Managers’ work and behaviour patterns in profitable growth SMEs

We investigated managers’ work and behaviour patterns in profitable growth small- and medium-sized Swedish companies, and considered how these patterns might be associated with good health outcomes. Specifically, we looked at hours worked by managers, proportion of time spent on working activities, and leadership behaviour orientation. We used a quantitative cross-sectional design and collected data via a standardized questionnaire that was answered by 133 top managers. The data were analyzed with descriptive statistics, linear regression, and compositional data analysis. Our results indicate that the managers worked long hours, which is a health risk both for them as individuals and for their organizations, but also that they engaged in work practices and leadership behaviours that were favourable for organizational health and for their employees. The managers spent a high proportion of their time in touring, which could be beneficial to organizational health, and exercised active leadership through behaviours that contribute to both employee health and company effectiveness. Comparing our results to other studies, we can observe that patterns of managers’ time use differ between small and large companies, confirming that the size of the firm is an important determinant of managerial work.

performance and health, drawing on the Healthy Work Organization (HWO) perspective (Sauter et al. 1996). The topic is important since there is increasing global interest in organizational health across small businesses, which represent a large share of many economies (Legg et al. 2014;Nowrouzi et al. 2016). In Sweden, smalland medium-sized enterprises (SMEs; ≤ 250 employees) constitute 99.9% of all companies and employ around 65% of all private-sector employees (Tillväxtverket 2019). We also know that organizational health in SMEs is poorly managed, and tends to be marginalized (Legg et al. 2014). Employees in small businesses have poorer working conditions and are exposed to greater health and safety risks (Walters et al. 2018;Hasle and Limborg 2006).
Daily managerial work is overlooked in the contemporary leadership discourse (Alvesson and Sveningsson 2003a;Hales 2001), management research (Mintzberg 2009;Tengblad 2012), and small business literature (O'Gorman et al. 2005). Building on the practice perspective on management (Tengblad 2012), we maintain that successful and experienced managers' daily work practices, behaviours, and activities can, in addition to the mainstream theories on leadership, be used "as the primary data for theorizing about good management" (Tengblad 2012:5). The present article therefore focuses on managerial work and behaviour patterns in the context of profitable growth SMEs. This context is important for two reasons. First, firm growth has been an important area of research in entrepreneurship both due to the significance of SMEs for the economy and because SMEs' growth has been shown to be critical for their survival, success, longevity, and financial performance (Pasanen 2007). Second, too little research attention has been paid to managerial work in small businesses compared with large companies (O'Gorman et al. 2005;Tengblad 2012).
We explored the nature and extent of managerial work by means of a questionnaire answered by managers of 133 SMEs with profitable growth. Our results show that these managers worked long hours, which is a health risk, but also that they engaged in "pro organizational health" practices such as touring and active engagement in health-promoting leadership behaviours. Yet we also found that the extent and nature of managerial work were not determined by managerial style, and that organizational factors may matter. Our research contributes to the body of literature on the nature of managerial work by augmenting empirical evidence on managers' work and behaviour patterns in the context of profitable growth SMEs. Furthermore, by theoretically discussing the health effects of managers' work and managerial styles, we contribute to the emerging work on organizational health in the context of SMEs. To our knowledge, there is no other study that has employed the interdisciplinary approach we take here, bringing together research streams within managerial work and leadership behaviour, and discussing occupational health in the context of growing SMEs.
2 Literature review 2.1 The nature of managerial work Researchers in management and work behaviour have tried to understand management practices by asking what managers really do (Hales 1986;Tengblad 2012). Managerial practices have been discussed in terms of roles, work content and form, communication patterns, and informal work aspects (Hales 2001).
Mintzberg's classic study from 1973 has traditionally served as a reference for description of managerial work. In that study, Mintzberg examined patterns of managerial activities using the categories deskwork, telephoning, scheduled meetings, unscheduled meetings, and tours in the organization; the similarities that this revealed in managerial practices led him to advocate the generic nature of managerial work (Mintzberg 1973). Other researchers have emphasized both commonalities and variations in managerial behaviours (Hales 1999;Tengblad 2012). Stewart (1976) argued that variations in managerial work may be explained by variations in managers' work demands, constraints, and choices. By studying managers' time allocation in profitable growth SMEs, we demonstrate the managerial work and behaviour patterns prevalent among this group, implying that managers' choices might be associated with higher organizational performance.
According to researchers, e.g. Hales (1986;1999;2001;Martinko and Gardner 1985;O'Gorman et al. 2005), although managerial work research has provided a good account of what managers do, there is still insufficient explanation why do managers do what they do and whether it matters for employees and organizations. In particular, the criticism concerns too little attention in investigating: what factors (organizational, individual, and contextual) explain variations and commonalities in managerial work, how managers' work and behaviours relate to organizational performance outcomes, and how managers' work affect employees. By incorporating context and exploring variations within the group of profitable growth SMEs, we address some of the questions raised in this criticism.
Research also suggests that the size of the organization influences the nature of managerial work and behaviours (O'Gorman et al. 2005). The few studies that have investigated managerial time use in small companies found that these managers spent less time in scheduled meetings, more time in informal communication, and more time touring compared with peers in larger companies (Choran 1969;Tell 2004, 2012). O'Gorman et al. (2005) studied growing small businesses and found the same pattern. Florén and Tell (2012) investigated managerial work in fast-growing versus slow-growing small businesses but found no substantial differences in time use, degree of formalization, or communication pattern that could explain the firms' growth.

Managerial work from an organizational health perspective
According to the HWO model, organizational performance and employee health are interrelated and together form organizational health, which in turn is influenced by managerial practices (Sauter et al. 1996). Our standpoint in this paper is that managers' working hours and time spent on touring (defined here as managers walking around the workplace and interacting with employees) are important for organizational health.
Classical studies of managers' time use (Carlsson 1951;Mintzberg 1973;Stewart 1988) view touring as i n s p e c t i o n t o u r s a n d a m ed i u m f o r v i s u a l communication, collecting valuable information, and seeing if everything is going well. Stewart (1976) notes that managers can also have informal discussions with employees, creating value for the employees in terms of increased morale. Peters and Waterman's (1982) highlighted some years later management by walking around (MBWA) as a characteristic behaviour for managers in successful US companies. MBWA was thus regarded as a way of bringing managers out of office to talk to employees and clients in an informal manner. A Swedish study of successful private and public organizations with good employee health and high effectiveness also identified MBWA as a key leadership behaviour (Larsson and Vinberg 2010).
In a study of effective managers, Kotter (1982) found that they often are engaged in short informal conversations. He interpreted these seemingly non-managerial, chaotic, and often not work-related activities as an efficient way of problem-solving, getting useful information, relating to employees and setting agenda. In Alvesson and Sveningsson' study (2003b), managers stressed the role of such ordinary activities as listening and informal chatting in everyday managerial work, and maintained that these activities have positive effects on employees (since employees feel they are seen, respected, and an important part of the team). Other studies show managers' being hands-on and accessible for contacts with employees to be important for employee health (Lundqvist et al. 2012;Skarholt et al. 2016;Poulsen and Ipsen 2017).
According to our interpretation, the activities referred to in these studies (i.e. MBWA, short informal conversions, managers being accessible, listening to/chatting with employees) relate to the concept of touring reflecting managers' being present and available for spontaneous informal interactions with employees. Touring also provides greater opportunities for managers to engage in relation-oriented leadership behaviour, which has been found to be associated with employee well-being (Skakon et al. 2010). Touring is thus of interest both as a management technique that may be common to effective companies, and as a behaviour important for employee well-being.
Researchers have suggested that managers' excessive workload might hinder their ability to handle their own working situation in the long run (Carlsson 1951;Tengblad 2006). Working long hours is associated with depression, anxiety, sleep disturbances, and coronary heart disease (Bannai and Tamakoshi 2014), thus making it a serious occupational health risk. Furthermore, systematic reviews of available research have reported a relationship between managers' and employees' health (Skakon et al. 2010) as well as between managers' wellbeing and their leadership behaviours (Kaluza et al. 2019).

Leadership behaviour orientation
The relationship between what managers do, who they are, and where they work is poorly examined, particularly regarding individual factors and leadership styles (Hales 1999). Despite the large body of research on managerial behavioural styles, to our knowledge, these styles have not been related to managers' practical work, for example, in terms of time use.
A well-established typology of leadership behaviours is presented in the three-dimensional leadership theory (Yukl et al. 2002) and its twin, the CPE leadership model Arvonen 1991, 1994), categorizing managerial behaviours into three broad categories: task/production-oriented, relation/employee-oriented, and change-oriented. Despite different naming, the dimensions in both models are similar, and describe leadership in terms of how much emphasis a manager places on each dimension. Task-oriented behaviours (e.g. organizing and planning work activities, setting goals and standards, monitoring operations and performance) are aimed at maintaining effective production and task fulfilment; relation-oriented behaviours (e.g. providing support and encouragement, recognizing contributions and accomplishments, socializing to build relations) emphasize cooperation and trust; and change-oriented behaviours (e.g. providing encouragement to view problems and opportunities in a different way, developing innovative new strategies, and encouraging and facilitating innovation in the organization) point at change, growth, and adaptation to the external environment (Yukl et al. 2002). Change-oriented behaviours may be regarded as entrepreneurial, since innovativeness, risk-taking, creativity, and commitment to change and growth have been discussed as being characteristic of entrepreneurial behaviour (Sadler-Smith et al. 2003). We have selected the three dimensional model for this study due to its relevance from the HWO perspective and since it is a well-established model in our field of research. The three dimensions within the model are associated in various ways with health, job satisfaction, sickness absence, disability pension, performance, quality, and effectiveness (Arvonen 2002;Kuoppala et al. 2008;Larsson 2010;Nyberg 2008;Nyberg et al. 2005;Skakon et al. 2010). Studies show strong empirical support for the association between relation-oriented behaviours and employee health (Skakon et al. 2010). Also, other researchers (Kaluza et al. 2019) have regarded task-oriented, relation-oriented, and change-oriented behaviours to be constructive leadership behaviours in relation to their consequences for followers and the organization, as opposed to destructive (e.g. abusive and passive) behaviours.
For increased understanding of leadership, it is important to address patterns of specific behaviours used by managers within each dimension and to study leadership behaviour in its context (Yukl et al. 2002). Therefore, in this study, we use the healthy and effective leadership behaviour (HEL) model (Larsson and Vinberg 2010) since it complements the threedimensional model allowing to analyze specific leadership behaviours that promote organizational health and effectiveness. Furthermore, it is based on the study of organizations in Sweden, making it interesting also from the HWO and contextual perspectives. The model includes nine groups of common leadership behaviours: a strategic and visionary leader role, communication and information, authority and responsibility, a learning culture, subordinate conversations, plainness and simplicity, humanity and trust, walking around, and reflective personal leadership (Larsson and Vinberg 2010). These behaviours espouse visibility, openness, communication, trust, and cooperation with the aim of promoting the development of one's subordinates. When behaviours in the HEL model are viewed through the prism of the three-dimensional model, high relationship orientation appears to be a universal component for successful leadership, while task and change orientation vary in this regard depending on situational variables (ibid.).
Leadership behaviour and managers' time use reflect different but potentially interlinked aspects of managerial practices. If different leadership behaviour dimensions influence managers' time use and working activity prioritization in practice, this may explain variations in managerial work. The effects of leadership style may vary depending on the organizational and individual factors constraining managers' work.

Research questions
As discussed above, few studies have addressed managers' time use in growing SMEs and the type of leadership behaviours involved across managerial practices. This study adds to the empirical evidence on managers' time use in profitable growth SMEs and explores patterns of work that might be associated with personal and organizational health. We do this by looking at hours worked by managers, the nature of their work, and the prevalence of leadership behaviours associated with good health outcomes. Our approach pays special attention to time spent on touring and managers' total working hours, since these aspects might have implications for employees'/managers' health as well as organizational performance. Furthermore, we investigate whether leadership behaviours, managers' characteristics, and organizational characteristics can explain the extent and nature of managers' work. Linking managers' time use, leadership behaviour styles, and managers' background may enhance the understanding of factors influencing managers' time use in profitable growth SMEs.
The research questions were as follows: a) What is the extent and nature of managerial work in profitable growth SMEs? b) Do differences in (i) leadership behaviour orientation, (ii) organizational context, and (iii) managers' background characteristics influence the extent and nature of managerial work in profitable growth SMEs?

Study population and sample
The data for this cross-sectional study were collected within the project "Successful Companies in Gästrikland" (SCiG). This project gives annual awards to the 50 most successful companies in Gästrikland, a province in central Sweden.
The study sample was specified via a two-step sampling process: the SCiG process followed by the study sampling process (Fig. 1). The SCiG inclusion criteria are companies registered in Gästrikland, ≥5 years of operation, ≥4 employees, and ≥ 4 million Swedish Crowns in net sales. All companies fulfilling these criteria are rated by SCiG according to the dynamics of economic indicators such as net sales, number of employees, equity ratio, income, pre-tax profit margin, return on assets, and return on equity. Selection and rating are performed by an independent auditing firm using data on the companies' last five annual financial statements (with the last report weighing heaviest) aggregated by a European company specializing in quality-assured business and financial information. The project thus selects companies demonstrating growth with retained profitability. Annually, the 120 highest-rated companies are nominated for the award. Managers of these companies are interviewed via a standardized questionnaire on health-promoting and effective leadership.
Managers of companies nominated for the award during 2015-2018 constituted the study population. In 2018, data from their paper questionnaires were screened and entered into SPSS. We used two additional criteria beyond those of SCiG: SMEs and top managers. Top managers were defined as owner-managers, executive directors, and similar-level managers; SMEs were defined as companies employing up to 250 persons. The study included responses from one manager at each company, and for companies nominated several times, only data from their first entry were included.
Answers from 133 managers were included in this study ( Table 1). The managers' average age was 48, 88% were male, and 31% had university education. They had worked at the company for 15 years on average (range: 1-42) and had almost as long managerial experience. Average company size was 21 employees (range: 4-150).
All cases with any missing data were excluded from the analysis via list-wise deletion. When we compared the cases with valid responses (n = 133) with the cases with some missing data (n = 15), the groups differed in terms of age and proportion of time spent on spontaneous meetings. Thus, the internal missing data might have affected the results of the study. However, there were no significant differences regarding other variables, including the main outcomes. We also checked that there were no differences between the groups regarding the financial performance ranking that was used to select the sample. Furthermore, no differences in financial performance ranking were found between companies that participated in the survey (n = 148) and those that did not (n = 139). We were not able to make sociodemographic comparisons because sociodemographic data were not available for the non-respondent group.
To test the validity of the questionnaire and ensure the quality of data collection, five pilot interviews were carried out to ensure that the respondents understood the questionnaire. While screening the answers, additional telephone calls were made to some of the respondents to clarify their answers.

Measurements of variables
The key characteristics included in our analysis were the extent and nature of managerial work, leadership behaviour orientation, managers' background characteristics, and organizational context. According to Maes et al. (2005), factors on the level of manager, company, and managerial practices are basic determinants of company performance.

Outcomes
The main outcome of the study was the extent and nature of managerial work, measured in terms of All companies in the province, n=5891 a Companies matching the inclusion criteria, n≈480-500 b 120 companies nominated (highest rated) 50 companies awarded SCiG inclusion criteria Step 1. Economic rating Step 2. Structured interviews with managers willing to participate.
Step 3. Results of step 1 and 2 are combined to select companies for award managers' working hours and distribution of time between activities (particularly tours) in line with the categories used in previous research Tell 2004, 2012;Kurke and Aldrich 1983;Mintzberg 1973;O'Gorman et al. 2005;Tengblad 2006). Participants were asked to specify the percentage of their working time spent in a typical week on deskwork (e.g. e-mail, general administration), telephoning, scheduled meetings, unscheduled meetings (e.g. meeting somebody in a corridor and holding a spontaneous meeting), and tours (walking around and talking to subordinates). We assessed total working hours by asking "How many hours per week do you normally work?".

Independent variables
Leadership behaviour, as the main exposure, was measured with four one-item questions assessing managers' overall orientation towards relations, task, and change according to the taxonomy in the three-dimensional model (Ekvall and Arvonen 1991;Yukl et al. 2002) and usage of behaviours in the HEL model (Larsson and Vinberg 2010). Respondents were asked to read descriptions of core behaviours in each dimension and assess the extent to which they practiced these behaviours in their daily work on a scale ranging from 0 (I never do this) to 100 (I clearly do this and could be a role model for other managers). For example, the description for relation-oriented behaviours was as a manager I provide support and encouragement to employees, express conviction that an employee can accomplish a difficult task, recognize achievements, provide coaching when needed, discuss, advise, check with employees and keep them updated in decision-making processes, and handle conflicts in a constructive way. With respect to managers' characteristics, we used managers' age, sex, education, managerial experience, and tenure in line with the upper echelons perspective (Hambrick and Mason 1984), employing observable managerial background characteristics as predictors of strategic choices and organizational performance. Age, education, and management experience are commonly used as manager background characteristics in the area of small business performance research (Maes et al. 2005). Age was measured on a continuous scale. Sex was measured as male = 0 and female = 1 and education as no higher education = 0 and higher education = 1. Managerial experience was the number of years that the respondent had worked in a managerial position in current and previous organizations. Organizational tenure, measured as the number of years worked in the current organization, referred to experience and understanding of how business was done in the companyspecific context.
The final two variables reflected aspects of the organizational context within which the managers operated, and which might affect their work. Control span measured number of direct subordinates, and number of employees (total number in the organization) measured company size.

Statistical analysis
The first step consisted of a descriptive analysis of total working hours and proportions of time spent on each activity category: deskwork, telephoning, scheduled meetings, unscheduled meetings, and tours. Measures of central tendency and spread were employed: n, mean, median, range, and standard deviation. We also calculated arithmetic means and standard deviations for proportions of time spent on activities and total working hours corresponding to low and high levels of leadership behaviour orientation, situational characteristics, age, sex, and education. For these purposes, all the independent variables and covariates were dichotomized around the median (below median = 0, above median = 1), and differences were assessed with the Mann-Whitney U test. Correlation analysis was performed between all dependent and independent variables.
The second step included a univariate and multiple linear regression analysis exploring whether total working hours were related to leadership behaviour orientation, organizational context, and managers' background characteristics. First, univariate linear regression was performed to assess the association between total working hours and each of the predictors, and then multivariate regression analysis was performed to explore the relationship between working hours, leadership behaviour, and the predictors that were significantly associated with the outcome in the univariate analysis.
Since the total proportion of time spent on the categories of managerial activities made up 100% of a total working day, the variables were inherently co-dependent. Conventional statistical methods may be inappropriate for finite and collinear data, where parts compose the whole and their variation depends on other components and is constrained by the constant sum (Pawlowsky-Glahn and Egozcue 2006). We therefore performed compositional data analysis (CoDA) to explore whether the proportion of time spent touring was related to leadership behaviour orientation, organizational context, and managers' background characteristics. First, the dependent variable (tours) was transformed into a proportion (i.e. bounded between 0 and 1): in this way, the effect of explanatory variables tends to be non-linear and the variance tends to decrease when the mean gets closer to one of the boundaries. In order to estimate the impact of exposure variables on the dependent one, marginal effects after fractional logit model were calculated. P values and 95% CI were also reported.
To examine whether any patterns of time allocation were characteristic of managers in profitable growth SMEs, we qualitatively compared our results with studies of managers in different-sized private companies that also used Mintzberg's categories: three concerning small companies (Choran 1969;Florén and Tell 2012;O'Gorman et al. 2005), one concerning intermediate companies (Kurke and Aldrich 1983), and two concerning large organizations (Mintzberg 1973;Tengblad 2006

Ethical statement
The study was approved by the Regional Ethical Review Board in Uppsala, Sweden (ref: 2016/208). All participants gave their written informed consent.

Descriptive analysis
The managers in these profitable growth SMEs worked an average of 52.4 h per week, spending 34% of their working time on deskwork, 17% on telephone calls, 16% in scheduled meetings, 12% in unscheduled meetings, and 19.6% on touring the organization and spontaneous interactions with employees (Table 2). Their leadership behaviour was oriented about 80% towards relationships, 80% towards change, 65% towards tasks, and about 80% towards the HEL behaviours. Table 3 compares our results with the results of previous studies of managers' time use in small, intermediate, and large organizations. The managers in our study spent nearly 20% of their total working time on tours, as compared with 6-12% in other studies of managers in small companies (Choran 1969;Florén and Tell 2012;O'Gorman al. 2005) and 1-3% in studies of larger companies (Kurke and Aldrich 1983;Mintzberg 1973;Tengblad 2006). Table 3 clearly shows that managers in smaller companies spend more time in touring and administrative work and less time in scheduled meetings as compared with larger companies.
Managers in this study worked longer hours (52.4 h/ week) than managers in slow-growing (45.5 h/week) and fast-growing (44.5 h/week) small businesses (Florén and Tell 2004). The picture is less clear when it comes to comparison with large companies, as the present managers worked longer hours than the largecompany managers studied by Mintzberg (1973) (45 h/ week) and Kurke and Aldrich (1983) (44 h/week), but shorter hours than the large-company managers studied by Tengblad (2006) Patterns of managerial activities and total working hours stratified by sex, education group, and all other predictor variables (dichotomized around the median) are given in Online Resource 1. There were small variations in working hours and time spent on activities depending on levels of predictors. Managers with more orientation towards relationships (≥ 81%) spent less time on deskwork, those with more task orientation (≥ 66%) spent less time in unscheduled meetings, and those working in larger companies (≥ 13 employees) spent more time on administrative work and scheduled meetings. Those with greater span of control (≥ 13 subordinates) worked 5 h more than managers with less. Women reported more time spent on deskwork, but this should be interpreted with caution, since only 12% of respondents were female. Managers with higher education spent more time on scheduled meetings and worked 6 h less than managers without higher education. Time spent touring showed no differences when the analysis was stratified by sex, education, and other predictors.
Correlations between all the variables are given in Online Resource 2. There were no significant correlations between the main outcomes (time spent touring and total working hours) and the predictors of interest (perceived degree of leadership behaviour orientation). Tours were not significantly correlated with leadership behaviour orientation, company context, or manager background characteristics.
Longer working hours were correlated with larger span of control, more time spent on telephone calls, and less time spent in unscheduled meetings. Time spent on deskwork was negatively correlated with orientation towards relationships. Higher proportion of time spent in scheduled meetings was correlated with higher number of employees in the organization. All the categories of managerial  activities were correlated with each other, possibly due to the compositional nature of the data. All four leadership orientations were also correlated with each other.

Regression analysis
Regression analysis was used to explore whether managers' total working hours were associated with their leadership behaviour orientation, the organizational context, and managers' background characteristics. There was no association between working hours and leadership behaviour orientation ( Table 4).
The univariate analysis showed associations between hours worked and age, education, and managerial experience. However, after controlling for age, education, and managerial experience in the multivariate analysis, only education was related to the outcome (although with a larger confidence interval).
The compositional analysis examining associations between proportion of time spent touring and all exposures (Table 5) showed no statistically significant associations between time spent touring and leadership behaviour orientation. Number of employees and manager's educational level were negatively related to time spent touring. The multivariate analysis confirmed the association between number of employees and time spent touring.

Managers' work and behaviour patterns in profitable growth SMEs
The managers in our study spent a high proportion of their working time touring; this may be a factor contributing to organizational health, as illustrated in other studies, particularly in the USA and Sweden, emphasizing that touring is a management technique common in healthy and effective organizations (Larsson and Vinberg 2010;Peter and Waterman 1982). Spending more time touring may signify a manager who is often present and available and who has greater opportunities for engaging in relation-oriented leadership behaviour, which is linked to employee well-being (Skakon et al. 2010). The managers worked 52.4 h/week, exceeding the average working week by 30%. Bannai and Tamakoshi (2014) define long working hours as ≥ 40 h/week. Although there is no general agreement on the exact thresholds for hazardous overtime work (Spurgeon et al. 1997), thresholds between 41 and 63 h have been regarded as risky in relation to different health outcomes (Bannai and Tamakoshi 2014). Thus, the managers in our study worked long hours, and this can be a health risk. As discussed earlier, managers' health is linked to leadership behaviours (Kaluza et al. 2019) and might be an important prerequisite for exercising healthy and effective leadership behaviours (Lundqvist et al. 2012).
Comparing our results with other studies, we can distinguish differences between small and large companies in regard to managers' time allocation, suggesting that organization size matters for managerial work. Taken together, our results point at opposite directions in relation to two of Mintzberg's (1973) propositions on managerial work in small companies, namely that managers spend little time on tours and that they are preoccupied with scheduled meetings.
Deskwork was the most time-consuming activity in smaller companies, while managers in larger companies spent most of their time in scheduled meetings. This might be because managers in small companies have fewer supporting functions within their organizations (e.g. HR, finance). Furthermore, when compared with large companies, managers in small companies spent a lower proportion of time on formal activities (deskwork and scheduled meetings) and a higher proportion on informal activities (telephoning, unscheduled meetings, and tours), indicating that the degree of formalization is lower in small companies. However, we do not see a clear pattern of differences or similarities between groups of small and large companies in relation to the extent of managerial work.
We found a minor relationship between managers' leadership behaviour and manager background and organization characteristics on one hand, and time allocation to managerial activities and total working hours on the other. The analysis stratified by low versus high levels of predictors showed stable patterns of time use with only moderate variations. When applying Stewart's (1976) concepts, our study might indicate that constraints and individual choice have a minor influence on managers' daily work in practice. This suggests that work demands related to managerial responsibility play a bigger role in defining how work is performed.  The results indicate that managers in profitable growth SMEs show a high degree of engagement in task-oriented, relationship-oriented, change-oriented (as categorized in the three-dimensional model), and HEL behaviours. As already mentioned, the dimensions in the models relate to health, effectiveness, job satisfaction, and performance (Arvonen 2002;Larsson 2010;Nyberg 2008;Nyberg et al. 2005;Skakon et al. 2010) as indicators for organizational health. Our results support Ekvall's and Arvonen's (1994) conclusion that successful managers use all three dimensions to a marked degree. The score for task dimension was somewhat lower, which indicates that the managers in this study used entrepreneurial, supportive, and dialogue-oriented leadership behaviours to a somewhat higher degree than structuring and planning behaviours. The study results regarding HEL are also in line with previous findings concerning characteristics of healthy and effective organizations (Larsson and Vinberg 2010). More research is needed to understand the specific behaviours that managers in effective SMEs commonly use. Our findings also indicate that managers in profitable growth SMEs use active leadership behaviours which contribute to organizational health.

Touring and managers' working time in relation to leadership behaviours, managers' background characteristics, and organizational context
The regression analysis revealed no associations between leadership behaviour orientation and time spent touring. There are several possible explanations for this. First, our results seem to confirm Mintzberg's (1973) proposition on the stability of managerial work, and may also suggest that managers have little individual choice and their practical work is predetermined by their tasks. Second, the results might have been influenced by the sampling. We selected SMEs in the top bracket of financial performance, which might have reduced variation. Moreover, the managers reported high usage of leadership behaviours in all dimensions, and so there was no large spread of values in the behaviours. Comparing companies with different levels of profitability and growth (high-low) might have shown different results. Finally, the factors may have been indirectly linked due to the compositional nature of the managerial work categories. Our analysis showed correlations between time spent on deskwork and relationship orientation (Online Resource 2) as well as between time spent on unscheduled meetings and structural orientation (Online Resource 1). In line with Stewart's (1988) suggestion that tours can be seen as residual activities that tend to be curtailed or dropped when new tasks arise, we can assume that managers with higher relationship orientation who spent less time on deskwork (possibly due to delegation and empowerment) spent more time in tours.
Organization size was related to proportion of time spent touring. We demonstrated above that small and large businesses differ in this regard, but can also conclude that company size creates differences even within the group of SMEs. It might be easier for a manager to interact with and relate to a smaller number of employees. This contrasts with Stewart's (1988) suggestion that a small number of subordinates reduce the need for inspection tours. Our results indicate that SMEs should not be regarded as a homogenous group in relation to managerial work. Size, industry, structure, and other factors may influence managers' work situation, behaviours, and availability for contacts with employees and engaging in relations.
Managers' working hours appeared to be related to their individual characteristics (education as a managerial resource), affecting their individual choice despite the common demands and characteristics of SME managerial work. However, the difference may also reflect other factors, for instance the company's field and core activity, which affect what managers do in practice. Furthermore, even though managers with more direct subordinates worked longer hours than managers with fewer subordinates, the associations with company size and control span were not significant. The multivariate regression analysis showed no association between working hours and leadership behaviour orientation.

Limitations
This study aimed primarily to explore managers' work and behaviour patterns in profitable growth SMEs without intending to assert associations with companies' financial performance. Company effectiveness, measured in terms of profitable growth, was a given factor in the context of our study.
Survey data have both strengths and limitations. The traditional method of studying managerial work is structured observation as introduced by Mintzberg (1973), but trying a different method gave us new possibilities.
Survey data on a larger number of managers, including a broader array of variables on leadership behaviours and background characteristics of managers and companies, allowed us to explore variability of managerial behaviour within the studied context as well as links with leadership behaviours.
It is important to emphasize that our sample might not be representative of all growing SMEs in Sweden. Although the overall situation (legislation, organizational culture) is similar, local and regional differences might make it difficult to generalize the results.
Our use of self-reported data might have influenced the accuracy and internal validity of the results. First, the answers might have reflected managers' perceptions of their work more than the actual situation. Second, the respondents might have found it hard to remember an accurate picture of a normal working day. However, some researchers maintain that perceived workload is a better predictor of psychological health than actual workload (Hobson and Beach 2000).
Since our questionnaire followed mainstream studies that build on Mintzberg's (1973) categories of working activities, we were able to compare our results with the available data. However, we might have missed other relevant categories not covered by these predetermined categories (e.g. managers' operative work).
We are aware of the fact that SMEs are not a homogeneous group. The companies included in our sample varied in terms of number of employees (range: 4-150) age, industry, and other characteristics which may affect managers' work and behaviours. Nevertheless, for the sake of simplifying comparisons, we have treated SMEs in a single approach. Finally, the cross-sectional design used in this study did not allow us to establish causality of the observed relationships.

Conclusions
Managers in profitable growth SMEs work long hours, which is a health risk for them as individuals and for their organizations, but they also engage in work practices and leadership behaviours that promote organizational health for their employees. They spend a high proportion of time in touring, which could be beneficial to organizational health, and they exercise active leadership by substantial use of behaviours oriented towards relationships, tasks, change, and the dimensions of the HEL model, which contribute to both employee health and company effectiveness. The extent and nature of managerial work do not seem to be associated with managerial style, but may be affected by organizational factors. A comparison of our results with those of other studies shows that patterns of managers' time use differ between small and large companies, confirming that the size of the firm is an important determinant of managerial work. Further qualitative studies are needed to better understand the content and meaning of touring in organizations. There is also a need for more research on managers' health in growing SMEs, and its role for organizational health.
Acknowledgements The authors gratefully acknowledge the insightful comments of the editor and two anonymous reviewers that substantially improved the article. The authors wish to thank Amelie Carlsson for digitizing the questionnaire answers, Mirko Di Rosa for performing the CoDA analysis, Hans Högberg for statistical support, and Cormac McGrath for proofreading the revised version of the manuscript.
FundingInformation Open access funding provided by University of Gävle.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Availability of data and material The datasets generated and analysed during the current study are not publicly available as they are part of a larger and still ongoing study, but are available from the corresponding author on reasonable request.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.