The Role of Mood in Managing Small Businesses–Evidence from Dairy Farming

The aim of this paper is to explore how the mood of managers of small family owned firms influence on involvement from other family members, relations with business consultants, and the decision to maintain production. We randomly drew a sample of 90 dairy farmers who were visited and interviewed. First we did a qualitative analysis of the data and then we merged the interview data with a database showing who had maintained dairy farming six years after the interviews. Our results show that farmers in a happy mood get more help from the rest of the family in farming activities than farmers in a neutral or bad mood. Farmers in a bad mood are less satisfied with the consultancy service than farmers in a happy or neutral mood. Six years after the interview more farmers in a happy mood maintained dairy farming as compared to farmers in a neutral or bad mood. Our findings are relevant to all managers of small family owned businesses.


Introduction
Although the experience of work is saturated with emotion, research has generally neglected the impact of everyday emotions on organizational life (Ashforth & Humphrey, 1995). However, emotions are an integral and inseparable part of organizational life, and they can profoundly influence humans' performances across a diverse range of tasks (Scwarz & Skurnik, 2003). Therefore we wanted to study the implications of different moods among managers drawing on case studies of Norwegian dairy farmers. We wanted to explore how different moods influence involvement from other family members in farming activities. Further, we wanted to explore if farmers' moods can explain why some farmers are more satisfied with the consultancy service than others, and whether moods influence farmers' decision to maintain dairy farming. Unlike most previous studies we did not want to use experiments to explore the effects of mood on performance. Instead we aimed at an informal setting with personally experienced problems and behaviors that actually occur in natural settings. To answer our research questions we selected a random sample of farmers all over Norway and conducted 90 in-depth interviews during farm visits. long-term venture survival. Studies of job satisfaction have established a positive relationship between emotional stability and job satisfaction (Tokar, Fischer & Subich, 1998;Judge, Heller & Mount, 2002).
Moods and emotions are states that have an experiential, cognitive, and physiological component. Emotions reflect the on-going, implicit appraisals of situations with respect to positive or negative implications for the individual's goals and concerns (e.g. Arnold, 1960). While emotions are specifically directed towards something, the concept of mood refers to the feeling state itself when the object or cause is not in the focus of attention. In contrast to object-specific emotions, moods are either experienced as directed at multiple objects or exist as objectless background dispositions. Moods can e.g. be classified as happy, neutral or bad (Scwarz & Skurnik, 2003). Moods are defined as relatively long lasting affective states that are experienced without concurrent awareness of their origins (e.g., Schwarz & Clore, 1988). People are often unaware of the causes of their moods, which may include minor events or background variables such as lack of exercise (Thayer, 1996). Hence, moods lack a specific referent and usually come about gradually, are of low intensity, and may endure for some time. In ordinary language we typically say that we are "in" a good mode, while we are angry "about" something. As compared to emotions moods remain more in the background, and this diffuse and unfocused property accounts for their pervasive influence (Scwarz & Skurnik, 2003). Research has shown that people do self-regulate their mood, and individuals differ in how they regulate mood. Thus Thayer, Newman and McClain (1994) found that the most common self-regulating strategies used to eliminate bad moods were; call, talk to, or be with someone (54%), control thoughts (51%), listen to music (47%), avoid the thing causing the bad mood (47%), and try to be alone (47%). The most effective self-regulating strategy was exercise (Thayer et al., 1994).
Traditionally psychological research on emotions has been problem oriented and focused on factors which bring about negative emotions related to e.g. stress or conflict. In contrast, modern psychological research has focused on how positive emotions play out in work life and how such emotions influence behavior in organizations (Kaufmann & Kaufmann, 2011). Similarly, research has focused more on how positive emotions influence pro-social behavior, e.g. like helping colleagues at work. In line with this tradition we are primarily concerned with the effects of positive moods, and less with what the causes of different moods are.
Emotions are critical for guiding interpersonal relationships (Keltner & Kring, 1998;Shiota et al., 2004), and emotional exchanges can either maintain and enhance positive relationships, or become a source of antagonism and discord (Harker & Keltner, 2001;Shiota et al., 2004). Friendly people are easy to get along with others, to get along in a group, do not attack others and easily establish friendly relations with others (Digman & Inouye, 1986). They believe that people are good and are well-intentioned. Contrary, people with low friendliness easily show their emotions, are more likely to show aggressive behavior and language, and are less likely to establish friendship with others. Thus research has shown that people who are in a happy mood get more contacts at work and get more social support from their colleagues as compared to people who are in a bad mood (George, 2000;Tsai, Chen & Liu, 2007). Further, people in a happy mood are valued more positively by others than people who are in a bad mood (Staw, Sutton & Pelled, 1994). In short we perceive people in a happy mood as more attractive than people in a bad mood (Kaufmann & Kaufmann, 2011). Like in other family owned small businesses family members such as spouses, parents or children are involved in running the farm, and we expect them to participate more in farm work the better the farmers' mood.
Hypothesis 1: Farmers who are in a happy mood get more help from other family members than farmers who are in a bad mood.
Similarly, we think the farmers' mood will influence their relationship to farm consultants. Consultants from TINE visit each dairy farmer several times a year and deliver services in dairy farming. Typically they interact with farmers in solving problems related to dairy farming, such as milk quality, animal health, feeding schemes, farming economics and so on. The interaction between consultants and farmers involves extensive knowledge transfer. Knowledge transfer can be defines as "the communication of knowledge from a source so that it is learned and applied by a recipient." (Ko, Kirsch & King, 2005, p. 62). There is evidence that such knowledge transfer is influenced by the personal relationship between consultants and clients, the degree of shared understanding between them, and both parties intrinsic motivation (Ko et al., 2005;Szulanski, 1996). Thus we think that if the farmer is in a bad mood, this can have a negative effect on the transfer of knowledge. The result might be that the farmer finds the consultancy service to be of little use. Contrary, if the farmer is in a happy mood, we think this will improve the interaction with the consultant.
Hypothesis 2: Farmers in a bad mood are less satisfied with the consultancy service than farmers in a happy mood.
To the best of our knowledge the knowledge on how managers' moods affect strategic decisions such as to close down operations is limited. The mood-congruent retrieval view on moods contends that positive affect serves as a retrieval cue for positive material in memory (Isen, Shalker, Clark & Karp, 1978;Teasdale & Fogarty, 1979;Teasdale & Russell, 1983). Negative affect, however, activates significantly less information. Such differences in the content of what information is brought to mind have been shown to have important implications for a broad range of judgments and evaluations (see e.g. Kahn & Isen, 1993;Kraiger, Billings & Isen, 1989). According to Schwartz's (1994) cognitive tuning model, negative moods are associated with a more effortful, analytic style of information processing, whereas positive moods are associated with a more carefree, heuristic style of information processing (Kaufmann & Vosburg, 1997). Other studies have shown that participants in a bad mood adopt a more local focus whereas those in happy moods apply a global focus (Clore & Huntsinger, 2007). Thus research has supported the broaden-and-build theory (Fredrickson, 2001; which hypothesizes that positive emotions broaden the scope of attention and thought-action repertoires. There is also evidence that a negative mood reduces the amount of attentional commitment to the task at hand and enhances the focus on task irrelevant personal concerns (Smallwood, Fitzgerald, Miles & Phillips, 2009). Similarly, there is evidence that positive mood both impairs working memory (Martin & Kerns, 2010) and improves working memory (Yang, Yang & Isen, 2012). Finally, there is also strong evidence that mood can influence sports performance and competitive achievement (Barkhoff, Pagano & Heiby, 2007;Skinner & Brewer, 2004;Cockerill, Nevill & Lyons, 1991). However, some findings are mixed, partly because mood has not been measured sensitively (Lowther & Lane, 2002).
Positive moods signals that an object of judgment is valuable, leading to a positive evaluation, and negative mood signals that it lacks value, leading to a negative evaluation (Clore, Wyers, Dienes, Gasper, Gohm & Isbell, 2001). Thus mood can influence satisfaction with a product or service (Sirakaya, Petrick & Choy, 2004). Related to job perseverance, Seligman and Schulman (1986) showed that insurance agents who scored high in optimism stayed twice as long in their jobs as compared to their pessimistic colleagues. Similarly, Tsai, Chen & Liu (2007) found that positive moods increased perseverance in job tasks. To conclude it is safe to assume that moods can influence decisions. The specific effect, however, depends on which decision the mood is brought to bear. We expect farmers' mood to influence how perseverant they are in dairy farming. Further, in hypothesis two we proposed that farmers in a happy mood will get more help from family members, including their children. Such involvement of children in farm work can increase the likelihood that they will take over the farm. Thus it is likely that farmers' mood can affect how willing their children are to take over the farm and maintain milk production. We therefore hypothesize a positive relationship between happy mood and how long milk production maintains on the farm. To explore this relationship we checked whether farmers or their successors still produced milk six years after the interviews, i.e. in 2013.
Hypothesis 3: Milk production maintains longer on farms where the farmer is in a happy mood, as compared to farms where the farmer is in a bad mood.

Methods, Data and Respondents
In this section we describe the respondents, sampling procedures and choice of variables and statistical methods. Finally, we comment on the validity and reliability of the study.

Research Design
To test the hypotheses we aimed at an informal setting with personally experienced problems and practical behavioral solutions. The farmers were allowed to speak about how they run their farm, their interaction with other family members, their interaction with consultants and the challenges they faced in their own words. Several of them commented that it was both more satisfying and more interesting to be interviewed compared to filling in a questionnaire. One of them stated: "Well that's one of the problems with all these questionnaires that pop up, it's quite difficult to answer, sometimes it is impossible, because the questions require more than one cross, you know…" The dialogue with the researchers gave the farmers the possibility to ask control questions to check their understanding of the questions. Similarly, for us as researchers it was useful to be able to ask follow up questions to make sure we had understood and interpreted the farmers in the right way. We could be flexible and depart from the interview-guide when necessary, tailoring the interviews to each farmer and their specific context.

Methodological Choices
We wanted to use the strengths of interviewing to collect data related to the farmer. To design the interviews we applied the critical incident technique (Chell, 2004), and we asked farmers to tell the interviewer about how they run their farm and how they interacted with consultants and family members in running the farming. Our aim was to see the research topic from the perspective of the interviewee, and to understand how and why they came www.ccsenet.org/ijbm International Journal of Business and Management Vol. 10, No. 2;2015 to have that perspective. We used a largely unstructured interview to capture the respondents' thought processes and their frame of reference. The critical incident technique facilitates the investigation of significant occurrences (events, incidents, processes or issues) which the respondent identifies and manages (Chell, 2004).
To analyze interviews we applied attributional coding (Silvester, 2004). Causal attributions refer to the explanations we make for our own behavior, the behavior of other people and the events that we observe or hear about from others. The Leeds Attributional Coding System (LACS: Munton, Silvester, Stratton & Hanks, 1999) is a method that involves coding spontaneously produced spoken attributions. Using this system allows researchers to explore and code attributions made by one person about their own behavior or the actions of other people or entities (Silvester, 2004). Attributional coding can be a rewarding method for exploring how individuals make sense of their world (Silvester, 2004). An advantage of the method is that the method focuses on naturalistic data and events rather than hypothetical scenarios created by researchers.

Respondents and Sampling
The aim was to generalize the patterns we found among the respondents to all Norwegian dairy farmers, i.e. across the population in various settings as e.g. geographical area and time. To ensure validity across geographical areas in our study we included farmers from each of five dairy regions in Norway, from Rogaland to Nordland. In each region we picked a geographical area that was manageable for the researchers to visit without spending too much time travelling. At the same time it should be representative for the region with respect to farm size and farming conditions. Within these geographical areas we selected respondents randomly among members in TINE Efficiency Analysis (TEA), a consultancy tool which consists of financial data from the farm accountancy supplied with biological data. We drew a number that corresponded to the whole regions' fraction of the total number of dairy farmers in Norway. For example in region South, which consists of Telemark, Agder and Rogaland we chose Rogaland. Similarly in region West, consisting of Hordaland, Sogn og Fjordane and Sunnmøre, we chose Sunnmøre. In region East we chose two different areas, Lillehammer-Øyer and Østfold to represent the region. We visited and interviewed altogether 90 dairy farmers on-farm from September until December 2007.
Only five farmers refused to participate in the study. The reasons they gave varied from lack of time to skepticism towards the study. Comparing the dependent and independent variables in our sample with population data we found that none of the variables from the farm databases deviated more than 2% between sample and population. Our sampling procedure produced a representative sample. Therefore, the results are valid for all TINE dairy farmers; they constitute 94% of all dairy farmers in Norway. In Table 1 we present some summary statistics for the farmers. We interviewed 85 male and 5 female farmers. Some of the male farmers' spouses took part in the interview, particularly if they participated in the farm work. The farmers in Sunnmøre are significantly older and have longer experience than those in Nord-Trøndelag.

Dependent Variables
In 2013 we checked in the dairy company's database whether the farmers or their successors still produced milk. On 17 farms, or 18.9 percent of the farms, there was no longer milk production. The other dependent variables we use are whether the farmers get help from other family members or not, and how satisfied the farmers are with the consultancy service. Other family members were engaged in farm work on approximately half of the farms, or 52.2 percent. Altogether 80.4 percent of the farmers were satisfied with the consultancy service, while 19.6 percent were dissatisfied.

Independent Variables
We apply three different categories of mood; happy, neutral and bad. Farmers who during the interview gave more than one positive statements about their farming, and no negative statements, were considered being in a positive mood. Contrary, farmers who gave more than one negative statements about their farming, and no positive statements, were considered being in a bad mood. Farmers who did not give any, or maximum one positive or negative statement regarding their farming, were considered as being in a neutral mood. Altogether 48.9 percent of the farmers were in a happy mood, 40 percent were in a neutral mood while only 11.1 percent were in a bad mood.
To illustrate what we mean by positive and negative statements about farming we quote some of the farmers. A negative male farmer in his thirties had considered investing in a fire alarm in the cowshed. However, he dropped the plans because he found it too expensive: Well that's about the only way of financing a new cowshed nowadays, that's if it burns down. Because there is no chance you can make it nowadays with the quota sizes we have around here….
The same farmer also commented on the quality system which is mandatory on all farms: I fill in the check lists just to get it done with,… it' s only to avoid losing money. Because it's a nonsense system you know.
In contrast to these negative statements we quote a happy male farmer in his fifties who really enjoyed both farm work and living on a farm. His wife and his son's family were also heavily engaged in the farm work. We asked him what he appreciated in farming: Everyone who visits this farm comments on the nice garden. And now in summertime, we can sit on the terrace and barbecue or have a cup of coffee …and people drop by and have a chat. That's when we feel comfortable…We emphasize a good quality of life you know… The happy farmer also appreciated other aspects of farming: I really appreciate being part of the farming community around here. I meet a lot of people and get involved in the professional environment, that's really rewarding….
A negative male farmer in his fifties was very much occupied with problems. First he commented on the farming economy: What influences the economy positively is how much we work outside the farm. Because on the farm there is not much we can do. We are tied on our hands and feet on everything, so that's it. But negatively we can influence it by investing on the farm.
This farmer feels that he has little control over his farm income, and attributes the causes externally. When we asked what was challenging in farming he answered: Well, that's to keep up the spirit and courage to be here on the farm every day all year around. Of course things can improve if we enter joint farming, but that's about the only advantage in joint farming I guess.
Contrary, we quote a happy male farmer in his forties who answered the following when we asked what he enjoyed in farming: The advantage is that I am very pleased to run my own farm, and that I am allowed to be my own boss. I can manage my working day myself, and I appreciate having a challenging work where I can decide things on my own.…And in winter I can go skiing in the middle of the day. …But perhaps the most important is to have a job you enjoy.
This farmer was very conscious how discussing with farmers in a bad mood could influence his own mood negatively: I prefer talking to those who are optimistic in farming, the negative ones I deliberately try to avoid.
As control variables we use farmers' age, region of residence and the number of discussion partners. To measure how motivated the farmers are we counted how many aspects of farming they enjoy and how many farming magazines they read. We also asked the farmers whether they had potential successors or not.

Statistical Methods
After coding the interview data, we transferred the codes to a data sheet. Then we added the data from the dairy company's database showing who had quitted dairy farming, and then merged the datasets in a SAS-JMP-database. First, we explored the hypotheses in a preliminary analysis. Our goal in this first step was to get to know the variables and how they are related. We started with descriptive statistics of our dependent and www.ccsenet.org/ijbm International Journal of Business and Management Vol. 10, No. 2;2015 independent variables. For continuous variables we performed correlation analysis, analysis of variance, t-tests and regression analysis. We also tested interaction terms and tested for non-linear effects.

Validity and Reliability
We started at the individual level and apply an interpreter's view. The first aim was to understand the meanings, reasons, purpose, values and behaviors of farmers from their perspective. Qualitative methods are often used when the aim is to understand individual actors, and we decided to use semi-structured in-depth interviews. When we understood the farmers' behavior from his point of view, we wanted to present models of cause and effect for the whole group of farmers. We moved from understanding individuals to explain the behavior of 90 farmers. Quantitative methods have advantages over qualitative methods when the aim is to explain behavior of whole systems. Therefore, we combined qualitative and quantitative methods.
At the beginning of our work we developed hypotheses based on theory and prior findings. Our only independent variable is farmers' mood. To ensure reliability we conducted the first 13 interviews together. In the interview setting it was important to establish an atmosphere of mutual confidence. We started with some small talk on matters outside the interview guide, sometimes related to farming and sometimes not. Then we asked the farmer if we could record the interview. This was an opportunity for the interviewer and the respondent to get to know each other before we turned on the voice recorder. Gradually, the conversation turned over to the themes in the interview guide, and the interview started almost without the farmer noticing it. During the conversation we noticed that confidence increased, and we experienced that farmers are open-minded and trustful. Then we could discuss more difficult subjects as problems they had faced and how they had solved them. At the end of the interview the farmers typically said "Well this was useful for me too.".
We collected all the transcribed interviews in one file and analyzed them with the qualitative software Hyper Research. We coded the data and mapped explanatory patterns. To ensure reliability in the coding process we constructed some preliminary codes together based on our interview guide and our research questions. We coded some interviews together while discussing both the coding and the interpretation of the respondents' statements, then each of us coded a number of respondents and the other checked the coding. We adjusted the coding as our knowledge of the different subjects in the interview data got deeper. We discussed and resolved possible differences in coding to secure inter-coder reliability.

Results
Happy farmers get more help from the rest of the family than farmers in a neutral or bad mood, see Table 2. The likelihood ratio test statistic is χ 2 (1, n = 90) = 6.55, p< 0.0105. The finding here supports hypothesis two. A happy male farmer in his sixties told us with great enthusiasm how much the whole family enjoyed farming and how much it meant to all of them: But what is most important to us, my wife and I, is that we have a family who support us. We have a very good network in our family. So we can go on holiday etc. …and the others give us a helping hand.
A happy female farmer in her thirties with three small children also got heavy support from the rest of the family: Yes my husband is also involved, but he works full time outside the farm. Still he helps me in the evenings and in the weekends, and when I'm on maternity leave (laughter)…And also my parents participate a lot. They often help me in the cowshed. Even though I tell them to have some time off, they still want to participate…. So I consider this as a family business.
Farmers in a bad mood are less satisfied with the consultancy service than farmers in a happy or neutral mood (Table 3).  The likelihood ratio test statistic is χ 2 (1, n = 87) = 5.45, p< 0.0196. From Table 3 we notice that only half of the farmers in bad mood are satisfied with the consultancy service, while 84 percent of the farmers in happy or neutral mood are satisfied. One of the farmers in a happy mood was very satisfied with the dairy consultant: Yes we are very satisfied with our advisor, she's a very clever professional. We keep in touch with her once a month. They should have had someone like her also further down south.
Contrary, one of the farmers in a bad mood commented the following when we asked him of his opinion on the consultancy services: Well I' m a bit tired of training new advisors all the time. I think I've had four different advisors over the last years, and just when I've finished training one, a new one shows up.
The finding in this section supports hypothesis three.
Milk production maintains longer on farms where the farmer is in a happy mood as compared to farms where the farmer is in a bad mood, see Table 4. The likelihood ratio test statistic is χ 2 (1, n = 90) = 8.79, p< 0.003. On farms where the farmer was in a happy mood in 2007, there was still milk production on more than ninety percent of the farms in 2013. In contrast, one third of the farmers who were in a neutral or bad mood had quitted dairy farming in the same period. One of the farmers in a bad mood commented the following when we asked her if her children had any interest in farming: No, none of them are interested, and I don't think we should encourage them to take over the farm either, when you see how the rest of the society treats us. Is there any future in farming? And become just as tied on our hands and feet as we have been? And to never have a day off, regardless of whether it is Easter or Christmas.
Similarly another farmer in a bad mood commented on the economy when we asked him if any of the children were interested in taking over the farm: If it really is something to take over then, I wonder….Our oldest son is interested but he earns money like grass in the industry now, so times must change, otherwise the transition will be too big for him I think.
Contrary a farmer in a happy mood told us how the whole family engaged in the farming activities: The most important thing for me and my wife is that we have a family who supports us. We have such a good network in the family. We can go on holiday and the others keep the wheels running while we are away…..It's truly a family enterprise.
The finding in this section supports hypothesis three.
We found no significant relationship between farmers' mood and farmers' age, region of residence, number of discussion partners or farming performance. Further, we found no relationship between motivation and how long the farmers maintained dairy farming. Finally, we did not find any relationship between whether the farmers had potential successors at the time of the interview, and how long they maintained dairy farming.

Discussion and Conclusion
Farmers in a happy mood get more help from other family members, are more satisfied with the consultancy service, and maintain dairy farming longer than farmers in a bad mood. The lifestyle of farmers is inextricably linked with their work. Not all farmers take holidays, and their lives are governed by often unpredictable forces such as weather, disease and problems with animals and farm machinery. In spite of this, our research reveals that only approximately 11 percent of the farmers are in a bad mood, and half of them are in a happy mode. The remaining are in a neutral mood. This is an important finding, considered the impression one frequently gets through media that farmers often complain. The high number of farmers in a happy mood can explain why more than half of them get help from other family members. Such help can work against some of the potential negative factors in the farmers' work environment, and our findings support the findings of George (2000) and Staw (1994).
According to our findings mood influences strategic decisions such as to maintain production. The finding that a happy mood influences job perseverance positively is somewhat contrary to the finding of Ciavarella et al. (2004). One reason why our results differ might be that Ciavarella et al. (2004) studied only new ventures and measured survival after eight years, while most of the respondents have taken over the farm from their parents and run it for on average 16 years. However, our findings are in line with the findings of Seligman and Schulman (1986) and Tsai et al. (2007). Farmers in a positive mood judge their farming activities as valuable, leading to a positive evaluation, while farmers in a negative mood interpret their farming activities as less valuable, leading to a negative evaluation. In fact our findings indicate that a negative mood signals to the farmer that farming activities lack value. Here the findings are in line with the findings of Clore et al. (2001) and Sirkaya et al. (2004). Our findings also connect to the mood-as-input model (Clore et al., 1994;Schwarz & Clore, 1988).
Taken together the findings show that bad mood has a significant negative impact on the different social relations farmers engage in, both with farming consultants and with other family members. Thus our findings are in line with the findings reported (Keltner & Kring, 1998;Shiota et al., 2004;Harker & Keltner, 2001;Shiota, Campos & Keltner, 2004;Ko et al., 2005;Szulanski, 1996). Farmers in a bad mood who want to benefit more from interaction with consultants should try to regulate their mood. For example they could try to increase their number of discussion partners.
Our findings indicate that farmers in a happy mood are more likely to get successors who are willing to maintain dairy farming, as compared to farmers in a neutral or bad mood. Thus farmers who consider it important to get a successor who maintains dairy farming should reflect upon their mood and consider techniques to eliminate bad mood (see Thayer et al., 1994). Further, they should be aware of how they talk about farming when they are together with the rest of the family, e.g. during dinner. Future studies could explore the relationship between farmers' mood and the likelihood of getting successors in more detail.
A weakness of this study is that we measure the farmers' mood only once. It is well known that moods can change over time, and this could represent a potential threat to the validity of the study. However, the relatively high number of participants counteracts this possible threat. Further, real-life stories are important because they contain important processes that can give in-sights to social and economic behavior. Thus our design ensures that the behaviors observed and recorded reflect behaviors that actually occur in natural settings, as compared to laboratory studies. Future studies could include personality traits to increase the understanding of the effects of different moods. Another possible avenue for future research could be to explore the effects of moods on other strategic decisions, such as to expand the farm.
To conclude the mood of managers of small family owned firms influences involvement from other family members and interaction with consultants. Further, managers' moods influence strategic decisions such as to maintain production, and the likelihood of getting successors.