Differences in farmers’ perception of opportunities for farm development

Highlights

•	Farmers operating in comparable context differ in perception of opportunities.
•	Farmers differ in three dimensions: diversifying, ending and maximising.
•	Four clusters were found that translate in different patterns of farm development.
•	Identifying differences opens avenues to study heterogeneity in farm development.
•	Farmers as entrepreneurial actor select a strategy in their socio-material context.

Abstract

This paper empirically identifies differences between dairy farmers in their perception of opportunities for farm development. The construct ‘perceived Room for Manoeuvre’ (pRfM) is used which is defined as: ‘the opportunities perceived as viable in order to obtain a (substantial part of) business income’. A unique case study of 79 dairy farmers operating in a highly comparable socio-material context at the level of the case study allows for an empirical analysis of differences in the pRfM using a mix of quantitative and qualitative methods exploratory research, a questionnaire and in-depth interviews with stakeholders in farm development. Differences in the pRfM will likely affect the early phase of strategic decision making and consequently explain heterogeneity in farm development. Data analysis revealed the following three dimensions of pRfM: perception of the opportunity (1) to diversify; (2) to end production; and (3) to maximise production. These dimensions proved useful to explain heterogeneity in farm development and thus showed the importance of a subjective approach towards opportunity identification in farm development. This paper shows the need to view the farmer as entrepreneurial actor in the process of strategic decision making who interacts with the socio-material context of the farm.

Keywords:

• Strategic decision making
• Opportunity identification
• Strategic awareness capability
• Entrepreneurship
• Farm development

1 Introduction

This paper is an empirical study of the differences in farmers’ perception of opportunities for farm development. For this purpose the construct perceived Room for Manoeuvre (pRfM) is introduced which is defined as: ‘the opportunities perceived as viable in order to obtain a (substantial part of) business income’. By including in this definition the viability to obtain an income, the focus is placed on the farm as an economic unit that serves the purpose of providing a farm family income. The range of opportunities for farm developments is broad and is not limited as such to either on- or off-farm developments and includes obtaining additional income from work outside the farm. Perception is in the definition of pRfM used as a noun and in the following definition: ‘the way you think about or understand someone or something’ (Merriam-Webster, 2015). Or in everyday words: ‘perceptions are shaped by what we know, by what we think we know, and what we do not know’ (Renko et al., 2012). The perception of a business owner is as well the driver of opportunity recognition (Renko et al., 2012) and is therefore important in the identification of opportunities in the early phase of the strategic decision making process (SDM).

In order to study the differences in perception of opportunities between dairy farmers (as small business owners), it is important to be able to identify differences in the pRfM. However, no studies were found that operationalise the pRfM in an empirical way. The lack (or even absence) of these empirical studies is understandable as the contexts of businesses usually vary in many ways and this complicates a comparative analysis of perceptions of business owners (Sutcliffe and Zaheer, 1998; Yanes-Estévez et al., 2010). For this paper we use a unique case study in dairy farming that does allow for an empirical analysis of differences between business owners with respect to their pRfM, as all 102 family-owned dairy farmers operate in a highly comparable socio-material context at the level of the case study area.

This paper aims to answer the question whether the construct pRfM is a useful construct to further study farm development strategies in relation to other characteristics of the farm and the farmer. In rural areas large parts of the land are in use by farmers (Berkhout and van Bruchem, 2006), making farm development of interest for stakeholders in regional development (Lauwere et al., 2006). Policy schemes and development programmes of local governments influence business development, e.g. by supporting the diversification of small business (Liberman-Yaconi et al., 2010) or in the adaptation to changes in the environment (Feola et al., 2015). However, the final strategic decision for the development of a farm is made by the farmer in the role of owner-manager who has a rather autonomous position in decision making (Culkin and Smith, 2000; Pietola and Lansink, 2001; Jocumsen, 2004; Hang and Wang, 2012). For policies and programmes to be effective in aiming to influence strategic choices at farm level, it is important to connect to the world of the decision maker (Pietola and Lansink, 2001). This means it is important to understand how the farmer perceives the opportunities for farm development as it likely affects the strategic choices. This paper contributes to this understanding with an empirical study on the differences between farmers in their pRfM and is an answer to the call to study entrepreneurship in the context of the everyday and real life situations of business owners (Bjerke, 2007; Johannisson, 2011; Watson, 2013). This paper answer as well to the call to integrate research on strategic management with research on entrepreneurship (Kor et al., 2007; Short et al., 2010) and is in line with the view that the domain of entrepreneurship research can contribute to the understanding of the changes in agriculture and rural areas (Alsos et al., 2011). The paper will first present the theoretical embedding of the construct pRfM and will answer the following questions:

(1)	How can the construct pRfM be measured?

(2)	Can the construct pRfM differentiate between farmers in a meaningful way?

(3)	Are differences in the pRfM linked to differences in farm development over a longer period of time?

(4)	Are differences in the pRfM related to the farmers’ personal preference for farm development?

Question 3 tests the validity of the construct pRfM using a time perspective. A consistency between pRfM and farm development in the past indicates that pRfM is (relative) stable over a longer period of time. Question 4 tests the validity of the construct in relation to the personal preference of the farmer given a situation without limitations. If pRfM and personal preference mostly overlap, the construct pRfM apparently does not offer added value to measuring mere personal preference of the farmer.

2 Theoretical embedding

First the context of small businesses will be defined followed by the connection of pRfM with recent theory on SDM in small businesses. As the construct pRfM is closely related to the identification of opportunities, the relation with this important topic in entrepreneurship research will be described.

There is no single definition of small businesses available in literature. For this paper the definition of the Bolton Committee (a committee on the role of small businesses in the UK economy) will be used: ‘a small firm is an independent business, managed by its owner or part owners and having a small market share’ (Culkin and Smith, 2000). Family-owned dairy farms meet this definition. Due to the low number of employees in family-owned dairy farms, they could even be classified as micro-firms (Liberman-Yaconi et al., 2010). Family-owned dairy farming is a specific kind of small business, yet shares important characteristics with small businesses as the farm is an independent business, is managed by its owner or part owners and has a small market share (Culkin and Smith, 2000). The farmer needs to fulfil different roles in the business as do small business owners (Chandler and Jansen, 1992) and the farmer personally learns from the experience of running the business (Atherton, 2003). In a literature review Liberman-Yaconi et al. (Liberman-Yaconi et al., 2010) describe that owner-managers of small businesses typically cannot delegate (parts of) the SDM to other people in the firm. Owner-managers usually do not have specialised expertise in all aspects related to SDM and need to combine SDM with their managerial tasks. The context in which entrepreneurs operate has been highlighted as very important (Watson, 2013; Shane, 2008). This context is not just the start-up business owner who is looking for opportunities to develop a new venture. Entrepreneurship is also a vital element of the continuous development of an ongoing business (Watson, 2013; Johannisson and Dahlstrand, 2009).

2.1 pRfM and the SDM process

‘Strategy’ is defined in this study as “a choice out of available routes and means in order to realise a goal” (Encyclo, 2012). In this view continuation of current activities is an opportunity as well, as it can be a route to realise the goal of the business owner. SDM in small business has been modelled by a number of studies, starting with a trigger caused by an internal or external development and up to the actual strategic decision. The personal characteristics of the business owner are known to be an important factor to explain the heterogeneity in small business (Liberman-Yaconi et al., 2010; Jocumsen, 2004; Hang and Wang, 2012). The small business owner’s perception of opportunities is found to be more decisive for the outcome of SDM than formal analyses (Parnell et al., 2000; García-Pérez et al., 2014). The perception is influenced by dominant paradigms, lock-in effects and path dependencies (Vanloqueren and Baret, 2009; Lamine et al., 2012; Cowan and Gunby, 1996). The opportunities for business development are by definition perceived and therefore impossible to objectify, different business owners will perceive the opportunities in a given situation in different ways. A useful concept to study the perceived opportunities for business development is the ‘evoked set of opportunities’ (Krueger et al., 2009): ‘the full set of possibilities perceived as opportunities by a decision maker’. The evoked set of opportunities is limited to the number of opportunities that are within the perceived reach of the business owner: their perceived room for manoeuvre (pRfM). The evoked set of opportunities implies that there are differences between decision-makers in their pRfM even when they operate in a comparable context. Differences in the capability to analyse the opportunities for business development are likely to affect the pRfM of business owners. This capability is described by Hannon and Atherton (Hannon and Atherton, 1998) as ‘strategic awareness capability’: ‘the process of continuously improving how one identifies and conceptualises one’s own world, recognises events in this world, interprets these events and makes decisions on taking appropriate action to achieve positive business outcome’. The perception of the business owner is therefore important for the SDM of small businesses.

A number of recent studies describe models to represent SDM in small businesses (Liberman-Yaconi et al., 2010; Jocumsen, 2004; Hang and Wang, 2012). The starting point of these models is a trigger that creates the need for a strategic decision and the end point is the moment when a strategic decision is made. All models imply an interaction between the different steps and depict the process as iterative. This interaction underlines that SDM in small businesses is not a clear-cut procedure following a set number of steps. The circular-iterative character of SDM is most evident in the model of Liberman-Yaconi et al. (2010), as it depicts three overlapping circles of activities: (1) informing; (2) option generating; and (3) deliberating. The activities lead in a circular-iterative way to a decision followed by the implementation of the decision and an evaluation of the results. This iterative process is recognised as well in the context of agriculture (Huirne, 2003). In all three models the starting point is where the need for a decision is triggered. Although the SDM may start at this point, the owner-manager does not start without a history: his experiences, limitations in his views and his biases will affect the SDM (Simon and Houghton, 2002). The owner-manager will from time to time assess the situation of his firm and the viability of opportunities for the development of his firm. This pRfM acts as a preliminary filter for the selection of opportunities for a formal assessment in an SDM process. In McMullen and Shepherd’s (McMullen and Shepherd, 2006) model of entrepreneurial action, an opportunity that is part of the pRfM of a farmer is referred to as ‘first-person opportunity’, an opportunity deemed viable for one self. Opportunities may as well be recognised as an opportunity that is valid in a general sense (‘a third person opportunity’). The name pRfM for the construct underlines the connection to the everyday language of entrepreneurs when deliberating on their range of opportunities, their ‘Room for Manoeuvre’. Using words that connect to the context of everyday life facilitates a common understanding between researchers and business owners (Gartner et al., 2003). An opportunity that is part of the pRfM is deemed viable for farm development, however, this does not necessarily mean that the opportunity is put in practice. Choosing to realise one opportunity may exclude the possibilities to realise another opportunity. There may as well be a gap between thinking about and actively realising an opportunity (Zwan, 2012).

2.2 Opportunity identification

The challenge for strategic entrepreneurship is to both exploit existing opportunities and to identify new opportunities (Ireland et al., 2001). Opportunity identification has a central place in entrepreneurship research (Short et al., 2010; Shane and Venkataraman, 2000; Ardichvili et al., 2003). For this paper we build on the following definition: ‘an opportunity is an idea or dream that is discovered or created by an entrepreneurial entity and that is revealed through analysis over time to be potentially lucrative’ (Short et al., 2010). The reference to potential lucrative in the context of farming refers to the farm as an economic unit that serves, amongst other goals, the purpose to provide in a farm family income. The range of opportunities is not limited as such to for example on or off-farm developments. Two important views exists concerning: ‘why’ and ‘how’ are opportunities identified (Short et al., 2010; Gartner et al., 2003). Looking at the question ‘why’, differences in motivation are important. A ‘pull’ motivation is driven by the entrepreneurs wishes and a ‘push’ motivation is driven by changing circumstances (Amit and Muller, 1995). A push motivation means that the opportunity identified is not necessarily a preferred development by the business owner. Regarding the question ‘how’ opportunities are identified, two perspectives can be described: ‘discovery’ or ‘enactment’ (Gartner et al., 2003). The discovery perspective takes opportunities as ‘there’ to discover for those with enough ‘entrepreneurial alertness’ (Gaglio and Katz, 2001). An opportunity is so to say waiting to be found. In the enactment perspective opportunities are an outcome of the sense-making activities of individuals (Gartner et al., 2003). The entrepreneur uses his experiences and recognises a given situation as an opportunity for him to develop the business. Renko et al. (2012) argue to use ‘opportunity perception’ to bridge the gap between the discovery view and the enactment view. The perception of opportunity is subjective and idiosyncratic to each entrepreneur. Different actors perceive different realities, e.g. because of different prior experiences or because of the way information is framed (Long, 2001). Reality is not about ‘seeing’ but about ‘sense making’ [Gartner et al. (2003)]. The pRfM of a business owner is his personal ‘perceived room for manoeuvre’. Another business owner in a comparable situation may or may not ‘see’ the same opportunities as viable. The development of small businesses is ‘enacted’; it influences and is influenced by the environment in which the businesses operate (Atherton, 2003). The subjective approach to entrepreneurship is supported as well by Kor et al. (2007) who state that an opportunity is not necessarily a completely new innovation to the economy. Being new to the specific business is sufficient to call a development option an opportunity.

3 The case study

The case study area is Kampereiland (the Netherlands), an area of 4200 ha of cultivated land owned by the town of Kampen who is the lessor of Kampereiland. The land is rented by 102 family-owned dairy farms, the average farm needs to strengthen farm income (Methorst, 2013). The highly comparable socio-material context allows for an accurate investigation of differences in perceptions on the level of the individual farmer. Farmers operate in challenging circumstances due to structural changes that increase the importance of SDM and entrepreneurship (Clark, 2009; Hansson et al., 2010; Verreynne, 2006; Bergevoet, 2006). The end of the EU production limitation has brought to an end the highly regulated and protected economic system (Potter and Tilzey, 2005; McElwee, 2006). An increase in price volatility is expected while production costs are rising due to legislation on environmental impact and animal welfare (Meulen et al., 2012; Samson et al., 2016). The predominant strategy in Dutch dairy farming is increasing the production capacity via scale enlargement and intensification of production (Schans and Keuper, 2013), a strategy in line with the focus of policies and markets on the production of large quantities of uniform products (Ploeg, 2003; Benvenuti, 1989). Diversifying their activities is another strategy for farm owners (Atterton and Ward, 2007; Ploeg et al., 2009). The owner of Kampereiland (the lessor) aims to develop opportunities for diversification of dairy farms, as scale enlargement and intensification endanger the natural and landscape values of Kampereiland. The ability to recognise opportunities has not been developed for a long time as the need for a critical assessment of the economic situation of the farm was low (McElwee and Bosworth, 2010). Moreover, farm development in the past had created a path dependency causing reluctance to cross the boundaries of agriculture and choose a diversification strategy (Wilson, 2008). In general farmers are more focused on improving current conditions rather than exploring new ideas (McElwee, 2006; Haugen and Vik, 2008; Morgan et al., 2010; Vesala and Vesala, 2010). The farmers in the case study operate in a context with high natural and cultural-historical value and are confronted by changing societal demand on rural and farm development (Wästfelt et al., 2012). The farmer needs to find a strategy to counter the cost price squeeze that is the result from increasing costs of resources and a decrease in the price of products (Ploeg, 2000). The predominant strategy in Dutch dairy farm development is scale enlargement and intensification of production, a strategy that negatively affects landscape and biodiversity values (Marsden, 2003; Wiskerke and Roep, 2007) and therefore has limited potential in the case study area. Alternative farm development strategies entail new value chains based on regional products and on the characteristics of the farm in the rural context such as farmers’ markets, care and recreation (Potter and Tilzey, 2005; Oostindie, 2015). This means there is a range of opportunities for farm development. The farmers’ perception of these opportunities will affect the SDM of the farmer as is the case for small business owners (Liberman-Yaconi et al., 2010).

The highly comparable socio-material context of the dairy farmers at the level of the case study of Kampereiland allows to study the individual differences. However, a closer look on the specific situation is needed as it may affect the generalizability of the findings. For this purpose a desk study was made on the characteristics of dairy farming in Kampereiland combined with interviewing experts who are well acquainted with both dairy farming in the Kampereiland area and in the Netherlands in general. The situation in Kampereiland for dairy farms was deemed to be highly specific in the tenancy context, leading to a limitation in the free land market and in the number of farmers still active. The number of farms still active is relatively high as young farmers need less foreign capital to take over the family farm and for senior farmers there is less incentive to stop farming as there is less value in the assets of the farm that can be made available by ending the farm. The expert opinion showed that many important characteristics of dairy farming were deemed to be like in other dairy farm regions.

4 Methodology

4.1 Approach

In this paper we follow Kor et al. (2007) and Renko et al. (2012) taking the subjective opportunity enactment perspective as the starting point for the construct pRfM. In the enactment perspective opportunities are an outcome of the sense-making activities of individuals (Gartner et al., 2003). The entrepreneur uses his experiences and recognises a given situation as an opportunity for him to develop the business. Research on farm development in relation to the context of labour requirements and product markets has revealed different ways of enactment which result in different ways of organising the farm (Ploeg et al., 2009). The construct pRfM is tested in a context of family-owned dairy farming which creates an opportunity to focus as well on the why and how of opportunity identification in this context. The research consists of three phases: (1) exploratory qualitative research; (2) quantitative research; and (3) confirmatory in-depth interviews with farmers and stakeholders.

In Phase 1, a combination of literature study and exploratory ethnographic research is used to operationalise the set of opportunities for dairy farm development that are considered to be opportunities in general, the so-called third-party opportunities (McMullen and Shepherd, 2006). Phase 2 entails a survey collecting data by sending a questionnaire to respondents in the case study area. Using the results of the exploratory research in Phase 1, the construct pRfM is operationalised in the questionnaire as ‘the opportunities perceived by the farmer as viable in order to obtain (part of) his income’. In the context of a farm as a running business, opportunities include the adaptation of ideas and activities already in practice on other farms. The list of 15 opportunities can be divided in 5 groups: (1) Opportunities related to the primary production process of milk; (2) Opportunities related to diversification involving direct contact with people; (3) Opportunities related to diversification – not involving direct contact with people; (4) Opportunities involving ending dairy production; and (5) Other opportunities for farm development. A 16th, blank, opportunity is added to provide a possibility for respondents to add an opportunity which was not included in the list. The details of the 15 opportunities can be found in Table 1. To ensure the quality of the questionnaire external expertise in the context of the case study and in the design of questionnaires is used and the questionnaire is tested on dairy farmers. In Phase 3, interviews with survey respondents and stakeholders of dairy farm development are conducted to check the results of the statistical analysis for validity and meaningfulness from a practical perspective.

Table 1 Average score on the presented opportunities for farm development for the four clusters of dairy farmers. Scale 1–5 (‘1-certainly no’ to ‘5-certainly yes’).

Clusters	1	2	3	4
Opportunities for farm development^1	Maximising production	Optimising resources	Diversifying production	Ending production
	n = 29	n = 21	n = 21	n = 8
Intensive dairy, high milk yield/ha and per cow	4.2	2.4	2.5	1.8
Joint farming with 1 or more other dairy farms	3.3	1.8	2.7	2.6
Relocating outside Kampereiland	2.6	1.6	2.2	3.1
Land-based dairy, mostly own feed	2.3	4.2	4.2	3.5
Organic dairy	1.3	1.7	2.5	1.6
Dairy plus another company	1.9	1.5	2.9	3.0
Dairy plus income from nature conservation	2.5	2.8	3.5	2.4
Average^2 of 4 opportunities: Dairy plus: −processing own milk, −on-farm costumers, −care, −recreation	1.5	1.3	2.9	1.4
Average^3 of 2 opportunities: End dairy plus: −other company, −other income source	1.4	1.4	1.9	3.4

The opportunities with their values represented in bold were used to identify the farm development strategy.

1 Two opportunities are excluded from the table: ‘dairy&off-farm job’ and ‘dairy&energy production’ as the differences between clusters for these opportunities were small.

2 The four opportunities on on-farm diversification are presented as average, differences within clusters were small.

3 The two opportunities on ending dairy farming are presented as average, differences within clusters were small.

4.2 Data

The questionnaire was sent in February 2013 via regular mail and e-mail to all 102 farmers (census) leading to 79 completed questionnaires (78%). The group of non-respondents were assessed by local experts as not markedly different in their characteristics from the respondents. Respondents were asked to indicate on a 5-point Likert scale their perception of the viability of each opportunity in their situation, the so-called first-person opportunity (McMullen and Shepherd, 2006). Data on production characteristics were asked for the years 1985 and 2012 creating data to answer question 3 on the relation between differences in pRfM and farm development over a longer period of time. Finally the farmers were asked to indicate their preferred opportunity provided a situation without limitations. This part of the data allows to answer question 4 on the relation between pRfM and the personal preferences of the farmer.

4.3 Data analysis

Data about the first-person opportunities were analysed using principal component analysis and two-stage cluster analysis. To establish the number of factors to be retained the Eigen values, the scree plot, and the theoretical sound interpretation of the factors will be used. To facilitate interpretation of the factors the factor solution was rotated (Oblique rotation). The second step was a two-stage clustering procedure, based on three factor scores of the first-person opportunity data, to identify clusters of farmers with a similar pRfM. The first stage clustering procedure is hierarchical to determine an appropriate number of clusters. A steep increase in the agglomeration coefficient indicates the merging of clusters that are too different, which therefore indicates the appropriate number of clusters. Several solutions were explored to find a parsimonious solution. The second stage clustering procedure is non-hierarchical to further minimise the heterogeneity within clusters. Cluster centres of the first stage are used as starting values in the second stage. Non-hierarchical procedures often result in clusters of more equal size than hierarchical procedures. To analyse the relation between preferred opportunities (given a situation without limitations) and the clusters of pRfM, the 15 preferred first-person opportunities were combined into five variables of preferred farm strategy. Each variable of preferred farm strategy consists of the preferred first-person opportunities that correspond with the preferred farm strategy. A fifth variable of preferred farm strategy called ‘other’ was needed, this variable combines five preferred opportunities that were not unique in their link with a preferred farm strategy: off-farm job, energy production, dairy and nature, relocation of the farm and joint farming. A cross table was made to relate the five variables of preferred farm strategy with the four clusters based on the pRfM (Table 3). The asymptotic significance of the Pearson Chi-Square test is added, indicating whether the preferred farm strategy and the cluster of pRfM are independent where a value of 0.05 or lower indicates that they are not independent. The validity and meaningfulness of the clusters in practice was tested by presenting the findings to 15 respondents (selected at random) and 16 stakeholders of dairy farming (Phase 3 of the research). The leading question was whether the clusters found were recognised and whether the interviewee missed specific (sub)groups of farmers in the results.

Table 3 Farm strategy as preferred by the farmers who are part of the four clusters of farm development strategies deemed viable by dairy farmers (based on perceived Room for Manoeuvre).

			Cluster	1	2	3	4
Farm strategy preferred		Assymp. Sign. Chi Square test	Total	Maximising production	Optimising resources	Diversifying production	Ending production
1	Maximising production	0.05	35	18	9	7	1
2	Optimising resources	0.00	17	1	10	5	1
3	Diversifying production	0.01	6	1	0	5	0
4	Ending production	0.01	3	0	0	1	2
5	Other opportunities chosen^2	^3	18	9^1	2^1	3	4
	Total		79	29	21	21	8

The values represented in bold are the number of farmers where the farm strategy preferred corresponds with the farm strategy deemed viable according to their pRfM.

1 Including 1 missing value.

2 Opportunities which may be combined with all four clusters.

3 0.25/0.22/0.32/0.93/0.42.

5 Results

In this section, the four research questions as stated in the introductory section will be answered:

5.1 How can the construct pRfM be measured?

Using principal component analysis three factors were found to explain 50.3% of the variation in the first-person opportunity data. The scree plot shows a clear heel at the fourth factor (with an Eigen value of 1.02), suggesting three or four factors. The most sound interpretation was possible with three factors as the three dimensions of the pRfM. The three dimensions are interpreted as: (1) diversifying dairy, (2) ending dairy and (3) maximising dairy. The farmers do differentiate in their perception of opportunities for farm development even when operating in a highly comparable socio-material context at the level of the case study.

5.2 Can the construct pRfM differentiate in a meaningful way between business owners?

The 79 farmers were plotted in a cube with the three dimensions of the construct pRfM as three axis. This resulted in 79 ‘dots’, each representing an individual farmer. Using two-stage cluster analysis four clusters of farmers were found within this set of 79 ‘dots’. These four clusters were characterized using the mean values per cluster for the first-person opportunities (Table 1). High mean scores were used to describe the characteristics of the clusters, which are as follows:

Maximising production (n = 29)

Focus on maximising production per ha using imports of feed next to own feed production. Joint farming is seen as ‘maybe possible’. Energy production may add to farm income.

Optimising resources (n = 21)

Focus on milk production using on-farm produced feed with limited inputs of (concentrated) feed. The use of the farm as production unit is optimised. Off-farm job and energy production are seen as possible opportunities.

Diversifying production (n = 21)

Focus on milk production possibly combined with on-farm diversification. The use of the place where the farm is located is optimised. Dairy farming is based primarily on own produced feed, comparable to Cluster 2.

Ending production (n = 8)

Focus on ending dairy farming in the near future, income from other activities (possibly retirement). Currently optimising on-farm resources, possibly an off-farm job or another company.

The question on meaningfulness is answered using interviews with stakeholders (Phase 3 of the research). The clusters found were presented to 15 farmers and 16 stakeholders of dairy farm development in Kampereiland. All interviewees recognised the four clusters as existent in the context of Kampereiland and in Dutch dairy farming in general. They found the clusters to be in line with their own views on the existing differences between farmers. For most interviewees the number of farmers with a positive perception of diversification (21 farmers in Cluster 3) exceeded their expectations.

These results show that the construct pRfM allows for a differentiation of business owners. The characteristics of the clusters found are meaningful for both farmers and stakeholders in dairy farm development in the region. The next part describes the relation of pRfM with the long-term strategic development of the business.

5.3 Are differences in the pRfM linked to differences in farm development over a longer period of time?

Table 2 presents the average production characteristics in kg of milk per cow, per ha and per farm for the four clusters for the years 1985 and 2012. The percentage of growth in this period is included in the table. Cluster 1, maximising production, shows the highest numbers for all three parameters, both in absolute numbers and in growth rate. Milk production per ha is nearly 18,000 kg in 2012, clearly showing a focus on maximisation of production using imported feeds. According to local experts a level of 14,000 kg of milk per ha is considered to be self-sufficient for roughage. Total milk production is by far the highest and doubled between 1985 and 2012. The focus on maximising production for Cluster 1 farmers in their pRfM is linked to their strategy over the past 25 years. For the other three clusters the data show a great similarity in the business for the year 1985 while there are clear differences between the clusters in 2012. Starting with 4 farms, ending production, a stable and limited growth for all three characteristics can be seen. The limited growth in total milk production and number of ha is the result of a policy of the lessor to redistribute milk quota and land among other tenant farms when farms ended their production. The decrease in milk production per ha for Cluster 4 farms shows they have not actively pursued to increase farm production. The focus on ending production for Cluster 4 farms in their pRfM is consistent with long-term farm development.

Table 2 Production characteristics for the four clusters 1985–2012 (×1000 kg).

	milk/cow			milk/ha			milk/farm
Cluster^1	‘85	‘12	delta%	‘85	‘12	delta%	‘85	‘12	delta%
1	6.7	9.0	34	14.3	17.9	25	437	893	104
2	6.8	8.3	22	12.4	14.1	14	346	543	57
3	6.7	7.5	12	12.2	11.9	−2	366	622	70
4	6.7	7.2	7	12.2	10.7	−12	334	410	23

1 1-Maximising production, 2-Optimising resources, 3-Diversifying production, 4-Ending production.

Both Cluster 2 farms, optimising resources, and farms in Cluster 3, diversifying production, have increased total milk production between 1985 and 2012 with respectively 57% and 70%. The clusters show a different development looking at production per ha and per cow. Cluster 2 farms, optimising resources, increased the production per cow and per ha to 14,000 kg per ha and 8300 kg per cow which fits the strategy of optimisation of available assets. This strategy differentiates Cluster 2 farms from the focus on maximising production of Cluster 1 farms. We can conclude that the focus of Cluster 2 farms in their pRfM, optimising resources, is consistent with long-term development.

Cluster 3 farms, diversifying production, combine an increase in total milk production with a decrease in production per cow and per ha. This points to a production strategy with lower production levels per unit, which decreases the pressure on the production units for maximum production. A decreased pressure on maximisation of production per unit increases the flexibility of the business owner to spend time and energy on other activities. The combination of less focus on production per unit and more attention for other activities for Cluster 3 farms is recognised by interviewees. For Cluster 3 farms we can conclude as well that their focus on diversifying production in their pRfM is consistent with the long-term development of their farms.

We can conclude that the coherent and meaningful clusters of farmers, which were found based on their pRfM, were not only clusters at the moment when the questionnaire was answered. There is a clear link with the general development of the farms over a longer period of time.

5.4 Are differences in the pRfM related to the farmers’ personal preference for farm development?

Table 3 presents the cross table with the five variables according to preferred farm strategy and the four clusters according to the pRfM (the strategies deemed viable). Each variable of preferred farm strategy is a combination of a number of first-person opportunities presented to the respondents in the survey. For each first-person opportunity the assymptotic significance was determined (Pearson Chi Square). The highest value for Maximising production was 0.05, for Optimising resources 0.00, for Diversifying production 0.01 and for Ending production 0.01. This indicates that the preferred farm strategy and the cluster of strategies deemed viable (according to the pRfM) are not independent from each other, they are linked. For the fifth preferred farm strategy this is not the case, this can be explained by the fact that the five strategies in this group can be combined with all four business strategies found. This result shows that the preferred farm strategy and the clusters of strategies deemed viable (based on the pRfM) are linked. This means that the relation needs to be analysed further.

The most preferred way of farming is ‘maximising production’ with 35 out of 79 farmers. Ending production is the least preferred (3 farmers) followed by diversifying production (6 farmers) and optimising resources (17 farmers). The diagonal of Group 1-Cluster 1 to Group 4-Cluster 4 represents those 35 farmers whose preferred strategy is the same as the strategy perceived as viable. For 25 farmers, however, there is a discrepancy between the preferred strategy and the strategy perceived as viable. 17 farmers would prefer maximising production as strategy, 7 farmers would prefer optimising resources and 1 farmer would prefer diversifying production. The construct pRfM measures the strategies that are perceived as viable and this measure is different from measuring the preferred strategy.

6 Discussion

As the results have shown, the construct pRfM allows to differentiate between family owned dairy farmers in their perception of the viability of opportunities for farm development. Using the construct pRfM, meaningful, coherent and consistent clusters were found that reflect different patterns of farm development. This section will discuss the usefulness of the construct to study strategic decision making in small businesses. The first question is on possible limitations of the methodology used followed by the question about the validity of the results outside the context of the case study. The central question is how the findings of this study affect the theoretical understanding of SDM in relation to farm development and entrepreneurship. Additionally the results will be discussed in relation to the question how and why opportunities are identified in the context of family-owned dairy farming. The results offer the chance to discuss opportunity identification in the specific context of family owned dairy farming.

6.1 Survey data collection

The exceptional high response rate of 80% is likely related to the ethnographic study of the case study area in the first phase of the research. This allowed for the design of the questionnaire to use the wording that connects to the every-day reality of the farmers. Using local experts it was possible to analyse the 20% non-respondents resulting in the conclusion that the non-respondents were not markedly different from the respondents. Two minor issues can be raised regarding the questionnaire: (1) does the person who answers the questionnaire represent the actual decision maker? and (2) are the answers on viability of opportunities influenced by presenting a list of opportunities? Regarding the first question, family-owned dairy farms predominantly have shared ownership with one or more family members. Different owners may differ in their perceptions on viability of opportunities for farm development, especially when ownership is shared between different generations. The covering letter for the questionnaire specifically noted that the questions were meant for the person most involved in SDM. However this cannot be guaranteed. Nevertheless, the overall consistency of the results supports the validity of the answers given. The second question is whether presenting a list of opportunities may have affected the answers given. To ensure that the list represents ‘the total playing field’ of dairy farm development the questions were based on the explorative, ethnographic study of the context of dairy farming in Kampereiland in Phase 1. The questionnaire was discussed and tested with both experts and farmers. The blank option presented to the farmers did not bring forward opportunities outside the list of opportunities presented.

6.2 Validity in a broader context

The first question is whether the results are valid for family-owned dairy farmers outside Kampereiland. In interviews with both dairy farmers and stakeholders of dairy farming, the results were deemed valid in general for family-owned dairy farming. The next question is on the validity for small businesses in general. In small businesses a very broad range of different sectors can be found, each with very different characteristics. Dairy farming as small business is characterised by working with natural resources, living animals and a capital intensive infrastructure which gives them a high resource dependency (Bjerke, 2007). The capital intensive infrastructure leads to a strong connectedness of dairy farms with the production location, it is not easy to move the business to another location. Dairy farming is therefore a specific type of small business and the results need to be interpreted with these characteristics in mind. However, there is no clear reason why the influence of the owner-manager’s perception of opportunities for SDM would be different in a different sector. Testing the construct pRfM in a different context of small businesses would be interesting. It would require an operationalisation that is fitting to the specific context of those small business owners.

The results of this study underline the need of a subjectivist view of entrepreneurship in a farming or small business context. Opportunity identification is an important part of strategic decision making and as this paper clearly shows, the farmers’ perception of the opportunities differs even in a situation with a highly comparable socio-material context. This finding contributes to the existing literature on farming styles that showed that heterogeneity in farm development can be explained in how farmers organise their farm based on a different approach towards labour and capital (Ploeg, 1994; Ploeg and Ventura, 2014).

6.3 The construct pRfM

The central question of this paper is whether the construct pRfM allows to study the role of the individual business owner’s perception of opportunities for SDM. Overall, the results in Table 1 and the interviews with respondents and stakeholders show that the construct pRfM allows for a meaningful, coherent and consistent way of distinguishing between small business owners in their pRfM. All dairy farms in the case study operate in comparable environmental conditions. Therefore, the differences in pRfM are related to either differences in the farmers’ perception of the context or to differences in the situation of the farm. And the latter, the differences in the situation of the farm, are related to the decisions made by (the predecessor of) the farmer while operating in comparable context. The differences between dairy farms in this study were therefore strongly related to differences between the farmers. The construct pRfM therefore allows a distinction to be made between business owners in their perception of the viability of opportunities. This means that the construct offers a tool to identify differences between business owners in their perception of their opportunities and thus allows for a further study on the relation with other characteristics of the business and/or the business owner. More knowledge on why business owners differ in their pRfM can increase the understanding of SDM in small businesses. An interesting avenue for further research is to investigate the aspects that drive the differences found between business owners and to study how the differences in pRfM are related to the embeddedness of the business in the context in which it operates.

The theoretical contribution of the construct pRfM to literature on SDM is the focus on the importance of the perception of the owner-manager of the opportunities for business development in the early phase of the decision making process. The pRfM spans the ́room for manoeuvré for business development as perceived by the business owner. When more opportunities are deemed viable, it leads to an increase in the pRfM. However, this is not necessarily a reason to change the business strategy. Especially not when the preferred strategy is successful in providing income. In theory there is no need for the pRfM to contain more opportunities than the one preferred. However, when the pRfM contains more than one business strategy the resilience of the business is better. A business is more likely to be able to absorb changes in the market or in the business situation when it has more than one opportunity for business development (Darnhofer et al., 2010). When the pRfM contains more than one opportunity the business owner is more free in making a decision about the business strategy to follow. The pRfM is not static and so new opportunities for a business strategy may be added to the pRfM. The combination of the preferred strategy with the pRfM may provide a theoretical construct to explain why small business owners do or do not change their business strategy.

The construct pRfM is of practical use as it allows the pRfM of business owners to be measured and thus to identify different subgroups of business owners based on their pRfM. Measuring the pRfM in the early phase of designing a policy or support programme is useful for gathering information about the perspective of the business owners regarding their opportunities for business development. Measuring changes in the pRfM after a policy is implemented or a support programme has finished can be useful for monitoring the effects. It may take several years before the effect of changes in business strategies are seen in practice.

6.4 Opportunity identification in the context of dairy farming

The aim of this study is to test the usefulness of the construct pRfM. The data can be interpreted from the viewpoint of ‘how’ and ‘why’ opportunities are identified in the context of family-owned dairy farming. For the question how opportunities are identified, the discussion in the literature focuses on either opportunity discovery or opportunity enactment (Gartner et al., 2003). The construct pRfM implicitly assumes that a general opportunity for business development becomes a concrete opportunity as a result of a sense-making process. In the perspective of the construct pRfM, opportunities are enacted by the business owner. The validity of the construct pRfM in this study supports the assumption that opportunity identification in family-owned dairy farming is a process of enactment. The assumption is further supported by the clear reasoning on the viability of the different opportunities for farm development shown by the farmers in the interviews. All opportunities presented were acknowledged as existing, yet not all opportunities were assessed as viable for their specific situation [third-party versus a first-party opportunity, 33]. The farmers named different aspects as important in the sense-making process, varying from economic to personal reasons. In the context of family-owned dairy farming, opportunities are assessed in the light of the current situation and the farmer's goals. When this leads to a positive perception then an opportunity becomes a ́reaĺ opportunity for the farm in its own situation.

The data on pRfM in relation to the preferred opportunity (Table 3) provides an insight into the question as to why opportunities are identified. Only 6 farmers preferred diversification as strategy, whereas 52 farmers preferred a full focus on dairy farming. Within these 52 farmers 35 preferred a focus on maximisation of production. In the context of Kampereiland the conditions for a full focus on maximisation of production are not optimal due to nature and landscape restrictions. The focus on maintaining dairy farming as core production is dominant. Opportunities for farm development are predominantly assessed on their contribution to obtaining an income by dairy production. This confirms other studies on the guiding factor to change the business strategy, revealing that personal motivation to look outside the dairy sector is not the most important factor (Vik and McElwee, 2011). So when farmers do consider opportunities that lie outside of their personal preference, this is driven by a need to change the strategy (push) rather than a desire to change the strategy (pull). In agriculture, the shift towards diversification is seen as a shift towards more entrepreneurial behaviour (Grande et al., 2011), though for small businesses in general, diversification of production is found to indicate a survival strategy (Robson et al., 1993). The context of dairy farming is likely to have affected this finding due to: (1) path dependency (Wilson, 2008), (2) the lack of urgency for a critical assessment due to protected markets (McElwee and Bosworth, 2010) and (3) the primary focus on improving current conditions rather than exploring new ideas (McElwee, 2006; Haugen and Vik, 2008; Morgan et al., 2010; Vesala and Vesala, 2010).

7 Conclusions and outlook

The construct pRfM has proven to be useful for distinguishing business owners based on their perception of the viability of business strategies. The construct pRfM has three dimensions in which business owners differentiate: (1) diversifying production, (2) ending production, and (3) maximising production. Using these three dimensions, four clusters of business owners were found: (1) maximising production, (2) optimising resources, (3) diversifying production, and (4) ending production. The clusters proved coherent and meaningful from the perspective of business owners and stakeholders. The clusters are as well consistent with the characteristics of the long-term development of the businesses in the cluster. The construct pRfM proved to be different from merely measuring the business owner’s preferred strategy. This conclusion leads to the question which driving factors can explain the differences found and how differences in the pRfM are related to differences in the embeddedness of the farmers in the context in which they operate. More research is needed to further the theoretical and explanatory potential of the construct pRfM in relation to SDM. One step would be to develop the construct into a tested measure and another would be to include the relation with the preferred business strategy in the framework.

This study on perception of opportunities confirms the need for a subjective approach towards farm development. The personal influence of the farmer as entrepreneurial actor cannot be separated from the identification of opportunities, just as the dancer cannot be separated from the dance (Sarason et al., 2010). In showing the personal influence of the farmer on opportunity identification this paper opens the field to further explore this relation and hereby supports the call to develop a sociology of entrepreneurship (Watson, 2013). In this sociology of entrepreneurship the decision-making process of the farmer as entrepreneurial actor needs to be studied in an integral way in relation with the socio-material context of the farm(er) (Welter, 2011; Hansson and Ferguson, 2011; McKeever et al., 2015). A better understanding of the perception of opportunities is important as the ability to identify different opportunities for farm development is vital to adapt to changes in the socio-material context of the farm (Feola et al., 2015; Darnhofer et al., 2010).

Regarding the practical importance, a better understanding of drivers for differences in the pRfM is useful in the design and evaluation of policies and support programmes on small business development. Due to the connection with the location, small businesses are an integral part of the area in which they are located. This is particularly true when the businesses use capital-intensive production factors linked to a specific location such as buildings and land. In such a situation the development of the businesses affects the development of the area and vice versa. For the design of effective support programmes it is important to understand the driving factors in the decision-making process regarding business strategies, especially when the aim is to stimulate diversification of small business development that involves new business strategies.

Acknowledgements

The authors wish to thank the anonymous reviewer for the valuable comments that allowed to refine the paper.

The authors wish to thank the farmers, the farmers union and the lessor of Kampereiland e.o. for their co-operation in this research. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Additional information

Ron Methorst finished a PhD on agricultural entrepreneurship in a changing environment. He teaches at Aeres University of Applied Science, has experience in farm advisory and in the management of a production company in an international setting.

Dirk Roep is Assistant Professor of Rural Sociology. He has expertise in agricultural, rural and territorial development, sustainable food networks, joint learning, and innovation and transition of the agro-food system. He has specific expertise in place-based approaches to development and institutional innovation.

Frans Verhees is Assistant Professor of Marketing and Consumer Behaviour. He has expertise in marketing of small firms, cooperatives, and food and agribusiness. He holds specific expertise in market orientation, entrepreneurship, innovation and branding.

Jos Verstegen is a senior researcher in entrepreneurship and innovation at LEI, part of Wageningen University and Research Centre. He has published on a broad range of topics including entrepreneurship, (open) innovation, experimental economics, information economics, and competence development.