenditure share is greater than the double of the mean (or median). More recent studies, (e.g. Hills, 2011, 2012) propose a Low-Income High Costs composite indicator, which counts individuals as energy poor if they spend more than 60% of the median of the disposable income distribution and they fall below a given income poverty line. Finally, affordability has been seen within a Minimum Income Standard framework that considers as energy poor those individuals lacking a minimum income required to satisfy primary needs after paying housing costs and energy costs (Moore, 2012). Close to Moore's indicator is the Residual Income Indicator Miniaci et al. (2014), which is aimed at understanding how many (not energy-related) goods an individual can purchase apart from energy.

    By taking a different perspective, the energy deprivation approach points to the importance of considering the different dimensions of EP, thereby paralleling the debate that characterizes the comparison between multidimensional approaches to poverty measurement and...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT