Energy Poverty Policy Predictions (EP3)

Energy Poverty Policy Predictions (EP3) | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Energy Poverty Policy Predictions (EP 3 ) Joseph Llewellyn, Titus Venverloo, Fábio Duarte This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8986845/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Energy poverty policies are now mandated by European Union regulations including energy renovations, energy coaching and energy subsidy interventions within homes. Yet, the long-term effects of such policies are usually only predicted and not evaluated. Thus, we visited 203 homes and used mixed methods to evaluate effects of these different policies implemented from 2020-2025. Subsidies have no long-term impact. Coaching reduces monthly electricity by 56kWh and gas consumption by 30.7m3 and percentage of income spent on energy, groceries and rent together by 3.48%. However, renovations lower gas (59.4m3) and energy bills (€110) with stable indoor temperatures. Lidar scans show relationships between indoor spatial characteristics and energy bills. Sociodemographic data finds that homes which were still energy poor post-policy, were more likely to live on social allowances and have chronic illnesses. Scientific community and society/Energy and society/Energy policy Scientific community and society/Energy and society/Energy security Scientific community and society/Social sciences/Economics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction The Energy Poverty Advisory Hub suggests multiple objective and subjective indicators to measure energy poverty, but there are policy gaps within all countries (EPAH, 2023). For example, most countries use expenditure indicators (i.e. % income spent on energy) and income-based policies (i.e. energy subsidies to cover the rising costs over winter). However, these policies are often reactionary rather than preventative (Pye et al., 2017) and more methods are addressing this using modelling (Garibay et al., 2023). Past research using affordability indicators in energy poverty predictive models finds 1.35 million homes could be lifted out of energy poverty in the EU (Vandyck et al., 2023). However, these models typically assume optimal policy conditions with rational actors. Nevertheless, because there is often a large gap between these policy predictions and actual outcomes, greater research is needed to explain how and why these gaps exist. Across EU member states, policies such as subsidies are by far the most prevalent, as they are easiest for governments to implement. They are effective in the short term (covering price spikes for 3 winter months) but with modest long-term effects (+1 year) where only 9% of homes escape energy poverty (Jove-Lopis & Trujilo-Baute, 2024). However, this does not usually address the root cause of the issue, mostly the building. The energy performance of buildings directive (EPBD) aims to renovate 35 million EU homes by 2030, focussing on efficiency and insulation (European Parliament, 2024). This directive can reduce energy poverty by 4% across all EU homes (Vivier et al., 2025). Other EU-wide studies focus on how such policies impact disabled and chronically ill homes, finding they are at higher risk of energy poverty (Ivanova & Middlemiss, 2021). They suggest energy policy research is needed inside these socially vulnerable homes. Longitudinal studies on expenditure helps to explain the persistence of energy poverty. Studies in Spain find the range of energy poor homes escaping energy poverty is greater than the range of income poor homes escaping income poverty (Phimister et al., 2015). However, their survey data could not state which policies caused these found effects. Studies in France, found no evidence of energy poverty traps with econometric models, but their models did not explain how energy policies may help (Chaton & Lacroix, 2018). Finally, a review across 27 EU countries finds that higher structural vulnerability indices, are associated with higher rates of long-term energy poverty (Recalde et al., 2019). They suggest EU governments should support more affordable social housing policies, such as the Netherlands, which has the most social housing (32%) in its total stock. In the Netherlands, social housing provides affordable housing for low-income citizens, with a rental cost upper limit of €880/month (Government of the Netherlands, 2026). Research on social housing managers influence on energy poverty finds they support renovations and behavioural interventions but lack data on resident’s poverty status, thus, other household costs must be considered in energy policies (Croon et al., 2024). A household is in poverty, if after paying for fixed expenses of rent, energy and health insurance, it has insufficient income to pay for other needs, such as food (CBS, 2025). But energy bills are not always fixed and the definition does not explain consumption, which is influenced by the insulation for energy, or the large expenses of grocery bills. For a single person home, the poverty line is €1,600/month and for a couple it is €2,145. For a couple with two children <13 years it is €2,625, and for a single parent it is €2,215. There are 550,000 people in the Netherlands below this poverty line (CBS, 2025). Delving deeper into the municipalities, Amsterdam has the highest poverty rate (6.6%). The ratio of homes in energy poverty (low-income + high bill + low energy quality home) in Amsterdam is 9.4% in all homes but rises to 15.1% in social housing (DEGO, 2024). Prior studies analyse effects of subsidies on income (Roosma, 2022) retrofits on energy efficiency (Dang et al., 2024) and coaching on consumption (Llewellyn et al., 2025). However, no studies have assessed multiple policy interventions in social housing with longitudinal effects that cover three major household needs of energy, food and rent. Globally, energy poverty studies rely on national household surveys (Guan et al., 2022), but researchers admit their findings are limited by assumptions on the indoor context. Advancements in Lidar technology have allowed building managers to measure the indoor spatial configurations, such as volume, which may contribute to energy demand. But to date, no study has used Lidar scans to try assess energy poverty within homes. Moreover, thermal cameras now allow measurements of cold and hot spots in homes. Cong et al., (2021) find heating inequalities between low to high income homes but again this was self-reported using national surveys rather than direct measurements. Finally, there is a need to complement policy modelling (Li et al., 2024; Xiao et al., 2024) with lived experiences of households effected by the policies (Bouzarovski et al., 2025). Research aims and approach This study measures: a) major household expenditures including energy, food and rent b) gas and electricity consumption c) spatial characteristics inside the homes, d) indoor temperatures, e) socio-demographics of the people inside the homes and f) evaluation of the policy intervention from the household perspective. We visited 203 homes, from February to July 2025. These homes have received either energy subsidies (n=50), coaching (n=51), renovations (n=51) or nothing at all (n=51). All homes received grocery vouchers, food packs and energy products after the study. The subsidies were financial transfers of €1800 in April 2022 and €800 in Oct 2023, to homes with low income and high energy bills to help them cover increased costs. The energy coaching was performed in the early months of 2022 and 2023, whereby a volunteer visits the home to provide energy advice designed to reduce consumption. The renovations also took place in the early months of 2022 and 2023 whereby buildings received uniform insulation upgrades in the ceilings, floors, walls and window areas. Detailed steps of data collection and analysis can be found in the methods section. This helps us answer the overall research question: ‘ What are the impacts of subsidy, coaching and retrofit policies on household poverty?’ Results Expenditure Results A staggered difference in differences model was applied to analyse causal effects. Importantly, these effects represent savings relative to the counterfactual scenario, rather than to their own pre intervention baselines. Coaching reduces energy bills by €90 a month (95% CI −112 to −69, p<.001), while renovations homes reduce by €110 (95% CI −134 to −85, p<.001) in figure 1 above. But, subsidy homes found no significant effect. Renovations, coaching and subsidies had no long-term effect on grocery bill spending, thus we do not find evidence that money saved on energy bills is displaced to groceries. However, subsidy homes report a short-term spike in monthly spending on groceries up to €700 per month on receiving a subsidy and then a significant decrease by €40 per month in the counterfactual (95% CI −62.6 to −16.7, p<.001) in figure 2 above. For rental bills, only renovation homes show a significant increase of €37 per month (95% CI −55 to −19, p<.001), due to home improvements. This is significant but is still €73 less than the money they saved on their energy bills. Next we use the expenditure indicator (% income spent) to account for their income. A Kruskal-Wallis test found no significant group income differences (H = 3.12, p =.373). Subsidy homes show no significant change in percentage income spent on all bills, whereas renovation homes do by 2.44% (95% CI −4.55 to −0.33, p<.001), however coaching home reduce by 3.48% (95% CI −6.02 to −0.94, p<.001) in figure 3 above. Consumption Results Coaching reports significantly reduced electricity consumption by 56kWh per month (95% CI −67.8 to −44.2, p<.001). Renovations and subsidies did not in. Homes receiving subsidies show no significant reduction of monthly gas consumption, whereas coaching homes do by 30.7m 3 per month (95% CI −38 to −23.4, p<.001) and renovation homes by 59.4m 3 (95% CI −70.6 to −48.3, p<.001). Heat and Health Results A Kruskal–Wallis test on regression residual indoor temperatures found no statistically significant differences between all the intervention groups, (H(3) = 0.65, p = .89). Therefore, we performed a slopes test on four intervention groups, in figure 4 below. We regressed indoor temperature on outdoor temperature with an interaction term. The slope was significantly lower for renovation Homes (Δb = −0.43 °C, p = .004), indicating reduced sensitivity of indoor temperatures to outdoor temperatures. Spatial and Socio-demographic Results Lidar scans within the homes helped to reveal spatial relationships to energy poverty. Significant differences were found with smaller total window area (m²) (U = 187.0, p = .00022) and window/wall ratio (U = 240.0, p = .00437) for coaching than renovations. Coaching had smaller window areas than subsidies too (m²) (U = 272.5, p = .02787). However, every other spatial variable was not statistically significant between groups. Thus, we analysed which spatial characteristics are common with higher energy bills. A univariate ordinary least squares linear regression found energy bills increased with number of windows (β=.059, p =.0002), number of rooms (β=.085, p =.0002), surface area (β=.0046 per m², p =0.0006), volume (β=.0015 per m³, p =.0046), and outside wall perimeter (β=0.0137, p =0.0062). This translates to 6% per extra window , 8% per room , .46% per m² , .15% per m³ , and 1.4% per meter of outside wall (using lidar in figure 6). Furthermore, we analysed which social characteristics persist with energy poverty. Homes on social allowance and with chronic illness were more prevalent in energy-poor than non-energy-poor homes (both χ², p < .05). Meanwhile, age is negatively correlated with grocery bills (ρ = −.44, p < .001), thus homes with older people spend less on food. These background spatial and social characteristics can be found in appendix A and B. Interview Results When asked to evaluate subsidies, homes said “ It helped massively but it was for such a short period of time. And afterwards we’re in the same situation again ”. But for others it was a much-needed intervention to aid them in a serious condition. “The rent is the first thing that goes out of my social allowance. Even if I don’t have nothing to eat, the rent must be paid. The last month I had €70 left. I am very thankful to the municipality. If it wasn’t for the money, maybe I’d commit suicide” . When asked to evaluate coaching, homes said “ I think the coaching is great. It is perfect. Slight small changes can make a difference. But you need to meet people in their own homes. If you just make a video no one will think about it ”. But others noted how homes who did not receive coaching, changed other behaviours. “I know that some people started having mould as soon as the energy prices went up. Basically, they stopped heating and they kept their windows closed as they didn't want things to cool down. So their behavior changed rather than the building ”. When asked to evaluate renovations, homes were happy with results. “ The big change for us was new windows. The heat now stays inside instead of leaking out ”. But, most homes were not happy about the process which provided little information. “ It was not a central communication, there were several companies working. So there came communications from all sides, and we were informed very poorly ”. “ There's little transparency to it. It was unclear who was who and why they were there. They would come in and out unannounced. It was difficult to reach them ”. Discussion and Conclusion Subsidies show no long-term effects for reducing energy consumption or energy bills. Prior studies find unstable prices can lead up to a 4.8% increase in energy expenditures and emphasise support for increased costs of groceries especially (Guan et al., 2023). The spikes in grocery bills were extremely pronounced during the intervention month, which is expected for an April subsidy, when energy bills are no longer as high as winter. Prior studies find homes struggle with eligibility for subsidies (Scheier & Kittner, 2022) and are often excluded from such policies due to socio-economic characteristics. Interviews revealed that control homes were actually eligible but unaware at the time. As in prior subsidy research, effects are short-lived (Jove-Lopis & Trujilo-Baute, 2024). For coaching homes there were significant effects for reduced electricity consumption, gas consumption, energy bills, percentage of income spent on all three bills together. This is consistent with previous energy coaching results (Llewellyn et al., 2025). But energy coaching appeared to have no significant impact on groceries or rental bills, suggesting the policy does not have any negative or positive spillovers to other needs. Savings from the energy bills were not necessarily displaced to groceries as expected. The coaching homes were on average significantly warmer than other treatment homes, but unlike renovation homes, this was due to behaviour change rather than the building. For renovation homes, there was a significant effect on gas consumption, energy bills and percentage of income spent on energy but not electricity or all bills combined. Interviews suggest that homes did not change their heating settings and behaviours as the building envelope was now providing enough warmth and cooling when necessary. However, they state the renovation process was overwhelming and they felt passive, with teams of anonymous strangers entering their homes without a personal touch. Furthermore, renovation homes also found a significant increase in their rental bills. Compared to coaching, percentage of income spent on all 3 bills was less pronounced. As in prior work, renovation policies tend to come with trade-offs (Vivier et al., 2025). The policy intervention timing is crucial so that positive effects can become long-term. Short-term subsidies work better in October than April to cover heating season costs, but many homes were unaware they were eligible until a neighbour had told them. Medium-term coaching may work better with renovations to multiply possible effects. This may involve coaches who help apply the renovations or visit very shortly after. Renovations are often easier implemented in summer, then we can deploy energy coaches in autumn and send out subsidies in winter when homes need them most. The main limitation here concerns the research design of studying policies in isolation, when in reality most homes are exposed to multiple policies throughout the year. Moreover, some homes may receive food insecurity policies and go to food banks, therefore their spending on groceries is less, which may impact energy bill spending. Meanwhile, data on the consumption of groceries could not be obtained, this means that we cannot state if they under/over consume enough healthy and nutritious food. Subsidy results here show it was not a conditional transfer that must be spent on energy and homes chose to spend it on another need, more akin to universal basic income. Policies for food insecurity favour cash transfers and food stamps (Teeuwen et al., 2022) but future studies could compare conditional versus unconditional financial transfers. Moreover, our sample had few under-consumers of energy (see supplementary data). However, many other studies do find patterns of under consuming (Cong et al., 2022). Global research finds poverty policies yield minor emissions (Wollburg et al., 2023), while climate policies have spillover effects that reduce energy poverty (Li et al., 2024). But, when implemented in cities, they may increase energy poverty (Xiao et al., 2024), especially if rising energy prices outweigh improved energy infrastructure of buildings. EU studies have largely relied on simulations, models, and annual household surveys, creating a knowledge gap in how homes experience EU-mandated energy policies. We find subsidies are helpful but only give transient relief and have limited effects. Until EU policies become bolder and implement universal basic income policies, there will forever be mismatch between the models and needs of energy poor homes. We find coaching reduces both gas and electricity, while reducing energy poverty. But EU energy policies prioritise technical interventions such as retrofits and heat pumps that reduce gas related emissions but not electricity (Vivier et al., 2025). We find retrofits are the most effective policy to reduce energy poverty indicators. But, when retrofit policy fails to consider energy savings being used to cover increased rental or grocery costs, then energy poverty effects will be limited by houseold poverty. We find energy poverty mostly effects homes with social allowance and chronic illness. Future research into household poverty should ensure that single mothers get help beyond the financial burdens of paying energy, grocery and rental bills altogether. Many interviewees started crying when discussing their strategies to deal with poverty. Issues ranged from eating less so kids would be fed, to managing mental health issues. Previous research also suggests that energy poverty measurement be compared with a formal measurement of food insecurity using a severity scale (Bednar & Reames, 2020). If governments take a proactive response to energy poverty as done with food insecurity, then the problem might be better measured and mitigated than it is being currently. The Netherlands has a wealth of energy production. The country is not energy poor. So why should the homes be? These interventions make small but significant impacts on facets of household poverty. Put together in a timely policy, they may help homes heat, eat and live healthier lives. Methods A mixed methods naturally occurring experiment was conducted to study the effects of three singular policy interventions that had already been implemented two years prior. All homes had to request or approve the policy intervention in their home themselves. Homes were opportunity sampled through various ways depending on the intervention. Renovation homes responded to a flyer in the mail asking for research participation. Coaching homes responded to a request by email from the energy coaching services. Subsidy homes requested coaching and agreed to take part in the study upon request. Control homes requested coaching or were found when visiting other treatment homes where neighbours recommended the research request through snowball sampling. This method helped to keep control home characteristics similar to treatment homes. All the homes were low income, with similar demographics living in social housing with low energy labels but even in the same building, within home differences can be vast. For example, Lidar scans show homes of the same dimensions but with varied design where some homes have couches in front of radiators, or others have thermal curtains. Each visit consisted of 1 hour of data collection, followed by 1hour of energy coaching. Homes received energy products, grocery vouchers, and food packs as renumeration. The subsidies, coaching and renovations serve as our independent variables while the bills, consumption and percentage of income spent, serve as our dependent variables. This included monthly data from January 2020 to December 2024 for 203 homes. Income statements, energy, grocery and rental bills were prepared at home beforehand. However, we also took lidar and thermal scans when visiting the homes to investigate spatial characteristics and heating characteristics that may relate to energy poverty. These variables are not treated as dependent variables as they are a single point in time, therefore it is more difficult to claim causality due to any of the policy interventions. All data is in additional material A. Lidar and thermal scans are available upon request. Moreover, interviews gathered socio-demographic information (i.e. age, income type), but also asked the same 8 open ended questions about their evaluation of the policy. This ranged from; ‘Can you tell me about your experience with the intervention’ to ‘What can the municipality do to improve the policy to make sure your needs are met?’. All data was anonymised and handled in accordance with EU GDPR rules (EU, 2016). Participants were given the right to withdraw their data at any time. However, none did. Under-consuming is defined as less than ½ the national median (2,800 kWh electricity; 1,040 m³ gas). The proportion of electricity under consumers did not differ significantly across intervention groups (χ²(3)=5.02, p=.171; V=0.16). However, non-energy-poor households were more likely to be under consumers than energy-poor households (14.0% vs 3.9%; χ²(1)=6.43, p=.011; φ=0.18; RR≈3.6; Δ=10.1 pp). The proportion of gas over-consumers differed across intervention groups (χ²(3)=9.33, p=.025; V=0.21), with renovations showing the highest share (23.5%) and Control the lowest (5.9%). Non-energy-poor households were more likely to be gas under-consumers than energy-poor households (20.0% vs 3.9%; χ²(1)=12.64, p<.001; φ=0.25; RR≈5.1; Δ=16.1 pp). A main strength of this study was to include 5 years of longitudinal data from 2020-2024 therefore we have 2 years of data pre-intervention and 1-2 years post-intervention. This means we can assess pre-trends to analyse whether indicators were rising before, and post-trends to analyse whether the intervention really has long-lasting effects. A difference in differences (DiD) model is typical for policy analysis on panel data, which was used to analyse and assess which intervention was the most effective. As interventions occurred in different months, which ranged from January 2022 to February 2023 we employed a staggered DiD policy analysis to estimate causal effects. In total, 12,180 observations were made across all homes and variables of interest. To observe the treatment effects we visualised counterfactual trajectories representing the expected outcomes for treated homes in the absence of any policy intervention. The counterfactual value (post-intervention) is equal to the treated groups mean (at the last pre-intervention period) plus the cumulative change found in the control group (from pre- to post-intervention). Background economic variables such as gas and electricity prices are controlled for. Rental prices were not controlled for, as they are a fixed cost determined each year. Grocery prices were not controlled, as we could not obtain this data from the homes. Parallel trends at the group and household level were conducted with assumptions met. A two-way fixed effects model with monthly and household effects was specified. Renovation homes’ pre-intervention gas use differed significantly from the others. Parallel trends were validated at both the group and household specific level (p>.05). However, renovation homes’ pre-intervention gas differed significantly from the others. A placebo test was run with no effect found pre intervention, improving the test validity. The tests were adjusted for horizon effects, so that we are reporting long-lasting effects beyond 1 year and not only short-term effects that may rebound (i.e. with subsidies). An event study analysis examined dynamic treatment effects and pre-treatment trends, a heterogeneity analysis on baseline consumption levels and a sensitivity analysis using winsorized outcomes to address outliers, were all performed for robustness checks. The model simplified is; Yit = αi + γt + βDit + εit Yit = dependent variable (i.e. energy bills) αi= household fixed effects (controls for each home’s time-variant traits) = month fixed effects (controls for macro shocks) = treatment indicator (1 if treated household in post-period) = DiD estimate of the policy effect (i.e. energy bills are reduced by €50 per month) εit = error term (i.e. unexplained variation such as job loss) However, we did observe heterogenous effects at baseline for the renovation homes. This treatment group had significantly higher gas, energy bills and income spent on energy than other groups, meaning that there post-intervention effects could be greater. This reflects a flaw in naturally occurring experimental designs, where interventions usually occurred due to some pre-conceived intent or governmental policy purpose. Renovations were done in these homes because they were in urgent need of them. Therefore, we also accounted for background spatial and social characteristics. However, lidar scans showed no significant differences in spatial characteristics between the intervention groups, nor between the energy poor and non-poor homes. Moreover, interviews found no significant differences in social characteristics between the intervention groups but did for income between energy poor and non-poor homes. Therefore, by investigating only social housing homes specifically in this sample, background characteristics are homogenous and intervention effects more causal. Furthermore, outdoor temperature was controlled for during the thermal scans. A delta temperature was considered for the thermal differences found between homes but comparisons would be unreliable as we measured from February to August. Therefore, we remove the linear effect of outdoor temperature by analysing residuals and the buildings sensitivity to outdoor temperatures, quantified by regression slopes. An interaction test confirmed that the renovation homes slope was significantly lower (0.11) than control homes (0.54), subsidy homes (0.51) and coaching homes (0.69). Lidar and thermal scans were done side by side and hand in hand during the home visit. This took approximately 10 minutes using a Flir thermal camera and an iPhone 12 pro. Declarations Ethics Approval: This study was granted approval by the ethics committee from the AMS Institute for Metropolitan Solutions. Participant Consent: All participants provided informed consent prior to data collection and were given the right to withdraw responses at any time. Data Availability All data generated during this study are included in this article and supplementary files. Competing Interests The authors declare no competing interests. Author Contributions J.L: Conceptualisation, Methodology, Validation, Formal Analysis, Investigation, Writing (Original Draft), Writing (Review and Editing) T.V: Validation, Writing (Review and Editing), Supervision F.D: Writing (Review and Editing), Supervision. References Bouzarovski, S., Cedano-Villavicencio, K. G., Delina, L. L., Martiskainen, M., Reames, T. G., & Simcock, N. (2025). Reframing how we talk about ‘energy poverty’. Nature Energy , 1-5. https://doi.org/10.1038/s41560-025-01794-w Centraal Bureau voor de Statistiek [Statistics Netherlands]. (2025, December 17). Welfare of persons; key figures [Statistical data]. Retrieved from https://www.cbs.nl/en-gb/figures/detail/86135ENG Chaton, C., & Lacroix, E. (2018). Does France have a fuel poverty trap?. Energy Policy , 113 , 258-268. https://doi.org/10.1016/j.enpol.2017.10.052 Chipunza, T., & Ntsalaze, L. (2025). Multi-dimensional poverty: a bibliometric analysis and content co-occurrence literature review. Humanities and Social Sciences Communications , 12 (1), 1-12. https://doi.org/10.1057/s41599-025-04924-7 Cong, S., Nock, D., Qiu, Y. L., & Xing, B. (2022). Unveiling hidden energy poverty using the energy equity gap. Nature communications , 13 (1), 2456. https://doi.org/10.1038/s41467-022-30146-5 Croon, T. M., Hoekstra, J. S. C. M., & Dubois, U. (2024). Energy poverty alleviation by social housing providers: A qualitative investigation of targeted interventions in France, England, and the Netherlands. Energy Policy , 192 , 114247. https://doi.org/10.1016/j.enpol.2024.114247 Dang, M., van den Dobbelsteen, A., & Voskuilen, P. (2024). A Parametric Modelling Approach for Energy Retrofitting Heritage Buildings: The Case of Amsterdam City Centre. Energies , 17 (5), 994. https://doi.org/10.3390/en17050994 Energy Poverty Advisory Hub. (2023, March 2). EPAH Handbook 1: A Guide to Energy Poverty Diagnosis . Directorate-General for Energy, European Commission. Retrieved from chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://energy-poverty.ec.europa.eu/system/files/2024-05/EPAHhandbook_diagnosis_finalpdf.pdf? European Parliament & Council. (2024, April 24). Directive (EU) 2024/1275 on the energy performance of buildings (recast) . Official Journal of the European Union, L1275. https://eur-lex.europa.eu/eli/dir/2024/1275/oj European Union. (2016). Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data (General Data Protection Regulation) . Official Journal of the European Union. https://eur-lex.europa.eu/eli/reg/2016/679/oj/eng? González Garibay, M., Primc, K., & Slabe-Erker, R. (2023). Insights into advanced models for energy poverty forecasting. Nature Energy , 8 (9), 903-905. https://doi.org/10.1038/s41560-023-01311-x Government of the Netherlands. (2026). Rented housing . Government.nl. https://www.government.nl/topics/housing/rented-housing Guan, Y., Yan, J., Shan, Y., Zhou, Y., Hang, Y., Li, R., ... & Hubacek, K. (2023). Burden of the global energy price crisis on households. Nature Energy , 8 (3), 304-316. https://doi.org/10.1038/s41560-023-01209-8 Ivanova, D., & Middlemiss, L. (2021). Characterizing the energy use of disabled people in the European Union towards inclusion in the energy transition. Nature Energy , 6 (12), 1188-1197. https://doi.org/10.1038/s41560-021-00932-4 Jové-Llopis, E., & Trujillo-Baute, E. (2024). Escaping the energy poverty trap: policy assessment. Environmental and Resource Economics , 87 (12), 3335-3355 https://doi.org/10.1007/s10640-024-00918-2 Li, J., Li, J., Guo, K., Ji, Q., & Zhang, D. (2024). Policy spillovers from climate actions to energy poverty: international evidence. Humanities and Social Sciences Communications , 11 (1), 1-12. https://doi.org/10.1057/s41599-024-03614-0 Llewellyn, J., Venverloo, T., Duarte, F., Ratti, C., Katzeff, C., Johansson, F., & Pargman, D. (2025). Assessing the impact of energy coaching with smart technology interventions to alleviate energy poverty. Scientific Reports , 15 (1), 969. https://doi.org/10.1038/s41598-024-80773-9 Phimister, E., Vera-Toscano, E., & Roberts, D. (2015). The dynamics of energy poverty: evidence from Spain. Economics of Energy & Environmental Policy , 4 (1), 153-166. https://www.jstor.org/stable/26189479 Pye, S., Dobbins, A., Baffert, C., Brajković, J., Deane, P., & De Miglio, R. (2017). Energy poverty across the EU: Analysis of policies and measures. In Europe's Energy Transition (pp. 261-280). Academic Press. https://doi.org/10.1016/B978-0-12-809806-6.00030-4 Recalde, M., Peralta, A., Oliveras, L., Tirado-Herrero, S., Borrell, C., Palència, L., ... & Marí-Dell’Olmo, M. (2019). Structural energy poverty vulnerability and excess winter mortality in the European Union: Exploring the association between structural determinants and health. Energy Policy , 133 , 110869. https://doi.org/10.1016/j.enpol.2019.07.005 Roosma, F. (2022). A struggle for framing and interpretation: the impact of the ‘basic income experiments’ on social policy reform in the Netherlands. European Journal of Social Security , 24 (3), 192-212. https://doi.org/10.1177/13882627221109846 Scheier, E., & Kittner, N. (2022). A measurement strategy to address disparities across household energy burdens. Nature Communications , 13 (1), 288. https://doi.org/10.1038/s41467-021-27673-y Teeuwen, A. S., Meyer, M. A., Dou, Y., & Nelson, A. (2022). A systematic review of the impact of food security governance measures as simulated in modelling studies. Nature Food , 3 (8), 619-630. https://doi.org/10.1038/s43016-022-00571-2 Vandyck, T., Della Valle, N., Temursho, U., & Weitzel, M. (2023). EU climate action through an energy poverty lens. Scientific reports , 13 (1), 6040. https://doi.org/10.1038/s41598-023-32705-2 Vivier, L., Mastrucci, A., & van Ruijven, B. (2025). Meeting climate target with realistic demand-side policies in the residential sector. Nature Climate Change , 1-8. https://doi.org/10.1038/s41558-025-02348-4 Vereniging van Nederlandse Gemeenten. (2025). Datavoorziening Energietransitie Gebouwde Omgeving (DEGO) . Retrieved October 28, 2025 from https://dego.vng.nl/?label=topo&layer=layer0#13.98/52.08743/4.32195 Wollburg, P., Hallegatte, S., & Mahler, D. G. (2023). Ending extreme poverty has a negligible impact on global greenhouse gas emissions. Nature , 623 (7989), 982-986. https://doi.org/10.1038/s41586-023-06679-0 Xiao, Y., Feng, Z., Li, X., & Wang, S. (2024). Low-carbon transition and energy poverty: Quasi-natural experiment evidence from China’s low-carbon city pilot policy. Humanities and Social Sciences Communications , 11 (1), 1-18. https://doi.org/10.1057/s41599-023-02573-2 Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryDataCleanVersion.xlsx Data set on household poverty indicators across 5 years RS.pdf Reporting Summary AppendixAandB.docx Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8986845","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":607029141,"identity":"d4702ca9-05d1-4593-8a23-62b9b9a4317c","order_by":0,"name":"Joseph Llewellyn","email":"data:image/png;base64,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","orcid":"","institution":"Amsterdam Institute for Metropolitan Solutions","correspondingAuthor":true,"prefix":"","firstName":"Joseph","middleName":"","lastName":"Llewellyn","suffix":""},{"id":607029142,"identity":"6bac62e0-3e4b-4266-9a3f-ff54a696dea5","order_by":1,"name":"Titus Venverloo","email":"","orcid":"https://orcid.org/0009-0009-4629-9152","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Titus","middleName":"","lastName":"Venverloo","suffix":""},{"id":607029143,"identity":"35d5e492-8b80-4491-ae7d-ecda64e00fc8","order_by":2,"name":"Fábio Duarte","email":"","orcid":"https://orcid.org/0000-0003-0909-5379","institution":"Massachussets Institute of Technology","correspondingAuthor":false,"prefix":"","firstName":"Fábio","middleName":"","lastName":"Duarte","suffix":""}],"badges":[],"createdAt":"2026-02-27 10:30:34","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8986845/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8986845/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105185654,"identity":"ab5bb0e8-b12b-4460-b112-6c3abf866b83","added_by":"auto","created_at":"2026-03-23 08:28:41","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":121821,"visible":true,"origin":"","legend":"\u003cp\u003eAverage monthly energy bills over time with yearly averages\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8986845/v1/a010572eaabb9455f0e11d26.jpg"},{"id":105185697,"identity":"c9ab59c8-13c5-433d-9ed4-b700839dff00","added_by":"auto","created_at":"2026-03-23 08:28:48","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":106833,"visible":true,"origin":"","legend":"\u003cp\u003eAverage monthly grocery bills over time with yearly averages\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8986845/v1/396179cfb9c6319ff56cd9ef.jpg"},{"id":105185722,"identity":"f6d9aebe-c696-4ef9-bb37-a971d0c9813f","added_by":"auto","created_at":"2026-03-23 08:29:04","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":125278,"visible":true,"origin":"","legend":"\u003cp\u003eMonthly percentage of income spent on all three bills with yearly averages\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8986845/v1/56e3357d6e87c7142031a8a7.jpg"},{"id":105185700,"identity":"20435aa9-adab-41df-ba3e-f669bbb8cfe8","added_by":"auto","created_at":"2026-03-23 08:28:51","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":127466,"visible":true,"origin":"","legend":"\u003cp\u003eSlope test of indoor versus outdoor temperature by intervention group\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8986845/v1/e15d50fe6af300a97c7f7456.jpg"},{"id":105185715,"identity":"4dfa1a70-6e16-4007-9beb-30aadf5f9139","added_by":"auto","created_at":"2026-03-23 08:28:58","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":119235,"visible":true,"origin":"","legend":"\u003cp\u003eControl home, renovation home, subsidy home and coaching home scans Post-intervention mould, damp or leaks were still present in 66.7% of subsidy homes, \u0026nbsp;56% of coaching homes and just 11.8% of renovation homes (in figure 5 above). \u0026nbsp;Average temperature residuals were significantly lower in homes with mould \u0026nbsp;(median −.09 °C) than without mould (median +.82 °C) (U=1297.5, p=.039).\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8986845/v1/eb937a091e5197e55b2a8ea9.jpg"},{"id":105185634,"identity":"20eb2a78-ad00-4560-be8d-a8af403dcfa3","added_by":"auto","created_at":"2026-03-23 08:28:33","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":128397,"visible":true,"origin":"","legend":"\u003cp\u003eLidar scans in 123 homes reveal 3d spatial features using Polycam\u003c/p\u003e","description":"","filename":"6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8986845/v1/fe951293a24e6cf0837b2aac.jpg"},{"id":105185725,"identity":"cc1bf50c-8290-4860-befd-3394f2c13ce9","added_by":"auto","created_at":"2026-03-23 08:29:09","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1274095,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8986845/v1/78825892-c976-4541-8d49-7147a136ec53.pdf"},{"id":105185652,"identity":"475942e1-6682-4a1a-a320-910b1fc39e01","added_by":"auto","created_at":"2026-03-23 08:28:40","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1779456,"visible":true,"origin":"","legend":"Data set on household poverty indicators across 5 years","description":"","filename":"SupplementaryDataCleanVersion.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8986845/v1/bf721fec1d9477f8c0ec751a.xlsx"},{"id":105185701,"identity":"e6115233-a1cd-45a6-ac17-51a556403a68","added_by":"auto","created_at":"2026-03-23 08:28:52","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":95785,"visible":true,"origin":"","legend":"Reporting Summary","description":"","filename":"RS.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8986845/v1/3abc6d7eba312afbe717856a.pdf"},{"id":105185696,"identity":"164821e8-7049-4afd-810c-7ae497c912c1","added_by":"auto","created_at":"2026-03-23 08:28:47","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":28556,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixAandB.docx","url":"https://assets-eu.researchsquare.com/files/rs-8986845/v1/f33915424db94491f8b76a46.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"\u003cp\u003eEnergy Poverty Policy Predictions (EP\u003csup\u003e3\u003c/sup\u003e)\u003c/p\u003e","fulltext":[{"header":"Introduction ","content":"\u003cp\u003eThe Energy Poverty Advisory Hub suggests multiple objective and subjective indicators to measure energy poverty, but there are policy gaps within all countries (EPAH, 2023). For example, most countries use expenditure indicators (i.e. % income spent on energy) and income-based policies (i.e. energy subsidies to cover the rising costs over winter). However, these policies are often reactionary rather than preventative (Pye et al., 2017) and more methods are addressing this using modelling (Garibay et al., 2023). Past research using affordability indicators in energy poverty predictive models finds 1.35 million homes could be lifted out of energy poverty in the EU (Vandyck et al., 2023). However, these models typically assume optimal policy conditions with rational actors. Nevertheless, because there is often a large gap between these policy predictions and actual outcomes, greater research is needed to explain how and why these gaps exist.\u003c/p\u003e\n\u003cp\u003eAcross EU member states, policies such as subsidies are by far the most prevalent, as they are easiest for governments to implement. They are effective in the short term (covering price spikes for 3 winter months) but with modest long-term effects (+1 year) where only 9% of homes escape energy poverty (Jove-Lopis \u0026amp; Trujilo-Baute, 2024). However, this does not usually address the root cause of the issue, mostly the building. The energy performance of buildings directive (EPBD) aims to renovate 35 million EU homes by 2030, focussing on efficiency and insulation (European Parliament, 2024). This directive can reduce energy poverty by 4% across all EU homes (Vivier et al., 2025). Other EU-wide studies focus on how such policies impact disabled and chronically ill homes, finding they are at higher risk of energy poverty (Ivanova \u0026amp; Middlemiss, 2021). They suggest energy policy research is needed inside these socially vulnerable homes. \u003c/p\u003e\n\u003cp\u003eLongitudinal studies on expenditure helps to explain the persistence of energy poverty. Studies in Spain find the range of energy poor homes escaping energy poverty is greater than the range of income poor homes escaping income poverty (Phimister et al., 2015). However, their survey data could not state which policies caused these found effects. Studies in France, found no evidence of energy poverty traps with econometric models, but their models did not explain how energy policies may help (Chaton \u0026amp; Lacroix, 2018). Finally, a review across 27 EU countries finds that higher structural vulnerability indices, are associated with higher rates of long-term energy poverty (Recalde et al., 2019). They suggest EU governments should support more affordable social housing policies, such as the Netherlands, which has the most social housing (32%) in its total stock. \u003c/p\u003e\n\u003cp\u003eIn the Netherlands, social housing provides affordable housing for low-income citizens, with a rental cost upper limit of \u0026euro;880/month (Government of the Netherlands, 2026). Research on social housing managers influence on energy poverty finds they support renovations and behavioural interventions but lack data on resident\u0026rsquo;s poverty status, thus, other household costs must be considered in energy policies (Croon et al., 2024). A household is in poverty, if after paying for fixed expenses of rent, energy and health insurance, it has insufficient income to pay for other needs, such as food (CBS, 2025). But energy bills are not always fixed and the definition does not explain consumption, which is influenced by the insulation for energy, or the large expenses of grocery bills. \u003c/p\u003e\n\u003cp\u003eFor a single person home, the poverty line is \u0026euro;1,600/month and for a couple it is \u0026euro;2,145. For a couple with two children \u0026lt;13 years it is \u0026euro;2,625, and for a single parent it is \u0026euro;2,215. There are 550,000 people in the Netherlands below this poverty line (CBS, 2025). Delving deeper into the municipalities, Amsterdam has the highest poverty rate (6.6%). The ratio of homes in energy poverty (low-income + high bill + low energy quality home) in Amsterdam is 9.4% in all homes but rises to 15.1% in social housing (DEGO, 2024). Prior studies analyse effects of subsidies on income (Roosma, 2022) retrofits on energy efficiency (Dang et al., 2024) and coaching on consumption (Llewellyn et al., 2025). However, no studies have assessed multiple policy interventions in social housing with longitudinal effects that cover three major household needs of energy, food and rent. \u003c/p\u003e\n\u003cp\u003eGlobally, energy poverty studies rely on national household surveys (Guan et al., 2022), but researchers admit their findings are limited by assumptions on the indoor context. Advancements in Lidar technology have allowed building managers to measure the indoor spatial configurations, such as volume, which may contribute to energy demand. But to date, no study has used Lidar scans to try assess energy poverty within homes. Moreover, thermal cameras now allow measurements of cold and hot spots in homes. Cong et al., (2021) find heating inequalities between low to high income homes but again this was self-reported using national surveys rather than direct measurements. Finally, there is a need to complement policy modelling (Li et al., 2024; Xiao et al., 2024) with lived experiences of households effected by the policies (Bouzarovski et al., 2025).\u003c/p\u003e\n\u003ch2\u003eResearch aims and approach \u003c/h2\u003e\n\u003cp\u003eThis study measures: a) major household expenditures including energy, food and rent b) gas and electricity consumption c) spatial characteristics inside the homes, d) indoor temperatures, e) socio-demographics of the people inside the homes and f) evaluation of the policy intervention from the household perspective. \u003c/p\u003e\n\u003cp\u003eWe visited 203 homes, from February to July 2025. These homes have received either energy subsidies (n=50), coaching (n=51), renovations (n=51) or nothing at all (n=51). All homes received grocery vouchers, food packs and energy products after the study. The subsidies were financial transfers of \u0026euro;1800 in April 2022 and \u0026euro;800 in Oct 2023, to homes with low income and high energy bills to help them cover increased costs. The energy coaching was performed in the early months of 2022 and 2023, whereby a volunteer visits the home to provide energy advice designed to reduce consumption. The renovations also took place in the early months of 2022 and 2023 whereby buildings received uniform insulation upgrades in the ceilings, floors, walls and window areas. Detailed steps of data collection and analysis can be found in the methods section. This helps us answer the overall research question: \u003c/p\u003e\n\u003cp\u003e\u0026lsquo;\u003cem\u003eWhat are the impacts of subsidy, coaching and retrofit policies on household poverty?\u0026rsquo; \u003c/em\u003e\u003c/p\u003e"},{"header":"Results","content":"\u003ch3\u003eExpenditure Results\u003c/h3\u003e\n\u003cp\u003eA staggered difference in differences model was applied to analyse causal effects. Importantly, these effects represent savings relative to the counterfactual scenario, rather than to their own pre intervention baselines. \u003c/p\u003e\n\u003cp\u003eCoaching reduces energy bills by \u0026euro;90 a month (95% CI \u0026minus;112 to \u0026minus;69, p\u0026lt;.001), while renovations homes reduce by \u0026euro;110 (95% CI \u0026minus;134 to \u0026minus;85, p\u0026lt;.001) in figure 1 above. But, subsidy homes found no significant effect. \u003c/p\u003e\n\u003cp\u003eRenovations, coaching and subsidies had no long-term effect on grocery bill spending, thus we do not find evidence that money saved on energy bills is displaced to groceries. However, subsidy homes report a short-term spike in monthly spending on groceries up to \u0026euro;700 per month on receiving a subsidy and then a significant decrease by \u0026euro;40 per month in the counterfactual (95% CI \u0026minus;62.6 to \u0026minus;16.7, p\u0026lt;.001) in figure 2 above. For rental bills, only renovation homes show a significant increase of \u0026euro;37 per month (95% CI \u0026minus;55 to \u0026minus;19, p\u0026lt;.001), due to home improvements. This is significant but is still \u0026euro;73 less than the money they saved on their energy bills. \u003c/p\u003e\n\u003cp\u003eNext we use the expenditure indicator (% income spent) to account for their income. A Kruskal-Wallis test found no significant group income differences (H = 3.12, p =.373). Subsidy homes show no significant change in percentage income spent on all bills, whereas renovation homes do by 2.44% (95% CI \u0026minus;4.55 to \u0026minus;0.33, p\u0026lt;.001), however coaching home reduce by 3.48% (95% CI \u0026minus;6.02 to \u0026minus;0.94, p\u0026lt;.001) in figure 3 above. \u003c/p\u003e\n\u003ch3\u003eConsumption Results\u003c/h3\u003e\n\u003cp\u003eCoaching reports significantly reduced electricity consumption by 56kWh per month (95% CI \u0026minus;67.8 to \u0026minus;44.2, p\u0026lt;.001). Renovations and subsidies did not in. Homes receiving subsidies show no significant reduction of monthly gas consumption, whereas coaching homes do by 30.7m\u003csup\u003e3\u003c/sup\u003e per month (95% CI \u0026minus;38 to \u0026minus;23.4, p\u0026lt;.001) and renovation homes by 59.4m\u003csup\u003e3\u003c/sup\u003e (95% CI \u0026minus;70.6 to \u0026minus;48.3, p\u0026lt;.001).\u003c/p\u003e\n\u003ch3\u003eHeat and Health Results\u003c/h3\u003e\n\u003cp\u003eA Kruskal\u0026ndash;Wallis test on regression residual indoor temperatures found no statistically significant differences between all the intervention groups, (H(3) = 0.65, p = .89). Therefore, we performed a slopes test on four intervention groups, in figure 4 below. We regressed indoor temperature on outdoor temperature with an interaction term. The slope was \u003cstrong\u003esignificantly lower\u003c/strong\u003e for renovation Homes (\u0026Delta;b = \u0026minus;0.43 \u0026deg;C, p = .004), indicating reduced sensitivity of indoor temperatures to outdoor temperatures. \u003c/p\u003e\n\u003ch3\u003eSpatial and Socio-demographic Results\u003c/h3\u003e\n\u003cp\u003eLidar scans within the homes helped to reveal spatial relationships to energy poverty. \u003cbr\u003e Significant differences were found with smaller total window area (m\u0026sup2;) (U = 187.0, p = .00022) and window/wall ratio (U = 240.0, p = .00437) for coaching than renovations. \u003cbr\u003e Coaching had smaller window areas than subsidies too (m\u0026sup2;) (U = 272.5, p = .02787). However, every other spatial variable was not statistically significant between groups. Thus, we analysed which spatial characteristics are common with higher energy bills. \u003c/p\u003e\n\u003cp\u003eA univariate ordinary least squares linear regression found energy bills increased with \u003cstrong\u003enumber of windows\u003c/strong\u003e (\u0026beta;=.059, \u003cem\u003ep\u003c/em\u003e=.0002), \u003cstrong\u003enumber of rooms\u003c/strong\u003e (\u0026beta;=.085, \u003cem\u003ep\u003c/em\u003e=.0002), \u003cstrong\u003esurface area\u003c/strong\u003e (\u0026beta;=.0046 per m\u0026sup2;, \u003cem\u003ep\u003c/em\u003e=0.0006), \u003cstrong\u003evolume\u003c/strong\u003e (\u0026beta;=.0015 per m\u0026sup3;, \u003cem\u003ep\u003c/em\u003e=.0046), and \u003cstrong\u003eoutside wall perimeter\u003c/strong\u003e (\u0026beta;=0.0137, \u003cem\u003ep\u003c/em\u003e=0.0062). This translates to \u003cstrong\u003e6% per extra window\u003c/strong\u003e,\u003cstrong\u003e \u003cstrong\u003e8% per room\u003c/strong\u003e\u003c/strong\u003e, \u003cstrong\u003e.46% per m\u0026sup2;\u003c/strong\u003e,\u003cstrong\u003e \u003cstrong\u003e.15% per m\u0026sup3;\u003c/strong\u003e\u003c/strong\u003e, and\u003cstrong\u003e \u003cstrong\u003e1.4% per meter\u003c/strong\u003e\u003c/strong\u003e of outside wall (using lidar in figure 6). \u003c/p\u003e\n\u003cp\u003eFurthermore, we analysed which social characteristics persist with energy poverty. Homes on social allowance and with chronic illness were more prevalent in energy-poor than non-energy-poor homes (both \u0026chi;\u0026sup2;, p \u0026lt; .05). Meanwhile, age is negatively correlated with grocery bills (\u0026rho; = \u0026minus;.44, p \u0026lt; .001), thus homes with older people spend less on food. These background spatial and social characteristics can be found in appendix A and B. \u003c/p\u003e\n\u003ch3\u003eInterview Results\u003c/h3\u003e\n\u003cp\u003eWhen asked to evaluate subsidies, homes said \u0026ldquo;\u003cem\u003eIt helped massively but it was for such a short period of time. And afterwards we\u0026rsquo;re in the same situation again\u003c/em\u003e\u0026rdquo;. But for others it was a much-needed intervention to aid them in a serious condition. \u003cem\u003e\u0026ldquo;The rent is the first thing that goes out of my social allowance. Even if I don\u0026rsquo;t have nothing to eat, the rent must be paid. The last month I had \u0026euro;70 left. I am very thankful to the municipality. If it wasn\u0026rsquo;t for the money, maybe I\u0026rsquo;d commit suicide\u0026rdquo;\u003c/em\u003e. \u003c/p\u003e\n\u003cp\u003eWhen asked to evaluate coaching, homes said \u0026ldquo;\u003cem\u003eI think the coaching is great. It is perfect. Slight small changes can make a difference. But you need to meet people in their own homes. If you just make a video no one will think about it\u003c/em\u003e\u0026rdquo;. But others noted how homes who did not receive coaching, changed other behaviours. \u003cem\u003e\u0026ldquo;I know that some people started having mould as soon as the energy prices went up. Basically, they stopped heating and they kept their windows closed as they didn\u0026apos;t want things to cool down. So their behavior changed rather than the building\u003c/em\u003e\u0026rdquo;.\u003c/p\u003e\n\u003cp\u003eWhen asked to evaluate renovations, homes were happy with results. \u0026ldquo;\u003cem\u003eThe big change for us was new windows. The heat now stays inside instead of leaking out\u003c/em\u003e\u0026rdquo;. But, most homes were not happy about the process which provided little information. \u0026ldquo;\u003cem\u003eIt was not a central communication, there were several companies working. So there came communications from all sides, and we were informed very poorly\u003c/em\u003e\u0026rdquo;. \u0026ldquo;\u003cem\u003eThere\u0026apos;s little transparency to it. It was unclear who was who and why they were there. They would come in and out unannounced. It was difficult to reach them\u003c/em\u003e\u0026rdquo;. \u003c/p\u003e"},{"header":"Discussion and Conclusion","content":"\u003cp\u003eSubsidies show no long-term effects for reducing energy consumption or energy bills. Prior studies find unstable prices can lead up to a 4.8% increase in energy expenditures and emphasise support for increased costs of groceries especially (Guan et al., 2023). The spikes in grocery bills were extremely pronounced during the intervention month, which is expected for an April subsidy, when energy bills are no longer as high as winter. Prior studies find homes struggle with eligibility for subsidies (Scheier \u0026amp; Kittner, 2022) and are often excluded from such policies due to socio-economic characteristics. Interviews revealed that control homes were actually eligible but unaware at the time. As in prior subsidy research, effects are short-lived (Jove-Lopis \u0026amp; Trujilo-Baute, 2024). \u003c/p\u003e\n\u003cp\u003eFor coaching homes there were significant effects for reduced electricity consumption, gas consumption, energy bills, percentage of income spent on all three bills together. This is consistent with previous energy coaching results (Llewellyn et al., 2025). But energy coaching appeared to have no significant impact on groceries or rental bills, suggesting the policy does not have any negative or positive spillovers to other needs. Savings from the energy bills were not necessarily displaced to groceries as expected. The coaching homes were on average significantly warmer than other treatment homes, but unlike renovation homes, this was due to behaviour change rather than the building. \u003c/p\u003e\n\u003cp\u003eFor renovation homes, there was a significant effect on gas consumption, energy bills and percentage of income spent on energy but not electricity or all bills combined. Interviews suggest that homes did not change their heating settings and behaviours as the building envelope was now providing enough warmth and cooling when necessary. However, they state the renovation process was overwhelming and they felt passive, with teams of anonymous strangers entering their homes without a personal touch. Furthermore, renovation homes also found a significant increase in their rental bills. Compared to coaching, percentage of income spent on all 3 bills was less pronounced. As in prior work, renovation policies tend to come with trade-offs (Vivier et al., 2025). \u003c/p\u003e\n\u003cp\u003eThe policy intervention timing is crucial so that positive effects can become long-term. Short-term subsidies work better in October than April to cover heating season costs, but many homes were unaware they were eligible until a neighbour had told them. Medium-term coaching may work better with renovations to multiply possible effects. This may involve coaches who help apply the renovations or visit very shortly after. Renovations are often easier implemented in summer, then we can deploy energy coaches in autumn and send out subsidies in winter when homes need them most. \u003c/p\u003e\n\u003cp\u003eThe main limitation here concerns the research design of studying policies in isolation, when in reality most homes are exposed to multiple policies throughout the year. Moreover, some homes may receive food insecurity policies and go to food banks, therefore their spending on groceries is less, which may impact energy bill spending. Meanwhile, data on the consumption of groceries could not be obtained, this means that we cannot state if they under/over consume enough healthy and nutritious food. Subsidy results here show it was not a conditional transfer that must be spent on energy and homes chose to spend it on another need, more akin to universal basic income. Policies for food insecurity favour cash transfers and food stamps (Teeuwen et al., 2022) but future studies could compare conditional versus unconditional financial transfers. Moreover, our sample had few under-consumers of energy (see supplementary data). However, many other studies do find patterns of under consuming (Cong et al., 2022). Global research finds poverty policies yield minor emissions (Wollburg et al., 2023), while climate policies have spillover effects that reduce energy poverty (Li et al., 2024). But, when implemented in cities, they may increase energy poverty (Xiao et al., 2024), especially if rising energy prices outweigh improved energy infrastructure of buildings. \u003c/p\u003e\n\u003cp\u003eEU studies have largely relied on simulations, models, and annual household surveys, creating a knowledge gap in how homes experience EU-mandated energy policies. We find subsidies are helpful but only give transient relief and have limited effects. Until EU policies become bolder and implement universal basic income policies, there will forever be mismatch between the models and needs of energy poor homes. We find coaching reduces both gas and electricity, while reducing energy poverty. But EU energy policies prioritise technical interventions such as retrofits and heat pumps that reduce gas related emissions but not electricity (Vivier et al., 2025). We find retrofits are the most effective policy to reduce energy poverty indicators. But, when retrofit policy fails to consider energy savings being used to cover increased rental or grocery costs, then energy poverty effects will be limited by houseold poverty. \u003c/p\u003e\n\u003cp\u003eWe find energy poverty mostly effects homes with social allowance and chronic illness. Future research into household poverty should ensure that single mothers get help beyond the financial burdens of paying energy, grocery and rental bills altogether. Many interviewees started crying when discussing their strategies to deal with poverty. Issues ranged from eating less so kids would be fed, to managing mental health issues. Previous research also suggests that energy poverty measurement be compared with a formal measurement of food insecurity using a severity scale (Bednar \u0026amp; Reames, 2020). If governments take a proactive response to energy poverty as done with food insecurity, then the problem might be better measured and mitigated than it is being currently. The Netherlands has a wealth of energy production. The country is not energy poor. So why should the homes be? \u003c/p\u003e\n\u003cp\u003eThese interventions make small but significant impacts on facets of household poverty. Put together in a timely policy, they may help homes heat, eat and live healthier lives. \u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eA mixed methods naturally occurring experiment was conducted to study the effects of three singular policy interventions that had already been implemented two years prior. All homes had to request or approve the policy intervention in their home themselves. Homes were opportunity sampled through various ways depending on the intervention. Renovation homes responded to a flyer in the mail asking for research participation. Coaching homes responded to a request by email from the energy coaching services. Subsidy homes requested coaching and agreed to take part in the study upon request. Control homes requested coaching or were found when visiting other treatment homes where neighbours recommended the research request through snowball sampling. This method helped to keep control home characteristics similar to treatment homes. All the homes were low income, with similar demographics living in social housing with low energy labels but even in the same building, within home differences can be vast. For example, Lidar scans show homes of the same dimensions but with varied design where some homes have couches in front of radiators, or others have thermal curtains. \u003c/p\u003e\n\u003cp\u003eEach visit consisted of 1 hour of data collection, followed by 1hour of energy coaching. Homes received energy products, grocery vouchers, and food packs as renumeration. The subsidies, coaching and renovations serve as our independent variables while the bills, consumption and percentage of income spent, serve as our dependent variables. This included monthly data from January 2020 to December 2024 for 203 homes. Income statements, energy, grocery and rental bills were prepared at home beforehand. However, we also took lidar and thermal scans when visiting the homes to investigate spatial characteristics and heating characteristics that may relate to energy poverty. These variables are not treated as dependent variables as they are a single point in time, therefore it is more difficult to claim causality due to any of the policy interventions. All data is in additional material A. Lidar and thermal scans are available upon request. \u003c/p\u003e\n\u003cp\u003eMoreover, interviews gathered socio-demographic information (i.e. age, income type), but also asked the same 8 open ended questions about their evaluation of the policy. This ranged from; \u0026lsquo;Can you tell me about your experience with the intervention\u0026rsquo; to \u0026lsquo;What can the municipality do to improve the policy to make sure your needs are met?\u0026rsquo;. All data was anonymised and handled in accordance with EU GDPR rules (EU, 2016). Participants were given the right to withdraw their data at any time. However, none did. \u003c/p\u003e\n\u003cp\u003eUnder-consuming is defined as less than \u0026frac12; the national median (2,800 kWh electricity; 1,040 m\u0026sup3; gas). The proportion of electricity under consumers did not differ significantly across intervention groups (\u0026chi;\u0026sup2;(3)=5.02, p=.171; V=0.16). However, non-energy-poor households were more likely to be under consumers than energy-poor households (14.0% vs 3.9%; \u0026chi;\u0026sup2;(1)=6.43, p=.011; \u0026phi;=0.18; RR\u0026asymp;3.6; \u0026Delta;=10.1 pp). The proportion of gas over-consumers differed across intervention groups (\u0026chi;\u0026sup2;(3)=9.33, p=.025; V=0.21), with renovations showing the highest share (23.5%) and Control the lowest (5.9%). Non-energy-poor households were more likely to be gas under-consumers than energy-poor households (20.0% vs 3.9%; \u0026chi;\u0026sup2;(1)=12.64, p\u0026lt;.001; \u0026phi;=0.25; RR\u0026asymp;5.1; \u0026Delta;=16.1 pp). \u003c/p\u003e\n\u003cp\u003eA main strength of this study was to include 5 years of longitudinal data from 2020-2024 therefore we have 2 years of data pre-intervention and 1-2 years post-intervention. This means we can assess pre-trends to analyse whether indicators were rising before, and post-trends to analyse whether the intervention really has long-lasting effects. \u003c/p\u003e\n\u003cp\u003eA difference in differences (DiD) model is typical for policy analysis on panel data, which was used to analyse and assess which intervention was the most effective. As interventions occurred in different months, which ranged from January 2022 to February 2023 we employed a staggered DiD policy analysis to estimate causal effects. In total, 12,180 observations were made across all homes and variables of interest. To observe the treatment effects we visualised counterfactual trajectories representing the expected outcomes for treated homes in the absence of any policy intervention. The counterfactual value (post-intervention) is equal to the treated groups mean (at the last pre-intervention period) plus the cumulative change found in the control group (from pre- to post-intervention). \u003c/p\u003e\n\u003cp\u003eBackground economic variables such as gas and electricity prices are controlled for. Rental prices were not controlled for, as they are a fixed cost determined each year. Grocery prices were not controlled, as we could not obtain this data from the homes. Parallel trends at the group and household level were conducted with assumptions met. A two-way fixed effects model with monthly and household effects was specified. Renovation homes\u0026rsquo; pre-intervention gas use differed significantly from the others. Parallel trends were validated at both the group and household specific level (p\u0026gt;.05). However, renovation homes\u0026rsquo; pre-intervention gas differed significantly from the others. A placebo test was run with no effect found pre intervention, improving the test validity. The tests were adjusted for horizon effects, so that we are reporting long-lasting effects beyond 1 year and not only short-term effects that may rebound (i.e. with subsidies). An event study analysis examined dynamic treatment effects and pre-treatment trends, a heterogeneity analysis on baseline consumption levels and a sensitivity analysis using winsorized outcomes to address outliers, were all performed for robustness checks. The model simplified is; Yit = \u0026alpha;i + \u0026gamma;t + \u0026beta;Dit + \u0026epsilon;it\u003c/p\u003e\n\u003cp\u003eYit = dependent variable (i.e. energy bills)\u003c/p\u003e\n\u003cp\u003e\u0026alpha;i= \u003cstrong\u003ehousehold fixed effects\u003c/strong\u003e (controls for each home\u0026rsquo;s time-variant traits)\u003c/p\u003e\n\u003cp\u003e\u003cimg width=\"14\" height=\"22\" src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABUAAAAhCAMAAAABIhAfAAAAAXNSR0IArs4c6QAAAF1QTFRFAAAAAAAAAAA6AABmADpmADqQAGa2OgAAOgA6OpC2OpDbZgAAZjoAZrbbZrb/kDoAkDo6kLaQkNv/tmYAtmY6tpA6ttv/tv//25A62////7Zm/9uQ/9u2//+2///bvSNduQAAAAF0Uk5TAEDm2GYAAAAJcEhZcwAAFiUAABYlAUlSJPAAAAAZdEVYdFNvZnR3YXJlAE1pY3Jvc29mdCBPZmZpY2V/7TVxAAAAlUlEQVQoU+2SwRLCMAhEoVVTrRq1VTGx+f/PLJgSyTjjzZucMuzy2GEC8K8vF0hXh1vRA+Kx+MbD1DcXgORxVw0PYgq4qYnE3eTb+2c3GGhW2Tv1NTR3af14GYY3nJpzxym47AjhEmpELomZucs8kAlCKievsqzQt8VGp4ttaCoHKChJo1jg7vOUI0RXFsc9rm6//Hcz0fAGom+4DWwAAAAASUVORK5CYII=\" v:shapes=\"_x0000_i1025\" alt=\"image\"\u003e = \u003cstrong\u003emonth fixed effects\u003c/strong\u003e (controls for macro shocks)\u003c/p\u003e\n\u003cp\u003e\u003cimg width=\"20\" height=\"22\" src=\"data:image/png;base64,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\" v:shapes=\"_x0000_i1025\" alt=\"image\"\u003e = treatment indicator (1 if treated household in post-period)\u003c/p\u003e\n\u003cp\u003e\u003cimg width=\"10\" height=\"22\" src=\"data:image/png;base64,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\" v:shapes=\"_x0000_i1025\" alt=\"image\"\u003e = DiD estimate of the policy effect (i.e. energy bills are reduced by \u0026euro;50 per month)\u003c/p\u003e\n\u003cp\u003e\u0026epsilon;it = error term (i.e. unexplained variation such as job loss)\u003c/p\u003e\n\u003cp\u003eHowever, we did observe heterogenous effects at baseline for the renovation homes. This treatment group had significantly higher gas, energy bills and income spent on energy than other groups, meaning that there post-intervention effects could be greater. This reflects a flaw in naturally occurring experimental designs, where interventions usually occurred due to some pre-conceived intent or governmental policy purpose. Renovations were done in these homes because they were in urgent need of them. Therefore, we also accounted for background spatial and social characteristics. However, lidar scans showed no significant differences in spatial characteristics between the intervention groups, nor between the energy poor and non-poor homes. Moreover, interviews found no significant differences in social characteristics between the intervention groups but did for income between energy poor and non-poor homes. Therefore, by investigating only social housing homes specifically in this sample, background characteristics are homogenous and intervention effects more causal. \u003c/p\u003e\n\u003cp\u003eFurthermore, outdoor temperature was controlled for during the thermal scans. A delta temperature was considered for the thermal differences found between homes but comparisons would be unreliable as we measured from February to August. Therefore, we remove the linear effect of outdoor temperature by analysing residuals and the buildings sensitivity to outdoor temperatures, quantified by regression slopes. An interaction test confirmed that the renovation homes slope was significantly lower (0.11) than control homes (0.54), subsidy homes (0.51) and coaching homes (0.69). Lidar and thermal scans were done side by side and hand in hand during the home visit. This took approximately 10 minutes using a Flir thermal camera and an iPhone 12 pro. \u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eEthics Approval: This study was granted approval by the ethics committee from the AMS Institute for Metropolitan Solutions.\u003c/p\u003e\n \u003cp\u003eParticipant Consent: All participants provided informed consent prior to data collection and were given the right to withdraw responses at any time.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eAll data generated during this study are included in this article and supplementary files.\u003c/p\u003e\n\u003ch2\u003eCompeting Interests\u003c/h2\u003e\n\u003cp\u003eThe authors declare no competing interests.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eAuthor Contributions\u003c/h2\u003e\n\u003cp\u003eJ.L: Conceptualisation, Methodology, Validation, Formal Analysis, Investigation, Writing (Original Draft), Writing (Review and Editing) T.V: Validation, Writing (Review and Editing), Supervision F.D: Writing (Review and Editing), Supervision.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBouzarovski, S., Cedano-Villavicencio, K. G., Delina, L. L., Martiskainen, M., Reames, T. G., \u0026amp; Simcock, N. (2025). Reframing how we talk about \u0026lsquo;energy poverty\u0026rsquo;. \u003cem\u003eNature Energy\u003c/em\u003e, 1-5. https://doi.org/10.1038/s41560-025-01794-w \u003c/li\u003e\n\u003cli\u003eCentraal Bureau voor de Statistiek [Statistics Netherlands]. (2025, December 17). \u003cem\u003eWelfare of persons; key figures \u003c/em\u003e[Statistical data]. Retrieved from https://www.cbs.nl/en-gb/figures/detail/86135ENG \u003c/li\u003e\n\u003cli\u003eChaton, C., \u0026amp; Lacroix, E. (2018). Does France have a fuel poverty trap?. \u003cem\u003eEnergy Policy\u003c/em\u003e, \u003cem\u003e113\u003c/em\u003e, 258-268. https://doi.org/10.1016/j.enpol.2017.10.052\u003c/li\u003e\n\u003cli\u003eChipunza, T., \u0026amp; Ntsalaze, L. (2025). Multi-dimensional poverty: a bibliometric analysis and content co-occurrence literature review. \u003cem\u003eHumanities and Social Sciences Communications\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(1), 1-12. https://doi.org/10.1057/s41599-025-04924-7 \u003c/li\u003e\n\u003cli\u003eCong, S., Nock, D., Qiu, Y. L., \u0026amp; Xing, B. (2022). Unveiling hidden energy poverty using the energy equity gap. \u003cem\u003eNature communications\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(1), 2456. https://doi.org/10.1038/s41467-022-30146-5 \u003c/li\u003e\n\u003cli\u003eCroon, T. M., Hoekstra, J. S. C. M., \u0026amp; Dubois, U. (2024). Energy poverty alleviation by social housing providers: A qualitative investigation of targeted interventions in France, England, and the Netherlands. \u003cem\u003eEnergy Policy\u003c/em\u003e, \u003cem\u003e192\u003c/em\u003e, 114247. https://doi.org/10.1016/j.enpol.2024.114247 \u003c/li\u003e\n\u003cli\u003eDang, M., van den Dobbelsteen, A., \u0026amp; Voskuilen, P. (2024). A Parametric Modelling Approach for Energy Retrofitting Heritage Buildings: The Case of Amsterdam City Centre. \u003cem\u003eEnergies\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(5), 994. \u003cstrong\u003ehttps://doi.org/10.3390/en17050994\u003c/strong\u003e\u003c/li\u003e\n\u003cli\u003eEnergy Poverty Advisory Hub. (2023, March 2). \u003cem\u003eEPAH Handbook 1: A Guide to Energy Poverty Diagnosis\u003c/em\u003e. Directorate-General for Energy, European Commission. Retrieved from chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://energy-poverty.ec.europa.eu/system/files/2024-05/EPAHhandbook_diagnosis_finalpdf.pdf? \u003c/li\u003e\n\u003cli\u003eEuropean Parliament \u0026amp; Council. (2024, April 24). \u003cem\u003eDirective (EU) 2024/1275 on the energy performance of buildings (recast)\u003c/em\u003e. Official Journal of the European Union, L1275. https://eur-lex.europa.eu/eli/dir/2024/1275/oj\u003c/li\u003e\n\u003cli\u003eEuropean Union. (2016). \u003cem\u003eRegulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data (General Data Protection Regulation)\u003c/em\u003e. Official Journal of the European Union. https://eur-lex.europa.eu/eli/reg/2016/679/oj/eng? \u003c/li\u003e\n\u003cli\u003eGonz\u0026aacute;lez Garibay, M., Primc, K., \u0026amp; Slabe-Erker, R. (2023). Insights into advanced models for energy poverty forecasting. \u003cem\u003eNature Energy\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(9), 903-905. https://doi.org/10.1038/s41560-023-01311-x \u003c/li\u003e\n\u003cli\u003eGovernment of the Netherlands. (2026). \u003cem\u003eRented housing\u003c/em\u003e. Government.nl. https://www.government.nl/topics/housing/rented-housing \u003c/li\u003e\n\u003cli\u003eGuan, Y., Yan, J., Shan, Y., Zhou, Y., Hang, Y., Li, R., ... \u0026amp; Hubacek, K. (2023). Burden of the global energy price crisis on households. \u003cem\u003eNature Energy\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(3), 304-316. https://doi.org/10.1038/s41560-023-01209-8 \u003c/li\u003e\n\u003cli\u003eIvanova, D., \u0026amp; Middlemiss, L. (2021). Characterizing the energy use of disabled people in the European Union towards inclusion in the energy transition. \u003cem\u003eNature Energy\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(12), 1188-1197. https://doi.org/10.1038/s41560-021-00932-4 \u003c/li\u003e\n\u003cli\u003eJov\u0026eacute;-Llopis, E., \u0026amp; Trujillo-Baute, E. (2024). Escaping the energy poverty trap: policy assessment. \u003cem\u003eEnvironmental and Resource Economics\u003c/em\u003e, \u003cem\u003e87\u003c/em\u003e(12), 3335-3355 https://doi.org/10.1007/s10640-024-00918-2 \u003c/li\u003e\n\u003cli\u003eLi, J., Li, J., Guo, K., Ji, Q., \u0026amp; Zhang, D. (2024). Policy spillovers from climate actions to energy poverty: international evidence. \u003cem\u003eHumanities and Social Sciences Communications\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(1), 1-12. https://doi.org/10.1057/s41599-024-03614-0 \u003c/li\u003e\n\u003cli\u003eLlewellyn, J., Venverloo, T., Duarte, F., Ratti, C., Katzeff, C., Johansson, F., \u0026amp; Pargman, D. (2025). Assessing the impact of energy coaching with smart technology interventions to alleviate energy poverty. \u003cem\u003eScientific Reports\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(1), 969. https://doi.org/10.1038/s41598-024-80773-9 \u003c/li\u003e\n\u003cli\u003ePhimister, E., Vera-Toscano, E., \u0026amp; Roberts, D. (2015). The dynamics of energy poverty: evidence from Spain. \u003cem\u003eEconomics of Energy \u0026amp; Environmental Policy\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(1), 153-166. https://www.jstor.org/stable/26189479 \u003c/li\u003e\n\u003cli\u003ePye, S., Dobbins, A., Baffert, C., Brajković, J., Deane, P., \u0026amp; De Miglio, R. (2017). Energy poverty across the EU: Analysis of policies and measures. In \u003cem\u003eEurope\u0026apos;s Energy Transition\u003c/em\u003e (pp. 261-280). Academic Press. https://doi.org/10.1016/B978-0-12-809806-6.00030-4\u003c/li\u003e\n\u003cli\u003eRecalde, M., Peralta, A., Oliveras, L., Tirado-Herrero, S., Borrell, C., Pal\u0026egrave;ncia, L., ... \u0026amp; Mar\u0026iacute;-Dell\u0026rsquo;Olmo, M. (2019). Structural energy poverty vulnerability and excess winter mortality in the European Union: Exploring the association between structural determinants and health. \u003cem\u003eEnergy Policy\u003c/em\u003e, \u003cem\u003e133\u003c/em\u003e, 110869. https://doi.org/10.1016/j.enpol.2019.07.005 \u003c/li\u003e\n\u003cli\u003eRoosma, F. (2022). A struggle for framing and interpretation: the impact of the \u0026lsquo;basic income experiments\u0026rsquo; on social policy reform in the Netherlands. \u003cem\u003eEuropean Journal of Social Security\u003c/em\u003e, \u003cem\u003e24\u003c/em\u003e(3), 192-212. https://doi.org/10.1177/13882627221109846\u003c/li\u003e\n\u003cli\u003eScheier, E., \u0026amp; Kittner, N. (2022). A measurement strategy to address disparities across household energy burdens. \u003cem\u003eNature Communications\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(1), 288. https://doi.org/10.1038/s41467-021-27673-y \u003c/li\u003e\n\u003cli\u003eTeeuwen, A. S., Meyer, M. A., Dou, Y., \u0026amp; Nelson, A. (2022). A systematic review of the impact of food security governance measures as simulated in modelling studies. \u003cem\u003eNature Food\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e(8), 619-630. https://doi.org/10.1038/s43016-022-00571-2 \u003c/li\u003e\n\u003cli\u003eVandyck, T., Della Valle, N., Temursho, U., \u0026amp; Weitzel, M. (2023). EU climate action through an energy poverty lens. \u003cem\u003eScientific reports\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(1), 6040. https://doi.org/10.1038/s41598-023-32705-2 \u003c/li\u003e\n\u003cli\u003eVivier, L., Mastrucci, A., \u0026amp; van Ruijven, B. (2025). Meeting climate target with realistic demand-side policies in the residential sector. \u003cem\u003eNature Climate Change\u003c/em\u003e, 1-8. https://doi.org/10.1038/s41558-025-02348-4 \u003c/li\u003e\n\u003cli\u003eVereniging van Nederlandse Gemeenten. (2025). \u003cem\u003eDatavoorziening Energietransitie Gebouwde Omgeving (DEGO)\u003c/em\u003e. Retrieved October 28, 2025 from https://dego.vng.nl/?label=topo\u0026amp;layer=layer0#13.98/52.08743/4.32195 \u003c/li\u003e\n\u003cli\u003eWollburg, P., Hallegatte, S., \u0026amp; Mahler, D. G. (2023). Ending extreme poverty has a negligible impact on global greenhouse gas emissions. \u003cem\u003eNature\u003c/em\u003e, \u003cem\u003e623\u003c/em\u003e(7989), 982-986. https://doi.org/10.1038/s41586-023-06679-0 \u003c/li\u003e\n\u003cli\u003eXiao, Y., Feng, Z., Li, X., \u0026amp; Wang, S. (2024). Low-carbon transition and energy poverty: Quasi-natural experiment evidence from China\u0026rsquo;s low-carbon city pilot policy. \u003cem\u003eHumanities and Social Sciences Communications\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(1), 1-18. https://doi.org/10.1057/s41599-023-02573-2\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8986845/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8986845/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Energy poverty policies are now mandated by European Union regulations including energy renovations, energy coaching and energy subsidy interventions within homes. Yet, the long-term effects of such policies are usually only predicted and not evaluated. Thus, we visited 203 homes and used mixed methods to evaluate effects of these different policies implemented from 2020-2025. Subsidies have no long-term impact. Coaching reduces monthly electricity by 56kWh and gas consumption by 30.7m3 and percentage of income spent on energy, groceries and rent together by 3.48%. However, renovations lower gas (59.4m3) and energy bills (€110) with stable indoor temperatures. Lidar scans show relationships between indoor spatial characteristics and energy bills. Sociodemographic data finds that homes which were still energy poor post-policy, were more likely to live on social allowances and have chronic illnesses.","manuscriptTitle":"Energy Poverty Policy Predictions (EP3)","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-23 08:25:17","doi":"10.21203/rs.3.rs-8986845/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"e24bb0a4-e156-4060-8d8f-8a311d858c60","owner":[],"postedDate":"March 23rd, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"This content is not available.","date":"2026-05-02T04:11:34+00:00","index":1,"fulltext":"This content is not available."}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":64951476,"name":"Scientific community and society/Energy and society/Energy policy"},{"id":64951477,"name":"Scientific community and society/Energy and society/Energy security"},{"id":64951478,"name":"Scientific community and society/Social sciences/Economics"}],"tags":[],"updatedAt":"2026-03-23T08:25:20+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-23 08:25:17","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8986845","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8986845","identity":"rs-8986845","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}