1 Introduction

Increased energy efficiency is one of the initiatives for reducing greenhouse gas emissions (Hesselink & Chappin, 2019). Its impact on climate change and the residential sector is a key contributor to reaching internationally defined climate objectives. The greater adoption of energy efficiency by households is required to counteract the consequences of a globally growing population and rising energy consumption. Energy efficiency is becoming increasingly important in Australia because, like the rest of the world, it strives to provide sustainable energy (Gerarden et al., 2017). Australia ranks 18th among the world’s 25 largest energy users, and its energy efficiency is low (Subramanian et al., 2022).

A particular issue is the proportion of low-income households,Footnote 1 who are ill-equipped to adopt energy-efficiency measures for several reasons. Prior research has extensively addressed diverse aspects of the issue, encompassing user experiences and comfort in low-energy housing. These highlight the significance of policy frameworks and incentives for energy savings (Berry et al., 2022; Li et al., 2022; Sherriff et al., 2019) and explore the complex relationship between energy efficiency, indoor comfort, health and affordability within housing policies. The barriers to adoption are also underscored, particularly in addressing lower-income households’ unique challenges and proposing policy adjustments to promote equitable access to energy-efficient solutions (Liu & Judd, 2017; Liu et al., 2019c). However, some studies have a limited direct focus on low-income households, along with a need for more rigorous causal evidence linking energy efficiency to health outcomes (Daly et al., 2018) and greater consideration of regional variations (Liu & Judd, 2019). While previous research provides valuable insights into the complexities of energy-efficient measure adoption in low-income households, there remains room for more targeted investigations and empirical evidence to inform effective policies and practices.

Accordingly, the present study aims to (1) examine the current energy-efficiency level of low-income households and the barriers they face in Australia and (2) evaluate the impact of energy-efficiency barriers on their adoption of energy-efficiency measures (EEM). This involves four hypotheses to determine the current level of EEM, the effect of demographic variables on EEM and barriers faced by low-income households, and how EEM adoption is affected by barriers faced by the households. A questionnaire survey of 212 low-income Australian households is used to develop a model for determining the link between barriers and energy-efficiency adoption and test the hypotheses. In finding that financial constraint is the main barrier to these households adopting EEM, recommendations are made to formulate evidence-based strategies for policymakers to overcome this to improve their standard of living and positively impact their health. This study provides a significant theoretical contribution to the existing body of knowledge regarding energy-efficiency barriers’ present scope and nature. The model links barriers and energy-efficiency adoption that may be utilised for future research in similar contexts.

The paper is organised as follows. The literature review in Sect. 2 offers an overview of the current literature relating to energy-efficiency barriers and the adoption of energy efficiency in low-income households, concluding with the proposal of four hypotheses. The third section explains the method used to achieve the research aims, then the questionnaire design, before presenting the data collection and analysis process. In the fourth section, the results of the survey questionnaire and hypothesis tests are presented, followed in the fifth section by a discussion of the results. The study’s limitations are identified in the sixth section. Finally, a summary is provided and the prospects for future research identified.

2 Literature review

2.1 Energy efficiency and housing performance

2.1.1 User experiences and comfort in low-energy housing

The pursuit of energy efficiency and sustainable housing has attracted significant attention due to its potential to mitigate environmental impacts and improve occupant comfort. This involves considering user experiences and comfort in low-energy housing, focusing on the impact of low-energy housing on occupants’ well-being, and the crucial role of user satisfaction in promoting the adoption of energy-efficient measures. Of particular relevance are three key studies that shed light on the experiences and comfort of occupants living in low-energy homes.

The study by Berry et al. (2022) addresses the challenges posed by historical market failures in delivering sub-optimal housing from an environmental and economic perspective. This emphasises the role of policy instruments, including minimum energy standards and energy performance disclosure, in driving energy and carbon savings. Additionally, the research features the potential benefits of complementary regulation, where building regulations and disclosure policies work in tandem. By analysing housing energy assessments in Australia, the study shows that combining these policies can enhance performance outcomes beyond minimum standards. This approach holds promise for creating a conducive environment for adopting energy-efficient measures, ultimately contributing to more sustainable housing and cities.

In a complementary manner, Sherriff et al. (2019) contribute by exploring user experiences of ‘low-energy’ homes, specifically in cooling-dominated climates. Through an oral history approach and interviews with householders in Lochiel Park Green Village, South Australia, the study examines the intricate relationship between occupants, the building and energy use. Contrary to expectations, the research reveals that transitioning to low-energy homes does not eliminate occupants’ active involvement in maintaining thermal comfort. This study therefore highlights the complexity of adaptive comfort practices and underscores the need to consider user experiences central to sustainable housing transitions.

Furthermore, Moore et al. (2019) investigate the impact of tenure on occupant experiences of low-energy housing. Recognising the significance of household energy use determinants, the study explores how tenure influences occupants’ interactions with their dwellings. The research highlights the frustrations of social housing tenants who face limitations in modifying low-energy dwellings compared to owner-occupiers. This distinction highlights the importance of control and customisation for ensuring user satisfaction in low-energy housing scenarios.

These studies provide valuable insights into the user experiences and comfort outcomes of low-energy housing. While Berry et al. (2022) underline the role of policy instruments in driving energy savings, Sherriff et al. (2019) highlight the complexity of adaptive comfort practices and Moore et al. (2019) emphasise the role of tenure in shaping occupant experiences.

2.1.2 Building energy standards and transition challenges

The transition towards sustainable housing and improved energy efficiency is a complex process that involves addressing building energy standards and overcoming various challenges. Berry et al. (2019) provide valuable insights into the role of building energy standards and their impact on housing performance in the Australian context. The study highlights the significance of understanding energy standards and their potential influence on housing options. Although the research does not directly address low-income households, it joins with the the outcomes of the Australian Housing Conditions Dataset (Baker et al., 2022), (which shows low-income housing tend to live in poorer condition/less energy efficient dwellings) and Australian Institute of Health and Welfare’s (2021), National Housing Survey in underscoring the relevance of building energy standards to adopting energy-efficient measures in such households. Furthermore, the analysis of the thermal performance of Australian housing underlines the importance of regulatory mechanisms in driving energy-efficiency improvements. Building energy regulation can be a powerful instrument for achieving improved performance outcomes, which could affect low-income households’ access to more energy-efficient housing options.

From a complementary perspective, Moore and Doyon (2023d) comprehensively explore the sustainable housing transition and its challenges. They effectively address the environmental, social and financial challenges associated with housing provision, shedding light on the need for an urgent transition towards sustainable housing. Additionally, they emphasise the broader implications of sustainable housing and present sustainability transitions theory as a robust framework for facilitating this transition. By discussing contemporary case studies encompassing various aspects of sustainable housing, from innovative financing to circular housing, the authors underscore the multifaceted nature of the transition process.

Expanding on these insights, Moore and Doyon (2023a, 2023b, 2023c) provide further insights into the prospects for a sustainable housing transition. Examining the challenges that need to be urgently addressed consistently highlights the importance of drawing upon sustainability transitions theory. They also stress the need for ongoing policy, practice and research innovations to facilitate the transition towards sustainable housing effectively.

2.2 Policy, health and well-being

2.2.1 Policy frameworks and incentives

The housing sector has emerged as a critical arena for achieving sustainable development goals. In this context, exploring policy frameworks and incentives to facilitate the transition towards sustainable housing is important. Li et al. (2022) explore policy frameworks and incentives for zero-carbon housing in Victoria, Australia. They propose a comprehensive framework to reduce carbon emissions in the residential sector. Despite global recognition, more progress has yet to be made in Australia. The study uses a rigorous three-phase approach to address the issue, including a literature review, stakeholder engagement and survey distribution. The study highlights the financial challenges in promoting zero-carbon housing adoption, highlighting builders’ lack of experience and the need for financial incentives from state governments to overcome higher capital costs. It emphasises the importance of policy frameworks that offer economic benefits to promote energy-efficient housing solutions. The research informs the development of a proposed policy framework for Victoria and other regions facing similar challenges.

While policy frameworks and incentives are necessary for transitioning to sustainable housing, the literature also highlights the importance of assessing the impact of such transitions on occupants’ well-being. Berry et al. (2019) contribute to this aspect by questioning the prevalent metrics used to evaluate the benefits of low-energy housing. The research shifts the focus from abstract energy and environmental outcomes to the personal user experiences of occupants living in purpose-built low-energy homes in the UK. The findings underline the deeply personal nature of these experiences, which are closely tied to health, well-being and family outcomes. This research suggests that a holistic approach to assessing sustainable housing should encompass energy savings and the human experiences and quality of life it affords.

Meanwhile, Moore and Doyon (2023d) present a comprehensive perspective on the urgency of transitioning to sustainable housing. The housing sector’s role in contributing to greenhouse gas emissions and environmental impact is underscored, emphasising the need for a swift and comprehensive shift towards low-carbon housing.

2.2.2 Indoor comfort, health and affordability

Integrating energy efficiency, indoor comfort and affordability within housing policies is a complex and multifaceted endeavour crucial for ensuring the well-being of residents, especially those from low-income backgrounds. In this context, Daly et al. (2018) emphasise the significance of mainstreaming low-carbon retrofits in social housing. They recognise the potential for a large-scale, aggregated approach to promote energy efficiency within Australia’s social housing sector. Notably, the authors highlight the vulnerability of low-income occupants, particularly social housing tenants, to energy price fluctuations and extreme weather conditions. Moreover, the study underlines the importance of addressing specific barriers low-income households face in accessing energy-efficient improvements. Additionally, it acknowledges the co-benefits of energy-efficiency interventions, including potential health benefits, and calls for rigorous research to establish causal links between energy efficiency and health outcomes.

In a related vein, Daniel et al. (2020) focus on the critical issue of energy hardship among low-income renters in Australia. This research highlights the prevalence of energy hardship and explores strategies and policy actions to alleviate its impact. Given that up to 40% of Australian renters experience energy hardship, this study underscores the financial challenges that low-income households face in maintaining adequate energy access. This theme directly addresses the economic aspect of promoting energy-efficient measures in low-income housing by examining solutions to enhance energy affordability.

Understanding the relationship between cold housing conditions and health effects is pivotal in promoting energy-efficient measures that improve indoor comfort. In this regard, Daniel et al. (2020) reveal that cold housing is not limited to countries with severe winters; even mild-climate countries like Australia experience this issue. The study presents evidence of unacceptably low indoor temperatures experienced by households and highlights the immediate health implications of such conditions. This study underlines the importance of indoor thermal comfort and stresses the need to address cold housing as a real and pressing concern for households’ well-being.

2.3 Programmes, adoption barriers and equity

2.3.1 Programmes and policy implications

Liu et al. (2019b) stress the need for effective policy options to combat energy poverty among low-income households. Their systematic review examines the mix and effectiveness of policy and programme options for improving energy efficiency in homes inhabited by low-income households. This comprehensive approach acknowledges low-income families’ unique challenges and the need for tailored solutions. Liu and Judd (2017) shed light on the challenges lower-income groups face in achieving low-carbon living. Through focus group discussions and stakeholder engagement, they identify financial and non-financial barriers that impede the uptake of energy-efficient technologies and advocate policy adjustments to promote equitable access and uptake of low-carbon living solutions. Liu et al. (2017a) further examine the impact of carbon reduction programmes on lower-income households. By assessing the barriers these households face in reducing carbon consumption and analysing the effectiveness of existing programmes, the study reveals the complexities surrounding carbon reduction among this demographic. This work highlights the need for targeted and effective policies that align with the diverse circumstances of lower-income households.

On the other hand, Liu et al. (2019c) examine the quality of housing for low-income households and propose policy avenues for improvement. The study advocates for reforms in building standards, empowerment of tenants and regulators, and addressing split incentives by examining housing quality issues and involving various stakeholders in a policy workshop. These reforms aim to ensure that housing quality is prioritised and improved across different income groups. Liu et al. (2017b) contribute by addressing barriers and potential policy avenues related to carbon reduction programmes for lower-income households. Their findings underscore the need for comprehensive and informed policy solutions to effectively bridge the gap between carbon reduction goals and equitable outcomes for lower-income individuals.

The significance of policy reform and empowerment strategies is also evident throughout the literature. For instance, Liu and Judd (2019) investigate regional variations in the experiences of energy poverty across Australia, stressing the need for localised responses to energy and housing policies. This approach recognises the nuanced impact of climate conditions, local policies and housing quality on energy poverty prevalence.

2.3.2 Barriers, housing quality and health relationships

Mansour et al. (2022) present an updated glossary describing the critical relationship between housing and health. This comprehensive glossary defines key terms and concepts that describe the housing-health nexus, focusing on affordability, suitability and security. The study emphasises the potential of housing to either support or harm health, providing a foundational understanding of the broader connections between housing and well-being. Daniel et al. (2018) explore the concept of housing affordability stress (HAS) and its impact on well-being. By analysing the relationship between HAS and material deprivation, the study reveals that both indicators capture distinct aspects of the experience of housing affordability problems. Understanding the well-being implications of housing affordability is essential, especially for low-income households, as energy-efficient measures can clearly contribute to lower utility costs and potentially alleviate financial stress.

Similarly, Liu and Judd (2017) and Liu et al., (2017a, 2019b) all address the importance of programmes and policies in enhancing energy efficiency and reducing energy poverty among low-income households. These studies underscore the significance of tailored policies to address housing quality, affordability and access to information. The research demonstrates the need for effective policy options considering lower-income households’ financial and non-financial barriers.

Liu and Judd (2016, 2023) and Liu et al. (2019a) focus on barriers to the adoption of energy-efficient practices and technologies. These studies feature financial and non-financial obstacles that hinder lower-income households’ transition to low-carbon living. The research underlines the significance of addressing these barriers to promote equitable access to energy-efficient solutions and improve overall well-being.

Liu and Judd (2016) and Liu et al., (2017a, 2019c), on the other hand, underscore the importance of housing quality in the context of energy-efficiency adoption. These studies emphasise the role of housing quality and tenure in influencing the ability of lower-income households to adopt energy-efficient practices. Poor housing quality is identified as a barrier to effective energy reduction initiatives. The research highlights the need for policies that address housing quality issues, advocate for better housing standards, and consider the financial and qualitative impacts on low-income households. Likewise, Liu and Judd (2019) and Liu et al., (2019a, 2019c) examine the profound impact of housing quality and energy poverty on the health and well-being of low-income households. These studies reveal the negative consequences of inadequate housing on health, financial stability and overall quality of life. The research emphasises regional variations in energy poverty experiences and the need for localised policy responses. The studies highlight the critical relationship between housing quality, energy efficiency, health and well-being for low-income households.

In addition, Liu and Judd (2016) and Liu et al. (2019a) focus on barriers to adopting energy-efficient practices and technologies. These studies feature financial and non-financial obstacles that hinder lower-income households’ transition to low-carbon living. The research underlines the significance of addressing these barriers to promote equitable access to energy-efficient solutions and improve overall well-being.

2.4 EEM types

The implementation of three EEMs is examined: 1) purchasing energy-efficient appliances, 2) investing in installing renewable energy, and 3) investing in energy-efficient renovations. There are several reasons for considering these three types of EEM, even though they are regarded as one-shot/one-off measures/behaviours or financial and non-financial factors that influence the reduction of energy consumption. First, investments in energy-efficient renovations and renewable sources usually require a great deal of financial commitment, time and the technical skills of professional contractors (Maller & Horne, 2011). On the other hand, unlike energy-efficient renovation measures, energy-efficient appliances are less expensive. They can easily be installed without a professional contractor’s help, which implies that they are more likely to be adopted. Second, while most appliances are not fixed to the house structure, energy-efficient renovations are. This means that if a family plans to move to another house, they may only be able to recoup their investment through a rise in property value (Achtnicht & Madlener, 2014).

2.4.1 Purchasing energy-efficient appliances

Azimi et al. (2023) highlight the challenges faced by low-income households in Australia in adopting energy-efficient practices, including financial constraints, decision-making difficulties, lack of information and split incentives. This aligns with Liu et al. (2019a), who emphasise the affordability of energy-efficient household products and the lack of reliable information. Financial factors significantly influence appliance purchasing decisions. Ren et al. (2021) highlight energy savings and reduced bills by replacing old refrigerators with energy-efficient models. Galarraga et al. (2011) explore consumers’ willingness to pay for energy-efficient appliances, emphasising price elasticities of demand. Wang et al. (2017) evaluate subsidies’ effectiveness in influencing consumer purchase intentions, emphasising policy incentives.

Moreover, behaviour change significantly influences energy-efficient practices in low-income households. James and Ambrose (2017) found that combining retrofit and behaviour change interventions leads to the most significant improvement. Trotta (2018) identified factors influencing energy-saving behaviours and investments, emphasising consumer and dwelling characteristics. The literature emphasises the importance of information provision and consumer perception in energy-efficiency labelling. Hammerle and Burke (2022) study the impact of energy-efficiency labels on appliance choices and behaviour, especially for low-income households. Zhang et al. (2020) study the role of perceived value in purchasing energy-saving appliances.

Furthermore, policy interventions are crucial for promoting energy efficiency in low-income households. Conway and Prentice (2020) suggest that price changes could influence electricity consumption, but more research is needed. Berry et al. (2022) suggest that building energy standards and disclosure regulations could improve environmental and economic outcomes. Tan et al.’s (2017) study highlights the influence of sociodemographic and environmental factors on consumers’ willingness to pay for energy-efficient appliances, providing insights into the potential impact on low-income Australian households’ purchasing decisions.

In summary, low-income Australian households’ purchase of energy-efficient appliances is influenced by financial constraints, behaviour change, consumer perceptions and policy interventions. Key factors include financial considerations, behaviour change interventions, information provision and policy incentives.

2.4.2 Investing in installing renewable energy

Financial barriers emerge as a recurring theme in previous studies. For example, Azimi et al. (2023) underscore the importance of recognising low-income households’ economic limitations in energy-efficiency programmes. Similarly, Ding (2013) highlights the role of government incentives and subsidies in making renewable energy technologies financially accessible to low-income segments.

Rosewarne’s (2022) work introduces the concept of ‘prosumers,’ marking a shift in household engagement with energy systems, empowering households to consume and contribute to the grid, and challenging traditional energy dynamics. In a different context, Streimikiene (2022) examines barriers and policies within the European Union, emphasising the need for targeted policy measures to combat energy poverty and facilitate low-carbon energy transitions.

Tailored information campaigns emerge as a promising strategy to bridge financial gaps and promote sustainable energy behaviours among low-income households. Podgornik et al. (2016) highlight the efficacy of customised consumption feedback in driving energy-efficient behaviours. Similarly, Hall et al. (2013) underline the significance of group discussions and social networks in fostering energy-saving actions.

Collectively, these studies illuminate the complex nature of investing in renewable energy for low-income households, suggesting that innovative policy measures and incentives can effectively overcome financial constraints. Moreover, the rise of prosumers challenges traditional energy paradigms, while tailored information campaigns enhance energy literacy and promote sustainable behaviours.

Further exploration by Rosewarne (2022) examines Australian municipalities’ strategic interventions in renewable energy adoption. On a global scale, Kiprop et al. (2019) shed light on Kenyan household consumers’ proactive responses to renewable energy sources. Additionally, Liu and Judd (2016) investigate the barriers faced by lower-income Australian households in reducing carbon consumption, underlining economic limitations as a hindrance to investments in energy-efficient measures, including renewable energy systems.

In short, the significance of addressing financial barriers is identified as the shift toward prosumers and the potential of tailored information campaigns. However, it is recognised that gaps in understanding persist and that the growing demand for sustainable energy necessitates further research to design effective policies and interventions that empower low-income households to adopt renewable energy for a more sustainable future.

2.4.3 Investing in energy-efficient renovation

Daly et al. (2018) highlight the potential of low-carbon retrofits in social housing, highlighting the need for widespread renovations due to centralisation. Building on this perspective, Li et al. (2022) propose adaptations in Victoria, Australia, promoting zero-carbon housing. Aligning with this trajectory, Moore et al. (2019) argue that enhancing building energy regulations is essential for a low-carbon future. Echoing this sentiment, James and Ambrose (2017) show that combining retrofit interventions with behaviour change improves energy consumption outcomes.

An overarching theme in multiple papers highlights the importance of behaviour change in energy-efficient renovations. In this vein, Berry et al. (2022) discuss the impact of building energy standards and disclosure regulations on housing energy assessments. Correspondingly, Trotta (2018) explores factors influencing energy-saving behaviours and investments in British households. Wilson et al. (2015) advocate for a situated approach in line with these insights.

The recurrent discussion on the effectiveness of retrofit interventions is a recurring topic. Moore et al. (2019) reveal that most housing meets only minimum standards, highlighting the need for more ambitious measures. Complementing this, Liu et al. (2019a) explore carbon reduction programmes for lower-income households in Australian cities, identifying barriers to reducing carbon consumption. In consonance with these observations, Charlier and Legendre (2021) stress the importance of retrofitting dwellings for energy efficiency, highlighting the significance of affordability and efficiency policies.

2.5 Hypotheses

The identification and systematisation of energy-efficiency adoption barriers have been approached from multiple perspectives (Painuly & Reddy, 1996); however, the categories of barriers were broadly generalised by Sorrell et al. (2000) in one of the major studies on the theory of barriers, which provides the most extensive classification of energy-efficiency barriers. The types of energy-efficiency adoption barriers vary and affect households to various degrees. Split incentives, for instance, are structural and largely outside the domain of influence of the household. Another barrier is financial: households may need more funds to invest in new energy technology or believe the initial cost is prohibitively expensive. When deciding, homeowners may have priorities other than energy efficiency, be unaware of energy efficiency or be unwilling to adapt, resulting in split incentive barriers. Finally, some barriers arise because of a need for more information or awareness. When deciding whether to embrace a new energy technology, people may consider the activities of their social counterparts.

The taxonomy of barriers was based on categories often used in the literature concerning energy efficiency. The Sorrell taxonomy covered most issues preventing energy-efficiency adoption (Sorrell et al., 2011). Although the most important factors that prevent households from accepting technical EEM are those linked to decision making, information and finance, there is a significant difference in the degree to which these barriers influence the adoption decisions of low-income households. Furthermore, another major barrier to adopting energy conservation measures by such households is social-economic constraints and split incentives (Michelsen & Madlener, 2010; Murphy, 2014; Pape, 2013; Trotta, 2018; Wilson et al., 2015). Each barrier in this study was considered a hypothesis that might explain why energy efficiency is neglected in the residential sector.

To a large extent, these findings illustrate that the common barrier that thwarts potential energy-efficiency improvement is related to information (Bukarica & Tomšić, 2017). One of the information barriers is scepticism or lack of awareness of existing resources. In small towns and rural communities, the residents’ decisions to improve their homes are usually influenced by verbal recommendations from messengers (third parties without any particular motive of selling services or products) and neighbours. Residents may be unaware of existing financial assistance programmes. Erroneous perceptions of households towards energy savings or consumption were revealed by Ameli and Brandt (2015), who suggested that residents possessed little or no knowledge of the efficiency of various energy-saving techniques. This lack of knowledge may affect investment decisions regarding energy conservation and renewable energy. According to Wilson et al. (2015), when households are unaware of such energy-saving attributes as prospective savings, they are less likely to invest in energy-saving measures (Gillingham & Palmer, 2014). Davis and Metcalf (2016), Newell and Siikamäki (2014) and Ward et al. (2011) all found that the information barrier causes under-investment in energy efficiency. This suggests that an information barrier might lead to a lower rate of adoption of energy-saving measures (Ameli & Brandt, 2015; Prete et al., 2017), prompting the hypothesis.

H1

The adoption of EEM is negatively associated with the information barrier.

The financial barrier is a major impediment to adopting EEM. A household’s inability to obtain enough capital to invest in such EEM as renewables, energy-efficiency renovation or energy-efficient appliances is the consequence of financial factors (O’Malley et al., 2003). The huge upfront cost required for energy upgrades makes it difficult for low-income households to optimise their energy consumption (MacDonald et al., 2019). For many years, financial barriers, especially the lack of funds or discontentment with the payback period, have marred households’ decisions to improve their energy performance, prompting the second hypothesis.

H2

EEMs are negatively related to the financial barrier.

Past studies have shown that people’s decision making is affected by their perceptions and, most of the time, they cannot decide on effective ways to conserve their energy consumption, even when they have access to the necessary information (Wilson et al., 2015). As Sorrell et al. (2000) explain, human decision making sometimes involves discovering and selecting optimal alternatives; however, in most cases, it deals with finding and choosing satisfactory alternatives. Compared to satisfactory decisions, optimal decisions require a more complex process involving several orders of magnitude. According to Murphy (2014), measuring implementation creates too much stress, which is a factor that contributes to the failure of the EEM adopted. Similarly, in a study that examined Dutch households’ reluctance to adopt EEM, the respondents refused to take any action towards installing different technical energy measures due to the time and effort needed to carry out such renovation measures in existing buildings (Murphy, 2014). Sorrell et al. (2000) identified technical risk as another essential factor that affects decision making to enhance energy efficiency. Technical risk concerns individual technologies’ technical performance and unreliability (O’Malley et al., 2003). Bukarica and Tomšić (2017) highlighted that decision-making barriers hurt the implementation of EEM. The third hypothesis is therefore.

H3

There is a negative relationship between adopting EEM and the decision-making barrier.

According to Pape (2013), the tenant-landlord split incentive is a persistent and significant market failure identified as a primary impediment to households’ decisions to implement energy-saving measures as it hampers investment in energy-efficient fixed appliances and renovation. Ameli and Brandt (2015) pointed out that rental apartment owners are less likely than owner-occupied apartments to invest in energy-saving measures. As Wilson et al. (2015) note, low-income homeowners and tenants are less likely to invest than high-income households and homeowners. In a survey conducted by Murphy (2014), the failure of energy-efficiency adoption measures was attributed to uncertainty about the period of residence at a particular dwelling. Based on empirical results, therefore, split incentives were found to be a strong predictor of a household’s energy-efficiency adoption, suggesting the hypothesis.

H4

The implementation of EEM has a negative relationship with the split incentives barrier.

Figure 1 represents a theoretical model highlighting the factors that affect EEM in low-income households. The model comprises four constructs (information barrier, financial barrier, decision-making barrier and split incentives) identified in the literature and shows how these constructs are related.

Fig. 1
figure 1

Research model Note. H1, H2, H3 and H4 denote the four hypotheses concerning the negative effects of barriers on the adoption of EEM by low-income residents

Full size image

3 Methodological approach

3.1 Questionnaire design

The questionnaire was designed to explore the relationship between energy-efficiency barriers and adopting EEM. The questionnaire is divided into three sections. Section 1 has 13 questions eliciting information concerning housing features, home tenure and household characteristics; Sect. 2 has 11 questions concerning EEM adoption and knowledge of energy efficiency; and Sect. 3 has 26 questions regarding financial, information, decision making and divided incentives obstacles. The information is intended to validate the factors and barriers in the literature. The last section collects information about the households’ energy-efficiency plans for energy savings.

3.2 Administering the survey

An online questionnaire survey was conducted with low-income households in Australia. It was marketed by a research panel firm with vast networks of low-income families. Based on information from the Australian Taxation Office (2021), low-income households were defined as those with a taxable income of less than AUD 69,999 (AUD 1 = USD 0.64 on 24 October 2022). The research panel forwarded the online questionnaires to low-income Australian households and those who passed the screening questions to gauge participant eligibility were invited to complete the questionnaire. The first page of the survey website detailed which low-income households were eligible, and the second page required the household’s eligibility to be verified. For this, first, to be a household resident; second, to be at least 18 years old; and third, to be the household’s financial decision maker (the individual responsible for paying the rent and utilities). The questionnaire could only be completed if these questions were answered. It was essential to capture the appropriate demographic category of the respondents in terms of their income and age bracket, including the power they had in making decisions in their households. The research panel, therefore, ensured that the online survey was sent to the correct target group to address the most difficult challenges and produce valuable results. After completing the online survey, the participants were given incentives most appealing to low-income families.

The respondents were asked to identify and rank several obstacles (both financial and otherwise) that impeded their household’s adoption of energy-efficiency programmes and, in the end, indicate whether they planned to adopt an energy-efficiency programme. Information about the household members’ age, gender, education, employment status, housing situation and personal income was collected. Other relevant information collected included the property characteristics.

The survey was performed in May 2021 and adhered to all ethical procedures. It was randomly conducted across cities in Australia to ensure that the findings were generalisable and accurate to the greatest possible degree. Two hundred and twelve completed questionnaires were collected. This was considered sufficient to obtain valid results as, according to Hair (2009), any common estimation procedure requires a sampling size of at least 200 for the intended structural equation modelling (SEM) and standard statistical analysis.

To obtain valid results, any common estimation procedure requires a sampling size of at least 200, as recommended for SEM and standard statistical analysis (Hair, 2009). A sample of 212 households was therefore randomly selected to represent the different demographic groups for this study. The study, which was performed in May 2021, included an online survey and adhered to all ethical procedures.

3.3 Data analysis

Descriptive statistics (such as mean scores) and inferential statistics (such as analysis of variance (ANOVA)) are used to help understand the current level of EEM, the effect of demographic variables and barriers faced by low-income households. Studies have shown that the precision of the operationalisation of the model’s construct and hypothesis specification is higher in SEM, which can be applied within and across sample groups, and justified its application here. Maximum likelihood is employed as the estimation method.

Previous research established that the multiple correlation coefficient within the original sample used to assign values to regression weights provides an optimistic sense of the regression equation’s predictive effectiveness when applied to subsequent data. For this study, the cross-validation approach is used to study this phenomenon. Cross-validation is a valuable tool for selecting the best model among competing structural models (Yuan et al., 2002). This involves taking a calibration sample and validation sample from the same population. For example, the calibration sample calibrates the regression equation or assigns values to the regression weights. The weighted linear composite of predictors is associated with the criteria using previously calibrated weights in a second sample; the validation sample. The resulting cross-validation index provides a genuine view of the linear composite of tests’ predictive performance.

A split sample technique is used to verify the model’s stability. Sub-samples for calibration and validation are chosen at random. Tight cross-validation is performed by setting all parameters throughout the calibration and validation samples to determine whether a model is replicated in samples other than those from which it was acquired.

4 Results

4.1 Respondent profiles

Table 1 shows the demographics of the respondents. Most of them are at least 55 years old, with education at diploma level or below. Approximately half are retired homeowners living with one to two people in the household. More than one-third are tenants.

Table 1 Demographics of the 212 Australian low-income household respondents
Full size table

4.2 Descriptive analyses

Table 2 shows the descriptive statistics for energy-efficiency adoption and the barriers faced. In general, the current level of energy-efficiency adoption by the low-income household is low, with an average of 3.4 on the 5-point Likert scale. The respondents are more receptive to energy-efficient devices but less willing to install renewable energy in their homes. Regarding the four types of barriers the households faced for energy-efficiency adoption, financial barriers scored the highest (3.66). High upfront costs are rated as the most important financial barrier.

Table 2 Means and standard deviations of the factors affecting the adoption of EEM
Full size table

4.3 Empirical model testing

Table 3 shows the goodness-of-fit statistics, indicating that the model adequately fits the calibration and the validation sub-samples.

Table 3 Goodness-of-fit for the structure model
Full size table

Figures 2 and 3 provide the full structural equation models for calibrating and validating the sub-samples.

Fig. 2
figure 2

Structural equation model of the calibration sub-sample of 2021 Australian low-income households Note: The arrows show the direction of influence in the interrelationships between EEM adoption (EE), the four barriers and the variables in Table 2

Full size image
Fig. 3
figure 3

Structural equation model of the validation sub-sample of 2021 Australian low-income households Note: as Fig. 2

Full size image

The KMO measure of sampling adequacy is 0.921 above the commonly recommended value of 0.6, Bartlett’s test of sphericity is significant χ2 (325) = 4000.380, p < 0.001) and all the communalities are above 0.4, further confirming that the data are suitable for factor analysis.

The initial eigenvalues indicate that the first five factors have eigenvalues just over unity. Applying parallel analysis to validate the number of factors results in four factors in total. Using the component correlation matrix to identify the correlation between the components indicates Principal Components Analysis and Varimax with Kaiser normalisation as the most appropriate method.

Three items are excluded because they do not contribute to a simple factor structure and fail to meet the minimum criteria of having a primary factor loading a minimum of 0.4 or cross-loaded. For example, doubts about transaction costs have factor loadings between 0.5 and 0.6 on both financial and decision making. Lack of information about energy-efficiency programmes has factor loadings between 0.4 and 0.7 on both financial and information. In contrast, physical distance from resources had similar factor loadings between 0.4 and 0.5 on information and decision making. The remaining items have primary loadings over 0.5.

Cronbach’s alphas are moderate: 0.65 for energy-efficiency adoption (3 items), 0.9 for information (6 items), 0.8 for financial (7 items), 0.9 for decision making (7 items) and 0.7 for split incentives (2 items). Eliminating more items would not result in a significant boost in alpha for any of the scales. The mean of the items with their primary loadings in each component is used to construct composite scores for the remaining four factors. The skewness and kurtosis are within a reasonable range for assuming a normal distribution. Although varimax rotation is used, which indicates that the factors are not correlated, there is an average correlation of 0.5 between information and decision making. There are small correlations between the rest of the composite scores.

Overall, the indications are that financial has the most significant impact (0.45) on adopting energy-efficiency measures. The two factors of split incentives (0.18) and decision making (0.14) have a moderate impact, and information has an insignificant impact (0.9).

4.4 Hypothesis testing

The research hypotheses for this study were developed based on the literature. This section explores the nature of the four primary barriers to energy-efficiency adoption. Figure 4 shows an SEM of the calibration and validation sub-samples using Sorrell et al.’s (2004) taxonomy. The values represent the standardised path coefficients, with the values from the validation sub-sample in parentheses. The error terms and disturbances are excluded for clarity.

Fig. 4
figure 4

Empirically tested structural equation model Note. All paths are significant at the 0.01, 0.05 and 0.10 levels. Insignificant path (information) is excluded

Full size image

Table 4 shows the results of the hypothesis tests using the estimated path coefficient value, critical ratio (C.R. equal to t-value) and p-value. The standard decision procedures (t-values ≥ 1.96 and p < 0.05) are used to determine the path coefficient’s importance. However, some researchers in this field consider p < 0.10 to be a marginal level of significance (Anderson & Weitz, 1992; Kim & Frazier, 1997; Kwon & Suh, 2004). R2 (explanatory power) measures the model’s predictive accuracy, depicting the exogenous constructs’ combined effect on the endogenous construct (Hair, 2009). R2 ranges from 0 to 1, with higher values indicating better prediction accuracy (Hair, 2009). Although the explanatory power (R2) of the calibration sample (R2 = 0.54) and validation sample (R2 = 0.52) are very close, this comparison is important to shed light on any path affected.

Table 4 Results of the structural model and hypothesis tests
Full size table

Hypothesis 1

EEM adoption is negatively correlated with information barriers.

This model does not support the first hypothesis, which indicates that EEM adoption is negatively correlated with information barriers. The standardised estimated path coefficients for the relationship are not statistically significant in the calibration and validation samples (β = − 0.09, t = 0.04, p = 0.305). The hypothesised relationship between information barriers and EEM adoption is therefore rejected.

Hypothesis 2

Financial barriers negatively affect the adoption of EEM.

The standardised path coefficients from EEM adoption and financial barriers in the calibration and validation sub-samples are − 0.45 and − 0.47, respectively, which indicates that the financial barrier plays a key role in adoption of EEM (β = − 0.45, t = − 2.12, p < 0 0.01). When the financial barrier increases by one standard deviation, EEM adoption decreases by 0.45 standard deviations, which means a one unit increase in the financial barrier leads to an approximately 0.4 unit decrease in adoption of EEM.

Hypothesis 3

EEM adoption is negatively correlated with decision-making barriers.

As shown in Fig. 4, the path coefficient from decision-making barriers and EEM adoption in the calibration and validation sub-samples are − 0.14 and − 0.12, respectively, with a marginal level of significance (p < 0.10). This indicates that the decision-making barrier may play a moderate role in adopting EEM (β = − 0.14, t = − 1.85, p < 0.10). When the decision-making barrier increases by one standard deviation, EEM adoption decreases by 0.14 standard deviations, which means a one-unit increase in the decision-making barrier leads to an approximately 0.1 unit decrease in the adoption of EEM.

Hypothesis 4

The implementation of EEM has a negative relationship with split incentive barriers.

The estimated standardised β value in the calibration sample (− 0.18) and validation sample (− 0.13) are significant to provide support to hypothesis four (β = − 0.18, t = − 1.98, p < 0.01). The results indicate that split incentives are a key factor that affects EEM adoption. When the split incentives barrier increases by one standard deviation, EEM adoption decreases by 0.18 standard deviations, which means that a one-unit increase in the split incentives barrier leads to an approximately 0.1 unit decrease in the adoption of EEM.

5 Discussion

This section explores the implications of the study's findings concerning the formulated hypotheses, juxtaposing empirical results with the anticipated relationships between Energy-Efficiency Measure (EEM) adoption and identified barriers. While examining the hypotheses, it is crucial to consider the limitations acknowledged in the study, particularly concerning the capturing of decision-making processes and barriers through the questionnaire form.

Hypothesis 1

EEM adoption is negatively correlated with information barriers:

According to this hypothesis, adopting EEMs would be adversely linked with information barriers. However, the hypothesis testing findings do not support this association. The standardised calculated path coefficients are not statistically significant (β = − 0.09, t = 0.04, p = 0.305), indicating that this hypothesis should be rejected.

When compared to the literature, the data show an interesting distinction. The literature emphasises the relevance of information hurdles and a lack of trustworthy information as impediments to low-income families adopting energy-efficient practices (Azimi et al., 2023; Liu et al., 2019a). The hypothesis testing results, however, imply that knowledge obstacles do not have a statistically significant negative influence on EEM adoption. One possible explanation for this is that, while information barriers have been identified as an issue in the literature, they may be less prevalent or impactful in the unique context of the present study. Other variables, such as financial hurdles, decision-making barriers and divided incentives, may overcome the impact of information barriers.

These consequences underline the importance of developing a more comprehensive understanding of the context in which energy-efficient practices are implemented. While information hurdles are acknowledged to be significant, there may be other impediments to EEM implementation. This emphasises the importance of customising treatments and policies to the unique barriers most relevant in each situation. In addition, that the hypothesis testing results do not support the anticipated negative association, questions the significance of information barriers in the unique context of the present study. This prompts reflection on the usability of the questionnaire options and raises awareness of the potential limitations in capturing multifaceted decision-making processes. Notably, the discrepancy between literature emphasis and empirical findings underscores the complexity of the energy-efficiency landscape and suggests a need for nuanced questionnaire design to better capture the interplay of various factors influencing EEM adoption.

Hypothesis 2

Financial barriers negatively affect the adoption of EEMs:

According to this concept, financial hurdles impede EEM adoption. The hypothesis testing findings validate this link as evidenced by the standardised path coefficients for EEM adoption and financial obstacles (− 0.45 and − 0.47 in the calibration and validation sub-samples, respectively). These coefficients are statistically significant (t− 2.12, p 0.01), demonstrating that financial obstacles have a substantial effect.

When these findings are compared to the literature, they significantly agree. Previous research has continuously highlighted financial restrictions as a major hurdle (e.g., Azimi et al., 2023; Ding, 2013). The research also emphasises the significance of government incentives and subsidies in making renewable energy technology affordable to low-income people (Ding, 2013). The consequence is that financial aspects heavily impact EEM adoption decision making.

The implications highlight the essential role of financial constraints in determining low-income families’ energy-saving habits. The statistically substantial inverse association implies that, as financial obstacles grow, so does EEM adoption. This is consistent with the notion that low-income households typically have limited financial resources – making upfront investment in energy-efficient appliances, renewable energy systems or renovations difficult, if not impossible. To address this, governmental interventions that reduce financial limitations or provide financial incentives could be critical in encouraging low-income households to adopt energy-efficient practices. Policymakers could also increase the use of EEMs and contribute to energy savings and improved living conditions by lowering financial obstacles.

Hypothesis 3

EEM adoption is negatively correlated with decision-making barriers:

According to this hypothesis, EEM adoption is adversely connected with decision-making hurdles. The hypothesis testing findings support this association to some extent, as evidenced by the standardised path coefficients for the relationship between decision-making obstacles and EEM adoption (− 0.14 and − 0.12, respectively, in the calibration and validation sub-samples). These coefficients are marginally significant (p 0.10), indicating that decision-making hurdles have a minor influence on EEM adoption.

When compared to the literature, these results indicate agreement with debates on the effect of decision-making obstacles on energy-efficient practices. The importance of decision-making challenges is highlighted in the literature (Azimi et al., 2023; Liu et al., 2019a). The importance of attitude modification in energy-efficient practices, especially in the setting of restorations, is consistent with debates on the influence of behaviour change interventions on energy consumption outcomes (James & Ambrose, 2017).

The implications suggest that, while decision-making hurdles may not be as powerful as financial barriers, they nevertheless significantly define EEM adoption. The marginal significance of the standardised coefficients in hypothesis testing implies that decision-making challenges, although to a smaller extent than financial obstacles, may contribute to a drop in EEM adoption. This conclusion emphasises the need for providing low-income households with accessible and practical information to help in decision making concerning energy-efficient practices. Effective communication and focused interventions that address decision-making constraints may result in higher EEM adoption. Furthermore, the marginal relevance shows that further study is needed to better understand low-income families’ distinctive decision-making processes in the context of EEM adoption.

Hypothesis 4

Implementing EEMs has a negative relationship with split incentives barriers:

According to this hypothesis, installing EEMs has a negative association with split incentives barriers. According to the estimated standardised values for the association between split incentives barriers and EEM adoption (− 0.18 in the calibration sample and − 0.13 in the validation sample), the hypothesis testing findings support this relationship. These statistically significant values (t = − 1.98, p 0.01) imply that split incentives barriers are important in EEM adoption.

When these results are compared to the literature, they are consistent with debates on the impact of divided incentive barriers on energy-efficient practices. The research emphasises the difficulty of divided incentives as a barrier to implementing energy-efficient solutions, particularly in cases where the property owner and renter have different motives and duties (e.g., Azimi et al., 2023). The importance of retrofit interventions and policy changes in enhancing housing quality and solving divided incentives is echoed in talks on upgrading houses for energy efficiency (Charlier & Legendre, 2021; Liu et al., 2019a; Moore et al., 2019).

The implications highlight the significant impact of split incentives barriers on EEM adoption. The inverse link between the two suggests that, as split incentives barriers grow, so does the adoption of energy-efficient solutions. This conclusion emphasises the necessity of matching property owners’ and renters’ incentives and motivations to stimulate the installation of energy-efficient practices in rental buildings. Policy interventions that bridge the gap between property owners and renters, such as incentive programmes that benefit both parties or rules encouraging energy-efficient upgrades, might assist in overcoming split incentives barriers. Furthermore, landlord-tenant involvement, clear communication and education campaigns can help to overcome this obstacle.

In summary, this section evaluates the study’s findings in connection with the four hypotheses, offering insight into the critical impediments to EEM adoption in low-income Australian homes. While the empirical data provide insufficiently convincing evidence for the predicted negative link between information barriers and EEM adoption, they highlight financial restrictions’ importance as a substantial deterrent. Decision-making hurdles (although having a minor impact) and split incentives are shown to be significant and influence adoption. The findings highlight the need for targeted interventions such as financial assistance for affordability, decision-making assistance, and aligning incentives between property owners and tenants. Such initiatives could successfully reduce obstacles and support the implementation of energy-saving solutions, encouraging sustainable practices in low-income housing environments.

6 Limitations

The study’s geographical focus solely on Australia introduces a limitation in directly extrapolating its findings to other countries and various cultural contexts. While a comparative study involving different countries could contribute to a clearer understanding, it is important to acknowledge that distinctive cultural nuances and policy variations might lead to varying outcomes in each context. In addition, the study’s scope is confined to investigating the energy-efficiency barriers encountered by low-income households. However, it needs to analyse the broader range of factors influencing energy-efficiency adoption comprehensively. As well as introducing some form of robustness estimation to validate the result of the study further, a more holistic understanding could be achieved by examining a wider spectrum of elements, such as drivers and critical success factors, that might enhance the adoption of EEM among this demographic. This would provide a more complete picture of the complex decision-making landscape. Moreover, by focusing solely on barriers, the study overlooks the potential counterbalancing effect of drivers that could mitigate the negative influence of barriers on energy-efficiency adoption. Expanding the investigation to explore the interplay between positive driving forces and existing barriers would offer a more nuanced perspective on the decision-making process of low-income households.

The conclusions drawn from this study are limited by the amount of data collected through survey responses and interviews. Although the study met the minimum requirement of 212 survey responses for conducting data analysis, a larger and more diverse dataset could enhance the representativeness and generalizability of the findings. A broader dataset would better capture low-income households’ multifaceted experiences and viewpoints. Similarly, although the qualitative insights gained from interviews provide depth to the study, the number of interviews conducted might only partially encompass the entire range of perspectives and experiences. Increasing the interview sample size could offer a more comprehensive qualitative understanding, allowing for more robust conclusions. Recognising the potential influence of social desirability bias on respondents’ answers is also important. Participants might provide responses that align with societal norms or expectations, impacting the accuracy of reported barriers and drivers and potentially affecting the reliability of the findings. Likewise, the demographic composition of the sample might only partially represent the diversity within the low-income household’s demographic. Age, education and employment status could significantly influence energy-efficiency adoption behaviours. A more comprehensive demographic representation would provide a more nuanced view of the various factors involved.

Finally, the study’s findings are within a specific period and might not account for potential shifts in attitudes, behaviours or external factors influencing energy-efficiency adoption. Longitudinal perspectives or broader temporal contexts offer a more dynamic understanding of the trends and changes over time.

7 Conclusion

In conclusion, this study sheds light on the complex terrain of energy-saving practices among low-income Australian homes. A detailed investigation of the correlations between Energy-Efficiency Measures (EEM) adoption and different hurdles offers useful insights. While the study sought to identify the elements that influence EEM adoption, the major findings indicate a complex interplay of barriers affecting this acceptance.

The data do not substantially support the first hypothesis, which proposed a negative relationship between information barriers and EEM adoption. This contrasts with prior studies, which emphasise information barriers, highlighting the need to consider contextual variables in their influence. The data substantially supports the second hypothesis, stressing the negative impact of financial obstacles, and is consistent with the literature’s persistent emphasis on low financial resources impeding energy-efficient practices. The third hypothesis, which proposes a negative relationship between decision-making obstacles and EEM adoption, receives some support, emphasising the significance of easily accessible and actionable information in promoting energy-efficient decisions. Finally, the data significantly confirm the fourth hypothesis, which indicates a negative link between split incentives barriers and EEM adoption, underlining the need for harmonising incentives in rental properties.

These findings have consequences for both policy and practice. Because financial restrictions are such a significant obstacle, tailored actions that ease these costs are required, possibly through financial incentives and support programmes. Furthermore, the modest effect of decision-making obstacles emphasises the importance of specialised information campaigns and easily available resources to help make decisions. The significant impact of split incentives barriers highlights the need for novel measures that connect the interests of property owners and renters, eventually promoting EEM deployment in rental properties.

However, the study has some limitations. Methodological restrictions, sample characteristics and other unobserved influences might have influenced the results. Furthermore, the study’s emphasis on low-income Australian households may restrict its applicability to other situations. Future studies should examine specific subgroups of low-income individuals and investigate dynamic connections between different obstacles.

In summary, this study enriches our understanding of EEM adoption in low-income households, revealing the intricate interplay of constraints influencing energy-efficient practices. It offers insights for crafting targeted interventions, informing policy decisions, and advancing our understanding of sustainable practices by acknowledging the diverse roles of information, financial, decision-making, and split incentive obstacles. It also makes a significant theoretical contribution in elucidating the nuanced influence of various barriers, expanding beyond traditional conceptualizations. By examining the intricate relationships between information, financial, decision-making, and split incentive hurdles, this study contributes theoretically to the comprehension of energy efficiency dynamics in low-income housing contexts. As nations strive for greener and more energy-efficient futures, these findings make a crucial contribution to addressing energy efficiency disparities in low-income housing contexts.