PHOTOS: ISTOCK

This chapter provides an assessment of energy poverty in the Nordics. The analysis is based on existing survey-based data obtained through the EU Joint Research Center (JRC)

The assessments are based on JRC data analysis over the EU SILC and HBS data (output JRC135908, part of the data catalogue “Just Energy and Transport Data Inventory”). The views expressed in the report are purely those of the authors and may not in any circumstance be regarded as stating an official position of the European Commission.

. Specifically, the statistics presented are based on data from the EU statistics on income and living conditions (EU-SILC) survey

https://ec.europa.eu/eurostat/web/microdata/european-union-statistics-on-income-and-living-conditions

and the EU Household Budget Survey (HBS)

https://ec.europa.eu/eurostat/web/microdata/household-budget-survey

. Given that these surveys are collected on an EU level from each Member State (and some countries outside of EU), they allow for a harmonized assessment across the Nordic countries

It should be noted that the surveys are susceptible to uncertainty in sampling. It has not been possible for us to test the statistical significance of the findings.

. The assessment is based on some of the most frequently applied indicators on energy poverty (see section 4.5 for a more elaborate discussion on indicators).

The chapter is structured in three major sections. Section 5.1 conducts an overall assessment of energy poverty in the Nordics through time, based on selected indicators of energy poverty. Sections 5.3 and 5.4 take stock of how the prevalence of energy poverty varies on a range of background variables. Specifically, section 5.3 is concerned with demographics whereas section 5.4 is concerned with living conditions such as dwelling and tenure types.

This section assesses the level of energy poverty in the Nordics through time

Due to data availability and comprehensive computational time, it has not been possible to provide complete time series. The selection of years to be presented from the SILC survey is based on critical socio-economic importance as well as the wish to be able to compare with HBS data. Thus, EU-SILC survey data are presented for the years 2015, 2016, 2018, 2020, 2021, and 2022. HBS data are available in 5-year intervals and thus presented for the years 2010, 2015 and 2020. The Nordic (and EU27) average reported in a given year is based on countries with available data in the given year.

. The assessment is based on five indicators in a combination of consensual and expenditure-based approaches (see section 4.5). The selection of indicators has been guided by data availability and the wish to triangulate the assessment of energy poverty. While no indicator is perfect on its own, as discussed in section 4.5, the assessment of energy poverty using multiple indicators and adopting different approaches may help identify important dimensions of energy poverty. The five indicators are:

Table 2 provides an overview of the identified and discussed pros and cons in chapter 5:

As will be seen, the five indicators provide a different picture of energy poverty in the Nordics, both across the Nordic countries and in comparison with the EU27. These results underscore the importance of triangulation and a thorough discussion of the contribution of each indicator. Also evident is the fact that data are missing on some indicators for some of the Nordic countries, reflecting the importance of investigating the potential for using alternative data sources or establishing new data sources to measure energy poverty in the Nordic countries.

One of the most referenced indicators of energy poverty is the self-reported inability to keep home adequately warm (warm) indicator stemming from the SILC survey. While this indicator can be criticized for being subject to individual preferences varying with age, gender, and other socioeconomic factors, it does provide important insights into the mechanics at play, like hidden energy poverty and deprivation, especially when comparing across time.

Figure 10 shows the proportion of individuals who report being unable to keep their houses adequately warm. The figure presents the share for each of the Nordic countries, the average across the Nordic countries, and the EU27 average. The figure suggests that all Nordic countries report fewer issues with keeping their housing adequately warm than the average across the EU27. While the Nordic average does not exceed 3.3% of individuals being unable to keep their home adequately warm, the EU27 average varies between 6.8% and 9.6%. This tendency of variation was a subject of criticism during the workshop because, to some extent, it constitutes uncertainties about the variable’s credibility.

The figure also suggests that the proportion of individuals reporting issues with keeping their housing adequately warm increased from 2021 to 2022 for both Denmark and Sweden, the Nordic average

Note that Norway and Iceland data were unavailable for 2021 and 2022. The Nordic average in these years comprises Denmark, Finland, and Sweden.

, and the EU27. This increase in thermal discomfort is probably due to the energy crisis that hit towards the end of 2021, causing a surge in energy prices. The reality and consequences of rising energy prices is seemingly reflected in the ability of individuals’ to afford essential energy services (such as being able to keep their home adequately warm) in line with the extra financial burden of increased energy prices. Somewhat surprisingly, the share of individuals being unable to keep their home adequately warm did not increase in Finland from 2021 to 2022 as opposed to in Denmark and Sweden

Energy efficiency at what cost? Unjust burden-sharing of rent increases in extensive energy retrofitting projects in Sweden, von Platten et al., 2022 and Member state reports on energy poverty 2019, EPOV, 2020

. This may be explained by the disconnection protection policy that Finland has in place to protect households in debt from being disconnected from the supply. Furthermore, Denmark has a higher share than both Sweden and Finland in the 2022 survey. This may also be explained by the fact that energy costs are included in the rent in Finland and, usually but not always, in Sweden too

Member state reports on energy poverty 2019, EPOV, 2020

, which shifts the decision of cutting down on energy services in response to price increases away from the individual household. Consequently, Swedish and Finnish households may have less control over and incentive to reduce energy costs.

Note: No available data for IS after 2018, and for NO after 2020. “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. *Years: 2017 and 2019 are missing in the presented time series.

The second self-reported indicator included in this assessment (and suggested by EPOV, see section 4.5.1) is the arrears on utility bills (arrears) indicator. The measure offers a focused and specific approach to addressing energy poverty, which can be advantageous when estimating its prevalence. However, the indicator may also reflect a prioritisation of paying energy bills over maintaining a warm living environment. In addition, the indicator has not been adjusted for general buying-power and levels of inflation underlining the indicator’s nature of self-regulation.

Figure 11 summarises the development in this indicator over time for the Nordic countries and the EU27 average. Struggles to pay the utility bills in time is generally more prevalent in the EU27 average than in the Nordic countries. During the entire time series, however, Finland has the highest share of individuals with arears on utility bills among the Nordic countries, which is on par with the EU27 average. This is in stark contrast to the warm indicator, for which Finland was found to be among the lowest shares. The discrepancy may be due to the Finnish disconnection protection policy ensuring that missed payments do not automatically lead to disconnections from supply. Norway and Sweden have the lowest average levels across the observed years, which may be due to the universal energy service obligation and relatively low energy prices in Norway.

The figure also suggests that all countries with available data in 2021-2022 (except Finland) experience an increase in poverty from 2021 to 2022, pointing once again to the energy crisis. Interestingly, the energy crisis does not seem to have had as big of an impact on the arrears indicator as it does on the warm indicator. There are several potential explanations for this. Firstly, the Nordic countries have strong social security mechanisms in place where households can apply for support when paying their energy bills. Secondly, the relationship between the two indicators could indicate a trade-off effect, where consumers are prioritising paying their bills over keeping their accommodation warm. Thirdly, with energy being a basic good, households might cut down consumption elsewhere in order to meet their basic needs for energy services. Lastly, the modest increase in poverty may also reflect that although increasing energy prices put a strain on household budgets, the price increases might have been manageable. This is considering the various policies, which have been implemented in the Nordics to support households against the financial stress during the energy crisis.

Note: The figure shows the proportion of individuals having arrears on utility bills during the past year. No available data for IS after 2018, and for NO after 2020. “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. *Years: 2017 and 2019 are missing in the presented time series.

Source: JRC analysis based on Eurostat data, EU-SILC (2015-2022)

The expenditure-based indicators are usually applied in tandem because they each capture distinct issues related to energy poverty and affordability. The expenditure-based indicators originate from HBS and are presented for the years 2010, 2015, and 2020

HBS data are not available for NO and IS. HBS 2020 data are only available for DK. The Nordic average is based on countries with available data, which in 2010 and 2015 were DK, FI and SE. The Nordic average is not reported for 2020, because DK is the only Nordic country with available data in this year.

. Unfortunately, the data do not cover the energy crisis as the most recent available data are from 2020

Some countries do collect these data on a more frequent basis, e.g. Denmark collects the HBS annually. However, the EU wide collection of HBS data via Eurostat, on which we base this assessment, follows 5-year intervals.

.

The twice the national median share (2M indicator) provides an insight into the budgetary burden of energy bills relative to households’ disposable income, using the national median as point of reference

EPOV Indicator Dashboard: Methodology Guidebook, Thema and Vondung, 2020

. This relative nature of the indicator causes the shares to be affected by the underlying distribution of income and energy expenditures within each country (see section 4.5 for a discussion on indicators), creating a moving target. On the other hand, the indicator may depict the relative hardship of households within a country. With regard to archetypes, the indicator may misidentify high-income households with proportionally high energy expenses while not identifying low-income households who underconsume energy due to general financial constraints. 

Figure 12 shows the proportion of households that have a share of energy expenses relative to income, which is equal to or higher than the 2M indicator. The data indicate that all the Nordic countries (with available data) have a higher prevalence of energy poverty according to the 2M indicator than the EU average. The estimated Nordic average is above 20 percent in both 2010 and 2015, indicating that more than a fifth of the Nordic households included in the analysis are energy poor according to this indicator. The result is especially driven by Sweden, for whom 29% of the households had more than twice the national median share of energy expenditures to income in 2015. Evidently, this high prevalence of energy poverty according to the 2M indicator stands in stark contrast to the warm and arrears indicators from above. The shares in the Nordic countries have increased over time, while the EU average has decreased slightly over time (note, however, that Denmark is the only Nordic country with data for 2020). Such increases may be driven either by increased energy expenses moving households above the 2M threshold or by a decrease in the 2M threshold and as such prove very hard to interpret.

Note: Income and energy expenditures are equivalized according to household size. No available data for IS and NO in all years, and for FI and SE in 2020. “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. EU figures as the average of countries with available data are inconsistent for 2010 and 2015 because of the inconsistencies in the sampling weights of some countries.
Source: JRC analysis based on Eurostat data, HBS (2010, 2015, 2020)

In order to avoid some of the issues related to the 2M indicator, we provide numbers for the 10% indicator

The 10% indicator is the absolute equivalent of the 2M indicator, Energy poverty indicators: Conceptual issues. Part I: The ten-percent-rule and double median/mean indicators, Schuessler, 2014

, which identifies the proportion of households with energy expenses of 10% or more of net household income. Unlike the previous indicator, the 10% cut-off is not dependent on the population distribution. Instead, the indicator is expected to reflect the general energy price level of the countries compared to other goods and services. Unfortunately, data from Iceland and Norway are not available for this indicator. In addition, it is common for certain property types in Finland and Sweden to have the energy expenditures included in the rental

Member state reports on energy poverty 2019, EPOV, 2020

. For this reason, when respondents are asked to indicate how much they spend on energy costs, it may not be possible to differentiate this number from the housing costs. This was evident in the Swedish data, where a lump of households specified that they had zero energy expenditures

This is also evident in the EPOV methodology report published in 2020: EPOV Indicator Dashboard: Methodology Guidebook, Thema and Vondung, 2020

. This highlights the importance of examining energy poverty within national constraints. It also underlines the problems associated with the 2M indicator, as the very low number of households in Finland and Sweden that identified with the 10% indicator does not nearly meet the high number of households identified with the 2M indicator. This is likely due to a downward bias in the median threshold applied by the 2M indicator.

Figure 13 shows that the proportion of households identified by this measure varies considerably across the three included countries. While Finland and Sweden have less than 10% of households spending more than 10% of their budget on energy, Denmark has 14-19% of households spending more than 10%. While the difference between the countries in part may be explained by higher energy prices in Denmark as described in section 4.4, some of the discrepancy may – as argued – be attributed to the fact that energy bills are included in the rent in Finland and, usually but not always, in Sweden as well.

Contrary to the 2M indicator, the results show a decrease in the percentages of the population that may be considered at risk of energy poverty. While some households have decreased their energy expenditure to below 10%, a number of households also moved above the 2M threshold. Due to the decrease in the absolute 10% indicator, it is likely that the increase in the relative 2M indicator above can be explained by a lower 2M threshold rather than increased energy expenditures. Nonetheless, the figure shows that of the three Nordic countries measured, Denmark is the only country where the identified share of households exceeds the EU27 average by a considerable margin (in 2010 and 2020), indicating a level of economic strain similar to the 2M measure above. Lastly, the EU27 share decreases considerably from 2015 to 2020. This reflects a decrease in the economic strain across the EU27 countries with available data in 2020.

Note: Income and energy expenditures are equivalized according to household size. No available data for IS and NO in all years, and for FI and SE in 2020. The figures in “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. EU figures as the average of countries with available data are inconsistent for 2010 and 2015 because of the inconsistencies in the sampling weights of some countries.
Source: JRC analysis based on Eurostat data, HBS (2010, 2015, 2020)

In contrast to the two previous indicators, Figure 14 illustrates the proportion of households with energy expenses in euros that is half or less than the national median level (M/2 indicator). This measure is referred to as hidden energy poverty, because it aims to measure the number of households with abnormally low energy expenditures, who potentially are failing to meet their basic needs. The pitfalls of this indicator are that it may also capture households living in highly energy efficient homes as well as also being referenced to as a changing target, where changes in the energy expenditure distribution translates into a new M/2 threshold.

Considering the shares of households based on this indicator, Finland and Sweden are observed to have an overall higher level of potentially hidden energy poverty than Denmark and the EU27 average. Interestingly, the distribution and development of this indicator resembles quite closely the M2 indicator.

Note: Income and energy expenditures are equivalized according to household size. No available data for IS and NO in all years, and for FI and SE in 2020. “Nordic average” and “EU27” refer to the average of the Nordic countries and EU Member States with available data in the given year. EU figures as the average of countries with available data are inconsistent for 2010 and 2015 because of the inconsistencies in the sampling weights of some countries.
Source: JRC analysis based on Eurostat data, HBS (2010, 2015, 2020)

The five indicators presented in the previous sections are summarised in Figure 15 in 2015 levels as 2015 is the most recent year where all Nordic EU Member States have available data in HBS. The figure shows a stark contrast between the consensual-based and expenditure-based indicators. In general, the expenditure-based measures indicate substantially higher levels of energy poverty in the Nordics. The same conclusion applies to the EU average. However, the difference is not nearly as large.

Note: The HBS indicators are based on income and energy expenditures equivalized according to household size. No available data for IS and NO in HBS. The figures in “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. EU figures as the average of countries with available data are inconsistent for 2010 and 2015 because of the inconsistencies in the sampling weights of some countries.
Source: JRC analysis based on Eurostat data, SILC (2015) and HBS (2015)

While providing an assessment of energy poverty on the aggregate level in the Nordics, this section also underlines some of the inherent challenges associated with using these indicators to assess energy poverty in the Nordics as the indicators do not identify a specific group of energy poor citizens.

Most notably, the structure of energy bills poses severe limitations on interpretations from the expenditure-based indicators. Furthermore, the 2M and M/2 indicators are subject to changing targets and do not lend themselves very easily to comparisons across countries. While it may seem that there is a high prevalence of energy poverty in relative terms, it might not be the case when comparing across countries in absolute terms. The expenditure-based indicators are also sensitive towards purchasing power. Other issues may include under-reporting of vulnerability (on consensual-based indicators), survey bias caused by sample sizes as well as the citizens’ space for action, e.g. consume energy when it is cheap, heat up specific areas of the residence while leaving others cold and similar. Irrespective of the root cause, these differences high­light the need for further efforts to triangulate the measurement of energy poverty and perhaps investigate alternative indicators suited for the specific national context.

This section takes stock of energy poverty using the indicators suggested in the Energy Efficiency Directive (EED). Given that the EED suggests using a general poverty indicator to indicate energy poverty, the section further investigates the intersection between energy poverty and general poverty. The aim of that discussion is also to provide some insights into the question of whether energy poverty is, in fact, an issue serving its own merit in the Nordic countries.

As discussed above in section 4.3, the Energy Efficiency Directive (Directive (EU) 2023/1791) requires the Member States to use the arithmetic average share of the following four indicators to measure the extent of energy poverty in their National Energy and Climate Plans, if they have not already committed themselves to other indicators:

All four indicators are collected in the SILC survey. Figure 16 shows the four indicators for each country in 2018

2018 is the latest year where data on all Nordic countries are available in the SILC survey.

. The warm and arrears indicators follow the same tendencies as described above, that the Nordic countries generally have lower shares than the EU27 average on both measures. The share of the population living with leak, damp or rot varies significantly across the Nordic countries with Iceland and Denmark having higher shares than the EU27 and Finland having the lowest share. The leak, damp or rot indicator can be interpreted as a proxy for building quality or poor conditions in dwellings. However, the measure is best used to support other measures for energy poverty. In Iceland, for example, it can be argued that the comparatively low energy prices (as shown in section 4.4) might somewhat offset energy poverty associated with poor housing quality. Although not disclosed here, the leak, damp or rot indicator remains relatively stable over time, which is to be expected given that changes to the building stock is often slow to implement.

The at risk of poverty (AROP) indicator identifies the share of the population with an equivalized disposable income (after social transfers) below 60% of the national equivalized median income

Glossary: At-risk-of-poverty rate - Statistics Explained, Eurostat

. Figure 16 shows that, for all countries, a larger share of the population is at risk of poverty than having trouble keeping their homes adequately warm or experiencing arrears on utility bills. As such, it seems that energy poverty is less prevalent than general poverty in all the Nordic countries. Recall, however, that the AROP measure is relative to the national median whereas the warm and arrears indicators are ‘absolute’ measures without reference to a changing target.

Note: The leak, damp or rot in the dwelling indicator shows the proportion of individuals living in a dwelling with a leaking roof, damp walls, floors or foundation, or rot in window frames or floor. The at risk of poverty (AROP) indicator identifies the share of the population with an equivalized disposable income (after social transfers) below 60% of the national equivalized median income. The figures in “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. 2018 is the latest year where data were available for all Nordic countries.

Figure 17 shows the development in the AROP indicator over time. Evidently, the share of individuals at risk of poverty is relatively stable in all Nordic countries and the EU27. This contrasts with the development in the energy poverty (warm and arrears) indicators shown above, which tend to fluctuate more and increased significantly in 2022. Although high correlation between poverty and energy poverty is to be expected, the development of the two are not perfectly correlated. Some of this discrepancy may be ascribed to the relative nature of the AROP indicator (where the share is dependent on the national median) versus the ‘absolute’ nature of the warm and arrears indicators.

Note: The at risk of poverty (AROP) indicator identifies the share of the population with an equivalized disposable income (after social transfers) below 60% of the national equivalized median income. No available data for IS after 2018, and for NO after 2020. The figures in “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. *Years: 2017 and 2019 are missing in the presented time series.
Source: JRC analysis based on Eurostat data, SILC (2015-2022)

Although the warm and AROP indicators do not perfectly align, there are correlations. Using 2018 data, these correlations are illustrated in Figure 18, where the share of individuals unable to keep their home adequately warm is calculated for individuals above and below the AROP threshold. Recall that the AROP threshold is 60% of the national equivalized median income. The third panel illustrates the ratio between the two groups. The panel shows that individuals at risk of poverty have 3.3 times the like­li­hood of being unable to warm their houses compared to individuals not at risk of poverty. Given that access to essential energy services is largely dependent on financial capabilities, this relationship is to be expected. Norway is an outlier in this respect with a ratio as high as 16.8, indicating that people at risk of poverty have almost a seven­teen times higher risk of not being able to keep their home adequately warm. This result is mainly driven by the very low share of individuals above the AROP thres­hold being unable to keep their home ade­quately warm. Ultimately, the AROP indicator is not necessarily an appro­priate indicator for identi­fying energy poor individuals if being used isolated. Rather, it may be useful for identifying a group of individuals at risk of energy poverty.

Note: The at risk of poverty (AROP) indicator identifies the share of the population with an equivalized disposable income (after social transfers) below 60% of the national equivalized median income. The figures in “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. 2018 is the latest year where data were available for all Nordic countries.
Source: JRC analysis based on Eurostat data, SILC (2018)

This section aims to investigate the prevalence of energy poverty across demographic groups in order to qualify a discussion of whether there are groups that are especially vulnerable towards energy poverty. The parameters considered include income levels, economic activity, health, household types, children in the household, age, and gender.

In October 2023, EPAH published a report on updated indicators divided into four primary topics: Climate, Facilities/Housing, Mobility, and Socioeconomic aspects

Energy Poverty National Indicators Uncovering New Possibilities for Expanded Knowledge. Energy Poverty Advisory Hub. October 2023.

. The topics are supplemented by subtopics, indicators and a comprehensive set of suggested variables applied for disaggregation of each indicator in the EPAH dashboard solution

Energy Poverty Advisory Hub. National Indicators.

. The inherent complexity of the concept of energy poverty merits an investigation of several indicators coupled with disaggregation of these indicators on secondary variables to get a comprehensive understanding of the issue. The following section is based primarily on the inability to keep home adequately warm indicator disaggregated on various secondary variables

While we have argued for a triangulation of indicators in general, this section focuses primarily on the warm indicator. This is due to the low reliability of expenditure-based indicators in the Nordic countries. The warm indicator is prioritised over the arrears indicator in order to limit the scope of the analysis. The warm indicator was chosen because some countries in the Nordics, e.g. Finland, have policies that guard against disconnections due to arrears. In such settings, the arrears would overestimate the level of energy poverty.

.

In general, income is one of the main explanatory components of energy poverty as described in section 4.1. This merits a thorough look into the dynamics of how energy poverty affects different groups in the income distribution. In this context, it is worth noting that energy poverty may be a subset of vulnerable groups, and if so, it is essential to describe the connections between these groups. In the Nordics, there is some overlap between vulnerable groups in general and those experiencing energy poverty. However, it should be emphasized that energy poverty does not automatically imply vulnerability. This underscores the necessity of clearly distinguishing between the two groups and elucidating how they are interconnected (see chapter 8 for policy recommendation).

Figure 19 shows the proportion of individuals in each income decile who are unable to keep their home adequately warm in 2018

2018 is the latest year where data on all Nordic countries are available in the SILC survey.

. Interestingly, the distribution of energy poverty according to the warm indicator varies quite a lot across the Nordic countries. There seems to be a strong correlation between income and energy poverty in Denmark and Norway and to some extent in Sweden, where the first- and second-income deciles have the highest shares of energy poverty on this measure. The distribution of energy poverty across income seems more equal in Finland and especially in Iceland.

Note: Income deciles are based on equivalized disposable household income. “Nordic average" refers to the average of the Nordic countries with available data in the given year. The EU27 panel is left out to be able to see the nuances in the Nordic countries. *Data are missing for the 5th income decile in IS and the 6th, 8th, and 10th income deciles in NO. 2018 is the latest year where data were available for all Nordic countries.

Figure 20 illustrates the proportion of individuals in each income decile who have experienced arrears on utility bills in 2018. Unsurprisingly, the overall pattern is that lower income deciles have a higher risk of experiencing arrears for all Nordic countries and the EU27. The larger the difference in proportions for high and low deciles, the more explanatory power income presumably has in terms of explaining the ‘arrears on utility bills’ indicator. The data indicate that individuals in the lowest income decile in the Nordic countries on average are 2.4 times more likely to have arrears on utility bills compared to the average individual.

Note: Income deciles are based on equivalized disposable household income. “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. 2018 is the latest year where data were available for all Nordic countries.

Figure 21 indicates the share of households in each income decile with a ratio of energy expenditure relative to net household income twice as high or more than the national median in 2015. The distribution of energy poverty according to the 2M indicator varies a great deal across countries. Income seems to correlate strongly with energy poverty according to this measure in Denmark and the EU27 average, with lower income deciles having a higher share of individuals with energy expenditures above the 2M threshold. In Finland and Sweden, on the other hand, the share of energy poor households is the highest in the 5^th income decile. This result underlines the structural difference in energy bills mentioned in section 5.1.1. Because energy bills are included in the rent in these countries, some households may wrongly indicate that they have (close to) zero energy expenditures, which translates into a downwards bias in the median. This caveat underlines the severe problems with using these expenditure-based indicators in Sweden and Finland as they do not appropriately reflect energy poverty.

Note: Income and energy expenditures are equivalized according to household size. No available data for IS and NO. “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. EU figures as the average of countries with available data are inconsistent for 2010 and 2015 owing to the inconsistencies in the sampling weights of some countries.

Although income is an important factor, other variables may provide alternative insights into what constitutes and drives energy poverty. Figure 22 illustrates the proportion of individuals with different economic status, who are unable to keep houses adequately warm. The figure shows that unemployed individuals are more exposed to not being able to warm their houses. However, the category ‘other’, which comprises individuals outside of the four other categories, also experiences a consistently higher risk of not being able to warm their houses. A key insight from this is that there is also a significant group of ‘other inactive’ who may not be active on the job market, who are shown to be exposed to potential energy poverty. Hence, any measure of energy poverty must also account for proportions of the population that slip through standard measures of the labour market.

Note: Data are based on the population aged 18 and above. Other inactive includes individuals unable to work due to long-standing health problems, individuals fulfilling domestic tasks (housewife/​husband), individuals in compulsory military or civilian service and other. *IS data are from 2016 due to missing data on unemployed in 2018. “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. 2018 is the latest year where data were available for all Nordic countries.

One of the consequences of energy poverty is poor health. There already exists comprehensive academic literature investigating the effects of energy poverty on health. Some studies even document how energy poverty increases morbidity and mortality

What are the effects of energy poverty and interventions to ameliorate it on people's health and well-being?: A scoping review with an equity lens, Ballesteros-Arjona et al., 2022

. Figure 23 illustrates the proportion of individuals unable to keep home adequately warm disaggregated based on self-perceived health status. It is apparent that having self-perceived poor health is highly correlated with not being able to keep your home adequately warm. Denmark and Norway have the highest correlations where the risk of not being able to heat your home is respectively 6 and 5.7 times higher for people with self-perceived poor health. Such connections may also reinforce the previous figure, where poor perceived mental or physical health has the risk of resulting in long-time unemployment.

Note: Data are based on the population aged 18 and above. General health status is defined by self-perceived general health status. ’Good health includes ’Good’ and ’Fairly good’. ’Bad’ includes ’bad’ and ’very bad’. “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. 2018 is the latest year where data were available for all Nordic countries.

Another key variable to investigate is household type. Figure 24 provides a break-down of the inability to keep home adequately warm indicator for different household compositions. The figure shows that single parent households struggle to keep their houses adequately warm to a higher degree than the other household compositions across all countries. Single adult households come in second in all countries except for Sweden. These shares stand in stark contrast to couples with and without children, who have very low shares of individuals unable to keep their home adequately warm. Seemingly the number of adults in a household correlate very strongly with the access to adequate heating.

Note: “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. 2018 is the latest year where data were available for all Nordic countries.

Delving further into household types, Figure 25 shows the share of households unable to keep their home adequately warm disaggregated by the number of children in the household in 2018. In Denmark and Sweden, the highest shares of energy poor households are households with one child. In Finland and Iceland, the highest shares are among households with three children. For the EU average, the highest incidence of energy poverty is observed among families with either no children or more than four children. While the pattern across the examined countries is ambiguous, it is challenging to definitively establish the significance of children in relation to energy poverty. Nonetheless, having children living at home constitutes a factor to consider when addressing the consequences of rising energy prices through mitigation measures.

Note: Total number of children in the household refers to children aged 0-17. *Data for households with 4+ children are unavailable for DK and IS. **The shaded bar indicates low reliability due to small number of observations. “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. 2018 is the latest year where data were available for all Nordic countries.

Interestingly, when the indicator of inability to keep home adequately warm is cross-tabulated with age groups as in Figure 26, there does not appear to be large differences. All countries show a similar distribution, where younger adults seem to struggle the most, which then tapers off with age. According to these numbers, older individuals do not seem to be especially vulnerable, which may come as a surprise since elderly people have been highlighted as a vulnerable group in literature and in the qualitative data in particular

Energy poverty – New insights for measurement and policy

.

Note: “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. 2018 is the latest year where data were available for all Nordic countries.

Previous work has examined aspects of energy poverty in relation to gender, specifically in terms of the gender pay gap; the gender pension gap; and women's more limited possibilities to work compared with men due to their disproportionate burden of care for children and other close relatives

Zamfir I., Gender aspects of energy poverty, EPRS, February 2023 (https://www.europarl.europa.eu/thinktank/en/document/EPRS_ATA(2023)739349)

. Taking inspiration from this work, Figure 27 shows the proportion of men and women who are unable to keep their houses adequately warm. The share of women is slightly higher for all Nordic countries (and the EU27 average) except for Norway. Like previous figures, this seems to suggest a well-functioning social welfare system. However, it is worth noting that the relative difference between men and women are similar between the EU27 average and DK, SE and FI.

Note: Data are based on the population aged 18 and above. “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. 2018 is the latest year where data were available for all Nordic countries.

This section aims to investigate the prevalence of energy poverty associated with living conditions. The parameters considered include dwelling type, tenure status, and the presence of leak, damp or rot.

The relationship between housing stock and ability to keep dwellings warm appears to be an important dynamic. Figure 28 shows the share of individuals unable to keep their home adequately warm within each type of dwelling. The figure indicates that individuals who live in apartment buildings with less than 10 dwellings (and to some extent in apartment buildings with 10 or more dwellings) appear to be at the highest risk of being unable to keep their home warm. This result may reflect several factors. Firstly, an inability of individual owners (or renters) of apartments to renovate in order to make their dwelling more energy efficient. Secondly, there could also be a correlation with income, where lower income or single income individuals for whom it is more difficult to afford adequate heating, may be more likely to reside in apartments as opposed to houses in the Nordic countries. Thirdly, it may also reflect the rental-owner situation, where renters may be over-represented in apartments while also having smaller disposable incomes and thus less likely to be able to afford adequate heating. Fourthly, with respect to the size of the apartment blocks, small blocks with less than 10 dwellings are more at risk than larger blocks in all cases (except for in Sweden). This may reflect that larger apartment blocks are typically more recent and, as a result, better insulated than smaller apartment dwellings. Some of the oldest apartments may have such poor insulation that adequately heating them becomes challenging irrespective of the resident’s income.

Irrespective of the underlying dynamic, it is an important finding that individuals living in houses have a very low risk of being unable to keep their home warm compared to individuals living in apartments. This may reflect that energy efficiency renovations are more available to individuals living in houses, and as such may stress the importance of targeting energy efficiency towards individuals living in apartments.

Note: “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. Germany is excluded from the EU27 average due to unavailable data on dwelling type in some years. 2018 is the latest year where data were available for all Nordic countries.

Figure 29 below shows the indicator for ability to keep home warm in terms of individuals’ tenure status. These results show a far less homogenous picture across the different Nordic countries, reflecting different (social) housing standards and regulations. For example, in Denmark, reduced rent and rent-free tenures are not reported, whilst in Sweden, reduced rate rentals are also not seen

“In a situation where there is no clear distinction between a ‘prevailing rent’ sector and a ‘reduced rent’ sector, all renters would be classified as ‘tenant or subtenant paying rent at prevailing or market rate’.” Methodological guidelines and description of EU-SILC target variables, 2021

. In contrast, rent-free tenures in Finland are seen to contain a very high proportion of individuals unable to keep their dwellings adequately warm. As mentioned, this seems to reflect differing housing tenure systems among the different Nordic countries. The implications are that energy poverty should be considered to a large degree in tandem with the individual countries’ housing tenure systems. Across the Nordic countries, however, owners both with and without outstanding mortgage report very low shares of individuals unable to keep their home adequately warm. Again, this may correlate with latent variables such as income and competencies to implement energy efficiency improvements.

Note: *Data for tenants renting at reduced price are unavailable for DK and SE. Data for rent-free tenants are unavailable for DK and NO. “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. 2018 is the latest year where data were available for all Nordic countries.

Finally, Figure 30 below gives an indication of how the building quality is associated with the ability to keep the dwelling adequately warm. Although it is somewhat expected that damp, leak, or rot would be correlated with the ability to keep the dwelling warm, the figure shows a very strong correlation between the two. The data indicate that the likelihood of being unable to keep the home adequately warm is 4.7 times as high when living in a dwelling with leak, damp or rot.

Note: “Nordic average" and "EU27" refer to the average of the Nordic countries and EU Member States with available data in the given year. 2018 is the latest year where data were available for all Nordic countries.