Effects of trust and perceived benefits on consumer adoption of smart grid technologies: a mediation analysis

ABSTRACT

End-customers are expected to play an increasingly important role in the future of smart grids. However, customer adoption of smart grid technologies has been hindered by a lack of trust in the energy sector. This study investigates the relationships between customer trust in companies in the energy sector, the perceived benefits of smart grid technologies and customer intention to use smart grid technologies. We conducted a cross-sectional online survey of Finnish households, and using structural equation modelling, examined the effects of competence-based trust, integrity-based trust, perceived financial benefits and perceived environmental benefits. Our results show that trust, perceived benefits, and usage intention are positively associated with each other. Neither competence-based nor integrity-based trust had a direct effect on usage intentions, indicating that the effects of trust on usage intentions are fully mediated by perceived benefits. Therefore, increasing customer trust in can indirectly improve the adoption of smart grid technologies.

KEYWORDS:

• Smart grid
• technology acceptance
• trust
• perceived benefits
• structural equation modelling
• mediation

1. Introduction

As a part of the 2040 climate target, the European Commission has proposed a plan to reduce greenhouse-gas emissions within the European Union by at least 90% below 1990 levels by 2040 (European Commission 2024). Europe’s transition away from fossil fuels is further accelerated by increasing concerns regarding energy independence and high energy prices brought about by the war in Ukraine (European Commission 2022a). Renewable energy sources are expected to play a key role in the transition towards sustainable power systems. The International Energy Agency (IEA) estimates that in 2025, renewable energy sources will surpass coal as the largest source of global electricity generation (IEA 2023). Wind and solar power are projected to account for 95% of global renewable energy generation increase during 2023–2028 (IEA 2023). As Europe transitions towards a more renewable-based energy portfolio, there is a growing need to transform and upgrade its power systems to accommodate these changes (Vasiljevska et al. 2021). The intermittent nature of renewable energy sources such as wind and solar power is going to strain power systems’ capability to maintain the balance between generation and consumption (Riesz and Milligan 2015). Generation capacity of wind and solar power varies over time which means that system flexibility is required to balance supply and demand when the generation capacity of these energy sources is low (Riesz and Milligan 2015). As the penetration of variable renewable energy resources increases, so does the amount of flexibility needed (Riesz and Milligan 2015). Traditional fossil fuel-fired power plants are both environmentally undesirable and relatively expensive as flexibility resources, emphasising the importance of finding alternative ways to balance the supply and demand of electricity.

One promising alternative to traditional balancing resources is the demand flexibility of end-customers. Indeed, the European Commission expects that customers will play an increasingly important role in the smart grids of the future (Vasiljevska et al. 2021). A smart grid (SG) may be defined as an

electricity network that can integrate in a cost efficient manner the behaviour and actions of all users connected to it, including generators, consumers and those that both generate and consume, in order to ensure an economically efficient and sustainable power system with low losses and high levels of quality, security of supply and safety (Vasiljevska et al. 2021).

In SGs, customers are envisioned to become active participants within their networks rather than being merely passive consumers (European Commission 2021; Nielsen, Reisch, and Thøgersen 2016). Customers may participate in SGs for example by generating and storing energy and by adjusting their energy consumption at strategic times (Consumer Electronics Association (CEA) 2011). Customer participation in SGs is made possible by various smart technologies, such as smart meters, in-home displays (IHDs), smart appliances and microgeneration technologies, as well as by the services that are enabled by these technologies, such as demand response (Dileep 2020; Kabeyi and Olanrewaju 2023; Schappert and von Hauff 2020). With their active involvement, customers can help reduce the dependency on coal and gas-fired generating capacity and to lessening the need for network investments (European Commission 2022b; 2016). Therefore, widespread customer adoption of SG technologies is one of the key requirements for the successful development of SGs and the decarbonisation of power systems.

Despite the growing importance of the role of customers in SGs, there are still numerous challenges related to customer participation. Previous research has identified customer distrust in companies and organisations in the energy sector as one of the key barriers to the adoption of SG technologies and services (Li et al. 2021; Parrish et al. 2020). As will be discussed in more detail in the next section, trust is a multifaceted concept that can influence technology adoption both directly and indirectly by influencing customer perceptions. Trust can to act as a heuristic to reduce complexity in decision making (Lewicki, McAllister, and Bies 1998) and customer trust in organisations responsible for managing and selling different technologies can directly influence the public acceptance of those technologies (Greenberg 2014). Furthermore, a lack of customer trust may lead to suspicions of claims organisations make about the benefits of the technologies they promote (Balta-Ozkan et al. 2013; Kahma and Matschoss 2017). Consequently, trust can influence the perceived benefits associated with various technologies (Siegrist and Cvetkovich 2000). Since perceived benefits are among the most important determinants of technology acceptance (Davis 1989; Visschers and Siegrist 2014), there is also an indirect link between trust and technology acceptance. Indeed, numerous previous studies have demonstrated both the direct effects of trust (Cologna and Siegrist 2020; DiPersio et al. 2020; Firestone et al. 2020) as well as its indirect effects via perceived benefits (Bronfman and Vázquez 2011; Huijts and Van Wee 2015; Terwel et al. 2009) on the acceptance and adoption of various different technologies. Consequently, increasing public trust in the energy sector may be an effective way to facilitate the adoption SG technologies, and therefore, it is imperative to understand the role of trust and the mechanisms by which it influences customer behaviour.

However, much of the previous trust research relating to SG technologies has focused on the negative effects of distrust. Although trust and distrust are closely related, they are not the opposite ends of a single continuum (Lewicki, McAllister, and Bies 1998) and as such, the potential positive effects of trust cannot necessarily be inferred from the negative effects of distrust. Even though the positive effect of trust has been recognised in other technological contexts, only a few studies have investigated the role of trust as a factor positively influencing adoption of SG technologies. Namely, Chen, Xu, and Arpan (2017) and Warkentin, Goel, and Menard (2017) found that customers’ trust in utility companies was positively associated with the intention to adopt smart meters. Similarly, Chen et al. (2020) and Mulcahy et al. (2019) found that trust was positively associated with intentions to use smart home technologies. Though these studies have demonstrated the potential positive influence of trust, they have not considered the separate effects of different antecedents of trust, nor have they investigated the relationship between trust and different perceived benefits of SG technologies. Therefore, the mechanisms by which trust influences SG technology adoption are not yet clear. The present study seeks to address this research gap by answering the following research question: How are trust and perceived benefits associated with customer intentions to adopt SG technologies?

This study will investigate the relationships between customer trust in companies in the energy sector, the perceived benefits of SG technologies and customer intention to use different key SG technologies. The main contributions of this study are twofold. Firstly, by measuring customers’ intention to use multiple different technologies, we aim to provide more generalizable results on SG technology adoption. And secondly, we will elucidate the mechanism by which trust influences customers’ behavioural intentions. More specifically, we will examine the separate effects of different antecedents of trust and consider both their direct effects and their indirect effects through perceived benefits. Therefore, the results of this study aim to inform efforts to promote adoption of SG technologies by offering new insights into the factors that influence technology adoption, as well as to complement previous research by extending the understanding of the role of trust.

2. Theoretical background and hypotheses

2.1. Role of trust in technology adoption

Rousseau et al. (1998) define trust as the willingness to accept vulnerability based on positive expectations of the intentions or behaviour of another. There is evidence to suggest that trust acts as a heuristic (i.e. a decisional shortcut) in many decision-making processes. Trust can serve as a mechanism to reduce complexity when making judgements and decisions, particularly in contexts characterised by risk and uncertainty (Lewicki, McAllister, and Bies 1998; Siegrist and Cvetkovich 2000). Trust is not a behaviour per se, but a psychological condition that may result in behaviour (Mayer, Davis, and Schoorman 1995; Rousseau et al. 1998). Trust is a multifaceted concept that encompasses different beliefs and perceptions of the other party and many different characteristics affecting trust have been identified in the literature. Mayer, Davis, and Schoorman (1995) classified these different antecedents of trust into three major categories: ability, benevolence and integrity. Ability (or competence) is the set of skills or characteristics that allow a party to have influence in some specific domain. Benevolence is the trustee’s sincere motivation to do good to the trustor, even in absence of a profit motive. Integrity refers to the trustee’s adherence to a set of principles that the trustor finds acceptable. These three factors represent unique and distinct elements of trust and together they explain a major portion of perceived trustworthiness (Mayer, Davis, and Schoorman 1995). Others have adopted a two-factor model, with morality relevant information (i.e. benevolence and integrity) and competence relevant information as the main antecedents of trust (Earle and Siegrist 2006; Terwel et al. 2009). In this study, we adopt the two-factor model with competence-based trust and integrity-based trust as the two main types of trust. From here on, where we refer to characteristics related to ability as competence and characteristics related to morality as integrity.

There is a large body of literature investigating the influence of trust on the public acceptability of different technologies and products. Trust in organisations and institutions has been shown to influence the public attitude towards technologies such as nuclear power (Greenberg 2014; Siegrist, Cvetkovich, and Roth 2000; Whitfield et al. 2009), carbon capture and storage (Cologna and Siegrist 2020; Midden and Huijts 2009; Terwel et al. 2009) and renewable energy (DiPersio et al. 2020; Firestone et al. 2020; Kalkbrenner and Roosen 2016; van Prooijen 2019; Yun and Lee 2015). Furthermore, trust is associated with acceptance of and purchase intention towards many consumer products and services, such as genetically modified foods (Costa-Font, Gil, and Traill 2008; Frewer et al. 2004; Zhang et al. 2018), personal health technologies (Arfi et al. 2021; Fox and Connolly 2018) and online commerce (Gefen, Karahanna, and Straub 2003; Hoffman, Novak, and Peralta 1999; Luarn and Lin 2005). Given its important role in human behaviour, researchers investigating customer behaviour have integrated trust into theoretical model such as Ajzen’s theory of planned behaviour (Wu and Chen 2005; Yang, Lee, and Zo 2017) and Davis’ technology acceptance model (Gefen, Karahanna, and Straub 2003).

There are various trust-related barriers hindering widespread adoption of SG technologies as well as other smart homes technologies and services. Previous research has shown that customers have concerns pertaining to different aspects of smart technologies and services, and that the distrust may be directed towards the device or service itself, or towards parties associated with these technologies, such as manufacturers or utility companies (Li et al. 2021). In a UK consumer trust survey conducted in November 2022, only 23% of respondents reported trusting the energy sector, whereas nearly half of the respondents reportedly did not (Which? 2023). High energy prices and prioritising profits in the time of crisis were cited as reasons for this lack trust (Which? 2023). Furthermore, there are potential security issues related to SG technologies (Bigerna, Bollino, and Micheli 2016) and concerns related to data security and privacy have become major sources of customer distrust. Among these concerns are, for example, the collection of too much or too detailed information, the use of data for unintended purposes, the possibility of unauthorised access by third parties, and loss of control over household devices (Balta-Ozkan et al. 2013; Gerpott and Paukert 2013; Karjalainen 2013; Lineweber 2011; Tu et al. 2021). Such concerns may stem from the other party’s perceived inability or unwillingness to live up to customers’ expectations, and accordingly, may be related to both competence-based and integrity-based trust. Importantly though, no previous studies have investigated the separate effects of these two types of trust in the context of SG technologies, and accordingly, we hypothesise that both competence-based trust and integrity-based trust influence customer intention to adopt SG technologies.

H1: Competence – and integrity-based trust are positively correlated with intention to adopt SG technologies.

There is evidence to suggest that characteristics related to integrity may influence perceived trustworthiness to a greater extent than characteristics related to competence. For instance, Martijn et al. (1992) found that a trustee with a combination of negative integrity traits and positive competence traits will likely be perceived as less trustworthy than a trustee with a combination of positive integrity traits and negative competence traits. Furthermore, previous studies have found that, compared to competence characteristics, perceived integrity characteristics have a more significant effect on impression formation (De Bruin and Van Lange 1999a; 2000), degree of cooperation displayed towards the trustee (De Bruin and Van Lange 1999b; Earle and Siegrist 2006) and on the acceptability of renewable energy projects (Liu et al. 2020). As such, we hypothesise that integrity-based trust will have a larger effect on the intention to adopt SG technologies compared to competence-based trust.

H2: Integrity-based trust has a greater effect on the intention to adopt SG technologies compared to competence-based trust.

2.2. Role of perceived benefits in technology adoption and the link to trust

Perceived benefits are a key determinant of technology adoption; people show greater acceptance towards (Siegrist 1999; Visschers and Siegrist 2014) and more frequent usage of (Davis 1989) technologies that they perceive as being useful and beneficial. Importantly though, customers’ perceptions are based on subjective evaluations and do not necessarily reflect the objective benefits (Davis 1989). Consequently, customers may adopt technologies based on overestimations of their benefits and reject technologies based on underestimations, emphasising the importance of perceived benefits. From the customer’s point of view, the most obvious benefits are personal ones, such as enhanced work performance (Davis 1989), improvements to health and well-being (Karahoca, Karahoca, and Aksöz 2018), financial gains (Venkatesh, Thong, and Xu 2012; Wilson, Hargreaves, and Hauxwell-Baldwin 2017) and enjoyment derived from use or possession (Gao, Li, and Luo 2015; Venkatesh, Thong, and Xu 2012). Wider benefits like societal and environmental impacts are also often taken into consideration by customers (Huijts, Molin, and Steg 2012; Siegrist 1999) and have been shown to influence on customer attitudes and purchase intentions (Chen, Lin, and Weng 2015; Claudy, Peterson, and O’Driscoll 2013).

Financial and environmental benefits are among the most important, though not the only, benefits by which customer evaluate SG technologies (Ianole-Călin and Druică 2022; Li et al. 2021; Sovacool and Del Rio 2020; Wolske, Stern, and Dietz 2017). SG technologies deliver financial benefits to customers by allowing them to effectively manage and monitor their energy consumption. For example, by combining feedback from IHDs with dynamic tariffs, customers can save money by choosing to shift their energy consumption to times of lower electricity prices (Darby 2020; Darby et al. 2015). Change in energy consumption patterns can be further facilitated via the use of smart appliances that can be remotely controlled or automated (Kobus et al. 2015; Stamminger and Anstett 2013; Vanthournout et al. 2015). Microgeneration technologies, such as solar PV and heat pumps, can allow customers to meet some or all of their own energy needs, increasing households’ energy self-sufficiency and reducing fuel costs (Balcombe, Rigby, and Azapagic 2013; Jager 2006; Leenheer, Nooij, and de Sheikh 2011). Environmental benefits of SG technologies are related to reduced fossil fuel consumption and reduced need for network investments due to increased grid stability (European Commission 2021; 2013). Therefore, we hypothesise that adoption of SG technologies is influenced by customers’ perceptions of the financial and environmental benefits of these technologies.

H3: Perceived financial and environmental benefits are positively correlated with intention to adopt SG technologies.

Customers’ trust in different authoritative sources of information, such as companies, experts and institutions has been shown to influence the perceived benefits associated with various different technologies (Siegrist and Cvetkovich 2000). When customers lack trust in companies and institutions that manage and sell new technologies, they may be suspicious of the claims made about the benefits of these technologies. In regards to SG technologies, there may be doubts about the actual financial and environmental benefits of these technologies (Balta-Ozkan et al. 2013; Hargreaves, Wilson, and Hauxwell-Baldwin 2018; Kahma and Matschoss 2017). For instance, customers may be worried that they are not getting fair deal when buying smart technologies or believe that the potential financial savings would be too small to matter (Balta-Ozkan et al. 2013; Kahma and Matschoss 2017). Even if customers believe in the financial feasibility of the technologies themselves, lack of trust in organisations or companies associated with those technologies can still negatively influence customer perceptions of these technologies. Customers may believe that the SG technologies serve to benefit only the utility companies and that the financial benefits will not be passed on to the customers (Balta-Ozkan et al. 2013; Stenner et al. 2017). Moreover, the incidence of corporate greenwashing, that is, misleading consumers about a company’s environmental performance or the environmental benefits of a product or service, has dramatically increased in the last two decades (de Freitas Netto et al. 2020; Delmas and Burbano 2011). Customers have become increasingly aware of this practice and sceptical towards claims made about the environmental benefits of a product (Chen, Lin, and Chang 2014; Lyon and Montgomery 2015). In the context of SGs, the true long-term environmental impacts of both domestic SG technologies and SGs as a whole are yet to be determined (Moretti et al. 2017). In practice, it may be difficult or impossible for customers to assess the authenticity of claims made about the environmental impact of a product or service. Consequently, whether or not customers believe such claims is often dictated by trust (Chen, Lin, and Chang 2014). Accordingly, we hypothesise that both types of perceived benefits are influenced by both types of trust.

H4: Perceived financial and environmental benefits are positively correlated with both competence – and integrity-based trust.

Due to the relationship between customers’ trust and their perceptions of different technologies, perceived benefits likely act as a mediator for trust. This means that positive effects of trust on the intention to adopt different technologies are likely at least partially due to increases in perceived benefits. Indeed, numerous studies have investigated the effects of trust and found indirect effects via perceived benefits on different outcomes, such as acceptability of technological and environmental hazards (Bronfman et al. 2008; Bronfman and Vázquez 2011), acceptability of technologies (Huijts and Van Wee 2015; Siegrist 1999; Terwel et al. 2009), intention to adopt technologies (Chen, Xu, and Arpan 2017) and intention to act in favour technologies (Huijts, Molin, and van Wee 2014). As such, perceived benefits likely play an important role in mediating the effect of trust on the customer intention to adopt SG technologies. However, due to the lack of previous studies in this area, it is unclear whether the effect of trust is fully mediated or if trust has both direct and indirect effects on behavioural intention. Furthermore, it has also not been investigated whether financial and environmental benefits have similar or different roles in mediating the effect of trust on intentions to adopt SG technologies. Because financial considerations may constitute a more immediate barrier to adoption of SG technologies than environmental ones (Balcombe, Rigby, and Azapagic 2013; Li et al. 2021), we hypothesise that financial benefits will be a stronger mediator than environmental benefits.

H5: Effects of competence – and integrity-based trust on the intention to adopt SG technologies are mediated by perceived financial and environmental benefits.

H6: Effects of competence – and integrity-based trust on the intention to adopt SG technologies are mediated more by financial benefits than by environmental benefits.

3. Method

3.1. Procedure and participants

The research was conducted as a cross-sectional, anonymous online questionnaire using an online survey software (Webropol). This research complies with the Finnish National Board on Research Integrity (TENK) ethical principles for research with human participants (TENK 2019), the European Union’s General Data Protection Regulation as well as the Finnish Data Protection Act (Office of the Data Protection Ombudsman 2023). The questionnaire was initially created in English, after which the questionnaire was translated into Finnish by one researcher. Four researchers independently analysed the translated questionnaire items and agreed on the final version of the survey. The questionnaire was then translated back into English to ensure translation equivalence. The survey instrument was then pilot tested for comprehension on a convenience sample of 10 Finnish-speaking adults. Based on the pilot test, the survey was deemed reliable and comprehensible.

The survey was conducted in Finland in 2022. In order to ensure a sufficiently large sample size for our model, the survey was in cooperation with a Finnish electricity retail company. Potential participants were recruited by distributing the survey link via email to the customers of the electricity retail company. Recruitment emails contained a short description of the purpose and contents of the questionnaire, an explanation of how the data would be used as well as a hyperlink to the online questionnaire. The survey ran from 24th of February 2022 through 6th of March 2022. To incentivise participation in the survey, participants were offered an entry to a raffle to win one of ten smart plugs (a remote-controlled electrical outlet that can be used manage and monitor electrical devices attached to it) upon completing the survey. The questionnaire was available in both Finnish and English.

A total of 2489 participants started the survey, of whom 1468 completed the survey, yielding a completion rate of 59% for those who started the survey. 34 participants were excluded due to missing data. As a result, responses from 1434 participants were included in the analysis. As such, the number of participants is substantially more than minimum required sample size, which, following Kline’s (2016) recommendation of at least 20 participants per estimated parameter, was determined to be 300 for our model. The majority of participants were male (71.8%), while 26.2% were female and 0.5% identified as other. Twenty-one participants (1.5%) chose not to report their gender. Participant age distribution was skewed towards older age groups: 0.8% of participants were between 19 and 29 years old, 5.1% were between 30 and 39 years old, 12.9% were between 40 and 49 years old, 21.6% were between 50 and 59 years old and 56.7% were 60 years old or older. Fourty-two respondents (2.9%) chose not to report their age.

3.2. Measures

All items were measured using a 5-point Likert scale (1 = Fully disagree, 2 = Disagree, 3 = Neither agree nor disagree, 4 = Agree, 5 = Fully agree). The Appendix shows the full measurement items used in the study. Items marked with an asterisk (*) were reverse scored.

3.2.1. Behavioural intention

In this study, the dependent variable of interest is behavioural intention to adopt SG technologies. Behavioural intention is an indicator of a person's readiness to perform a behaviour, and as such, it can be considered to be the most proximal determinant of behaviour (Fishbein and Ajzen 2009). Due to its close relatedness with behaviour, intention plays a central role in different behavioural models such as Theory of Planned Behaviour (Fishbein and Ajzen 2009), Social Cognitive Theory (Bandura 1991), and Unified Theory of Acceptance and Use of Technology (Venkatesh et al. 2003).

Attitudinal measures are able to more reliably predict behaviour in a given domain if behaviour is measured by an aggregate of multiple related actions, rather than by a single arbitrary action (Fishbein and Ajzen 2009). Therefore, in order to provide a more accurate, reliable and generalizable measure of customers’ intention to adopt SG technologies, we used an aggregate measure consisting of three items. More specifically, behavioural intention to adopt SG technologies was measured as the mean of three items. Participants were asked to indicate their intention to use three energy technologies: IHDs, smart appliances, and microgeneration technologies. These technologies were selected on the basis the participants likely having some familiarity with them due to their good availability on the market. A brief explanation of each technology was provided to the participants. There is evidence to suggest that intention questions have the strongest predictive validity when phrased as ‘I intend to [perform a behaviour]’ (Fishman, Lushin, and Mandell 2020), and as such, similar wording was used in this study.

3.2.2. Perceived benefits

Perceived financial and environmental benefits of SG technologies were measured with a total of eight items. Financial and environmental benefits were chosen for this study due to the number of previous studies demonstrating their importance. Perceived financial benefits were measured as the mean of four items relating to participants’ opinions about the potential monetary savings associated with the use of SG technologies and the extent to which these technologies were seen as prudent investments. Perceived environmental benefits were measured as the mean of four items that addressed participants’ beliefs about the ability to reduce greenhouse gas emissions and save natural resources using SG technologies as well as about the wider role of these technologies in combating climate change. Participants were asked to indicate their level of agreement with each of the eight items in relation to the three SG technologies introduced in the questions measuring behavioural intention. Measurement item 3 of the perceived financial benefits scale and measurement item 3 of the perceived environmental benefits scale were adapted to the context of this study from Kahma and Matschoss (2017).

3.2.3. Trust

We measured participants’ competence-based and integrity-based trust in actors in the energy sector with a total of 12 measurement items. Competence-based trust was measured as the mean of six items addressing participants’ perceptions about the ability and expertise of actors in the energy sector as well as about the quality of service provided by them. Integrity-based trust was measured as the mean of six items relating to the perceived ethicality and honesty of actors in the energy sector as well as to the extent to which participants felt like they were treated fairly and with respect by these actors. Participants were asked to indicate their level of agreement with each of the 12 items following the statement ‘In general, most companies in the energy sector … ’. Measurement items 1 and 2 of the competence-based trust scale and measurement items 3 and 4 of the integrity-based trust scale were adapted to the context of this study from Liu et al. (2020).

3.3. Reliability

The reliability of a measurement scale represents how reliability the individual items of a scale measure a common construct (Flora 2020). Reliability was assessed using the omega coefficient. The omega coefficient is more accurate and less prone to bias than the commonly reported Cronbach’s alpha, and it is therefore recommended to measure internal consistency using the omega coefficient (Dunn, Baguley, and Brunsden 2014; Revelle and Zinbarg 2009; Zinbarg, Revelle, and Yovel 2007). More specifically, we used the omega hierarchical which is an estimate of the general factor saturation of a test (McDonald 1999; Zinbarg et al. 2005). Out of the different omega coefficients, the hierarchical omega was chosen because it has been shown to produce the most reliable results for continuous items (Kelley and Pornprasertmanit 2016). Omega hierarchical was calculated using a bias-corrected and accelerated bootstrap method, as recommended by Kelley and Pornprasertmanit (2016). Point estimates and confidence intervals for the omega coefficients are reported in Table 1. While there exists no commonly accepted cut-off value for the hierarchical omega coefficient, some researchers have suggested a minimum benchmark value of 0.5 (Reise, Bonifay, and Haviland 2013; Zinbarg, Revelle, and Yovel 2007). As can be seen from Table 1, the point estimate of the omega coefficient for all measurements used in this study exceeds 0.5. Therefore, all measurement scales used in this study were found to be reliable. Lastly, to test for the presence of multicollinearity between study variables, variance inflation factors (VIFs) were calculated. VIFs ranged from 1.78–2.08, which is lower than the threshold value of 3.3 proposed by Kock and Lynn (2012), indicating that multicollinearity was not detected.

Table 1. Point estimates and 95% confidence intervals of omega hierarchical for survey measurement scales.

Construct	Point estimate	Confidence interval (95%)
Behavioural intention	0.687	0.650–0.720
Perceived financial benefits	0.703	0.672–0.732
Perceived environmental benefits	0.844	0.762–0.860
Competence-based trust	0.794	0.762–0.818
Integrity-based trust	0.834	0.815–0.852

3.4. Statistical analysis

All statistical analyses were carried out using R version 4.2.0. R package lavaan (Rosseel 2012) was used for the mediation analysis. Descriptive statistics for all measurement constructs were calculated. To test for hypotheses H1, H3 and H4, Pearson correlations between behavioural intention, trust (competence-based and integrity-based) and perceived benefits (financial and environmental) were calculated. To test whether the effects of competence-based and integrity-based trust on the intention to adopt SG technologies are mediated by financial and environmental benefits (H5), a path analysis was conducted. Path analysis is a special case of structural equation modelling that allows for evaluating whether a data set fits a proposed causal model (Fairchild and McDaniel 2017). Path analysis was chosen because it allows for all model parameters to be estimated simultaneously. This improves the reliability of the parameter estimates, particularly in the presence of multiple mediators, compared to methods estimating parameters separately (Preacher and Hayes 2008). Omitting important variables from a statistical model may lead to biased estimates and as such it is important to take measures to limit this bias (Tomarken and Waller 2005). Path analysis allows for all model parameters to be estimated simultaneously, which reduces the risk of omitted variable bias (Preacher and Hayes 2008). Furthermore, following the recommendations of Preacher and Hayes (2008), the residuals covariances both among the independent variables and among the mediator variables were allowed to covary. This accounts for any unmodeled sources of covariation between financial and environmental benefits as well as between competence-based and integrity-based trust. Lastly, age and gender were added to the model as control variables to eliminate any confounding effects of these demographic factors. For statistical analysis, age and gender were treated as dummy-coded categorical variables. Paths from these control variables to all other model variables were estimated, and consequently, the tested model is saturated (that is, it has zero degrees of freedom) and the model fit is therefore necessarily perfect. Maximum likelihood estimation was used for estimating model parameters. As recommended by Hayes and Scharkow (2013), bias-corrected bootstrap confidence intervals (using 10,000 bootstrap samples) were used to test for mediation effects. Mediation effect is considered statistically significant at the 0.05 level if the confidence 95% interval does not contain zero (Preacher and Hayes 2008).

To compare the sizes of the effects of competence-based and integrity-based trust on behavioural intention (H2) and to test whether their indirect effects on behavioural intention are mediated more strongly by financial benefits than by environmental benefits (H6), the indirect effects were formally contrasted. Contrasts represent comparisons between the unique abilities of each mediator to account for the effect of an independent variable on a dependent variable (Preacher and Hayes 2008). To test for differences between the two specific indirect effects, pairwise contrasts were estimated using bias-corrected bootstrap confidence intervals as described above.

4. Results

4.1. Descriptive statistics and correlations

Descriptive statistics and Pearson correlations for all survey constructs are shown in Table 2. As hypothesised (H1), both competence-based and integrity-based trust were positively correlated with behavioural intention. Therefore, individuals who expressed higher levels of trust in actors in the energy sector were more likely to intend to use SG technologies in the future. Furthermore, both perceived financial and environmental benefits were also positively correlated with behavioural intention. These findings support our hypothesis (H3) postulating that perceived benefits play an important role in determining the intention towards the adoption of SG technologies. Finally, in accordance with hypothesis H4, both types of trust are positively correlated with both types of perceived benefits. All reported correlations were statistically significant at the 0.0001 level.

Table 2. Descriptive statistics and correlations between survey measure constructs.

Construct	Mean	SD	1. Behavioural intention	2. Perceived financial benefits	3. Perceived environmental benefits	4. Competence-based trust
1. Behavioural intention	3.356	0.792
2. Perceived financial benefits	3.361	0.633	0.52
3. Perceived environmental benefits	3.597	0.795	0.40	0.64
4. Competence-based trust	3.241	0.582	0.21	0.38	0.39
5. Integrity-based trust	3.120	0.650	0.20	0.37	0.41	0.71

n = 1434, SD = standard deviation, all reported correlations are statistically significant at p < 0.0001.

4.2. Mediation analysis

The path coefficients representing the estimated change in the outcome variable for each unit change in the predictor variable for out path analysis model are shown in Tables 3 and 4. Path coefficients for the direct effects between the survey measure constructs are shown in Table 3. Table 4 shows the path coefficients of the indirect effects of competence-based trust and integrity-based trust on behavioural intention and the pairwise contrasts between the indirect effects. Furthermore, Table 4 also shows the total effects of competence-based trust and integrity-based trust as well as the contrasts between the two. A diagram of the tested path analysis model is shown in Figure 1.

Figure 1. Path analysis model depicting the effects of trust and perceived benefits on behavioural intention to use smart grid technologies and the path coefficients for the direct effects between survey measures.

Table 3. Unstandardised model path coefficients with standard errors in parentheses.

Predictor variable	Outcome variable
	Perceived financial benefits	Perceived environmental benefits	Behavioural intention
Competence-based trust	0.230* (0.044)	0.286* (0.057)	0.011 (0.053)
Integrity-based trust	0.229* (0.037)	0.320* (0.053)	−0.027 (0.047)
Perceived financial benefits			0.548* (0.041)
Perceived environmental benefits			0.126* (0.036)

*p < 0.0001.

Table 4. Unstandardised indirect and total effects of competence – and integrity-based trust and contrasts between these effects.

Effect on behavioural intention	Point estimate	Confidence interval (95%)
Indirect effect of competence-based trust through
Financial benefits (PB_fin)	0.126*	0.077–0.181
Environmental benefits (PB_env)	0.036*	0.016–0.066
PB_fin – PB_env	0.090*	0.039–0.149
Indirect effect of integrity-based trust through
Financial benefits (PB_fin)	0.126*	0.085–0.173
Environmental benefits (PB_env)	0.040*	0.018–0.071
PB_fin – PB_env	0.085*	0.039–0.141
Total effect of
Competence-based trust (TRUST_comp)	0.173*	0.065–0.287
Integrity-based trust (TRUST_int)	0.139*	0.036–0.239
TRUST_comp – TRUST_int	0.034	−0.158–0.237

*Effect statistically significant at the 95% level.

We hypothesised that integrity-based trust would have a greater effect than competence-based trust on the intention to adopt SG technologies (H2). However, as can be seen from Table 4, the 95% confidence interval of the pairwise contrast between the total effects (TRUST_comp – TRUST_int) includes zero. Therefore, contrary to our hypothesis, there was no statistically significant difference between the total effects of the two types of trust on behavioural intention. Furthermore, we hypothesised that the effects of competence-based trust and integrity-based trust on behavioural intention are mediated by perceived financial and environmental benefits (H5). As can be seen from Table 4, zero was not included in the confidence intervals of the indirect effects of either type of trust through either type of benefits. This finding confirms that the effects of trust on the intention to adopt SG technologies are mediated by the perceived benefits of these technologies. Moreover, Table 3 shows that the direct effects of neither competence-based trust nor integrity-based trust on behavioural intention are statistically significant. Therefore, even though both types of trust are positively correlated with behavioural intention, these relationships become non-significant when the effects of perceived benefits are accounted for. These findings indicate that trust influences behavioural intention solely through changes in perceived benefits.

We also hypothesised that the effects of trust on behavioural intention would be mediated more strongly by perceived financial benefits compared to perceived environmental benefits (H6). Table 4 shows the comparisons between the indirect effects through perceived financial benefits and perceived environmental benefits (PB_fin – PB_env) for both types of trust. Zero is not contained in the confidence interval of either of the pairwise contrasts, and therefore the effects of both competence-based trust and integrity-based trust are mediated more strongly by perceived financial benefits than by perceived environmental benefits.

5. Discussion

The present study investigated the relationships between customer trust in companies in the energy sector, the perceived benefits of smart SG and customer intention to use SG technologies. We carried out a questionnaire survey of Finnish households and analysed the data using Pearson correlations and structural equation modelling. We examined the separate effects of integrity-based trust and competence-based trust and their indirect effects through financial and environmental benefits. To our knowledge, this is the first study to investigate the separate effects of different antecedents of trust and their direct and indirect effects in the context of SG technology adoption. In agreement with our hypotheses (H1, H3 and H4) and previous studies (Bronfman et al. 2008; Bronfman and Vázquez 2011; Huijts and Van Wee 2015; Huijts, Molin, and van Wee 2014; Siegrist 1999; Terwel et al. 2009), we found that trust, perceived benefits and behavioural intention were all positively correlated with each other. Also in agreement with previous research (Chen, Xu, and Arpan 2017; Huijts, Molin, and van Wee 2014), we found that trust indirectly influences behavioural intention through perceived benefits. Moreover, as hypothesised (H6), we found that the effects of trust were mediated more strongly by perceived financial benefits than by perceived environmental benefits. In fact, financial benefits were by far the strongest predictor of behavioural intention to use SG technologies. This finding is largely in agreement with previous studies investigating the motivational power of financial benefits (Li et al. 2021; Marikyan, Papagiannidis, and Alamanos 2019).

Contrary to our hypothesis (H2), the effect of integrity-based trust did not have a stronger influence on the intention to adopt SG technologies compared to competence-based trust. In fact, there was evidence to suggest that the opposite may be true; the point estimate of the total effect of competence-based trust on behavioural intention was nominally larger than that of integrity-based trust, though the difference between the two effects was not statistically significant. This finding also runs counter to previous research showing that perceived integrity characteristics have a greater effect on various outcomes compared to competence characteristics (De Bruin and Van Lange 1999a; 1999b; Earle and Siegrist 2006; Liu et al. 2020). This apparent discrepancy may be explained by the fact that whereas for example Liu et al. (Liu et al. 2020) investigated public acceptance of industrial technologies or activities, our study focused on behavioural intention to use consumer technologies. According to Ajzen (2001) and others, attitudes towards objects are based on affect and cognition, and the extent to which individuals rely on either one or the other is context-dependent. Terwel et al. (2009) have argued that cognition-based judgements may be associated with competence-based trust, whereas affect is more associated with integrity-based trust. It is possible that when estimating their own intentions to use technology, people are more likely to engage in rational and effortful decision-making (i.e. cognition-based evaluation) compared to when estimating the general acceptability of technologies. This may cause customers to place additional emphasis on competence-based trust when evaluating their intention to use SG technologies. Furthermore, although the questionnaire items pertaining to trust were not directed towards any specific actor in the energy sector, it is likely that the survey responses were influenced by participants’ prior experiences with companies, as well as by the communication and marketing strategies used by these companies. It has been demonstrated that people tend to place more importance on positive competence-related information than negative competence-related information, whereas they place more importance on negative integrity-related information than positive integrity-related information (Kim et al. 2004). It can therefore be easier for companies to improve competence-based trust with their actions and marketing compared to integrity-based trust, which may lead them to emphasise their competence characteristics. This may ultimately influence which characteristics are important to customers.

Our findings also show that when controlling for the effects of perceived financial and environmental benefits, neither competence-based trust nor integrity-based trust had a direct effect on behavioural intention. Instead, the effects of trust are fully mediated by the perceived benefits of SG technologies. This finding confirms our hypothesis H5. This finding also supports the causal chain model of trust, in which trust in an organisation influences the perception of risks and benefits of technologies, which in turn influences acceptance and adoption of those technologies (Eiser, Miles, and Frewer 2002). Our findings suggest that the role of trust in the public adoption of SG technologies is to act as a decisional shortcut to help reduce the complexity faced by customers when evaluating these technologies. When evaluating the benefits of SG technologies, customers seem to consider both the competence and integrity of actors in the field of energy, suggesting that both cognition and affect contribute to this evaluation process. Therefore, the perceived benefits of SG technologies, and subsequently customers’ usage intentions, are likely influenced by both rational and emotional factors.

6. Implications and limitations

6.1. Practical implications

End customers are expected to play an increasingly important and active role in the future of SGs. The active involvement of customers is facilitated by various smart technologies and the services enabled by them, and as such, more widespread adoption of these technologies is imperative to the development of SGs. Based on our results, we argue that in order to increase the adoption and use of SG technologies, the actors in the energy industry should strive to improve trust between the industry and its customers. As is evident from the lack of direct effect, trust alone will not be sufficient to bring about widespread adoption of these technologies. However, building trust with customers can help dispel doubts and misconceptions about SG technologies and make customers more receptive to learning about the benefits their benefits. As such, trust can serve as a starting point to increase the adoption of SG technologies by favourably influencing customers’ perceptions of the benefits of these technologies. Furthermore, it is worth noting that trust may play a greater role in influencing technology adoption in some contexts than in others. Siegrist and Cvetkovich (2000) found that the relationship between trust and perceived benefits was moderated by knowledge such that this relationship was stronger for technologies on which participants had lower levels of knowledge. People are more likely to rely on trust to inform their decision-making when they have limited prior knowledge or low confidence in their own judgement, and as such, trust can have a particularly important role in fostering the adoption of novel or complex technologies. Similarly, trust may be particularly important for those with lower levels of education. Therefore, it may be advisable to devote greater effort to increasing trust among such populations. However, research directly investigating the moderating roles of knowledge and education as they relate to SG technologies is needed before any conclusions can be drawn.

According to our results, perceived financial benefits appear to be a more significant driver of SG technology adoption compared to environmental benefits. It is possible that when assessing the benefits of SG technologies, customers simply give financial aspects more consideration compared to environmental ones. In the context of our study, this seems a likely explanation, given the unprecedented rise of energy prices in Europe over the course of 2021 (European Parliamentary Research Service 2022). It should be noted that the start of the data collection for this study coincided with Russia’s invasion of Ukraine, which has only further exacerbated the issues of high energy prices and energy poverty in Europe (European Parliamentary Research Service 2022). One would expect that these developments would lead to customers placing increased importance on financial considerations.

Despite the larger effect of financial benefits compared to environmental benefits found in this study, it is important to note that different types of benefits may play different roles in technology adoption. Though extrinsic motivators (e.g. financial benefits) have been shown to strongly influence intentions to purchase and use energy technologies (Balcombe, Rigby, and Azapagic 2013; Li et al. 2021), there is evidence to suggest that long-term change in energy consumption behaviour depends mostly on intrinsic motivation (e.g. environmentalism) (Handgraaf, Jeude, and de Appelt 2013; Mi et al. 2021). Therefore, it may be that extrinsic motivators serve to create an initial interest in SG technologies or services after which customer engagement and long-term commitment is mostly driven by intrinsic motivation (Siitonen et al. 2023). This is an important consideration, since realising the wider benefits of many customer-side SG technologies require active engagement from the customer in addition to the initial investment. Moreover, placing too much emphasis on extrinsic rewards may actually decrease intrinsic motivation within the target population, an effect known as crowding-out (Bolle and Otto 2010; Frey and Jegen 2001). Marketing these technologies by emphasising a single attribute or benefit may therefore ultimately hinder their widespread adoption.

6.2. Theoretical implications

There are several factors to consider when interpreting the findings of this study. It remains to be determined if the findings reported here are influenced by socioeconomic or other contextual factors. For instance, it is possible there exists an interaction effect such that the magnitudes of the effects of perceived financial and environmental benefits on intentions to adopt SG technologies are moderated by factors such as income. Individuals with higher levels of income are typically less concerned with meeting their material needs, and are thus more likely to shift their focus towards non-material values, such as environmental ones (Franzen and Meyer 2010; Gifford and Nilsson 2014). When it comes to adopting SG technologies, environmentalism may be a privilege of the wealthy, whereas less affluent people need to be concerned first and foremost with fulfilling their physiological needs. As such, it is possible that in populations with higher economic status, perceived environmental benefits may influence behavioural intentions to a greater extent than suggested by our findings. It may be important for future research to investigate whether financial barriers such as low levels of income are associated with greater importance placed on the financial benefits of SG technologies compared to the environmental ones. Additionally, even if certain benefits could be derived from the use of SG technologies in theory, not all customers will be able to do so in practice. The ability to use and benefit from SG technologies is strongly influenced by daily routines and habits as well as by personal skills and competencies (Christensen et al. 2020; Siitonen et al. 2023; Strengers 2014; Verkade and Höffken 2017). Therefore, it may be prudent for researchers investigating the importance of different benefits to include some of these factors as into their statistical models. Furthermore, although financial and environmental benefits are among the most important benefits by which customers evaluate SG technologies, they are not the only ones. Other benefits identified in previous research but not included in our model include for example convenience, novelty, improvements to comfort and health, as well as the social communities formed with other SG technology users (Li et al. 2021; Siitonen et al. 2023). Future research may consider exploring the effects of trust on these benefits not included in our study. Lastly, it is worth noting that the survey did not explicitly ask about participants’ willingness or intention to participate in SGs, but rather their intention to use technologies that qualify as SG technologies. As such, it is unclear how generalizable our results are to the broader context of SG participation.

6.3. Limitations

One potential limitation of our study is that our survey was distributed via email to the customers of a particular electricity retailer. This may have affected our sample (and consequently the generalizability of our findings), as it has been shown that survey modality may influence the response rates of different sociodemographic groups differently (Rittase et al. 2020). Different sociodemographic groups may also respond at different rates based on the type of incentive used to encourage participation in the survey (Guo et al. 2016). It is also possible that customers who had more trust in the retailer were more willing to participate in the survey compared to those who had less trust. It is also important to note that this study used behavioural intention as the estimate of behaviour, and although plays a central role in many behavioural models and can predict behaviour well in various contexts (Fishbein and Ajzen 2009), it does have some limitations. Namely, correlations between intentions and behaviours are typically around 50% (Armitage and Conner 2001; Randall and Wolff 1994), indicating that behavioural intention is not a perfect measure of behaviour. The predictive ability of behavioural intention also tends decrease with time, meaning that it may be less accurate in predicting temporally distant behaviour (Mahardika et al. 2020; Venkatesh, Maruping, and Brown 2006). Moreover, although the model tested in this study is a causal model, the findings of the path analysis are correlational in nature. As such, causal inferences cannot be drawn from our findings. Lastly, although we attempted to reduce the risk of estimation bias caused by omitted variables by choosing the model variables based on existing theory, by estimating all model variables simultaneously, by including control variables and by freely estimating residual covariances, the possibility of estimation bias cannot be completely eliminated. Notwithstanding these limitations, we believe that the results of this study offer valuable insights into the factors influencing the adoption of SG technologies.

7. Conclusions

This study investigated the relationships between customer trust in companies in the energy sector, the perceived benefits of SG technologies and customer intention to use SG technologies. We conducted a survey of Finnish households and received over 1400 responses. The data were analysed using Pearson correlation coefficients and structural equation modelling. We examined both the direct effects of competence-based and integrity-based trust as well as the indirect effects through perceived financial and environmental benefits. Our results show that neither competence-based trust nor integrity-based trust had a significant direct effect on behavioural intention. Instead, the effects of both types of trust on behavioural intention to use SG technologies were fully mediated by the perceived financial and environmental benefits of these technologies. The results therefore suggest that while there is no direct link between trust and the adoption of SG technologies, there is an indirect one through perceived benefits. This finding indicates that increasing customers’ trust in actors in the energy sector may indirectly improve adoption of SG technologies.

This study contributes to the SG literature in several ways. Firstly, by measuring behavioural intention using an aggregate measure of three different technologies, the results of this study are more likely to be generalizable to various SG technologies. Secondly, by considering different antecedents of trust and different benefits of SG technologies, this study elucidates the mechanism by which trust influences the adoption of SG technologies. Finally, by more closely examining the indirect role of trust, this study helps inform future efforts to promote the adoption of SG technologies. It is important that both governmental and private organisations and companies in the energy sector find ways to improve customer trust in the sector and to educate customers about the benefits of SG technologies. Since the effects of trust are mediated by perceived benefits, it may be beneficial to tie together these two themes when communicating to customers. Both competence and integrity-based trust influence benefit perceptions, and as such, customers are likely to consider characteristics related to both competence and integrity in their trust assessments. Consequently, efforts to increase trust may be more successful if they are directed at improving both types of trust.

CRediT author statement

Petteri Siitonen: Conceptualisation; Data curation; Formal Analysis; Funding acquisition; Methodology; Software; Visualisation; Writing – original draft. Samuli Honkapuro: Conceptualisation; Funding acquisition; Methodology; Writing – review & editing; Supervision. Salla Annala: Conceptualisation; Methodology; Writing – review & editing; Supervision. Annika Wolff: Conceptualisation; Methodology; Writing – review & editing; Supervision.

Acknowledgements

The authors thank Koneen Säätiö for funding this research. The authors also want to thank Pohjois-Karjalan Sähkö for distributing the survey.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The authors do not have permission to share data.

Additional information

Funding

This work was supported by Koneen Säätiö under grant 202201616.

Appendix: Questionnaire

All items were measured using a 5-point Likert scale (1 = Fully disagree, 2 = Disagree, 3 = Neither agree nor disagree, 4 = Agree, 5 = Fully agree). Measurement item 3 of the perceived financial benefits scale and measurement item 3 of the perceived environmental benefits scale were adapted to the context of this study from Kahma and Matschoss (2017). Measurement items 1 and 2 of the competence-based trust scale and measurement items 3 and 4 of the integrity-based trust scale were adapted to the context of this study from Liu et al. (2020).

Behavioural intention

1. I intend to use an in-home display in the future.

2. I intend to use smart appliances in the future.

3. I intend to use technologies for self-production of energy in the future.

Perceived financial benefits

1. Novel energy technologies can help me save money.

2. Novel energy technologies are good long-term investments.

3. Novel energy technologies might cost more money than they save. (*)

4. It is possible to reduce my electricity and heating bills using novel energy technologies.

Perceived environmental benefits

1. Novel energy technologies can help me reduce my carbon footprint.

2. The environment would benefit if everyone used novel energy technologies.

3. Natural resources cannot be saved by using novel energy technologies. (*)

4. Novel energy technologies can help us reach the international climate goals.

Competence-based trust

In general, most companies in the energy sector … 

1. have experience in working with new energy technologies.

2. have knowledge about new energy technologies.

3. do not keep up to date with the latest science and information. (*)

4. are not able to keep my personal information secure. (*)

5. offer high-quality appliances and equipment to their customers.

6. offer high-quality service to their customers.

Integrity-based trust

In general, most companies in the energy sector … 

1. care about their customer.

2. treat their customers unfairly.

3. take customer feedback into account.

4. are dishonest when advertising their services and products. (*)

5. try to be ethically responsible in their business conduct.

6. do not care about the environment. (*)