Do employees hold the key to environmental sustainability in tourism businesses? Empirical evidence from a field study

Abstract The tourism industry must reduce the negative impacts of its operations on the environment to secure its own prosperity into the future and to contribute to humanity’s collective aim of more sustainable production and consumption. An increasing number of studies in sustainable tourism have attempted to develop and test in the field the effectiveness of behavioural change interventions aimed at enticing tourism stakeholders in behaving more sustainably. These efforts have focused primarily on tourists. Employees have been largely neglected as potential agents of change, despite the substantial environmental consequences of their behaviours. This article pioneers this area of investigation. In a quasi-experimental field study conducted during regular operations of two hotels in Europe, we demonstrate that an equity-theory based behavioural intervention can successfully reduce the number of single-use shampoos dispensed by hotel cleaning staff during daily routine room cleans. Results are of immediate value to managers of tourism businesses – especially those of small and medium sized accommodation providers who do not have the financial means to make major infrastructure changes – by equipping them with a practical measure they can easily deploy to reduce the negative impact on the environment of their operations while also reducing their operating cost.


Introduction
The environmental sustainability of tourism businesses is predominantly seen as the responsibility of management (Kirk, 1995).As a result, investigations into how tourism businesses could be transformed to operate in more sustainable ways have focused on managerial action (e.g.Demeter et al., 2021), including the implementation of technical solutions (Butler, 2008) and participation in environmental accreditation programs.
The potential of targeting staff with behavioural interventions has largely been neglected, although staff behaviour is closely linked to resource consumption.In a hotel, for example, staff replace towels, use electricity to vacuum clean, use water to clean bathrooms, replace soaps and shampoo bottles and order, prepare, and serve food.Sharma (2002) acknowledges that the role of the individuals in affecting environmental change in organisations has been under researched (p.11).In addition, the current literature has little to say about the potential of targeting employees with the aim of increasing environmental sustainability of organisations (Robertson & Barling, 2013).Existing research focuses mostly on companies or groups of employees, neglecting individual managers and employees (L€ ulfs & Hahn, 2014).Yet targeting individual employees has the potential to reduce the negative environmental impacts of the organisation by changing their behaviours (O'Brien, 2015).Only few studies investigated the behaviour of employees as a pathway to reducing the negative environmental impacts of tourism (e.g.Pham et al., 2019;Su & Swanson, 2019;Zhang et al., 2021).The core construct in those studies is the motivation of employees to display pro-environmental behaviour at the workplace with a focus on changing their private behaviour, such as taking stairs instead of the elevator, switching off the lights in the office etc.(Chan et al., 2017;Fatoki, 2019;Luu, 2019a, Luu, 2019b, 2020;Pham et al., 2019;Rezapouraghdam et al., 2018;Zhang & Huang, 2019;Zientara & Zamojska, 2018).
The present study aims to contribute to this relatively neglected area of investigation.Specifically, we conduct a field study in which we test whether-using equity theory as the theoretical basis-sharing savings from more environmentally sustainable staff behaviour is effective in changing staff behaviours.If this approach emerges as promising, it may represent a win-win situation for staff (bonus payment), the tourism business (reduced expenses) and the environment (reduced negative environmental impacts resulting from tourism business operations; Demeter et al., 2021).
The result from this study has direct practical implications for tourism businesses.If the equity-based intervention tested in this field study is successful, it can immediately be implemented by tourism businesses to save operating cost and reduce their negative impact on the environment, while educating employees about and rewarding them for their environmentally friendly behaviour.

Theoretical background
The number of theories that could be used as the basis for this investigation is very large, as illustrated by Sovacool and Hess (2017), who extracted 96 theories from 22 different disciplines when asking social science experts about which theories could form a useful basis for the study of technology innovation.
Workplace pro-environmental behaviours are defined as any action taken by employees that she or he thought would improve the environmental performance of the company (Ramus & Steger, 2000, p. 606), or as changing the organisational practices to more environmentally sound ones (Ramus & Killmer, 2007, p. 557).Both definitions cover a wide range of behaviours and relate to individual-level actions within the organisation.Workplace behaviour can be divided into categories: behavior that is required and contributes to core business goals and behavior that is voluntary and contributes instead to the organizational, social, and psychological environment that provides the context for task performance (Norton et al., 2015, p. 105), or, alternatively, into the following major behavioural categories consisting of working sustainably, conserving resources, influencing others, taking initiative, and avoiding harm (Ones & Dilchert, 2012).
Theoretically, workplace pro-environmental behaviour is a type of prosocial behaviour (Ramus & Killmer, 2007), relating to individual actions that promote the welfare for individuals or organisations (Brief & Motowidlo, 1986).This behaviour is not formally required by employers (Mischel, 1973;Shamir et al., 1993) but has the potential to affect functional change in organisations and impact the value creation process within an organisation (Ramus & Killmer, 2007).Because employees often find themselves in situations without clear goals or certainty about the organizational rewards associated with the optional behaviour of promoting environmentally beneficial changes (Ramus & Killmer, 2007, p. 555), traditional organisational motivation theories are unlikely to be effective in triggering this type of prosocial behaviour (Shamir, 1991).Instead, well-established theories of human behaviour-such as the theory of reasoned action (Fishbein & Ajzen, 1975), the theory of planed behaviour (Ajzen, 1991), value-belief-norm theory (Stern et al., 1999)-can serve as the basis to explain workplace pro-environmental behaviour.
Because studying workplace pro-environmental behaviour from the perspective of cognitive attitudes only could unnecessarily limit the development of new effective behaviour change approaches, future research attention should be directed towards emotional aspects as individual-level behaviour motivators (Tian & Robertson, 2019).In addition, because people's behaviour in an organisation is not driven solely by individual-level motivators such as cognitive or affective personal intentions, supervisory support and organisational norms and values should be considered (Daily & Huang, 2001;Ramus & Killmer, 2007;Ramus & Steger, 2000).
Individual and organisational motivators of workplace pro-environmental behaviour have been studied extensively.Individual-level factors that impact on employees' voluntary pro-environmental behaviour include subjective norms (Bissing-Olson et al., 2013), conscientiousness (Kim et al., 2017), motivation (Graves et al., 2013) and empathy (Tian & Robertson, 2019).Employee attitudes and norms (individual-level motivators) and existing regulations, employee welfare and competitive pressures (organisational-level motivators) all emerge as important predictors of workplace pro-environmental behavioural intentions (Silverman et al., 2005).Perceived corporate commitment towards environmental protection predicts staff intention to display pro-environmental behaviour (Andersson et al., 2005).Attitudes of environmental managers towards pollution prevention and their perceptions of environmental regulation are highly associated with their past activities related to reuse of resources in their organisations (Cordano & Frieze, 2000).The intentions of small and medium enterprises to behave in environmentally friendly ways are driven by the economic motivation to improve business performance (Williamson et al., 2006).Personal norms towards the environment also emerge as predictors of energy-conservation behavioural intentions of employees (Scherbaum et al., 2008).A study from Canada shows that the pro-environmental behavioural intentions of leaders are likely to predict the passion of employees for the environment and their intention to behave in an environmentally friendly way at the workplace (Robertson & Barling, 2013).Awareness of the organisation's written environmental policy as well as the support of supervisors of environmental actions further increases the likelihood of employees expressing their intention to engage in environmental initiatives at the workplace (Ramus, 2002).
A review of 69 empirical studies on employee pro-environmental (green) behaviour (Norton et al., 2015) differentiates between required and voluntary pro-environmental behaviour of employees and develops a conceptual framework that postulates that five types of outcomes can result from such behaviour: institutional outcomes (e.g.gaining a competitive advantage), organisational outcomes (e.g.cost savings), leader outcomes (e.g.leader effectiveness), team outcomes (e.g.positive social norm), and employee outcomes (e.g.future intentions).Both required and voluntary pro-environmental behaviour by employees are defined by three groups of factors: contextual (institutional, organisational, leader, team), personal (between-person factors, such as personality, environmental attitudes, job factors, behaviour; within-person factors: motivation, intentions), and motivational states (controlled, as "I have to", and autonomous, as "I want to").From their review, Norton et al. (2015) identify gaps in existing research, including: a lack of understanding how to encourage pro-environmental behaviour among employees without strong environmental attitudes; a lack of research into how employee personality can be accounted for; and a lack of consideration of the financial (e.g.reducing the cost) and non-financial outcomes (e.g.motivation of the employee) for organisations.
Other research has explored the influence of organisational factors on pro-environmental behaviour of employees.Results show that organisational support for the environment (e.g.Cantor et al., 2012;Lamm et al., 2015), strategic human resource management practices (e.g.Paill e & Raineri 2015), environmental infrastructure (e.g.Holland et al., 2006;Houten et al., 1981) and incentives (e.g.Graves et al., 2013;Tam & Tam, 2008) are significant predictors of workplace pro-environmental behaviour.
In prior work investigating employee pro-environmental behaviour, actual behaviour is rarely measured (Norton et al., 2015).Among the few exceptions are studies concluding that a relaxed working atmosphere and a positive attitude of staff towards the environment motivates employees to perform their tasks in more environmentally friendly ways (Robertson & Barling, 2013).Employees with less positive attitudes towards the environment are more likely to engage in pro-environmental behaviour when they are feeling emotions such as enthusiasm and excitement.Pro-environmental behaviour was measured by observing employees in their daily tasks and taking notes on their behaviour.Similarly, Tudor et al. (2008) measured workplace pro-environmental behaviour by using self-reported behavioural intention and actual behaviour (measured by analysing and observing waste bins).They conclude that pro-environmental attitudes, beliefs, motivation, awareness (at individual level) and organisational focus, structure, type, size, and culture (at organisational level) predict sustainable waste behaviour.Installing visible recycling facilities (implementation intervention) significantly improved the recycling behaviour of the employees (Holland et al., 2006).
Studies that investigate employee behaviour in the tourism context use self-reported behaviour measures to find significant impact of environmental knowledge, environmental awareness, environmental concern (Chan et al., 2017), environmental values, green organisational climate, affective organisational commitment (Zientara & Zamojska, 2018), employee's work ethic, hotel's environmental benefits sharing system, ecological embeddedness of the employee (Peng & Lee, 2019), the role of corporate social responsibility, organisational trust, organisational identification, employee's wellbeing, leadership, institutional support, workplace spirituality, national park goal identification and attitude toward environmental corporate social responsibility (Fatoki, 2019;Luu, 2019b;Su & Swanson, 2019;Zhang et al., 2021) on employee's pro-environmental selfreported behaviour.
We conclude from the analysis of prior work that the management literature distinguishes between individual level (cognitive or affective) and organisational level (supervisor behaviour of organisation norms and values) motivators of workplace pro-environmental behaviour.Behaviour is typically measured as self-reported behavioural intention, reducing external validity of findings.Overall, past research suggests that cognitive behavioural theories can be used to explain proenvironmental behavioural intentions in the workplace, but no prior study attempted to implement a practical measure to motivate employees to improve their self-reported or actual pro-environmental behaviour at work.Our study fills this gap, using equity theory as the theoretical basis.
Equity theory (Adams, 1963) is a social exchange theory.The basic idea of social exchange theory-first proposed by Homans (1958) as part of his investigations into human behaviours in small groups-is that social interactions are an exchange in which the participants seek to maximize their benefits (the rewards they receive minus the costs they incur) within the limits of what is regarded as fair or just (American Psychological Association, n.d.).The notion of reciprocity is central to the fairness of exchange; people expect to receive the same amount of net benefits from a social interaction as they offer.
Equity theory postulates specifically that people involved in a social interaction experience a feeling of tension if there is an asymmetry in benefits to their advantage or disadvantage.Such tensions are overcome by taking corrective action: if the imbalance is to the person's disadvantage, they adjust their behaviour and give less.If the imbalance is to their advantage, they give more to establish equity.
Despite being critiqued by many scholars for a range of reasons, including lack of consideration of heterogeneity, equity theory is popular among workplace researchers because it is parsimonious and has high predictive validity for constructs such as performance and job satisfaction (Al-Zawahreh & Al-Madi, 2012).The implications of equity theory have widely been adopted by employers to maximise the performance of their employees.The use of equity theory to increase employee productivity assumes that: employees want to be fairly rewarded for their contributions; employees use inputs and outputs to "calculate" reward; employees determine fairness through social comparison with other employees; employees determine fairness specifically for their circumstances; and employees adjust their efforts to establish what they perceive as an equitable balance of benefits gained by themselves and their employer (Carrell & Dittrich, 1978).
Equity theory has also been used extensively in the tourism literature as a theoretical lens for the analysis of social interactions between various stakeholders.Equity theory has proven its potential as a theoretical basis for behavioural interventions targeting tourists: offering hotel guests to share savings from waived daily room cleans significantly reduced the number of room cleans without reducing guest satisfaction (Dolnicar et al., 2019b).Equity theory has not been used as the basis for behavioural interventions targeting employees.We pioneer this approach in the present study.

Methodology
We conducted a field study in two hotels in Ljubljana, the capital of Slovenia.Both hotels are four star properties focusing on business tourists belonging to the same hotel chain: Grand Hotel Union (120 rooms) and uHotel, formerly known as Grand Hotel Union Business (176 rooms).
We chose business hotels because prior work has revealed that business guests typically care less about (environmentally unsustainable) extras in hotel rooms because, to them, the hotel stay is not an experience, but just part of their work life (Knezevic Cvelbar et al., 2017).In both hotels management was supportive and fully involved in the process, which was critical to the success of field studies.
Before the field study was conducted, both hotels used small single-use plastic bottles for shampoos and body lotions and small individually wrapped soap bars.The policy at both hotels was to replace only empty shampoo and body lotions bottles and to add new soap bars only if open soap bars had been used.Regarding towels exchange, cleaning staff were instructed to replace towels left on the bathroom floor, while replacing those left on the drying racks only every three days.Implementing the policy was at the discretion of cleaning staff and hotel management reported that cleaning staff were not diligently complying.Management believed that motivating cleaning staff to improve their environmentally friendly behaviour could help substantially to reduce negative impacts of the hotel on the environment.In our experiment the following hotel services with sustainability implications were used as dependent variables: shampoo bottles, body lotion bottles, soap bars and towels.
Our equity-theory based intervention involved messaging that emphasised the shared benefits of staff and hotel resulting from more pro-environmental staff behaviour.During the regular weekly housekeeping staff meeting, the housekeeping manager communicated the following message: Every week hotel guests use … bottles of shampoo, … soaps; … body lotions and … towels.If you can reduce the use of cosmetics and towels by 20% or more without impacting on guest satisfaction, we will pay a 50 EUR salary bonus at the end of the month.This means we will equally divide among us the savings from handing out fewer items.In addition, staff across both hotels were informed that: This program will run for two weeks and you will be informed on savings.Remember that the necessary condition is that customer satisfaction does not drop.Staff was reminded that guest satisfaction is a priority.Therefore, additional shampoos, body lotions, soaps and towels could be provided upon request by the guest.
The incentive payment of 50 EUR was selected in consultation with hotel management, who believed that this was sufficient to trigger behavioural change.The 50 EUR salary bonus was awarded only to staff who achieved the goal of reducing the use of cosmetics and towels by 20% or more.Study-specific guest satisfaction could not be measured because both hotels have a standardised guest satisfaction questionnaire they were unable to modify.Neither hotel management nor employees observed an increase in complaints during the field study.
The first 14 days of the field study served as control condition; both hotels operated as usual.The communication with staff during the 14 days of the control condition was as usual and we did not include any additional communication regarding the reduction of items dispensed or replaced towels with the employees.During the following 14 days, we implemented the equitytheory-based intervention.Management was instructed to communicate with employees the same cleaning procedure during the control and experimental condition (timing for room cleaning; cleaning routine etc).
Both hotels use the same housekeeping software system.Both hotels used this software prior to the experiment and staff were familiar with the software.Cleaning staff in both hotels use mobile devices to report the time of entering and exiting the room, the room number, the number of towels changed, and the number of shampoo bottles, body lotions and soaps replaced.The software also allows entry of missing items in the mini bar and repair requests.Data are automatically saved and were shared with the research team along with aggregate anonymised guest mix data.The system is also used for assigning and providing the housekeeping attendants with the daily room cleaning schedule.Both hotels use the same environmental signs inviting tourists to reuse towels and guests request towel exchange or indicate that they are willing to re-use their towels.Soaps at both hotels are not in bottles, but rather soap bars.
We collected data pre-COVID on 3173 tourist overnights, 1480 in Grand Hotel Union and 1693 in uHotel.Ethical clearance for this fieldwork approach was obtained from the University of Queensland's Human Ethics Committee (approval number 2015001475).Table 1 summarises the number of tourist overnights (Nights) for each hotel separately for each experimental condition (Condition) and states whether the nights are the last night of the guest party before check out (Last Day).The occupancy rate was higher during the experimental conditions than during the control condition for both hotels.Cleaning staff, however, have clear instructions to keep a fixed procedure for room cleaning.Variation in occupancy, therefore, should not influence the results.Guests can also be assumed to behave in the same way in both experimental conditions.Table 1 also contains the number of guest parties generating these overnights and the number of different cleaners servicing the rooms for these nights.Each night is associated with one room clean; Table 1 shows the average number of towels, shampoo bottles, soaps and lotion bottles dispensed per room clean, along with the standard errors of the means in parentheses.As can be seen, the average number of dispensed items varies across item type-towel, shampoo, soap, or lotion.The average number of dispensed items is higher when the room is cleaned after the guest party checks out because all items must be replaced in preparation for the next check in.We therefore exclude from the subsequent data analysis the last night.
In experimental studies the assignment of subjects to experimental conditions must be random.This is not easily achieved in field experiments which take place during normal operations of a business (which is why this study is a quasi-experimental field study).However, both hotels where the study was conducted do not have a system of assigning specific rooms to specific guests.Receptionists assign the rooms based on availability and clearing status.While not perfectly random, it is reasonable to assume that there is no systematic bias in the room assignments that could create an artificial association between certain types of hotel guests and specific cleaners assigned to certain rooms.
The hotel data base contains the following guest socio-demographics and travel characteristics: check in date and check out date, length of stay, age, gender, country of origin, number of persons in room, type of guest (business, leisure or not recorded) and room rate.Table 2 summarises these characteristics by hotel and experimental condition.For metric variables Table 2 shows means and standard deviations (SD).For categorical variables Table 2 shows relative proportions.Results indicate that guest composition varied considerably across experimental conditions.Statistical hypothesis tests for equality of means (two-sided Welch t-tests) across the experimental conditions for each of the hotels for metric variables and homogeneity of distribution (v 2 -tests) for categorical variables (Hothorn & Everitt, 2014) reject the null hypothesis at the 5% significance level for all variables except age for both hotels and gender for Grand Hotel Union only.To account for these differences in guest composition we include these covariates as control variables in the subsequent analysis.
The effect of the intervention was assessed separately for each of the four items potentially dispensed during a room clean for all nights except the last night before check out.We fitted a mixed-effects Poisson regression model using maximum likelihood estimation with a log link (Fahrmeir et al., 2013), including the following variables in the regression analysis: number of dispensed items as dependent variable, the experimental condition as independent variable and hotel, logarithmised number of people in the room, length of stay, age, room rate, purpose, gender, and origin (with categories as in Table 2) as control variables as well as a random effect for the cleaner.A likelihood ratio test compares this model to the same model without the experimental condition as independent variable to assess the effect of the intervention.

Results
Figure 1 illustrates the average number of towels, shampoo bottles, soaps and lotion bottles dispensed per person and room separately across hotels and experimental conditions (for room cleans that did not occur at the day of check out).In addition to the average numbers indicated by the bars, approximate 95% confidence intervals for the average counts are superimposed via a vertical line.As can be seen, the average number of items dispensed per person and room either remained approximately the same or decreased.The decrease is particularly pronounced for shampoo bottles where the confidence intervals do not overlap.This indicates that the observed decrease is larger than expected by chance under the assumption that the experimental intervention has no impact.Figure 1 is only an exploratory analysis of the results controlling for the hotel and the differences in guest party size, but not for other differences in guest party composition across the two experimental conditions.The results of the regression analysis (that accounts not only for hotel and guest party size, but also for length of stay, age, room rate, purpose, gender and origin as control variables) indicate that the equity-theory-based intervention was effective in reducing resources use in some instances, but not others.Not significantly reduced were towel replacement (coefficient b ¼ À0.0561, p-value ¼ 0.31), soap replacement (coefficient b ¼ À0.2280, p-value ¼ 0.11), and body lotion replacement (coefficient b ¼ À0.2685, p-value ¼ 0.18).The intervention did successfully reduce shampoo replacement (coefficient b ¼ À0.5026, p-value < 0.001).
These results are promising; they demonstrate that the equity-theory-based intervention tested in this study indeed has the potential to increase pro-environmental behaviour in hotel staff and warrants further investigation.The reason that only shampoo replacements dropped significantly is explained by the nature of the other dependent variables: towel reuse is actively encouraged by the hotel and guests already indicate their preference regarding exchange or keeping the towels in both conditions which prevents cleaners from being able to reduce the towel re-use further.Given that guests indicate their preference for towel re-use, cleaners in the experimental condition did not have the opportunity to decide not to replace the towel.Soaps at both hotels are dispensed as soap bars and guests typically open and use one soap bar.It is very difficult to use more than one soap bar during the relatively short stay of an average traveller.As a result, soap bar use in the control group use was already so low that further improvements were not possible.Similarly, guests do not display much interest in using the body lotion provided.As Table 1 shows for the control condition, one lotion is only replaced for every third room for both hotels in case the room clean is after the guest party departed and replacement is even less likely for room cleans during the stay at the hotel.The potential for improvement, therefore, is low for body lotion also.
As opposed to towels, soap bars and body lotion, shampoos are used by many guests with a shampoo bottled being replaced for seven to eight out of ten rooms for both hotels in case the room clean is after the guest party departed and for four to six out of ten rooms in case the room clean is not on the last day of stay.As a result, the potential for reducing replacements when it is not necessary is much higher for this dependent variable.The equity-theory-based intervention harvests this opportunity effectively.

Discussion
This paper addresses the influence of organisational factors on pro-environmental behaviour of employees in the hotel industry.Aligned with previous research, our results show that strategic human resource management practices (e.g.Paill e & Raineri 2015) and incentives (e.g.Graves et al., 2013;Tam & Tam, 2008) are significant predictors of workplace pro-environmental behaviour.Our results also show how important managerial actions are to transform tourism to operate in more sustainable ways (Demeter et al., 2021).Adding to the body of literature that focuses on motivating employees to display private pro-environmental behaviour at the workplace (Chan et al., 2017;Fatoki, 2019;Luu, 2019a, Luu, 2019b, 2020;Pham et al., 2019;Rezapouraghdam et al., 2018;Zhang & Huang, 2019;Zientara & Zamojska, 2018), our study shows how to successfully motivate the employees to display work related pro-environmental behaviour.These are in particular important findings for the hotel industry, as pro-environmental behaviour of hotel staff, due to the contact between staff and consumers which is more personal and direct than in other industries, also impacts on tourists' perceptions and behaviour (Batstic & Gojcic, 2012).
Existing studies tackling work related pro-environmental behaviour of employees were using the theory of reasoned action (Fishbein & Ajzen, 1975), the theory of planed behaviour (Ajzen, 1991) and value-belief-norm theory (Stern et al., 1999).No prior study attempted to implement a practical measure to motivate employees to improve their self-reported or actual pro-environmental behaviour at work using equity theory as the theoretical basis.Our results show promising results and potential for using equity theory in future research in this area.
This study has several practical implications.Most importantly, it offers a tangible practical measure that can easily be adopted by accommodation providers to entice their cleaning staff to behave in ways that are more environmentally sustainable while at the same time reducing hotel expenses for single use amenities.Due to privacy reasons, hotels did not agree to share data on staff that received the incentive.Management reported a monetary benefit due to intervention, yet the major motivation for management were not monetary benefits but more conscious consumption and behaviour of the housekeeping staff.
The proposed approach is of particular value to the many small and medium accommodation providers who are not in the position to implement expensive infrastructure changes.We expect also that the tested intervention will be more effective in small and medium sized accommodation businesses where owners have closer personal relationships with their cleaning staff.This assumption, however, should be empirically tested by running a comparative intervention study across both small and medium sized accommodation providers and large hotels.
Using equity theory as the basis for such behavioural interventions is particularly attractive because it leads to benefits for all involved: employees gain knowledge on how to behave environmentally friendly and benefit from a monetary bonus for displaying such behaviour.The business benefits from reduced operating cost.The environment benefits from increased sustainability of business operations.The intervention can easily be adapted to other employee behaviours, including reducing food waste, reducing electricity use, reducing water use, and increasing recycling.

Conclusions
This study investigates the potential to target behavioural interventions at staff working in tourism businesses in the pursuit of increasing the environmental sustainability of business operations.The equity-theory based intervention tested in this study involved inviting staff to reduce the dispensing of unnecessary items during daily routine room cleans and offering to share with them the resulting monetary savings.The intervention proved effective for the resources category that has a relatively high usage: shampoos.For other items, such as towels, soap bars and body lotions the intervention could not be empirically shown to be effective.This is likely to be due to the very low baseline use of these items in business hotels.These results lead to two key conclusions: (1) employees in tourism businesses do represent an important target group for behavioural change interventions aiming at increased environmental sustainability of business operations, (2) significant improvements can be achieved for items that-under normal operating procedures-are heavily used, (3) organisational factors have a strong impact on workplace proenvironmental behaviour and (4) managerial initiatives and strategic human resource management practices can significantly improve workplace pro-environmental behaviour of employees.Use is determined by the nature of the item (in our study shampoos were heavily used, whereas soap bars and lotions were not).s The key limitation of this study is that the fieldwork was conducted in two business hotels.Previous study shown that business travellers are not prone to pro-environmental behaviour, yet they use less energy, water or towels than the leisure guests (Knezevic Cvelbar et al., 2017), making it more difficult to achieve an improvement.The fact that our intervention was successful under these most difficult of experimental circumstances does, however, point to the substantial potential of interventions targeted at staff to materially improve the environmentally sustainable business operations.We expect that replicating this field study in a hotel catering primarily to leisure tourists would achieve substantially higher savings.
Another limitation is that we did not separate out the effect of the intervention and the effect that may have been caused merely by running any experiment and, in so doing, drawing attention to staff in relation to pro-environmental action.To isolate the two possible effects a future study would need to run a second experimental condition without offering an incentive.While this would be interesting from a theoretical point of view, findings from such a study would not change the practical implications of this study that a simple equity-theory based intervention has the potential to alter employee behaviour to be more sustainable.
Finally, study findings are limited to short-term effects of the intervention, as the ongoing long-term effect on pro-environmental behaviour of cleaning staff was not measured.This is a limitation of all field experiments and quasi-experimental studies testing sustainability interventions in tourism.Future studies could set up measurement systems, optimally automatic measurement systems that record behaviour automatically and on a continuous basis.Data of this kind would permit conclusions to be drawn about both short-and long-term effects resulting from interventions.Previous studies have drawn conflicting conclusions about whether financial incentives translate into desired behaviours only on the short term, potentially undermining long-term behavioural change (e.g.Moller et al., 2012), or whether the short-term behavioural change triggered by the incentive translates into long-term behavioural change (e.g.Mitchell et al., 2020).Measurement systems that capture behaviour continuously over time would help resolve this question.There is also a risk of perverse negative spill over effects of incentivising pro-environmental behaviour; staff may choose to stop behaving environmentally friendly unless there is an incentive.Furthermore, future studies should include more communication with staff to explore how changing the cleaning routine impacts on the daily routine and long term proenvironmental behaviour of staff.

Disclosure statement
No potential conflict of interest was reported by the authors.

Notes on contributors
Ljubica Kne zevic Cvelbar is professor of tourism at the School of Economics and Business, University of Ljubljana, and a visiting professor at more than 30 universities worldwide.She holds a PhD in Economics at University of Ljubljana, while her expertise is in the fields of tourism economics and sustainable development.Ljubica has 15 years of academic and professional experience as a professor, researcher and consultant.She also act as a president of Strategic Partnership for Research and Innovation in Tourism, a cluster of more than 60 key stakeholders in tourism in Slovenia.
Bettina Gr€ un is an assistant professor at the Vienna University of Economics and Business.Her research interests include model-based clustering, statistical computing and statistical applications in particular in economics, business and tourism.
Sara Dolnicar is a professor of tourism at the University of Queensland.Her current research program develops and experimentally tests measures that trigger tourist pro-environmental behaviour.She has conducted most of her experimental studies in collaboration with her Slovenian colleagues in hotels in Slovenia.Professor Dolnicar was awarded the Travel and Tourism Research Association Distinguished Researcher Award in 2017, and named the Slovenian Ambassador of Science in 2016, the highest honour the Republic of Slovenia bestows on expatriate Slovenian researchers in recognition of global excellence, impact, and knowledge transfer.

Figure 1 .
Figure 1.Changes in resources use across hotels and experimental conditions.

Table 1 .
Descriptive statistics of the data set.

Table 2 .
Descriptive statistics of the guest parties by hotel and experimental condition.