Managers’ and employees’ views of critical aspects for alcohol abuse prevention at small and medium enterprises – the case of ski resorts in Northern Sweden

ABSTRACT Seasonal workers at tourism resorts, particularly young employees, represent a high-risk population for alcohol abuse and other negative lifestyles. This study explores how managers and employees at ski resorts in Northern Sweden experience critical aspects for hazardous alcohol use at small- and medium-sized enterprises. This analysis is based on one open-ended question about hazardous alcohol use prevention at the workplace answered by 153 employees, and semi-structured interviews about being a manager at a ski resort conducted with 13 managers. The analysis showed that managers’ views regarding hazardous alcohol use prevention were double-edged, which was confirmed by the employees’ answers. The managers contribute to the preservation of existing alcohol norms but simultaneously enforce policies and administer tests to control employees’ sobriety at work. Three themes emerged: reproducing existing alcohol norms, an alcohol-controlled working environment and care-taking managers. The results identified a need for early prevention initiatives focusing on the norms and work cultures of small- and medium-sized ski resorts. In relation to the theory of planned behaviour, the analysis illuminates how evaluations of behaviours related to alcohol consumption are communicated in policies and through leadership. In addition, the importance of social pressure regarding alcohol use is clarified in the material.


Introduction
Seasonal workers at tourism resorts represent a highrisk population for alcohol abuse and other negative lifestyles (Kelly, Hughes, & Bellis, 2014;Vinberg & Warne, 2015). Seasonal employment is more common among young adults than older adults (Larsson, 2014), and more men (8%) than women (5%) are employed in seasonal work compared with short-term employment in general (Statistics Sweden, 2014). In Sweden, this age group also has the highest rates of hazardous alcohol use, with 31% of men and 50% of women assessed as at-risk consumers based on the Alcohol Identification Test (AUDIT; Källmén, Wennberg, Ramstedt, & Hallgren, 2015). Younger people are more vulnerable to hazardous alcohol use than older ones, and this behaviour has been identified as a major health risk factor in terms of disability-adjusted life-years (Gore et al., 2011). Alcohol is a major risk factor for the global burden of disease (Casswell & Thamarangsi, 2009) and contributes to absence due to sickness, accidents and work-related injuries among employees (Hermansson, 2004). Based on these facts, alcohol consumption has a detrimental effect on the working environment. Hotels and restaurants are common at tourist resorts, and seasonal workers are overrepresented in this industry (Larsson, 2014); this situation sets the stage to study the possibilities for managers to reduce alcohol consumption among their staff. In Sweden, the Working Environment Act (SFS 1977(SFS :1160(SFS , 1977 regulates employer responsibility; this act states that it is employers' duty to prevent ill health, injuries and accidents at work and to promote a healthy work environment. This responsibility is independent of company size. Many enterprises at ski resorts are small-or medium-sized, and the managers play important roles in how well occupational health and safety issues are addressed by the organisation (Mattson, 2015). However, being a manager of small-and medium-sized enterprises (S.M.E.s) means that the work is often hectic, and the content is varied and fragmented. Furthermore, many activities are reactive, interactions involve a great deal of oral communication, decision processes are disorderly, and most planning is informal and adaptive (Yukl, 2009). These circumstances can hinder the implementation of policies aimed at addressing occupational and health issues, including alcohol-related problems (Whysall, Haslam, & Haslam, 2006). On the other hand, facilitating factors can also exist in S.M.E. s, including supportive managers, local control over budget spending for health and good communication among managers and employees (Whysall, Haslam, & Haslam, 2006). Some researchers even argue that S. M.E.s feature positive characteristics for implementing improvements: the work of the employer and his or her employees is less hierarchical, information is generally comprehensive and easily accessible, and a positive information flow exists (Meggeneder, 2007).
In relation to this background, the study's purpose is to understand the critical aspects of hazardous alcohol use and drug use prevention as well as the role that subjective norms play in hazardous alcohol use prevention at ski resorts by questioning managers and employees.

Previous research and theoretical framework
As mentioned above, hazardous alcohol use can have detrimental effects on the working environment and employees' health if preventive measures are not implemented. Babor et al. (2010) divided alcohol use prevention into seven areas: pricing and taxation; regulating physical availability; modifying the drinking environment; drink-driving countermeasures; early prevention; restrictions on marketing; and education, persuasion and treatment (Babor et al., 2010, pp. 243-248). For example, early intervention can incorporate screening for risk consumption, brief advice and the implementation of policies and technical equipment to hinder drunk driving (World Health Organization, 2012). Therefore, it is reasonable to believe that many problem drinkers can be reached via early intervention at the workplace. However, previous research does not provide clear advice regarding how early prevention should be developed. For example, studies of the effects of alcohol policies on alcohol consumption have yielded contradictory results. Ames, Grube, and Moore (2000) compared two similar organisations in a large industrial company. They found that in the organisation with a developed alcohol policy and more consistent social control, significantly less alcohol was consumed in connection with work than in the organisation where that policy was lacking. In contrast, Anderson, Chisholm, and Fuhr's (2009) review revealed little evidence that workplace alcohol policies result in changes to drinking norms or reduce harmful drinking behaviour. Another problem is that policy evaluations have traditionally focused on the establishment of policies and not the outcomes or knowledge of why and how they work (Treno, Marzell, Gruenewald, & Holder, 2014).
Brief intervention programmes take a broader perspective on prevention. A review by Tanner-Smith and Lipsey (2015) on the effects of alcohol consumption among adolescents and young adults showed that a brief intervention programme, including multi-component programmes not exclusively focused on alcohol, could produce significant results. These results indicate that to reduce hazardous alcohol use, it is important to adopt a holistic view of alcohol and address how hazardous use can be prevented. In a brief intervention study conducted in a high-risk workforce of young adults in Australia, Pidd, Roche, and Fischer (2015) found that alcohol abuse was reduced and that mental wellbeing increased in the intervention group compared with the control group. The intervention was a brief training programme that included stress reduction, coping and communication skills, and the second part sought to reduce alcohol and drug harm (Pidd et al., 2015). Some researchers have suggested that prevention programmes using motivational approaches with brief individualised feedback are effective at reducing hazardous drinking (Anderson & Larimer, 2002;Baer, Kivlahan, Blume, McKnight, & Marlatt, 2001).
Living and working at a ski resort engenders special circumstances for both managers and employees. For example, work and leisure time tend to be mixed. The business and employment schedules are temporary and linked to the season. Thus, health interventions (e.g. those for hazardous alcohol prevention) might be a bigger challenge for managers in this vocation than for those in other organisations. For many young women and men, seasonal employment at a tourist resort denotes a lifestyle choice. In addition, seasonal employment is often a first job, and it might be associated with the first time staying away from home without one's parents for a longer period. Many young and old seasonal workers return each season, and some employees continue with a summer season after being at a ski resort (Möller, Ericsson, & Overvåg, 2014). A study of the motives for seasonal work in the Swedish mountains found that social aspects such as skiing and being in a popular place with many people and parties were motivating (Möller et al., 2014). Research has also shown that seasonal workers in tourist resorts constitute a risk group because seasonal employment is often associated with a culture characterised by high alcohol consumption, drug availability and risky sexual behaviours (Briggs, Tutenges, Armitage, & Panchev, 2011;Kelly et al., 2014). The party culture, and sometimes occasional risk-taking behaviours associated with ski resorts might be considered as a construction of masculinity (Willott & Griffin, 1997). Men and woman are involved in normalisation processes involving alcohol and binge drinking (Brown & Gregg, 2012), and if these behaviours are based on social constructs more than biological causes, then men and women will act in accordance with the ways in which femininity and masculinity are outlined in the cultural context (West & Zimmerman, 1987).
Social change theories (e.g. the theory of reasoned action [T.R.A.]; Fishbein & Ajzen, 1977) are one way to understand how culture and norms affect hazardous alcohol and drug use, focusing on interaction between individuals and in relation to the social environment. The T.R.A. is based on the assumption that individuals usually change their behaviours based on intention. Intentions indicate the motivation of people (i.e. how much people try and the level of effort they apply) to change their behaviours (Ajzen, 1991). Ajzen (1985) developed the theory further to create the theory of planned behaviour (T.P.B.) and added "perceived behaviour control" as the third social-cognitive variable to the original that included attitudes and subjective norms. Together, the theoretical components of the T. P.B. are (1) attitudes towards a behaviour (i.e. a person's evaluation of his or her behaviour, for example, those related to alcohol consumption); (2) subjective norms, a person's evaluation of the social pressure to change (e.g. their alcohol consumption); and (3) perceived behaviour control (i.e. the degree to which a person feels capable of performing the behaviour; Ajzen, 1991). This theory has been used in relation to health behaviours (Godin & Kok, 1996) and to predict alcohol consumption and binge drinking in companies (Hagger et al., 2012); to our knowledge, however, it has rarely been used in qualitative research. In this study we use T.R.A. and T. P.B. to understand the role that subjective norms play in hazardous alcohol use prevention.
As described above, it is important to adopt a holistic view of hazardous alcohol and drug use (Tanner-Smith & Lipsey, 2015) to better understand how S.M.E.s can work to reduce this behaviour among their seasonal employees. In this study the T. P.B. (Ajzen, 1991) is used to interpret the results. The study's purpose was to understand the critical aspects of hazardous alcohol use and drug use prevention, as well as the role that subjective norms play in hazardous alcohol use prevention, at ski resorts by questioning managers and employees.

Methods
A qualitative design was selected to achieve the aim of this study. Our assumptions about the reality and roles of people in this context are based on the idea that people are largely formed by and part of their surrounding environment. A holistic approach (i.e. one in which the whole is more than the sum of its parts) is appropriate to examine the complex interactions that might occur in organisations. This methodological approach enables us to consider whether phenomena exist in a certain context, in order to try to understand why people act and think as they do. Based on a holistic perspective with an empiricalholistic viewpoint, knowledge about the phenomenon under study should be obtained by questioning people about how they perceive reality and their roles within this reality. From these aspects, it is important to grasp how people from different levels within the organisation view their roles and the study phenomena. If the researcher wants to understand the research area in a holistic way, then it is recommended that he or she use more than one method (Morse & Chung, 2003); therefore, different types of data (i.e. interviews and open-ended questions) were collected.

Participants and procedure
This study was performed between 2014 and 2015 as part of a hazardous alcohol and drug use prevention project funded by the Public Health Agency of Sweden. The project was conducted in a municipality in the northern part of Sweden where several ski resorts are located, and was led by a project leader together with managers from different tourism industries and the health sector. Key industries in this area are the S.M.E.s in the tourism sector. Each year, approximately 2000 individuals work in seasonal employment, and approximately 60% live temporarily in the municipality during the season.
Two groups of participants were included in this study. The first group was a sample of all staff (n = 1313) across 48 enterprises. These participants were invited to complete a questionnaire about alcohol, drugs, health and psychosocial working conditions. The employees were employed in S.M.E.s of various sizes (ranging from 5 to 250 employees). In total, 611 participants (46 [4%] men and 53 [6%] women) completed the questionnaire. The response rate for the total sample was 47%. The largest age group was 18 to 24 years old (51%), and this group included significantly more women (61%) than men. To complete the survey, a questionnaire with a cover letter was delivered to the company managers, by the project leader (from the municipality). In this study, the term "manager" is used to denote a person who has a responsibility to lead staff in the company, including preventing ill health, injuries and accidents at work and promoting a healthy work environment. The managers then distributed the questionnaire to their employees. Participation was voluntary. Employees who participated could choose to leave the questionnaire in a sealed envelope with their manager or send it via post to the researchers. The questionnaire addressed areas such as employment relationships, social aspects, psychosocial working conditions, health, alcohol consumption and preventive measures. In the present study, one openended question from the questionnaire was selected: "Do you think that your employer needs to do more in [the prevention of] alcohol and drugs?" A total of 153 participants (70 male, 83 female) answered this question by writing short-or medium-length responses addressing the managers responsible for staff.
The second group consisted of managers representing a variety of companies in the Swedish tourism industry. An invitation to participate was sent to 22 managers via e-mail a week or two before the author called them. Thirteen managers (five women and eight men) agreed to participate. These participants were middle managers who had worked in the tourism industry for between 5 and nearly 30 years. Semistructured interviews were conducted to investigate their experiences of managing seasonal employees at a ski resort. Semi-structured questions were asked about hazardous alcohol use prevention, the presence and use of alcohol and drug policies, managers' responsibilities with regard to prevention, the role of alcohol in the company and in employees' lives, and the opportunity for employees to enter rehabilitation programmes. The interviews were performed in the participants' offices or another place at their company, and every interview lasted between 30 and 90 minutes. The interviews were audio recorded, and an independent researcher transcribed the recordings verbatim. One interview was conducted on the telephone per the informant's desire, and their narration was written down and not audio recorded. One limitation of this method was the quality of the audiorecorded data. Some of the managers decided to be interviewed in a public area (e.g. a lounge or office shared with others). This decision led to distracting sounds from the surrounding environment, which in turn led to poor sound quality. Data from inaudible parts of the recordings were not transcribed.

Ethical considerations
The managers were given information about the study and were informed that their participation was voluntary before the interviews were conducted. They were informed that their participation in the study could be ended at any time without giving any reason. Written consent was received from each participant. The questionnaire was distributed by the project leader to the managers of the enterprises, who handed them out to their employees. Participation was voluntary. Employees who participated could choose to hand in the questionnaire in a sealed envelope that was given to their manager or to send it by post directly to the researchers. Data was handled confidentially, meaning that the content is presented in a way that makes it impossible to link actual individuals to the results.

Analysis method
During the first stage of the study, the interviews with the managers were analysed following Braun and Clarke's six phases for thematic analysis (Braun & Clarke, 2006). Initially, the authors became familiar with the data; the scripts were read several times, and ideas that arose were written down. Short sentences were then marked and sorted into nodes using the QSR NVivo 10 software. After the initial coding process, the next step was to search for themes and check the relationships between the themes and codes to develop an initial thematic map of the candidate themes (Braun & Clarke, 2006). In the second step, the employees' answers to the open-ended questions were analysed. Their answers were sorted into categories, which were then compared with the codes and categories developed from the interview scripts. The categories were then included in a thematic map to obtain a visualisation of the relationships among them and to search for patterns (Braun & Clarke, 2006).

Results and analysis
In this section, we present the results from the thematically analysis based on the questionnaire and interviews. Three themes were developed: reproducing existing alcohol norms, an alcohol-controlled working environment and care-taking managers (see Table 1). These themes are discussed in relation to the T.P.B. (Ajzen, 1991) with focus on attitudes towards behaviours and subjective norms and in relation to previous research. Every theme then concluded with summary reflections. Some categories are dominated by data provided by managers, and some are dominated by data provided by employees. The name of the ski resort was replaced with an X when mentioned by the respondents.

Reproducing alcohol norms
Under the theme "reproducing alcohol norms", categories were included about the normalisation of drinking behaviours and managers sending mixed messages, not only in following an alcohol and drug policy but also encouraging drinking or looking the other way when employees are under the influence of alcohol at work.

Normalisation of alcohol
The managers' backgrounds as seasonal workers at ski resorts with their own experiences of a party culture complicated their approach to hazardous alcohol use prevention. Some of them stated that hazardous drinking was more common when they were young seasonally employed workers and therefore downplayed the problem. Others saw the problem more clearly and said, "It is easy to become an alcoholic in this environment". Thus, in relation to the T.P.B. (Ajzen, 1991), managers' own experiences influenced the evaluation of behaviour related to alcohol consumption. Alcohol was presented as an important part of ski resort culture, and some managers simply said that they did not want their staff to abstain. The managers were also aware that the main reason for young people coming to the ski resort was not to work: it was for the opportunity to party and ski, and the work was merely a way to pay for them to be there. One of the managers said, Just look at the village: All those young people don't come up here to work; that's not the major driving force for going to X [the ski resort] to work. They go to X to ski and because it's fun, to party. Work is something you do so that you can afford to stay up here, and somewhere along the line we need to try to combine these interests (laughing). (IP7) This passage demonstrates the normalisation of alcohol use, where the attitudes towards the specific behaviour (i.e. alcohol use) are made sense of in relation to one's own experiences of working at a ski resort. The normalisation of alcohol is an on-going process in which abuse is explained in relation to age ("they are young", "socially insecure", "they want to meet a partner") or as a treatment, an exuse to use alcohol to handle various health issues ("he had a severe headache"). The managers said that most of the employees who came to the resort had fun, behaved well and did not drink too much. In contrast with drugs, alcohol was observed as a part of the culture; from this point of view, however, hazardous alcohol use was considered as more problematic to manage in certain aspects. The manager quoted below found it easier to have a zero-tolerance policy for drugs than to interpret the limits of "too much" alcohol: To catch an alcohol addiction and try to work with that, it's kind of a long process, not something that you manage to do in the first 4-5 weeks when you haven't even met a person, and then 25% of the time has passed. While there's a zero-tolerance policy for using drugs here, there is a significantly higher [tolerance] for alcohol. It's kind of socially accepted in a different way. Here, you can just sit with them and ask, "Are you drinking?", "Are you happy about how much you consume or do you feel that it's too much?" While if it involves drugs, you can make this clear right off the bat: "Look, this is not acceptable, it's not even open for discussion, whether it's too little or too much, because we have a zerotolerance policy." So that's easier for us to approach. (IP3) Investigating normalisation as well as illuminating attitudes and subjective norms was also conducted by asking the managers, "Would it be acceptable if the employee had no interest in alcohol?" (Ajzen, 1991). The material revealed that abstaining required an excuse but was not considered as unthinkable by this manager: This quotation also shows that attitudes and subjective norms among managers can lead to an increased intention to continue to drink (Ajzen, 1991). The context in which we live and the socioculture factors that dominate this context have a large effect on drinking patterns (Gordon, Heim, & MacAskill, 2012). However, Gordon et al. found in their review that drinking cultures have changed dramatically over the past 30 years, are now more homogenous among countries and are strongly influenced by Anglo-US culture. The former division between wet and dry culture (in which Sweden was categorised as wet) is no longer useful (Savic, Room, Mugavin, Pennay, & Livingston, 2016). The norms and culture surrounding alcohol are related to gender, age, masculinity, class, place and manner; when and how much people drink change not only across entire cultures but also within different subcultures (Savic et al., 2016). Thus, it can be argued that the managers maintain an old drinking culture that no longer exists in society as a whole.

Mixed messages
This category was developed after analysing employee answers to the questionnaire because the same result did not explicitly appear in the managers' narrations. Some employees questioned the liberal alcohol norm and made suggestions about restrictions. They implied that their manager should not encourage alcohol use at staff parties and should be more restrictive in giving away free alcohol. One male employee said, It becomes odd when the employer offers alcohol at staff parties when at least half of the staff has to get up early for work the next day and some people work evenings. Clearer alcohol-free alternatives and less of an emphasis on partying and drinking is what's strived for.
Another employee made suggestions about better restrictions to prevent alcohol problems, including drinking and driving: Stop selling alcohol on credit (i.e. you as an employee can have a glass of wine after work and then it's deducted from your salary). I think that would benefit those who drink too much. Perform an alcohol test in the morning if someone has to drive a snowmobile.
These quotations from the employers' pinpoint the subjective norms described in the T.P.B. (Ajzen, 1991); thus, the discussions about the mixed messages from the managers illustrate that the employees perceived a social pressure with regard to alcohol use. Interestingly, the social pressure includes both an expectation that the individual should drink alcohol in their spare time but remain sober at work. The problem with the amorphous boundary between one's leisure time and work time is significant for seasonal work because employees live and work with each other at the ski resort. Seasonal workers are "tourists", employees and guests all at the same time. A female employee asked for guiding principles about workers' drinking even during their spare time: "Maybe it needs to be clarified that you should not be at your workplace if you're under the influence of drugs/alcohol. [I] mean that you shouldn't be there, even if it's your day off." Attitudes towards alcohol and social pressure reproducing alcohol norms at the ski resort In summary, the theme reproducing alcohol norms illustrates that cultural norms across ski resorts, including those of managers, hinder changes in drinking habits such as binge drinking. The aforementioned aspects can be related to the T.P.B. (Ajzen, 1991) and the component attitudes towards behaviours and subjective norms. For example, it can be assumed that the managers experience a social pressure not to actively change alcohol consumption, and they do not have the capability of applying behaviours focused on alcohol/drug prevention among their employees. Similary, John and Alwyn (2010) found in their review of alcohol-related social norms among university students that a drinking norm that included the senior management level places pressure on students to adopt the university's drinking culture. In the current findings, employees told us about managers who send mixed messages by having an alcohol and drug policy while also ignoring hangovers at work or encouraging drinking. That is, sometimes employees express other norms, than the attitudes towards behaviours that are approved, and even encouraged, by the managers. In this way, employees are pressured to drink, albeit unconsciously. Sending such mixed messages decreases the managers' credibility with regard to hazardous alcohol use prevention. Based on the concluding remarks above, it is possible to change a drinking culture because it is a social construct. Managers have the opportunity to introduce healthier lifestyles in their organisations, not only those related to sports activities but also other lifestyle behaviours such as drinking patterns and eating healthy food.

An alcohol-controlled working environment
The second theme found in the material, "alcoholcontrolled working environment," indicated that the managers used a policy that can be described as an "alcohol and drug testing policy" for preventive measures. It also describes the managers' ambivalence to testing workmates and employees who would prefer that professionals perform the tests.

An alcohol and drug policy to control employees
Nearly all of the managers stated in their interviews they had some type of drug and alcohol policy at their company or some type of document or guideline regarding alcohol use and hazardous alcohol use. Occasionally, the policy was local, developed together with the union; more often, however, it was developed by the head office, and the same policy was used for the entire company. The role of policies is especially interesting in relation to the T. P.B. (Ajzen, 1991); thus, a policy can be understood as a formally expressed evaluation of behaviour and social pressure. The current study found that alcohol and drug policies were primarily observed as documents linked to the employment contracts that were expected to be signed by employees upon recruitment. This contract might also give the company the right to test or control the staff with regard to the abuse and misuse of drugs. One female manager used the policy frequently when employment began. The policy was signed by the employee and then displayed afterwards several times a month at the table in the canteen, mentioned on the organisation's website or on Facebook, and used to remind employees suspected of having problems. Managers representing S.M.E.s stated that they have few opportunities to both educate and introduce employees to work and talk about hazardous alcohol use prevention. One manager said, . . .we're not bigger and we don't have more power than. . . before the season starts in December. Possibly, [we] take a half-day when everybody is gathered. At that time, my co-worker and I go over things before the season starts, and one of the things we discuss are the rules at the workplace. . .. it's not our own alcohol and drug policy but more of a staff guideline. I'd like to emphasise that what the staff receives are 5 sentences concerning alcohol and drugs. (IP1) Managers also mentioned that they have other restrictions regarding, for example, how many glasses of alcohol are served at a personal event. This type of restriction was not a part of the written policy, but it can be considered as an expression of social pressure when it comes to alcohol use within the working environment (cf. Ajzen, 1991).
The basic concept I have kind of involves your position as [a manager]; I've never said, "No, we're not going to supply any alcohol," but it will be very limited. Some companies kind of have an unlimited supply, but I stand by that. . . I think that 2 units is enough, and no hard alcohol. (IP2) In addition to a written policy, supervisor training, employee assistance programmes and drug testing are important components. One problem consistent with the managers' comments provided in this study is that managers have limited opportunities to educate employees about hazardous alcohol use prevention issues, despite the regulation concerning work adjustment and rehabilitation (AFS 1994:1, paragraphs 10 and 13). According to Frone (2013), five components are important in comprehensive workplace substance use programmes: supervisor training should include information about specific drugs, the employer's written substance use policy, potential workplace performance, behaviour problems associated with substance use and impairment at and outside of work.

Ambivalence regarding testing workmates
Several managers were clear about the necessity of controlling employees' drug and alcohol use, not only illustrating their attitudes towards testing but also illuminating a unity regarding their evaluations of the risks of alcohol use at work (cf. attitudes towards a behaviour, Ajzen, 1991). Vehicles and cars were often equipped with a breath-alcohol test. The employees also provided suggestions for a breath test for snowmobiles and piste-bashers via the questionnaire. One company conducted random drug tests in which drugs found in urine were tested with "nano sticks". The team managers performed the tests, and the informants discussed this policy as being somewhat problematic. The relationship between the team manager and the tested employee is sometimes close, and the situation was perceived as troublesome. Another problematic aspect was testing older employees when the test manager is young, especially when the result is positive. Respondent: …it's a slower process [to conduct tests] in the departments that have a higher average age….
[but] it's pretty simple for us when we're working with rather young people. I think it's perceived as more embarrassing and that it's harder for… well for those [who are older]. It's when I have been out there performing drug tests, that when I need to perform a drug test on people over the age of 60, they think I'm just some young punk, and "why the hell are you doing this?" Interviewer: mm Respondent: …kind of… but at the same time, it's my job so they have no choice but to [take the test]. (IP5) One of the managers wanted to leave the testing to a professional at the healthcare centre but knew that this suggestion was not realistic because of the associated higher costs. Therefore, they had to address it themselves. Others wanted the opportunity to test the worker, but the union did not approve general testing.
In the open-ended questions, the employees also suggested that the tests should be administered by external staff instead of the team managers because of the risk that these tests would be frivolously handled: ". . .more unannounced drug and alcohol tests, preferably by outside companies or businesses; otherwise the tests will most likely become too unreliable and chummy." Some employees noted the problem that although the policy mentioned drug tests, they were conducted infrequently. Thus, they were perceived as an empty threat.

The role of policy and testing in an alcoholcontrolled working environment
A report from the European Commission (European Commission, 2011) based on a systematic analysis of research concluded that a clear written policy is most likely of great importance for preventing alcohol problems in the workplace. However, few studies have investigated the effects of such policies. One study that included 10,000 employees (Ames et al., 2000) found that the policy should be written, clear and list rules with consequences if those rules are not followed. A Swedish report (Sverige Företagshälsor, 2016) emphasised that a policy must also include routines for rehabilitation treatment for employees who engage in the risky consumption of alcohol and that the policy should be part of the regulations included in a systematic work environment document (AFS, 2001:1, paragraph 11).
In summary, nearly all of the managers reported having an alcohol and drug policy. These policies were developed in different manners; sometimes the policy was considered as a separate document, and other times it was included in a handbook for workers. Most often, it was included as document to be signed by the employee as a concession to follow the company's rules. The document also included an agreement to be subjected to drug tests. The results of this study indicated that possibilities exist to increase managerial competencies and skills via developed training activities. Furthermore and consistent with brief intervention research (Tanner-Smith & Lipsey, 2015), managers should use a holistic approach to drug prevention (Anderson et al., 2009). A policy written from a holistic perspective in line with Tanner-Smith and Lipsey (2015) would provide an opportunity to promote employee health, prevent problems related to alcohol and drugs, provide advice when problems arise and have a plan when rehabilitation occurs.
Several managers were clear about the necessity of alcohol and drug tests; however, they also concluded that these tests are problematic when close relationships with employees exist and that it is cost prohibitive to assign this task to professionals outside the enterprise. Importantly, alcohol and drug testing can be used properly to minimise false positives or negatives (Frone, 2013). To achieve this goal, one can use personnel from occupational health services, who might also engage in early prevention activities such as screening employee health behaviours and providing feedback to employees with alcohol or drug problems (Sveriges Företagshälsor, 2016).

Care-taking managers
"Care-taking managers" is a theme that describes how managers act like parents with control and love while also problematising this style of managership from the aspect of an employee's age and position.
Acting like a parent When managers were asked whether they could describe any other type of preventive work or activities related to alcohol or drugs, aside from policies or drug or alcohol testing, most of them could not. However, several of the informants described different types of caring activities performed for their employees. They also pointed out the importance of staying healthy by eating fresh food and exercising. One manager encouraged a young employee who too often sat in the bar drinking beer after his shift to take a ski tour instead, reminding him of his intention of be out skiing when the season began. Others, upon witnessing certain employees spending too much time partying, took them out to climb a mountain and tour a beautiful place in the environment. Another example of caring included being a mother more than a manager; this manager said: . . .as an employer at X, you're not an employer; you're somewhat of a semi-psychologist, a replacement mother. Since a lot of these people are away from home for the first time, they're living in staff quarters together with people who are also in similar situations, which makes you have a somewhat more personal relationship with the employees, compared to what you would have in some large company at a different location. This results in you sometimes having more of a "mother dialogue" with your employees, rather than a "boss-employee dialogue". (IP6) Taking care was mentioned in relation to meeting and speaking with employees every day, watching their faces and eyes and even smelling their breath. Some managers were also aware that a great pressure exists in living together in company housing. They have no mother or father to come home to, and it is difficult to stay away from a party in your own house. Ski resorts lack enough apartments to house all of the seasonal workers, so the companies often have company housing available for their staff. The workers rent a room of their own in an apartment or share a room in a corridor with others, sharing a common kitchen and living room. Some managers said that they monitored company housing or even made inspections; others said restrictions existed regarding alcohol in their housing units. Managers who referred to themselves as being parent-like figures were also more involved in company housing: Even if you feel that you don't want to drink, it's not that easy to just stay on your bed in your room if ten other people who live in the house start a party. . .. In those situations, we try to help suppress [the problem] in our free time in the staff quarters if we're the ones supplying the accommodations. (IP6) The manager above discussed how employees behave like others in the same situation, which might be related to the T.P.B. (Ajzen, 1991) and the two interacting components regarding subjective norms: beliefs about how other employees would like them to behave and judgements about these beliefs (i.e. if the employee has strong control over his or her drinking behaviour and the feeling of being able to say no to partying). If the subjective norms are too strong, then a manager might increase the pressure by offering another accommodation. This arrangement with shared apartments also shaped certain social controls within the staff. A female employee living at one of the company houses said, [I] work nights now, so [I] never see large parts of the staff. But I know [that] drug use often occurs in some staff quarters, so they should have drug tests and breathalyser tests in the morning since some of the staff clearly have problems with alcohol and are out every night, regardless of whether or not they're working early the next day.
From the aspect of being a caring manager, age was discussed, and the caring approach was not useful. The managers said that it was often more problematic to address the situation when an older employee exhibited problematic drinking behaviour than when a younger one did. Older employees sometimes denied that they were drinking too much, in contrast with the awareness of young employees, who often agreed and apologised for their behaviour; ". . .let's use an older person as an example, say 35 years old; there's a denial when you talk to them, they would prefer to act unknowingly. . .". That is, the managers noted that it can be problematic to address older seasonal employees and that managers take more responsibility for permanent employees and put more rehabilitation effort into this latter group. This result is consistent with earlier research showing that non-permanent employees are subject to more pronounced task-oriented leadership than permanent employees, who are subject to more relationshiporiented leadership (Gossett, 2006;Svensson, Vinberg, & Larsson, 2015). In relation to the T.P.B. (Ajzen, 1991), this distinction can be understood as attitudes towards a specific behaviour (i.e. drinking alcohol) that vary between various groups of employees. That is, the manager's attitudes towards employee drinking behaviours are related to the age of the person, the length of the employment and the role of the employer (i.e. based on whether the employer is a key person).
Being a manager at S.M.E.s involves long and irregular working hours, great work demands and a lack of resources, all of which can increase the risk of stress and role conflicts (Davidsson, 2004;Gunnarsson, 2010;Nordenmark, Vinberg, & Strandh, 2012). Task delegation, an option for managers at larger enterprises, is one method for shifting time demands, but it is not possible to the same extent for those working at S.M.E.s. However, the managers in this study expressed several care-taking activities because they acted like parents and encouraged employees to perform physical activities instead of drinking at a bar or spending time at a party. Other similar activities include talking to employees every day and examining how they feel and whether they have been drinking alcohol in company housing. Research has shown that these relationship-oriented components of leadership are associated with employees' job satisfaction and self-rated health outcomes (Larsson & Vinberg, 2010;Nyberg et al., 2005;Skakon et al., 2010). Interestingly, acting like a parent obviously includes various strategies based on norms (cf. Ajzen, 1991). This scenario not only includes leadership and formal alcohol prevention but also social pressure (Ajzen, 1991) expressed through care.
The managers who care about the (young) employees clearly try to motivate this group of employees to drink less. We assume that this behaviour, when performed regularly, also reduces employee alcohol consumption and contributes to other healthy behaviours. The results of our study indicate that acting like a parent is ineffective with older and permanent employees. However, research on trust has found that using different leadership strategies for different categories of employee risk taking creates a culture of mistrust that can be negative for the involvement of employees in different company activities (Rothstein, 2005;Sztompka, 1999).
Caring for key staff In contrast to the above descriptions of caring for young seasonal workers, other managers said that they took greater responsibility for permanent employees because of the problems caused by seasonal workers' too-brief employment period. Distinctions made between permanent staff and seasonal staff or between key staff and "more ordinary" staff were a recurring theme among the managers throughout the interviews.
Naturally, I feel that an employer has a responsibility, absolutely; but I think that the problem is that you're responsible for this person for so few weeks that I think that sometimes you need to look the other way because of the fact that there are so few weeks in a season. I think that, as an example, in a company that has a lot of people who are employed until further notice, a lot of people who work there all year round, you feel a greater responsibility towards your employees than you might perhaps feel with this type of employment. I can't say for sure that this is the case, but it's what I think. (IP1) Some managers made distinctions when abuse was detected with regard to a key person versus less important staff members, illuminating the underlying attitudes towards alcohol use among various groups of employees (cf. Attitudes towards a behaviour, Ajzen, 1991). The difference was described as follows: if the chef was a heavy drinker, then managers more often than not turned a blind eye because it was not possible to recruit a new chef in a brief period of time. By contrast, a key person, depending on his or her position, might obtain treatment more easily.
. . .the gems, you have you try to give more hours to or employment during the off-season. And sure, you can try to stop these people from getting sick of things; you make an extra effort to solve problems or if they need to take time off to do [something] and be here. In that case, you try to find a solution. While when Bob, who's here for the first time, comes to you, you say, "Nope, we're not going along with this, you have to do this," meaning you. . . well yeah, you treat people differently. (IP2) One manager described the feeling of pressure and sadness when she was prompted by the business owner to dismiss an employee instead of offering rehabilitation. She meant that the reason was not lack of knowledge but rather lack of money: . . .we're in an. . . industry that's unbelievably shortterm in terms of staff. It's only a matter of a few months. To make everything run as efficiently as possible and try to make what little money there is to be made during the few months that we operate in. (IP1) The manager continued, I've worked and have seen a world where you can work in an entirely different way, long-term with staff and problems. Now I'm in a position in a company where I really have a manager and an owner who actually feel that they don't want to cover the cost of Ann receiving company healthcare when we have her on the payroll for 5 more weeks. In situations like that, the order from them will be, "Get rid of Ann!" Then, it will be my responsibility to take care of that, and it's really tough a lot of times. A lot of the time, I feel really badly about what I need to do to get the person to quit. It's hard, it's not really about your own knowledge, it's about money. . . the people who own the business really don't want to spend any money on that [staff member]. (IP1) Even the seasonal employees asked for more equal opportunities for rehabilitation if hazardous alcohol use, drug use, or both were discovered. Employees also noticed differences between how staff groups were treated compared with permanent or temporary employees. A couple of employees noted that offers were unequal, depending on the terms of employment: "Offer the seasonal employees help if they are abusing drugs and alcohol too, not just those who have permanent employment." Others asked for better discussions about alcohol and drug abuse and wanted their manager to provide help to workers with problems.

Care-taking as a strategy for prevention
In summary, many of the employees in our study mentioned that the managers successfully addressed aspects related to alcohol and drugs. This result is positive, given that research regarding small and medium-sized tourism businesses shows that "employee-friendly managers" with characteristics such as empathy and fairness can positively influence employee motivation and satisfaction (Peters, 2005). Management attitude is also one of several factors that influence the willingness of seasonal employees in the ski industry to return to seasonal employment (Ismert & Petrick, 2004). John and Alwyn (2010) found that alcohol interventions should incorporate numerous social cognition components, of which personal feedback is one important aspect, together with drinking prototypes and normative feedback. Educated managers aware of alcohol norms and how to prevent risk behaviours might be able to reduce risky consumption and drug use among employees.

Discussion
Seasonal employees, particularly young men and women at ski resorts, are a high-risk group for developing hazardous drinking, drug use and other negative health behaviours. Therefore, it is important to develop early prevention initiatives at both societal and organisational levels that focus on norms and work culture concerning hazardous alcohol and drug use. Critical aspects of alcohol and drug prevention identified from the present study included managers having a double-edged relationship with alcohol by acting to increase alcohol consumption while maintaining the alcohol norms and policies based on drug testing without a holistic perspective. Managers were also a critical aspect related to the relationship with different groups of employees.
Of particular importance is changing the culture and standards regarding alcohol and drugs. Employers who present an alternative to the current standard of the "party culture" have opportunities to influence prevailing norms. One option might be related to percieved health trends in today's society, mentioned in the interviews with the managers. Another important issue is employees' living conditions, because a drinking culture can be created in company housing where many temporary employees stay during the season. These characteristics can be understood in relation to the T.P.B. (Ajzen, 1991) where we have seen several examples of how evaluations of behaviours related to alcohol consumption are communicated in policies and through leadership. In addition, the importance of social pressure regarding alcohol use is illuminated in the material. The third component in the T.P.B. regarding the degree to which a person feels capable of performing the behaviour is also relevant in relation to the study results. For example, the manager's complex roles across many different work tasks and his or her limited knowledge about how to handle different groups of alcohol and drug problems can be hindering factors.
This study's purpose was to understand the critical aspects of hazardous alcohol use and drug use prevention as well as the role that subjective norms play in hazardous alcohol use prevention at ski resorts. The conclusions from this study are that the managers express ambivalence towards alcohol and drug policies as well as towards the testing of the employees, which is also reflected in the attitudes towards drinking behaviours and the social pressure related to alcohol (cf. Ajzen, 1991). These subjective norms are intertwined in how managers perform caring activities as part of their leadership and, through those behaviours, try to change negative attitudes among employees. A review analysing the correlations among the T.P.B., the intention to drink alcohol and actual alcohol consumption found that attitudes and subjective norms had the strongest relationship (Cooke, Dahdah, Norman, & French, 2016), and the authors recommended that alcohol consumption reduction interventions should focus on these variables.
The intention of this study was to deepen the understanding of the critical aspect regarding hazardous alcohol use prevention using qualitative methods. Thematic analyses (Braun & Clarke, 2006) are useful for interpreting both the interviews and the answers to open-ended questions. This approach was selected because the interpretive aim was to deepen the understanding of the critical aspects associated with hazardous alcohol and drug use prevention at S.M.E.s (specifically ski resorts) by asking questions of both managers and employees. The research design suited the study, and it was possible to answer the research question about the critical aspects of alcohol and drug prevention by including both manager narrations and employee replies. Through the data comparison, emerging themes were strengthened based on the employees' answers.
The concepts of credibility, originality, resonance and usefulness outlined by Charmaz (2006) make it possible to judge the trustworthiness of the present study. The method and analysis were described in detail to allow the reader to form an independent assessment of the study's credibility. The resonance was assessed through presentation and a discussion of the early results to a group of S.M.E. managers at the ski resort. The study was motivated by its original approach to interview managers in this enterprise at a ski resort because this group has special assignments as managers. The results are important and hopefully usable, even for other managers with young seasonal employees at different types of resorts.

Implications
Given that many workplaces at ski resorts are S.M. E.s indicates that one can successfully integrate work-related hazardous alcohol and drug use into occupational health and safety issues. Therefore, a need for more collaboration exists between enterprises and occupational health services that can assist managers with different issues related to hazardous alcohol and drug use and with the education of both managers and employees. One strategy might be to create manager networks that have been successful in other working environment areas, focusing on training and discussions about alcohol and drugs in the workplace. Another strategy might be to involve employees in developing a local policy for a healthy working life. Importantly, managers should think about their own attitudes concerning these issues to improve their leadership behaviour. The limited knowledge of alcohol and drug problems among young seasonal employees at ski resorts indicates the need for more quantitative and qualitative research to foster a deeper understanding of the mechanisms that explain high alcohol and drug consumption. For example, we need to know more about motivating factors for being seasonal employed at a ski resort, especially regarding the role of individual alcohol consumption. Also, it is important to illuminate the recurrent seasonal workers and their living and working conditions.

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

Funding
This work was supported by the Public Health Agency of Sweden [grant number 03198-2015-6.2].