Exploring Diversification Strategies among Italian Farms

Abstract

The multifunctionality model is receiving more and more attention from policymakers as a result of recent initiatives to build more resilient and sustainable food systems as well as the potential for increased farm revenue. This paper explores the role of multifunctional farming in the Italian agriculture viewed through the lens of an entrepreneurial strategy grounded on-farm diversification. Farm diversification strategies, which broaden the farm’s traditional boundaries to include additional activities at the farm level, help the evolution towards multifunctionality. A policy-driven transition towards multifunctional farming has been noticed in Italy during the past few decades, which has prompted a strategic reconfiguration of the farm’s business models. Drawing on the identified activity of portfolio diversification, this study provides an overview of the analyzed 49,429 Italian farms, by articulating diversification strategies into four entrepreneurial activities, which are related to on/off-farm/farm-related or farm-diverse diversification strategies. This article has attempted to verify the presence of farm types that responded to portfolio diversification management strategies through the use of a cluster analysis on data from the general census of Italian agriculture. Supporting new patterns in the adoption of business models focused on multifunctionality should be considered in European rural development policies.

1. Introduction

Multifunctional agriculture is at the core of the EU policy agenda since the mid-1990s, when the concept of multifunctionality emerged as a policy issue, due to the market failure in securing adequate compensation for non-commodity outputs generated by the primary sector [1]. As this prospect provides the foundation for a holistic evaluation of the synergies that exist between the ecosystems’ services and functions and the economic value that is generated by farms, its significance is growing when considering future challenges [2]. According to the Scott et al. [3], multifunctionality may not completely eradicate land use issues, but it clearly demonstrates that contemporary ambitions such as nature recovery and carbon sequestration, among others, do not necessarily have to come at the expense of food supply. Rather, the concept of multifunctionality, as well as the recent European strategies of the green deal and plans of the long-term vision for rural areas embrace a model of agriculture that goes beyond the production of food [4,5]. This model encourages socially desirable adaptation strategies and sustainable resource management and, in addition, calls for an overhaul of agricultural policies focusing more on rural development to improve European farm resilience and competitiveness [6]. In fact, the European Commission reports that a “rural proofing” will be implemented whereby EU policies will be reviewed through a rural lens [7].

Consequently, the new European Farming Model (EFM) acknowledges the multifunctionality and diversity of European agricultural systems.

Multifunctionality refers to the multiple roles of agriculture beyond food production, including environmental stewardship, cultural preservation, and economic development [8,9]. In accordance with the well-known Triangle of Van der Ploeg and Roep [10], the most recognized multifunctional practices include deepening (e.g., quality production), broadening (e.g., biodiversity conservation, rural tourism), and regrounding (economical farming) [11]. This model recognizes the essential public goods provided by farming and seeks to develop policies with clear objectives and targeted measures that are tailored to the farmers’ diversity. In fact, this model could represent a broad conceptual framework within which practices that promote sustainable agriculture are embraced. Firstly, it promotes the integration of sustainable practices (e.g., climate change mitigation practices, circular strategies, and agroecological and regenerative practices) and, by embracing non-agricultural activities, reduces the specific dependence on individual ways of cultivation or animal husbandry and thus the intensive use of resources. Secondly, it supports local communities by promoting long-term sustainable management of rural areas [12]. Therefore, multifunctional agriculture is conceived as the possibility of providing both commodity and non-commodity outputs from farming through securing economic and environmental effects, while contributing to the socio-economic viability of rural areas [11].

Set against this background, this paper tries to provide a contribution to the literature, by adhering to a “constituent” perspective of the multifunctional farming, meant to be a strategic approach [13], aimed to orient the farmer’s behavior through different entrepreneurial activities. This perspective emphasizes the involvement of other stakeholders in the process of building a multifunctional farming system, such as policy actors, local and regional institutions, non-governmental organizations, etc. Accordingly, the adoption of the multifunctionality view brings about the identification of a policy mix impacting rural areas through a diversified set of measures involving not only the individual level (the farm’s strategy towards multifunctionality), but also the territorial level, due to the high level of territorial embeddedness of multifunctional activities [14].

If viewed from the perspective of an entrepreneurial journey, the evolution towards multifunctionality can be supported by farm diversification strategies, which expand the farm’s traditional boundaries, to incorporate new activities at a farm level. Boosted by the policy action, a transition towards multifunctional farming has been observed in Italy over the last decades, which has prompted a strategic reconfiguring of the farm’s business models [4,15,16,17]. If, on the one hand, a wide literature has underlined the recent trends towards multifunctional farming, on the other hand, there is limited research on viewing the process through the lens of exploring the role of entrepreneurial behaviors in driving this transition.

This paper aims to fill this gap in the literature by pointing out that multifunctionality in agriculture is the outcome of an entrepreneurial behavior framed within a functional reposition of farming activity [18]. As a consequence, this paper rejects the ex-ante opinion that farmers are not entrepreneurial [19]. More precisely, our hypothesis is that the transition towards multifunctional agriculture is the outcome of an entrepreneurial journey triggered by [20]: (1) developing a new entrepreneurial identity, (2) expanding beyond the boundaries of agriculture, and (3) creating new opportunities for a family farm.

This entrepreneurial journey redesigns the farming activity through rural entrepreneurship representing those activities strongly rooted in rural contexts and drawn on local natural and human resources [21], which depict a new entrepreneurial identity of farmers, embedded in rural contexts. As pointed out by Marsden and van der Ploeg [22], this implies shifting from the conventional farms’ boundaries to enlarge the farm’s activities through diversification strategies.

The literature has underlined several factors driving the diversification of farming activities beyond conventional agriculture, for instance, the reduction in risk and the opportunity of using idle resources, but also the promotion of a different lifestyle [23]. The way through which farmers build up competitive strategies of territorial anchoring is the outcome of entrepreneurial activities aimed to setting up “mixed farms” combining both agricultural and non-agricultural activities through spatially embedded social mechanisms [14,24,25]. These strategies are centered around the farm’s boundary shift [10], leading to the development of the so-called portfolio strategies, which are analyzed through the lens of the rural enterprise school [26,27]. More precisely, this paper focuses on diversification strategies employed by the farms and amounting to activate other gainful activities. According to Vik and McElwee [28], additional income-generating activities in the agricultural sector can be classified into two categories, each contributing to the diversification of the farm household’s portfolio (Figure 1):

• On-farm and off-farm diversification: this includes further classification into farm-related activities (such as firewood, bioenergy production, etc.) and farm-diverse activities (such as tourism), which encompasses lodging, accommodation, tours, and other services.
• Off-farm diversification: This can be divided into off-farm and farm-related and off-farm and farm-diverse. Off-farm and farm-related category includes activities such as organizational rural service, machine contracting, snow cleaning, etc. Meanwhile, off-farm and fam-diverse category includes activities such as consulting and accounting service, fishery, etc. These diverse gainful activities not only enhance the economic viability of farm households but also contribute to the resilience and sustainability of rural economies by integrating traditional farming practices with broader economic and ecological functions [28].

This paper is set against this background with the purpose of exploring diversification strategies in the Italian agriculture. A cluster analysis on the farms of the Italian census allowed the identification of well-defined groups characterized by specific diversification strategies. The paper presents a Materials and Methods Section with an overview of the additional gainful activities considered and a representation of the multivariate analysis used. Then, a Results Section is provided, first in the form of a descriptive analysis of the sample and then a representation of the groups identified. Finally, a Discussion and Conclusion Section identify the highlights of the presented research.

2. Materials and Methods

The empirical analysis is drawn on the last Italian census of agriculture [29]. The 7th General Census of Italian Agriculture surveys the main structural characteristics of agricultural holdings at the national, regional, provincial, and municipal levels. The collection is required by Regulation (EU) 2018/1091 [30] of the European Parliament and the Council of 18 July 2018 concerning integrated statistics on agricultural holdings and Commission Implementing Regulation (EU) 2018/1874 [31] of 29 November 2018 on data to be submitted for the year 2020 pursuant to Regulation (EU) 2018/1091 [30] of the European Parliament and the Council of 18 July 2018. The questionnaire was administered from 7 January to 30 July 2021 and can be found published online in the official ISTAT documentation and the aggregated results on the census website [29].

The data are extracted from the section D of the questionnaire, related to other gainful activities (OGAs) related to the farms. Multifunctional farms are accordingly considered with regard to the presence of OGAs within the farm. Extracted data have been processed and classified according to

Table 1. Other gainful activities connected to the farms.

Macro-Category	Activity
On-farm diversification + farm-related activities (ONFR)	Processing of agricultural products (vegetable and animal products)
	Wood processing
	Aquaculture
	Forestry
On-farm diversification + farm-diverse activities (ONFD)	Social farming
	Teaching farm
	Agritourism
	Handcraft
	Production of renewable energy
Off-farm diversification + farm-related activities (OFFFR)	Farm contract manufacturing
	Services for animal farms
	Management of gardens and parks
	Feed production
Off-farm diversification + farm-diverse activities (OFFFD)	Non-farm contract manufacturing
	Other activities

, which is structured based on the classification proposed by Vik and McElwee [28].

A cluster analysis allowed us to classify homogeneous farms according to typologies of OGAs through the Ward method (ascendant hierarchical). Cluster analysis is a statistical technique used to group objects or cases into clusters based on their similarities. Among various clustering methods, ascendant hierarchical clustering (AHC) is one of the most used approaches. AHC is a method that builds a hierarchy of clusters in a bottom–up manner. This process starts with each object or data point as a single cluster and then successively joins pairs of clusters until all data points are grouped into a single cluster. The result of AHC is a tree structure known as a dendrogram, which illustrates the arrangement of clusters and the distances at which they join. The Ward method is a specific strategy for determining which clusters to join at each stage of the hierarchical clustering process [32,33]. Within AHC, the Ward method is a popular choice due to its focus on minimizing variance within clusters. By minimizing variance, it is possible with this method to create compact clusters even with low observation numbers. Moreover, with respect to other classification techniques, it allows to take into account specific units evidencing extreme values for each single variable. One of the limitations of the method is the sensitivity to outliers based on variance. In our case, the high numerosity of cases eliminates the incidence of outliers. The following active and illustrative variables have been taken into account to classify the farms (

Table 2. Active and illustrative variables for cluster analysis.

Type	Variable
Active variables	Adhesion to a producers’ organization
	Adhesion to a network of enterprises
	Adhesion to other enterprises/organizations
	Rural development plan (RDP) areas
	Information communication technology (ICT) and/or innovation adoption
	Revenues from selling of products
	Farm size
	Gender of the head of farm
	Farm head’s working days
	Farm head’s age
	Farm head’s level of education
	Standard output
	On-farm diversification + farm-related activities
	On-farm diversification + farm-diverse activities
	Off-farm diversification + farm-related activities
	Off-farm diversification + farm-diverse activities
	Family members/relatives working in the farm
	Number of family members/relatives working in the farm
Nominal illustrative variables	Region
	Technical economic orientation—TEO
	Citizenship
	Time devoted to OGAs
	Livestock Unit
	Number of innovations
	Combination of innovations
	Combination of OGAs
Continuous illustrative variables	Utilized agricultural area—UAA
	Working days
	Farm head’s age
	Standard output

).

Both economic and sociodemographic variables are considered, with the purpose of better clarifying the cluster composition and the differences among the clusters. Particular attention is also devoted to the presence of family members working in the farm. As a matter of fact, diversification strategies are made possible thanks to the support of the family members and to the division of labor within the family group (for instance, in the agritourism activities, the family members work at the farm level, hospitality, and food provision level or directly sell farm products) [34].

3. Results

The sample analyzed is 49,429 farms, which captures the representation of Italian agriculture. Figure 2 represents the spread of the sampled farms by region.

In the Northern Italy, Trentino Alto Adige with the two autonomous provinces of Bolzano and Trento represent the most numerous region (with 4837 and 1041, respectively). Other northern regions such as Piedmont and Lombardy also have a significant number of farms, 4532 and 4172, respectively. In the center, it is Tuscany that shows the highest number of farms represented (5481), while in the South, the two islands and Apulia have the highest concentration. This representation is consistent with the Italian ISTAT census and thus best represents the distribution of Italian farms.

Analyzing the distribution by rural development plan (RDP) classification (Figure 3), it can be seen that areas C and D constitute three quarters of Italian farms, and the RDP categories are as follows:

• A: urban and peri-urban;
• B: rural areas with intensive agriculture;
• C: intermediate rural areas;
• D: rural areas with overall development difficulties.

The distribution per utilized agricultural area of the farms (Figure 4) clearly shows how small farms predominate in the Italian context. Overall, 54.6% of the farms have an average size of less than 10 hectares and only slightly more than 10% are more than 50 hectares. This figure perfectly reflects the structural polymorphism of holdings that has always characterized the Italian agricultural sector, as Vecchio et al. [35] reported.

Cluster Analysis

A cluster analysis was conducted that clearly delineated four homogeneous and different farm types with reference to the OGAs, distributed as reported in

Table 3. Clusters and related diversification strategies.

N.	Name	Farms	%
Cluster 1	Off-farm diversification strategies in non-rural contexts	15,217	30.8%
Cluster 2	On-farm diversification strategies in farm-diverse activities in remote rural contexts	8809	17.8%
Cluster 3	On-farm diversification strategies in farm-diverse activities adopted by women farmers in intermediate rural contexts	14,657	29.7%
Cluster 4	On-farm diversification strategies in remote rural contexts	10,746	21.7%
	Total	49,429	100%

.

Cluster 1—off-farm diversification strategies in non-rural contexts: The first cluster is the largest and comprises 15,217 farms, with a percentage share of 30.8%. The farms are predominantly managed by a mature male entrepreneur (41–64 years old) with a good level of education (high school diploma with specialization in agriculture). Furthermore, 69% of the farms in this cluster are located in Northern Italy, and considering the RDP areas, a significant portion of the farms in areas A and B are in this cluster, 36.4% of the farms in area A and 54.5% in area B, and together they represent about 42% of the farms in the cluster. The data show that the farms have a relatively medium-to-high physical and economic size and are specialized in arable and grain crops. In fact, almost all farms with a standard output (SO) over 100,000 are present in this cluster. In addition, about 60% of the farms declare to have more than 20 ha area, and 77% of the farms declare to have more than 50 ha area in the sample, and 84.4% of those declaring to have more than 100 ha area are concentrated in this cluster. With regard to diversification, the farms in the first cluster are characterized by non-agricultural diversification strategies (58.4%). More precisely, extra-agricultural activities are linked or not to agricultural activity (42% of the cluster), thus including farm and non-farm contract production, livestock services, garden and park management, etc. Focusing on this aspect, 71.5% of the total of this typology in the entire sample is present in the cluster.

Cluster 2—on-farm diversification strategies in farm-diverse activities in remote rural contexts: In the second cluster, 8809 farms are predominantly in the north and account for 78% of the farms, with the northeast alone accounting for 62.3%. The farms are located in rural areas with complex development problems (D areas): indeed, 74% of the farms are located in these areas, where the main motivations for diversification are related to the need to escape price–cost compression [22]. The farms in the cluster are managed by young or mature entrepreneurs with a vocational high school diploma. Analyzing the economic size, 56% of the farms have an average size SO between EUR 25 thousand and 100 thousand. Farms in this cluster have a small-medium size on average, with 57% of companies with an average size between 1 and 10 hectares. Business strategies focusing on intra-farm diversification strategies (82%), both towards farm-related and other activities, are implemented, either through the valorization of agricultural products at the farm level (through processing and direct sales) or through the diversification of agricultural activities (agrotourism, bioenergy production, etc.). Strategies are also supported by collective farming initiatives, based on the membership of producer organizations, which enable farmers to strengthen their strength in agri-food chains. Farms are technically and economically oriented towards herbivorous livestock and make a strong contribution to multifunctional agriculture due to their territorial location in marginal rural areas and the type of professional farming. In fact, farmer-entrepreneurs are engaged in farming for more than 300 days and demonstrate both the adoption of innovation (in the last 3 years) and digital solutions.

Cluster 3—on-farm diversification strategies in farm-diverse activities adopted by women farmers in intermediate rural contexts: The third cluster is the second most numerous, comprising 14,657 farms, with a percentage incidence of 29.7%. The cluster shows a relatively high presence of women farmers, with 60% of them being concentrated in this cluster. The farm is configured as a family business, where the woman is supported by other family members (almost two members). About 64% of the farms are located in the center and south of Italy, which is also 77% of the farms in the sample in these areas. The farms are essentially located in intermediate and remote rural areas (80%), with a relatively small size; in fact, 40% of the farms declare an SO between EUR 8 and 25 thousand, and 25% of the farms declare an SO between EUR 25 and 50 thousand. Moreover, 56% of the farms in the cluster have an average size between 1 and 10 ha, with a specialization in permanent and multiple crops. As far as diversification strategies are concerned, most of the farms (74.7%) focus on on-farm ones, favoring mostly diversification into off-farm activities.

Cluster 4—on-farm diversification strategies in remote rural contexts: The last cluster consists of 10,746 farms, with a percentage incidence of 21.7%. The farms are predominantly concentrated in the south (40.8%). By investigating the locations in RDP areas, more than three quarters of this cluster is characterized by farms in areas C and D. The farms in this cluster are mostly small and micro enterprises managed by elderly people. In fact, 64% of the farms in the cluster do not exceed EUR 8000 of SO, and 23% reach a maximum of EUR 25,000. This fact is confirmed by the average size of the farms; in fact, 84.2% of them do not exceed 5 ha in size, and going into more detail, 71% of the farms of the entire sample below 1 ha are present in this cluster. The cluster is characterized by on-farm diversification strategies (62%). The farms in this cluster utilize diversification strategies that are clearly linked to de-activation and de-structuring strategies, aimed at enhancing the resilience of the farm, as a preparatory condition before abandoning the activity.

Through Figure 5, the authors tried to position each cluster on the theoretical graph by Vik and McElwee [28], by differentiating on- and off-farm diversification and by considering farm-related and farm-diverse activities.

4. Discussion

This research represents an attempt to adopt a constituent perspective on building a multifunctional farming system through an entrepreneurial strategy aimed to remove a conventional farm’s boundaries and to take on new diversification paths. As evidenced from the empirical analysis, these strategies have brought about new value-capture approaches based on new (alternative) networks, thus confirming inferences of previous studies [36].

Empirical analysis has identified four entrepreneurial typologies of multifunctional farming. A first group of farms has been identified as having the strategic purpose of adding value to conventional farm products [37]. Off-farm diversification, the only group that has the focus that includes activities not directly related to agriculture such as animal farm services or park management, reflects a resilient strategy to cope with the volatility of agricultural markets and exploit external expertise and resources. Hansson et al. [23] state that a farmer diversifies for various reasons, including economic factors but also family and social factors, in line with the importance of human and social capital [19] in the management of diversified farms.

The second cluster is of particular importance, in terms of the territorial localization of farms, working in areas with complex problems of development. These farms are at the center of policy targets aimed at empowering them within initiatives aimed at making these areas prosperous, connected, and resilient [38]. Cluster 2 contributes mostly to building smart rural development strategies through the adoption of innovation and the presence of digital solutions [39], as well as through the activation of collective actions that can develop a remote territory [40].

The third cluster reflects a family business model in which women play a crucial role in management and strategic decision making. The business models of the farms of the cluster are embedded in rural contexts and are oriented to boost on-farm diversification with the purpose of supporting social farming and the valorization of local resources. As suggested by Giarè et al. [15], social agriculture and multifunctional diversification can promote social inclusion and sustainability, helping to keep rural communities dynamic and resilient. The literature also suggests that the empowerment of women on farms can improve the effectiveness and sustainability of diversification strategies [37]. The third cluster is strongly characterized by the presence of farms led by women, a rare farm characteristic of Italy that has often been analyzed as a variable of attention for the well-known gender gap in agriculture [41,42].

A final group (Cluster 4) is characterized by a reduced farm activity. Sociodemographic variables are the basis of this diversification strategy, with the farms being managed by elderly farmers that have limited physical and economic resources. Consequently, farms of Cluster 4, due to sociodemographic constraints, seem to be characterized by the lack of locus of control [43] and are managed by “individuals who cannot believe in their ability to control the environment through their actions (and are)… reluctant to assume the risks that starting a business would entail” [44] (p. 27). In this case, based on McElwee’s [19] typology, it is possible to classify a farmer as a farmer with a limited capability of diversification, which is in line with the marginalization of the farm enterprise, as described by Meert et al. [45], due to reduction in the farm activity or the introduction of part-time farming or semi-retirement. The differences in education levels, economic and physical sizes of farms, and demographic characteristics of farm entrepreneurs, such as age and gender, profoundly influence the adopted diversification strategies. Diversification, both on-farm and off-farm, appears to be closely linked to territorial specificities and available local resources [14].

5. Conclusions

The literature has emphasized a varied set of motivations for starting new business beyond conventional farming models. This paper has analyzed these issues through the perspective of the contribution to building a multifunctional farming system, acting as a driver for rural transformation and development. In this new vision, we have conceptualized the idea of a farmer as an entrepreneur [19] under the hypothesis that “farmers are much more seen as rural entrepreneurs who combine a number of farm and non-farm activities partly paid by commodity markets and partly by government markets” [46] (p. 22).

To the best of our knowledge, this analysis represents the first attempt to explore multifunctionality and diversification through the lens of an entrepreneurial strategy based on the official data of the Italian census [29].

The clear distinction among different clusters (that is, different strategies) indicates the need for a more articulated set of policy measures (multi goals, multi policies, and multi-level governments) [47]. The result of this empirical analysis provides us with an articulated set of diversification strategies that, to different extents, are tied to the activation of multifunctional farming.

These observations underline the need for tailored policy and development strategies that consider the multifunctionality of agriculture and the different needs of Italian farming communities, in line with the need to adapt agricultural policies to specific local contexts to promote resilience and sustainability [1,24]. Against this backdrop, the coherence of trajectories of farm diversification facing the policy target underlined in the long-term vision of rural areas up to 2040 needs to be discussed. “Diversified rural areas” are considered as one of the engines of rural development identified in the long-term vision. Nonetheless, the policy provision is different, and an access to rural development policies requires a different entrepreneurial attitude: survival strategies can be supported through “traditional” annual support granted through direct payments. On the other hand, reducing transaction costs [48] for investment measures is vital to trigger a higher access to policy measures whose purpose is to support diversification strategies and revitalize rural areas through endogenous and sustainable rural development paths. The results could help to populate the rural observatory that has been proposed by the European Commission in the Rural Action Plan of the Long-Term Vision, an initiative that will be supported by the Common Agricultural Policy and the Cohesion Policy to make rural areas stronger, connected, resilient, and prosperous. With a special reference to the specific diversification strategies that have been adopted, the value-creation strategies require a dedicated policy access that boosts synergies and cooperation among farms (for instance, by promoting producers’ associations), to expand the potential markets. Moreover, either on-farm or off-farm and farm-related/farm-diverse diversification strategies need to combine different policy instruments through renewed synergies between the micro level (the farm) and the meso level (the local context) [1]. This articulated set of opportunities may trigger and sped up the transition towards multifunctionality, by supporting new business models centered around the valorization of a basket of goods and services [49]. As pointed out by Jeannerat and Crevoisier [50] (p. 13), new territorial policies should support more entrepreneurial ecosystems “that are not only about developing new production and market opportunities but also about shaping the future social conditions of these opportunities”.

Even though this article makes use of a representative sample of the entire Italian territory, the study exhibits the limitation of drawing conclusions based on entrepreneurial behavior from secondary data, thus lacking a primary survey to delve into the behavioral and perceptual dynamics of entrepreneurs by investigating the sphere of the “self” that conditions business choices, in particular, innovative business management choices [35]. In any case, the secondary data offer the possibility of being able to interpret the behavior as an ex-ante consequence of the choices made, thus only lacking the motivation of the choices.

Future research is also necessary to better qualify diversification strategies, for instance, through considering the role of new disruptive technologies, such as digital technologies, as well as new relational configurations and territorial proximities [37], which may affect the development paths and redesign smart rural development trajectories.