Introduction

The concept of energy poverty and its gender implications vary depending on the context in which the phenomenon is analysed. In low- and middle-income countries, where energy poverty has been widely explored, barriers to energy access are often linked to poor infrastructure and low income levels. This leaves households dependent on wood and other forms of biomass, which triggers serious health problems, especially for women. Conversely, in high-income countries, energy poverty is often associated with affordability challenges, with the assumption that difficulties in accessing energy are gender-neutral1,2.

Although there have been numerous definitions of energy poverty, they all refer to an insufficient energy consumption level to meet basic energy needs3,4. In high-income countries, the first studies about energy poverty were developed in the 80s in the UK5 using the term “fuel poverty” which later evolved to the concept of energy poverty. Following publication of Brenda Boardman’s seminal book “Fuel Poverty”6, there have been numerous attempts to develop a proper definition of energy poverty and a good methodology to measure it. In this sense, the most relevant contributions have been from the UK7,8,9,10,11,12,13,14, but these attempts have also spread to other European Union (EU) countries4,15,16,17,18,19,20,21,22. However, this historical evolution of the concept has largely failed to integrate a gender perspective.

Indeed, despite the extensive literature available on this issue, little (if any) attention has been paid to gender. In fact, few studies have pointed to the gender-energy-poverty nexus in high-income countries23,24,25,26,27. Some of the shortcomings identified by these studies that properly integrate the gender dimension on energy poverty are the scarcity of gender disaggregated data and the general lack of awareness amongst politicians, advisors, and researchers23,28. A small but insightful body of qualitative research, including case studies across different countries, has been pivotal in highlighting that gender is a critical factor in experiencing energy poverty25,29,30,31. Their findings consistently show that women are often more vulnerable, yet this evidence has not been followed by systematic quantitative measurement.

This paper addresses these gaps by advancing the energy poverty literature in three significant ways. First, we introduce the Energy Poverty Gender Gap (EPGG), a novel metric to quantify gender disparities in energy poverty—a conceptual and methodological contribution absent in prior work. Second, building on the foundational work of qualitative studies, to the best of our knowledge, we present the first systematic measurement and comparison of these disparities across all EU Member States using harmonized microdata. Third, our analysis provides the first EU-wide comparison of energy poverty that explicitly integrates gender, revealing policy-relevant variations invisible to single-country or non-gender-disaggregated studies.

To achieve this, we define and calculate the Energy Poverty Gender Gap (EPGG) for the EU using 2019 Household Budget Survey (HBS) microdata (a year deliberately selected to capture pre-pandemic and pre-energy crisis conditions). This temporal focus enables us to establish a crucial baseline measurement of gender disparities in energy poverty before these major disruptions, while providing policymakers with reference values to assess how subsequent crises may have altered these gaps. Additionally, a linear regression analysis is conducted to identify the factors contributing to higher or lower EPGGs. Furthermore, the Survey on Income and Living Conditions (SILC) is used to calculate the gender health gap for individuals living in energy-poor households and determine if the outcomes of energy poverty are also gender-sensitive.

Analysing the gender dimension of energy poverty is particularly relevant to the design of effective energy poverty alleviation and just energy transition policies that avoid exacerbating existing inequalities. So, the results of the study are especially pertinent in the context of the European Parliament’s 2016 resolution on energy access (B8-1227/2016 – Resolution on access to energy in developing countries), which calls for gender mainstreaming in all EU energy policies, and subsequent policy developments. Despite these ongoing policy calls for the inclusion of a gender dimension in EU energy policies, a need further highlighted during the recent energy crisis, a 2023 European Commission report confirms that research and debate on this issue remain severely limited, explicitly identifying it as the “missing debate”32. Our findings provide the critical quantitative evidence needed to advance this debate and inform policymakers.

Results

The Energy Poverty Gender Gap: empirical evidence across EU27 households

Here, we explore the EPGG across EU Member States according to the Low Income High Cost (LIHC) metric for 2019 (Fig. 1). The EPGG reflects the difference in energy poverty levels between households led by women and those led by men. Following the WHO definition33, in this study, we understand gender as a social and analytical category (intersectional in nature) through which behaviours and emotional, affective and intellectual characteristics, that are erroneously assigned as proper and natural to men and women, are defined28. Before analysing the gender dimension, a comparative analysis of energy poverty in the EU27 has been carried out, using various metrics commonly analysed in the literature. This analysis is available in Section 1 of Supplementary Information. Here, the LIHC metric is used to assess the EPGG because it has been widely used in the literature, the data needed for its calculation are available for all the countries of the EU and it identifies low and middle-income households as energy poor (see Supplementary Information, Section 2). The EPGG results for the remaining metrics are available in Supplementary Information, Section 3.

Fig. 1
Fig. 1
Full size image

Energy poverty gender gap in the EU in 2019 (LIHC).

On one hand, our findings indicate that women (throughout the Results section, “women” refers specifically to women-headed households) in Estonia (8.5%) and Latvia (4.93%) bear a disproportionate burden of energy poverty. They present the highest EPGG, closely followed by Germany (4.08%) and Czechia (3.85%). Estonia, Latvia and Czechia share several similarities that contribute to the high gender inequality levels in terms of energy poverty. They are Central and Eastern European countries that, although they have experienced notable economic and social transitions in the past decades, still face several challenges related to income inequality, employment opportunities, and social welfare provision34,35,36.

Additionally, factors such as energy prices relative to income levels and household composition may contribute to the observed gender gap. Indeed, these three countries exhibit the most notable disparity in terms of the proportion of income allocated to household energy expenditures. Specifically, in Estonia, Latvia and Czechia, women spend an average of 2.79, 1.99, and 1.41 percentage points more of their income on domestic energy compared to men in their respective countries, whereas the average gender gap of the share of domestic energy in the EU is only 0.61 points (detailed data are provided in Supplementary Information, Section 4). This is also closely related to the average income gender gap, which is something they share with Germany.

One explanation for the elevated EPGG in Germany lies in income disparities. In our analysis, total household expenditure is used, since it is considered a better proxy for the permanent income of families as it undergoes a lower fluctuation than income both in the medium and long term37. Although Germany’s energy poverty level (6.74%) closely mirrors the EU average (6.65%), it stands as the third country with the widest gender gap. Germany, alongside Estonia and Latvia, exhibits the highest differences in average income between women and men: 6.56%, 8.89%, and 4.96%, respectively. This income inequality among countries with the highest EPGGs is further evidenced by pay gap data published by EUROSTAT for 2019. In fact, all four countries feature among the top five European countries with the highest gender pay gaps: Estonia (21.7%), Latvia (21.2%), Czechia (19.2%), and Germany (19.2%).

Top of form

On the other hand, among the countries with the lowest EPGG, the cases of Denmark (−0.76%) and Sweden (−0.74%) are particularly remarkable because they are the only EU Member States where women do not bear the brunt of energy poverty. In both cases, their established social welfare systems and their policies to reduce gender inequality could be contributing to their unique situation regarding EPGG. Moreover, energy prices relative to income are notably lower in these countries compared to other European Member States. Energy prices could be influenced by investments in renewables and energy efficiency and income by their relatively high levels of gender equality and women’s participation in the workforce. Notably, Sweden (alongside Finland) stand out as the only European Member State where men allocate a higher proportion (0.61 points higher) of their income to energy needs. In Denmark, men and women spend roughly the same percentage of income on energy needs (women expend just 0.38% more).

In Finland (0.39%), Poland (0.63%) and Italy (0.75%), women are slightly more affected than men, but they also share a good position regarding their EPPGs, well below the European average (2.04%). While the factors influencing Finland’s situation are closely aligned with those of the aforementioned cases, the low gaps in Italy and particularly Poland are more noteworthy. In Italy, the gap in the income allocated to household energy expenditure between men and women stands at 0.44, while in Poland, it reaches 0.82. These countries also rank among those with the smallest average income gaps and are among the nations with the lowest gender pay gaps in the EU: 4.70% in Italy and 6.50% in Poland, compared to the EU average of 12.69%.

To extend our analysis beyond the gender of the household reference person, we also analyse the feminization degree of the households. The feminization degree of the household represents the percentage of household members who are women over 14 years old38 and aims to capture intra-household consumption behaviours. For more methodological details, see the methods section. In the Supplementary Information, Supplementary Table SI5 in Supplementary Information collects the energy poverty levels according to the LIHC metric for each country alongside the feminization degree. Notably, our analysis reveals that in all EU countries, except Denmark, Poland, and Sweden, energy poverty levels are higher in the most feminized households (FD5), which is in line with our previous findings.

Measuring intersectionality: a multidimensional analysis of the EPGG

Intersectionality is a key factor when developing gender analysis. In its first conception, intersectionality referred to “the phenomenon by which each individual suffers oppression or holds privilege based on their belonging to multiple social categories”28. In practice, however, it has become an analytical tool that allows us to understand complex realities by analysing the articulation of different socioeconomic categories (such as, class, gender, race…) rather than considering them as independent forms of power relations. In fact, some studies have already shown that energy poverty is intersectional in nature26, where gender interacts with other axes of social difference, such as age or socioeconomic status, and others have highlighted the lack (and need) for studies related to energy poverty that integrate this approach39,40. Thus, in this section, we analyse the gender dimension of energy poverty according to other relevant variables commonly analysed in gender studies for 2019. Figure 2 shows the EPGG based on the LIHC metric in four of the most affected groups by energy poverty in the EU (detailed data are provided in Supplementary Information, Section 6).

Fig. 2
Fig. 2
Full size image

Energy poverty gender gap (LIHC) in the EU in 2019 for specific sociodemographic characteristics.

When examining energy poverty across income levels in the EU27, it becomes evident that Q1 (income quintile, representing the poorest 20%) households are the most affected across all countries. In nearly all countries, there exist important EPGGs within this group (Supplementary Table SI6), with exceptions observed in Finland, Denmark and Sweden, where households led by men are more affected. The largest EPGGs are observed in Latvia (9.7%), Luxembourg (8.81%), Slovakia (8.75%), and Ireland (8.22%). Furthermore, Q2 households are also notably affected, with this gender gap persisting across most countries, except in Cyprus (−0.32%), Denmark (−1.67%) and Luxembourg (−1.99%).

Another group significantly impacted by energy poverty across most European countries is rural households. Given their distinct energy requirements and their heightened reliance on off-grid fuels41,42,43,44,45,46,47 such as heating oil, analysing energy vulnerability in this group is crucial within the context of energy transition. Similar to findings in transport poverty48, we find the higher the level of rurality of households, the higher their energy vulnerability. This trend is consistent across all EU countries except for Estonia, Latvia, Lithuania, and Germany. In the former three, households located in intermediate areas are the most affected, while in Germany, urban households bear the brunt. When examining rural households from a gender perspective (Supplementary Table SI7), we find higher EPGGs in Greece (9.91%) and certain Central and Eastern European countries such as Estonia (8.61%), Czechia (7.96%), Bulgaria (6.54%), and Latvia (6.26%). Conversely, we observe negative gender gaps in Sweden (−0.52%), Denmark (−1.06%) and Luxembourg (−2.37%), indicating that men are slightly more affected than women.

When considering family types, single parents, along with adults living alone, emerge as one of the most impacted groups by energy poverty. It is notable that across all EU countries the majority of single-parent households are led by women, ranging from 74.12% in Sweden to 96.30% in Czechia. When examining the EPGG within single-parent households (Supplementary Table SI8), notable gaps are observed in Luxembourg (13.43%), Croatia (10.27%), Czechia (9.73%), and Estonia (8.54%). While it may be surprising to find substantial negative gender gaps in some countries such as Latvia (−20.21%) and Lithuania (−13.25%), it’s important to note that these indices are based on small number of observations in each country, highlighting a key barrier to in-depth intersectional analyses.

On the other hand, the category of adults living alone is dominated by elderly people, especially women living alone, so we delve deeper into the EPGG by analysing another of the most affected groups: households whose reference person is elderly (Supplementary Table SI10). In this case, the EPGG is higher in Luxembourg (7.8%), Estonia (7.72%), Ireland (7.69%) and Latvia (7%), while the gaps are negative and/or near 0 in Bulgaria (0.62%), Belgium (0.43%), Finland (−0.16%), Sweden (−0.8%) and Denmark (−1.24%).

Drivers of the EPGG: a regression analysis of socioeconomic determinants

Table 1 presents the results of the linear regression model built to analyse the potential drivers behind the EPGG. The analysis reveals several key findings. First, the EPGG tends to be higher in countries where overall energy poverty levels are elevated, as well as in countries with a notable general poverty gender gap. Additionally, the gender pay gap contributes to this issue, with the EPGG being more pronounced in countries where the pay gap is wider.

Table 1 Summary of the regression model
Full size table

Conversely, countries with higher housing cost overburden gaps tend to have a lower EPGG. A similar trend is observed with greater government spending on social protection policies. Both variables are directly related to countries with higher incomes, where EPPG is lower. Finally, in countries with a higher proportion of rural households and, the EPGG is lower.

Gender differences in energy poverty outcomes

Energy poverty has numerous negative impacts on the people experiencing it, with health (both physical and mental) being one of the most studied outcomes in the literature49,50,51,52,53. In fact, research by Thomson et al. demonstrated that in European countries, energy-poor individuals experience a significantly higher incidence of poor health. Therefore, in this section, we analyse the gender differences in health status among individuals identifying as energy poor according to the subjective indicators of the SILC.

Figure 3 shows the gender health gap (GHG) in households reporting difficulties in maintaining an adequate temperature at home or with arrears on utility bills (detailed data are provided in Supplementary Information, Section 7). The results reveal a gender gap in health status among energy poor individuals in 2019, with women being disproportionately affected in most EU countries. Indeed, among those living in households struggling to keep the house warm (Supplementary Table SI12), the GHG is significantly high in Portugal (11.43%), Hungary (10.6%), Lithuania (10.39%) and Czechia (9.16%). However, there are also some countries where men’s health is significantly more affected, such as in Austria (−6.86%), Germany (−2.3%) and Slovenia (−1.77%).

Fig. 3
Fig. 3
Full size image

Gender health gap in energy-poor households in 2019.

Similar trends are observed for people living in households with arrears on utility bills (Supplementary Table SI13). In most countries, a higher percentage of women report poor health, with the largest gaps in Denmark (16.45%), Estonia (12.73%), Portugal (11.9%), Sweden (10.3%) and Lithuania (10.09%). However, there are countries where more men report poor or very poor health, including Austria (−10.07%), Slovakia (−4.98%), Malta (−4.47%) and Germany (−4.46%). These findings highlight the noteworthy gender disparities in health outcomes among the energy-poor individuals across EU countries. However, these results may be influenced by gender-specific financial behaviours: women-headed households tend to prioritize paying utility bills on time, often at the expense of cutting essential expenditures like food or mobility, which could exacerbate reported health impacts54,55,56. While our current analysis cannot fully account for this behavioural dimension, it represents a valuable base for future research combining expenditure patterns with health outcomes.

Discussion

Energy poverty manifests differently for women and men. Its drivers and outcomes are different depending on various sociodemographic and physiological characteristics, so it is important to analyse the gender dimension of energy poverty from an intersectional approach. However, current policies to end energy poverty remain uniform. While the literature emphasizes the importance of identifying characteristics of the most affected households to address the issue57 and the EU has underscored the necessity of mainstreaming gender into all policies, insufficient attention has been given to gender and its intersections in this regard. Sometimes under the argument that “gender is not statistically significant”, researchers do not delve into the gender dimension of certain issues, but dismissing gender as statistically insignificant overlooks i) the structural inequalities that underpin energy poverty and other socioeconomic challenges and ii) the interplay between gender, socioeconomic status, cultural norms and institutional structures. In other words, we could be oversimplifying complex social issues and perpetuating existing inequalities. So, going beyond this simplistic notion, we can uncover the underlying mechanisms driving energy poverty and design more just energy policies.

While statistical analyses might not always show a clear gender bias, the in-depth analysis developed in this paper shows that women are disproportionately affected by energy poverty due to various intersecting factors. In fact, this paper confirms that there is an energy poverty gender gap (EPGG) in the EU. And since the effects of energy poverty are influenced by numerous economic, physiological and sociocultural factors it is necessary to go beyond the traditional quantitative metrics used in the literature to understand the scope of the phenomenon and its gender implications.

There are strong links between women’s vulnerability to energy poverty, economic inequality and labour conditions. As we have seen before, the average income gender gap and the gender pay gap are closely related to EPGGs. So addressing these issues could help to lower the EPGG of a country. Also, the unemployment gender gap makes the EPGG larger, and although it is not a significant variable in our model, the literature shows that labour market conditions are a source of gender inequality. According to Robinson (2019), women are more affected by energy poverty due to lower employment rates, part-time employment, provision of unpaid care and precarious jobs. These issues are rooted in the gender division of labour, which persists largely unaddressed even in the EU. In fact, approximately 81% of women are engaged in unpaid domestic work on a daily basis, a stark contrast to the 48% of men involved in similar tasks58. Moreover, in 2019, 31% of women held part-time jobs, whereas the rate decreased to 8% for men. Additionally, in line with our results, Clancy et al. point to the greater vulnerability of single-person households (where women are overrepresented), especially elderly women who live alone and have limited pensions to pay for their basic needs. Nevertheless, since the drivers of the EPGG could be different in each country specific analysis should be done in further research.

Moreover, people tend to perceive domestic energy costs (and other essential services, such as transport) differently depending on their economic situation. In the EU there are notable differences among countries in the share of men and women that perceive energy costs as a “very serious problem” (42% vs 46% on average), but in most countries women agree more with this affirmation than men59. In line with our results, the biggest gender difference is observed in Germany (32% vs 42%), which is among the countries with the highest EPGG. Additionally, there are only 2 countries where men are more worried about energy prices: Sweden and Finland, where EPGG is near 0, and in the case of Sweden men are slightly more affected than women.

On the other hand, the regression results show lower EPGG levels in countries with both higher housing cost burden gaps and greater social government spending. While the first finding might seem surprising, research indicates that women-headed household face the heaviest housing costs burdens, especially single-mother households common in higher-income countries60, which helps explain their lower EPGG in these countries. The second finding highlights how social policies don’t just reduce energy poverty overall, but specifically shrink gender gaps. Since women generally face greater economic insecurity, they benefit more from social programs like healthcare, childcare and pensions61,62. These supports boost their disposable income and reduce energy affordability problems.

The outcomes of energy poverty are also gender-sensitive. In this sense, the health dimension of energy poverty has been widely explored in the literature. Integrating the health dimension in this research area is crucial since research has shown that initiatives aimed at alleviating energy poverty have the potential to mitigate gender disparities in health, reducing women’s mortality rates to a greater extent than men’s63. Most of the literature analysing the health impacts of energy poverty come from low-and-middle-income countries where women are more exposed to indoor pollution due to cooking with solid fuels. But in high-income countries, the health impacts are also gendered. Exploring the health dimension of energy poverty, we observe that among energy-poor individuals, women are overrepresented among those reporting poor or very poor health. This finding aligns with the literature, confirming that women’s health is more adversely affected by energy poverty.

According to Stojilovska64 in the EU women are also overrepresented among vulnerable communities using solid fuels (such as wood) to cope with energy poverty. Women spend more time at home due to their caretaker role, being more exposed to cold homes65, which exacerbates conditions such as rheumatism66. In this sense, even if the time of exposure to extreme temperatures is the same, women are more sensitive to them due to biological reasons. For example, in summer, women, and especially elderly women, have a higher risk of dying in heatwave episodes due to differences in biological processes and capacity to reduce body temperature67,68. In the case of elderly people the problem is exacerbated by their reduced body-temperature regulating function68,69.

Beyond its effects on physical health, energy poverty can also impact individuals’ mental health. Thus, Romanello et al. suggest that the inability to keep home sufficiently warm or cold (depending on season) could lead to anxiety or depression. Additionally, energy poverty may affect mood, exacerbate behavioural disorders and increase suicide risk. There is also a gender dimension to this issue, as studies investigating cases of social exclusion linked to energy poverty more frequently depict women as being disproportionately affected70,71. Moreover, mothers with dependent newborns are also more vulnerable since they find it essential to protect their children, who are especially susceptible to ambient temperature variations72. Hence, within the EU, women are not only disproportionately affected by energy poverty (as the results of this study show) but also bear a greater burden of its health consequences. Additionally, the literature shows that individuals who spend more time at home are more frequently and extensively exposed to insufficient energy services73,74. And it is mothers (often responsible for childcare) and elderly people living alone (most of whom are women) that tend to spend more time at home25.

In addition, the adverse health outcomes analysed in the results may also be linked to broader psychosocial stressors. In fact, the stress generated by energy poverty can also contribute to the generation of (or exacerbate) an environment of domestic violence (both psychological and physical). Although this topic remains underexplored in the academic literature, some social organizations have observed how power dynamics determine access to and control of energy within households. Men may exercise control and manipulate women by threatening to deny access to energy services, such as heating or cooling, thus endangering women’s physical and psychological well-being and creating an atmosphere of fear and dependency75.

While the drivers and impacts of energy poverty have an important gender component, strategies to address it are also gender-sensitive. Our finding of a consistent EPGG across the EU suggests that a one-size-fits-all policy approach may be ineffective, as it fails to account for these differences in how men and women experience and cope with energy poverty. Within households, there is a gender divide in energy-related activities: women are more likely to engage in energy-saving behaviours (for example, by adapting the heating hours to the times when the whole family is at home), whereas men are more inclined to undertake larger-scale energy efficiency actions25. This association of certain activities as feminine or masculine is based on the sexual division of labour and has managed to remain relatively stable over time76. The pressure to conform to stereotypes of being a “good wife” or a “good mother” makes the emotional burden greater for women in energy poverty25,77. Additionally, as demonstrated in behavioural economics, there are also gender differences in decision-making. Women tend to be more risk-averse, particularly when making decisions with or on behalf of others, due to their concerns about inequality aversion78. Consequently, even when they can afford energy efficiency improvements, they require a high level of financial knowledge, digital literacy skills and internet access (another areas where a gender gap exists) to apply for subsidies or tax reductions79. So, to design just and inclusive policies to end or at least mitigate energy poverty it is necessary to mainstream gender into all stages of policymaking.

Conclusions

Despite numerous calls at the European level to integrate the gender perspective in all legislative proposals, and as recommended in the Gender Equality Strategy 2020–202580, in the EU, gender mainstreaming and intersectionality are absent from most analyses on energy poverty75. Since the majority of energy poverty analyses have been gender-neutral, this has perpetuated the invisibilisation of women’s realities. This paper addresses this gap by presenting the first comprehensive comparative analysis of the gender dimension of energy poverty in the EU27.

The study shows that an energy poverty gender gap exists in almost all EU countries, with households led by women and more feminized households being the most affected. Additionally, intersectional analysis reveals that within groups most affected by energy poverty (such as low-income households, single-parent families or single elderly households) women are overrepresented and significantly more affected. So, we can conclude that women have a significantly greater risk of energy poverty, further exacerbated when intersecting with other forms of oppression, discrimination or vulnerability. Nevertheless, since the effects of energy poverty are influenced by numerous economic, physiological and sociocultural factors it is necessary to go beyond the traditional quantitative metrics used in the literature to understand its gender implications.

There are strong links between women’s vulnerability to energy poverty, economic inequality and labour. Our results confirm that income inequality and the gender pay gap contribute to higher EPGGs. The unemployment gender gap also makes the EPGG larger, and although it is not a significant variable in our model, the literature shows that labour market conditions are a source of gender inequality. According to Robinson, women are more affected by energy poverty due to lower employment rates, part-time employment, provision of unpaid care and precarious jobs. These issues are rooted in the gender division of labour, which persists largely unaddressed even in the EU. In fact, ~81% of women are engaged in unpaid domestic work on a daily basis, a stark contrast to the 48% of men involved in similar tasks58. Moreover, in 2019, 31% of women held part-time jobs, whereas the rate decreased to 8% for men. Additionally, in line with our results, Clancy et al. point to the greater vulnerability of single-person households, especially elderly women who live alone and have limited pensions to pay for their basic needs. Nevertheless, since the drivers of the EPGG vary by country, specific analyses are needed for effective policy-making.

The outcomes of energy poverty are also gender-sensitive. Exploring its health dimension, we observe that among energy-poor individuals, women are overrepresented among those reporting poor or very poor health. This finding aligns with extensive literature confirming that women’s health is more adversely affected by energy poverty. Women spend more time at home due to their caretaker role, being more exposed to cold homes65, which exacerbates conditions such as rheumatism66. Biological differences also make women more susceptible to extreme temperatures.

So, since the drivers, the outcomes and the strategies to address energy poverty are gender-sensitive, simplifying the diagnosis or neglecting to incorporate gender mainstreaming at all stages of policymaking risks exacerbating existing gender inequalities. For instance, policies aimed at mitigating energy poverty through housing rehabilitation or energy efficiency programs may unintentionally overlook the fact that women are overrepresented among tenants and that they are reliant on their landlords’ decisions to improve their homes energy systems. Hence, we recommend the design of targeted policy interventions, such as establishing specific support schemes and legal protections to facilitate tenants access to energy efficiency upgrades; developing energy advice programs tailored to women’s specific coping strategies and decision-making roles within households; and introducing mandatory gender-sensitive impact assessments for all new energy poverty policies—rather than including them merely as recommendations, as has happened now… Our analysis provides useful information to improve decision making in this area, but further research is needed since classic energy poverty indicators often overlook the structural factors and complexity of gender inequality, as the data they rely upon are themselves gender-biased. In fact, in order to improve gender-sensitive analysis, more gender disaggregated data and in-depth and context-specific analysis are needed.

Methods

Data

Different databases are used for each of the 3 main sections of the paper. First, to calculate energy poverty metrics and the energy poverty gender gap, we use microdata from the Household Budget Survey (HBS) for 2015 across all European Union countries (EU27), with the exception of Austria and Denmark, for which 2020 data is employed. So, the final analytical sample comprises 279,211 households at the EU level. The HBS provides annual data on all the expenses of the households across different types of goods and services defined by the COICOP classification (Classification of Individual Consumption According to Purpose), and extensive information on the socioeconomic and demographic characteristics of the households. Moreover, this survey is conducted in a diverse array of countries globally81, enhancing the utility of our approach due to its potential for replication across different contexts.

To build the database that is used to calculate the energy poverty indices, initially, raw data from the households and member files of the HBS are matched to construct a unified database. Some adjustments are made to standardize and create new relevant variables, such as household quintiles and deciles and feminization degree. Subsequently, the expenditure data is adjusted to 2019 levels using consumer price indices at the highest level of disaggregation within the COICOP classification. 2015 data is updated to 2019 to avoid pandemic- and energy crisis-related distortions (present in 2020 data), as energy expenditure patterns show strong short-term stability, a finding consistent with both our validation against national HBS series and literature on energy demand inelasticity82.

Second, to understand the drivers behind gender differences in energy poverty levels, an extensive range of 2019 data from EUROSTAT and other sources related to energy and gender aspects is collected, ensuring temporal consistency across all indicators. The selection of these specific indicators was driven by their availability as secondary data (or their ease of calculation from such data) for all EU countries, and their direct relevance to key energy poverty drivers with a gender dimension. Hence, we selected indicators related to gender gaps in income, employment, and other social factors, along with indicators directly correlated with energy poverty, such as climate, energy, market, policy, and socioeconomic indicators. While some aggregated indicators used in the regression analysis are available directly from EUROSTAT, some required reprocessing (for example, calculating gender gaps or mean values) to align with our research objectives. Table 2 provides a summary of the indicators considered, their relevance to energy poverty drivers and their respective sources, including whether they incorporate original calculations or derive from adjusted public sources.

Table 2 Complementary data on gender related energy poverty drivers (2019).
Full size table

Third, we incorporate data from the 2019 Survey on Income and Living Conditions (SILC), which aims to collect up-to-date and standardized information across different time periods and population groups, focusing on income, poverty, social exclusion and living standards. In this case, subjective energy poverty indicators related to the percentage of the population reporting difficulties in paying utility bills on time or struggling to maintain their home adequately warm (which are considered subjective energy poverty metrics) and microdata on reported health are mainly used to identify and analyse significant gender disparities in energy poverty outcomes.

The HBS and SILC datasets, although subject to the common limitations of survey-based research, are useful for developing comparable analyses of fuel poverty across the EU, and have therefore been the main sources for these studies in recent years. Although primary data collection could facilitate more in-depth analyses, these harmonized datasets remain indispensable for generating policy-relevant information, as they are the only sources that allow longitudinal monitoring of fuel poverty trends, and benchmarking across Member States. However, this trade-off between increased granularity and comparability does not diminish the unique value of the datasets for generating knowledge that is relevant for European energy policy.

Energy poverty measurement

As mentioned before, in high-income countries, energy poverty has primarily been associated with affordability issues. Thus, the development of indicators to measure energy poverty has predominantly been focused on assessing this dimension. Since the first indicator was developed in the UK6, extensive literature has emerged to refine and enhance the ways to measure energy poverty12,13,83,84. In this paper, we employ some of the most widely used energy poverty indicators to analyse the gender dimension of energy poverty. Specifically, we calculate the following energy poverty indices using microdata from the HBS: 2 M (Twice Median), LIHC (Low Income High Cost), HEP (Hidden Energy Poverty) and HEP_LI (see Supplementary Information 2). Table 3 outlines the indices considered in this study and provides the formulas used for their calculation.

Table 3 Energy poverty metrics.
Full size table

Energy poverty gender gap

To evaluate gender disparities in energy poverty we calculate the energy poverty gender gap (EPGG). The EPGG reflects the difference in energy poverty levels between households led by women and those led by men, where the household lead or the household reference person is identified as the member contributing the highest income. For countries where this variable was not well recorded in the HBS member file, we determined it by combining two variables: selecting the member with the highest individual income and cross-referencing it with gender information from the same file. This indicator can be calculated based in the different energy poverty measures presented before (see the equations below), but in this paper we focus mainly on the analysis of the EPGG according to the LIHC metric, as it has been widely used in the literature and identifies middle- and low-income households as poor (see Supplementary Information, Section 2).

$${EPGG}={IW}-{IM}$$
(1)

where:

$${IW}=\frac{{{ep}}_{w}}{{{pop}}_{w}}x100$$
(2)
$${IM}=\frac{{{ep}}_{m}}{{{pop}}_{m}}x100$$
(3)

Where EPGG represents the energy poverty gender gap, IW and IM represents the energy poverty indices for households led by women and led by men respectively, epw represents the total number of households led by women identified as energy poor according to the selected energy poverty metric and epm represents the total number of households led by men identified as energy poor, popw represents the total number of households led by women in the country and popm represents the total households led by men.

Moreover, to extend our analysis beyond the gender of the household reference person, we also calculate the feminization degree of the households. This approach, previously used in studies on household carbon footprints85 and the distributional impacts of oil price shocks38, aims to capture intra-household consumption behaviours. The feminization degree of the household represents the percentage of household members who are women over 14 years old38. So, households are divided into 5 groups based on their feminization degree: FD1 (0–20%), FD2 (20–40%), FD3 (40–60%), FD4 (60–80%) and FD5 (80–100%). Thus, FD1 households would be the most masculinized, while FD5 households would be the most feminized.

Regression analysis

To understand which are the drivers of the EPGG, we build a linear regression model based on the indicators listed in Table 2. However, during the construction of the model we realised that the drivers of energy poverty are not the same as those behind its gender gap, so we eliminate variables that do not add value to the model. After analysing the correlations and ensuring that there are no multicollinearity problems, we define the following model:

$${{EPGG}}_{c}={\beta }_{1}{I}_{c}+{\beta }_{2}{{WG}}_{c}+{\beta }_{3}{{UG}}_{c}+{\beta }_{4}{{PG}}_{c}+{\beta }_{5}{{HG}}_{c}+{\beta }_{6}{R}_{c}+{\beta }_{7}{{SP}}_{c}$$
(4)

Where EPGG represents the energy poverty gender gap in country c, I represent the energy poverty levels, WG represents the gender pay gap, UG represents the unemployment gender gap, PG represents the poverty gender gap, HG represents the housing cost overburden rate gender gap, R represents the share of households in rural areas and SP represents the expenditure on social protection.

To validate our model, a comparison is made between the full regression model presented above and a reduced model focusing only on the gender pay gap (WG), which initially appeared to be the best predictor of the EPGG. Using the Akaike (AIC) and Bayesian (BIC) information criteria, the results show that the full model demonstrates superior fit despite its higher complexity (AIC = 89.5 vs. 99.9; BIC = 99.9 vs. 103.9). This comparison confirms that the inclusion of additional socio-economic variables provides significant explanatory power that justifies the use of the extended model.

Gender health gap

To analyse whether the outcomes of energy poverty exhibit gender differences, we explore one of the most analysed dimensions in the literature: health. To do so, we calculate the gender health gap (GHG) by identifying individuals experiencing energy poverty (using the subjective metrics from the SILC mentioned before) and measuring the difference between the percentage of women and men reporting poor or very poor health (see the equation below). Self-reported quality of life measurements have been in development for over 50 years86, with numerous studies finding that self-reported health has strong predictive value for mortality risks87,88.

$${GHG}={HW}-{HM}$$
(5)

where:

$${HW}=\frac{{{bh}}_{w}}{{{ep}}_{w}}x100$$
(6)
$${IM}=\frac{{{bh}}_{m}}{{{ep}}_{m}}x100$$
(7)

Where GHG represents the gender health gap, HW and HM represents the percentage of energy poor women and men with bad or very bad health, bhw represents the total number of women reporting bad or very bad health, bhm represents the total number of men reporting bad or very bad health, epw represents the total number of women identified as energy poor according to the selected energy poverty indicator and epm represents the total number of men identified as energy poor.

Ethics & inclusion statement

This study was conducted with a commitment to providing EU policymakers with quantitative evidence that is both comparable across MS and relevant for national policy interventions. Our analysis relies exclusively on harmonized, anonymized microdata from official EU statistical sources (Household Budget Survey and EU-SILC), which are designed specifically for cross-country comparative research.

The research design ensures local relevance through its EU-wide comparative framework, which allows for the identification of both common patterns and country-specific variations in the energy poverty gender gap. By analysing all MS, our methodology inherently values and incorporates the diverse socioeconomic contexts across European regions.

All data used in this study are publicly available for research purposes through Eurostat’s secure access protocols, ensuring transparency and replicability. As the study involved secondary analysis of fully anonymized EU-level data, it was exempt from additional ethical approval requirements. The research poses no risks of stigmatization or discrimination to individuals, as no personally identifiable information is accessible or analysed.

Our citation practices acknowledge and build upon relevant research from across the EU, including studies from both Western and Eastern European countries, ensuring our literature review reflects the regional diversity of energy poverty research.