Theoretical background

Drawing from the social cognitive theory, our study aligns with Bandura’s (Reference Bandura2000) concept of reciprocal causation, which posits that individual beliefs underpin behavioural judgements influenced by personal and contextual factors (Wood & Bandura, Reference Wood and Bandura1989). This theory suggests that behaviours, individual characteristics, and environmental stimuli interact within organisational settings, and personal factors interact with environmental cues to shape behaviour. According to Schwarzer and Luszczynska (Reference Schwarzer and Luszczynska2005), an individual’s thoughts significantly influence their behaviour while executing tasks assigned by an organisation.

The social cognitive theory offers valuable insights into interpreting the relationship between ethical climate and low-carbon behaviour. Organisations with a robust ethical climate are inclined to encourage employees to engage in low-carbon behaviour (Bandura, Reference Bandura2000). By nurturing a moral environment that emphasises environmental stewardship, organisations can shape the attitudes and actions of employees towards sustainability, consequently diminishing harmful emissions. A positive ethical climate serves as a model for employees to shape their beliefs and attitudes towards environmental sustainability. According to social cognitive behaviour (Bandura, Reference Bandura2000), this modelling process can lead to the development of personal factors, such as a strong sense of environmental passion and mindfulness, which promote low-carbon behaviour. Therefore, by promoting an ethical climate that emphasises environmental accountability, organisations can motivate and empower employees to embrace and uphold low-carbon behaviours, thus playing a pivotal role in advancing sustainability initiatives overall.

A conceptualisation of low-carbon behaviour

Low-carbon behaviour has received significant attention from researchers (Bai & Liu, Reference Bai and Liu2013; Wang, Su, Sun, Zhou, & Zhou, Reference Wang, Su, Sun, Zhou and Zhou2015; Wang, Zhang, Yin, & Zhang, Reference Wang, Zhang, Yin and Zhang2011). Prior studies initially investigated organisational citizenship behaviour (Anwar et al., Reference Anwar, Mahmood, Yusliza, Ramayah, Faezah and Khalid2020; Yin, Ma, Gong, Chen, & Zhang, Reference Yin, Ma, Gong, Chen and Zhang2021), green behaviour (Fawehinmi, Yusliza, Ogbeibu, Tanveer, & Chiappetta Jabbour, Reference Fawehinmi, Yusliza, Ogbeibu, Tanveer and Chiappetta Jabbour2022; Kamil, Nordin, & Matsat, Reference Kamil, Nordin, Matsat, Yusliza and Renwick2024), ecological behaviour (Farooq et al., Reference Farooq, Yusliza, Wahyuningtyas, Haque, Muhammad and Saputra2021), and pro-environmental behaviour(Colombo et al., Reference Colombo, Chiarella, Lefrançois, Fradin, Simione and Raffone2023), before shifting their focus to analysing low-carbon behaviour. Low-carbon behaviour involves employees adopting eco-friendly habits in their daily lives to help reduce carbon emissions and protect the environment (Zhang et al., Reference Zhang, Hu, Mu and Kong2022). Conceptualised low-carbon behaviour is a low-carbon lifestyle that reduces energy consumption and carbon emissions (Wei, Chen, & Long, Reference Wei, Chen and Long2016). It also refers to voluntary actions taken by individuals to protect the environment (Boiral, Paillé, & Raineri, Reference Boiral, Paillé and Raineri2015).

Low-carbon behaviour is a major determinant in shaping human behaviour towards the environment and is a prerequisite for coping with environmental disasters (Klöckner, Reference Klöckner2013). Such behaviours can be classified into four categories: practising energy conservation, taking public transportation, buying green-themed products, and other low-carbon behaviours (Moloney, Horne, & Fien, Reference Moloney, Horne and Fien2010). These categories of low-carbon behaviour change the status quo as activism for the environment(Wiernik, Ones, & Dilchert, Reference Wiernik, Ones and Dilchert2013).

Low-carbon behaviour can also be practised among employees in the workplace, as described by Bissing‐Olson, Iyer, Fielding, and Zacher (Reference Bissing‐Olson, Iyer, Fielding and Zacher2013). Therefore, predicting low-carbon behaviour has a substantial impact on understanding behavioural performance patterns as it acts as a guide in decision-making (Ingwersen, Garmestani, Gonzalez & Templeton, Reference Ingwersen, Garmestani, Gonzalez and Templeton2014). This means that employees can incorporate environmentally friendly behaviours into their daily routines while accomplishing their required tasks at the workplace, and they can also aggressively participate in larger environmentally friendly improvements in their organisation’s policies and procedures (Pichel, Reference Pichel, Wüstenhagen, Hamschmidt, Sharma and Starik2008).

Ethical climate and low-carbon behaviour

Ethical climate has been the focus of scholarly enquiry for over four decades, consistently emerging as a pivotal predictor of organisational behaviours (Schneider, Reference Schneider1975). Defined as employees’ shared perceptions of vertically correct behaviour and the handling of ethical issues within an organisation (Victor & Cullen, Reference Victor and Cullen1988), ethical climate is influenced by various factors such as organisational environment, structure, and developmental history. The determinants of an organisation’s ethical climate span multiple perspectives, including departmental dynamics, organisational culture, industry context, and community concerns (Bourne & Snead, Reference Bourne and Snead1999).

Ethical practices extend beyond mere compliance with legal standards in order to align with the underlying ethical ideas of the community (Valentine, Godkin, Fleischman, Kidwell, & Page, Reference Valentine, Godkin, Fleischman, Kidwell and Page2011). The organisational behaviour system, shaped by employees’ perceptions of organisational signals and behaviours, profoundly impacts ethical decision-making processes (Martin & Cullen, Reference Martin and Cullen2006). Studies examining intra-organisational aspects of ethical environments highlight the need to consider structural elements when shaping organisational climates (Ford & Richardson, Reference Ford and Richardson1994).

Drawing from earlier studies that demonstrate the positive impact of an ethical climate on job satisfaction, turnover intention, and ethical behaviour (Brown & Peterson, Reference Brown and Peterson1993; Ferrell & Skinner, Reference Ferrell and Skinner1988; Schwepker, Reference Schwepker2001; Weeks & Nantel, Reference Weeks and Nantel1992) as well as the correlation between organisational ethical climate and employee behaviours (Tsai & Huang, Reference Tsai and Huang2008), this study explores how ethical climate shapes employees’ low-carbon behaviour.

In an ethical climate that fosters trust, predictability, and accountability towards all elements of the organisational ecosystem (Rahimiaghdam & Niroumand, Reference Rahimiaghdam and Niroumand2021), employees are more likely to engage in environmentally responsible actions, which is crucial in promoting low-carbon behaviour. According to the social cognitive theory, ethical climate influences individuals’ behaviour regulation through ongoing assessment, comparing personal standards with environmental cues (Bandura, Reference Bandura1986). Domino et al. (Reference Domino, Wingreen and Blanton2015)also showed that individuals tend to enhance their behaviour based on prevailing ethical climates, regardless of whether it is perceived as ‘good’ or ‘bad’. Therefore, we assume that

Ethical climate and environmental passion

The significance of ethical climate lies in its role in fostering environmental passion among employees. According to Valentine et al. (Reference Valentine, Godkin, Fleischman, Kidwell and Page2011), employees’ environmental passion contributes to environmental preservation by aligning with the legal standards of conduct established in ethical climates. Vallerand et al. (Reference Vallerand, Blanchard, Mageau, Koestner, Ratelle, Léonard and Marsolais2003) indicate that employees are more likely to embrace an ethical climate when they feel pleased, inspired, and energised. Martin and Cullen (Reference Martin and Cullen2006) argued that these norms can influence organisational behavioural systems, specifically organisational signals and ethical decision-making among employees. The social cognitive theory suggests that the intricate interplay between environmental influences, such as climate, and individual factors, such as positive emotions, shapes the dynamics of moral reasoning and its impact on moral conduct. This perspective, rooted in the interactionist approach, posits that understanding how these elements converge is crucial for unravelling the complexities of human behaviour (Bandura, Reference Bandura1986). We found no studies indicating that employees’ environmental passion drives their increase in low-carbon behaviour. However, we believe that environmental passion can assist in developing low-carbon behaviours. According to Valentine et al. (Reference Valentine, Godkin, Fleischman, Kidwell and Page2011), an ethical climate is one of the antecedents of employee behaviour that can improve the overall nature of the workplace, particularly when it comes to minimising climate change. Therefore, we propose a link between an ethical climate and environmental passion by examining the following hypotheses:

Environmental passion and low-carbon behaviour

Robertson and Barling (Reference Robertson and Barling2013) describe environmental passion as ‘a positive emotion that results in an individual wanting to engage in pro-environmental behaviours’. We posit that this definition can foster positive emotional responses among employees, thus influencing their adoption of low-carbon behaviour within an organisation. Environmental passion is pivotal in driving substantial organisational change, catalysing inspiring actions that promote environmental preservation. Social cognitive theory is instrumental in fostering individuals’ positive emotive emotions, thereby enhancing their learning, knowledge, and experience to cultivate low-carbon behaviour. In the workplace environment, individuals acquire behaviours and cognitive strategies by observing the actions of colleagues, peers, and leaders (Green & Peil, Reference Green and Piel2015), in turn shaping their attitudes and behaviours towards sustainability. Therefore, an individual’s involvement in various environmental issues can form the basis of their commitment to change (Vidyapeetham, Reference Vidyapeetham2019).

Therefore, there is a plethora of studies on environmental passion as a vital influence on low-carbon behaviour among employees. Environmental passion is a positive emotion that expresses an individual’s desire to engage in constructive environmental practices (Akhshik, Ozturen, & Rezapouraghdam, Reference Akhshik, Ozturen and Rezapouraghdam2021). Passionate employees about the environment will exhibit positive emotions towards engaging in low-carbon behaviour. Additionally, employees with experience in contributing to environmental conservation will feel inspired, display positive emotions, and show enthusiasm for environmental efforts. This behaviour earns them the description of having environmental passion. This situation demonstrates their confidence in prioritising their interest in low-carbon behaviour (Vallerand et al., Reference Vallerand, Blanchard, Mageau, Koestner, Ratelle, Léonard and Marsolais2003, Reference Vallerand, Salvy, Mageau, Elliot, Denis, Grouzet and Blanchard2007). Environmental passion is a significant predictor of low-carbon behaviour in empirical studies. An earlier study by Afsar et al. (2016) on Thai software and banking employees demonstrated that environmental passion is a comprehensive construct that covers the assessment of activities and the desire for low-carbon behaviour. A study involving 431 employees from the manufacturing sector (Yin et al., Reference Yin, Ma, Gong, Chen and Zhang2021) also revealed a relationship between environmental passion and low-carbon behaviour. Based on the discussion above, we propose the following hypothesis:

Environmental passion as a mediator

An ethical climate significantly influences employees’ low-carbon behaviour by impacting their environmental passion. As employees share similar views on what constitutes appropriate behaviour in reducing carbon emissions and how to address ethical concerns, they feel better (Robertson & Barling, Reference Robertson and Barling2013; Victor & Cullen, Reference Victor and Cullen1988). Organisational environmental management policies predict and strongly influence low-carbon behaviour, which is a significant emotional experience related to the environment (Yin et al., Reference Yin, Ma, Gong, Chen and Zhang2021). The social cognitive theory posits that incorporating green standards into organisational ethics not only boosts employee passion but also offers increased opportunities for practising environmentally friendly behaviour, ultimately leading to enhanced organisational performance. Individuals are motivated to engage in environmental activities that align with their interests when their psychological state ignites passion within them (Vallerand et al., Reference Vallerand, Salvy, Mageau, Elliot, Denis, Grouzet and Blanchard2007). Therefore, we believe that employees’ environmental passion motivates them to act in ways that are advantageous to the environment, such as adopting low-carbon behaviour. However, there is currently a lack of studies demonstrating the mediation of environmental passion by ethical climate in influencing low-carbon behaviour. Based on the reasons mentioned above, we propose that environmental passion might serve a significant mediating role in establishing an ethical climate at the organisational level that encourages employees’ low-carbon behaviour, with the following hypothesis:

The moderating role of green mindfulness

Mindfulness has gained prominence in the Western literature for its potential as a preventative measure against irresponsible human behaviour (Geiger, Grossman, & Schrader, Reference Geiger, Grossman and Schrader2019). As environmental consciousness grows, there is heightened interest in influencing low-carbon behaviour (Barbaro & Pickett, Reference Barbaro and Pickett2016). This shift in awareness highlights the importance of green mindfulness, which aims to improve employees’ dedication to low-carbon practices, and underscores the intrinsic value of environmental stewardship. Defined as ‘a state of conscious awareness’, green mindfulness empowers individuals to grasp both the context and the content of environmental information and knowledge (Chen, Chang, Yeh, & Cheng, Reference Chen, Chang, Yeh and Cheng2015).

In the realm of environmental psychology, Dharmesti, Merrilees, and Winata (Reference Dharmesti, Merrilees and Winata2020) illuminate the concept of green mindfulness as a catalyst for enhancing environmental behaviour, echoing the sentiments of Kalyar, Ali, and Shafique (Reference Kalyar, Ali and Shafique2021), who underscore the profound emotional connection and environmental passion exhibited by mindful employees. According to the social cognitive theory (Bandura, Reference Bandura1997), green mindfulness signifies employees’ confidence in recognising opportunities to engage in low-carbon behaviour, leading to positive environmental outcomes. Previous studies have consistently highlighted the substantial relationship between environmental passion, green mindfulness, and low-carbon behaviour (Amemiya & Sakairi, Reference Amemiya and Sakairi2019; Panno et al., Reference Panno, Giacomantonio, Carrus, Maricchiolo, Pirchio and Mannetti2018; St-Louis, Verner-Filion, Bergeron, & Vallerand, Reference St-Louis, Verner-Filion, Bergeron and Vallerand2018; Yin et al., Reference Yin, Ma, Gong, Chen and Zhang2021). However, a notable gap exists in understanding how green mindfulness moderates the relationship between environmental passion and low-carbon behaviour. Therefore, we propose that cultivating green mindfulness may not only elevate positive emotions but also have a tangible impact on the uptake of low-carbon behaviours, prompting the formulation of hypotheses to further explore this connection:

The proposed research framework is illustrated in Fig. 1.

Figure 1. Research framework.

Sample and procedure

This study used a cross-sectional multilevel design, engaging a sample of 407 public servants across 37 departments within the Malaysian public sector, with one department per organisation. Data collection was conducted between October 2022 and January 2023, adhering to the requirements of a multilevel study as outlined by Maas and Hox (Reference Maas and Hox2005), which necessitates a minimum of 30 groups and 5 group sizes for robust analysis. This timeframe was selected to maintain the integrity and validity of the data-collection process within the specified parameters. Furthermore, Maas and Hox (2005) describe multilevel research that examines how individuals interact within their social context, how their social group or organisation influences them, and how a group is affected by its members. A common challenge in multilevel research is the disparity in the sample size between the individual and team levels. Three conditions – number of groups, group size, and intraclass correlation (ICC) – are typically used to define multilevel studies, with variations such as number of groups = 30, 50, 100; group size = 5, 30, 50; and ICC = 0.1,0.2, 0.3.

The public sector’s proactive stance in promoting low-carbon behaviour, esteemed above the private sector (Blok, Wesselink, Studynka, & Kemp, Reference Blok, Wesselink, Studynka and Kemp2015), inspired our decision to select it as our research sample. Therefore, we adopted purposive sampling to target Malaysian public servants. This sampling was chosen based on prior findings indicating low response rates and cost inefficiencies associated with simple random sampling for such designs(Snijders & Bosker, Reference Snijders and Bosker2012). By focusing on this approach, we aimed to ensure a more targeted and effective investigation into the behaviours and attitudes of public servants regarding sustainability initiatives.

We approached Directors in the Management Services Divisions in each public sector to undertake research in their organisations using a combination of snowball approaches and professional connections. Thirty-seven of the fifty-five organisations notified (with a response rate of 67%) agreed to participate. In general, the majority of participants in the current study were male (77.6%), aged between 31 and 40 years (46.7%), Malay (90.7%), had a range of educational backgrounds (secondary school 59%; diploma, 7.9%; bachelor’s degree, 18.2%; master’s degree, 15%), and had a total work experience of 1–10 years (68.1%), 11–20 years (23.3%), 21–30 years (7.9%), and more than 30 years (0.7%).

Measures

The initial construction of each item in this study was conducted in English. Following the common back-translation method (Brislin, Reference Brislin1970), we developed Malay versions of all items. The first translator translated the original English version into Malay. This version was then translated back into English by a second translator who was not seen in the original version to ensure that the meanings of the words were the same. The two translators are Malaysian citizens who speak both Malay and English fluently, and their expertise is in management.

Aggregation procedures

We evaluated the evidence for inter-rater agreement within groups to determine whether the ethical climate was appropriate for aggregation. To determine an appropriate level of agreement within the organisation, we used the mean r _wg and a value of >0.90 was considered appropriate (LeBreton & Senter, Reference LeBreton and Senter2008; Mathieu, Maynard, Taylor, Gilson, & Ruddy, Reference Mathieu, Maynard, Taylor, Gilson and Ruddy2007). We discovered a significant F value using one-way random effects ANOVA, F _(III) = 1.78, p < .01. This provides more evidence that there are differences between organisations in terms of ethical climate. The ICC (1) intraclass coefficient, which measures the variance in ethical climate between groups, was 0.07, demonstrating that 7% of the variance in ethical climate might be attributed to random team effects. Based on previous studies, between-group variation is sufficient for exploration at ICC values between 0.05 and 0.20(Bliese, Reference Bliese, Klein and Kozlowski2000; Peugh, Reference Peugh2010). It is preferable to aggregate ethical climate measures to the organisational level using the mean r _wg, F-value, and ICC (1) values.

Hypothesis testing

We used hierarchical linear modelling (HLM version 7) to evaluate all the hypotheses because the data for this study are multilevel (individuals are nested inside teams) (Raudenbush, Bryk, Cheong, Congdon, & Du Toit, Reference Raudenbush, Bryk, Cheong, Congdon and Du Toit2011). First, we tested the null model to confirm the degree of group variance in the measurements for the mediator and moderator (Model 1, Tables 2 and 3). We explored Level 2, representing aggregate team data, and Level 1, representing individual data. We followed the steps suggested by Mathieu and Taylor (Reference Mathieu and Taylor2007) to investigate the hypothesis of the cross-level effects from Level 2 (team) to Level 1 (individual). First, we tested Hypothesis 1, which states that an ethical climate is associated with low-carbon behaviour. We regressed lower-level low-carbon behaviour onto upper-level ethical climate using the following equation:

\begin{equation*}{\text{Low}} - \text{carbon behaviour} = {\beta _{\text{0}}}{\text{ + r}}\end{equation*}

\begin{equation*}{\beta _0} = {\Upsilon _{00}} + {\Upsilon _{01}}\left( \text{Ethical climate} \right){\text{ }} + {u_0}\end{equation*}

We tested Hypothesis 2 by asserting that ethical climate is related to environmental passion by regressing lower-level environmental passion on upper-level ethical climate, using the following equation:

\begin{equation*}\text{Environmental passion} = {\beta _0} + {\text{ }}r\end{equation*}

\begin{equation*}{\beta _0} = {\Upsilon _{00}} + {\Upsilon _{01}}\left( \text{Ethical climate} \right){\text{ }} + {u_0}\end{equation*}

For the lower-level direct effect (Hypothesis 3), we tested environmental passion in relation to low-carbon behaviour, using the following HLM equation:

Low-carbon behaviour = β₀ + β₁(Environmental passion) + r

\begin{equation*}{\beta _0} = {\Upsilon _{00}} + {\text{ }}{u_0}\end{equation*}

\begin{equation*}{\beta _0} = {\Upsilon _{01}} + {\text{ }}{u_0}\end{equation*}

To evaluate the mediation effect (Hypothesis 4), each component of the mediation path is used to estimate path a (X → M) (as indicated above). Then, the following equation was used to estimate M → Y, controlling for X (path b):

\begin{equation*}\text{Low} - \text{carbon behaviour} = {\beta _0} + {\beta _1}\left( \text{Environmental passion} \right){\text{ }} + {\text{ }}r\end{equation*}

\begin{equation*}{\beta _0} = {\gamma _{00}} + {\gamma _{01}}\left( \text{Ethical climate} \right){\text{ }} + {\text{ }}{u_0}\end{equation*}

\begin{equation*}{\beta _1} = {\gamma _{10}} + {\text{ }}{u_1}\end{equation*}

The Monte Carlo Method was used to formally assess the significance of mediation while considering paths a and b, as suggested in Hypothesis 4. According to MacKinnon, Lockwood, and Williams (Reference MacKinnon, Lockwood and Williams2004), Monte Carlo Method is appropriate for assessing mediation effectiveness in a multilevel model. We used a Monte Carlo Method with a 95% confidence interval and 20,000 replications, reporting the lower and upper levels.

Finally, we examined the moderating effect (Hypothesis 5) by implementing the procedures provided by Dawson (Reference Dawson2014) and computing the interaction variables using standardisation variables. Then, we plotted the interaction effects using the worksheet available at http://www.jeremydawson.co.uk/slopes.htm. We entered the coefficient values for the independent, moderator and product terms.

We present multilevel outcomes in accordance with the suggestions of Aguinis, Gottfredson, and Culpepper (Reference Aguinis, Gottfredson and Culpepper2013). The report included the intercept value (γ₀₀), the within-team Level 1 variance level (σ²), the intercept Level 2 variance (τ₀₀), the – 2 log likelihood (full information maximum likelihood estimation), the number of estimated parameters, and pseudo (R ²) value.