Content uploaded by Angelo Frascarelli
Author content
All content in this area was uploaded by Angelo Frascarelli on Sep 15, 2020
Content may be subject to copyright.
Received: 11 December 2018 Revised: 17 October 2019 Accepted: 2 November 2019
DOI: 10.1111/apce.12263
ORIGINAL ARTICLE
Drivers of participation in collective arrangements
in the agri-food supply chain. Evidence from Italy
using a transaction costs economics perspective
Stefano Ciliberti Angelo Frascarelli Gaetano Martino
Università degli Studi di Perugia, Italy
Correspondence
Stefano Ciliberti, Universitàdegli Studi di
Perugia, Borgo XX giugno, 74 – 06121 Perugia,
Italy.
Email: stefano.ciliberti@unipg.it
Abstract
The analysis of institutions in the agro-food sector is gain-
ing momentum since it represents a complex and rele-
vant area of study as concerns intermediate-product mar-
kets. Traditionally there has been a problem of organization
among farmers mainly due to the reluctance to pool deci-
sional and property rights on input and/or output. Follow-
ing the Transaction Costs Economics framework, the paper
aims to investigate which are the main drivers of the collec-
tive forms of organization in the Italian agro-food system,
paying particular attention to transaction costs’ attributes
and to the increasing role played by the institutional envi-
ronment as well. The choice to join a cooperative or pro-
ducer organization is therefore conceptualized as a gover-
nance structure choice. In this regard, based on the Italian
version of the Farm Accountancy Data Network, econo-
metric models are estimated in order to account for the two
organizational alternatives under investigation.
KEYWORDS
cooperation, producer organization, TCE
1INTRODUCTION
The modern agro-food system is characterized by the presence of numerous people involved in deliver-
ing food from the field to the fork. It has a funnel-shaped structure (or oligopoly/oligopsony) caused by
the large number of farmers that harvest crops and/or breed livestock. They provide the raw materials
© 2020 The Authors Annals of Public and Cooperative Economics © 2020 EMF
Ann Public Coop Econ. 2020;1–23. wileyonlinelibrary.com/journal/apce 1
2S. CILIBERTI, A. FRASCARELLI AND G. MARTINO
to a small number of processing companies and manufacturers that, in turn, sell their finished products
to a few large retailers (Sexton, Sheldon, Mccorriston, & Wang, 2007). In this context, market power
emerges due not only to imbalances in the bargaining power of the firms, relationship-specific invest-
ments, asymmetric information and incomplete contracts, but also to asymmetric costs of contract
enforcement (Gow, Streeter, & Swinnen, 2000; Renda et al., 2014).
It is widely recognized that agricultural production continues to be highly dispersed, since there are
many small, economically disadvantaged farms (i.e. with limited capital and market access) in a much
weaker negotiation position than their various contractors (McCorriston, 2002). Moreover, organiza-
tion among farmers is traditionally problematic, mainly due to their lack of ability to pool decisional
and property rights on input and/or output. This fact negatively affects coordination of the different
stages along the supply chains, usually causing a miscoordination between producers (farmers) and
buyers (intermediaries, processing industries or retail sector) which, in turn, causes price instability
and market imbalances.
In this regard, it must be noted that the European authorities have historically fostered integration
and cooperation (both horizontal and vertical) in the agro-food sector (Budiguel, 2016; Falkowski
& Ciaian, 2016). Hopes for a more balanced distribution of rents within the food supply chain are
also placed on producer organizations (POs). Literature gives various names to POs providing eco-
nomic and technical services. Apart from the term a ‘producer organization’, others are used, such as
a ‘farmer organization’, ‘producer group’, ‘marketing group’, ‘producer and marketing group’, ‘agri-
cultural cooperative’, ‘producer-owned enterprises’ or ‘member-owned firms’ (Harris, Stefanson, &
Fulton, 1996; Bouamra-Mechemache & Zago, 2015). A common denominator to all these types of
organization is their objective to improve their members’ welfare. In a broad sense, PO presents it as
a rural business, owned and controlled by producers, and engaged in collective marketing activities
(Penrose-Buckley, 2007). In this sense, according to Bijman et al. (2012), they can be approached as
user-owned, user-controlled and user-benefit organizations, in the same way as cooperatives.
POs can take different legal forms, including agricultural cooperatives, but it is mainly identified
based on the specification of its functions and goals. Therefore, a PO may have the legal form of a coop-
erative, but in many cases it does not, as the legal requirements for cooperatives pose many restrictions
on the activities and the structure of the PO (Bijman et al., 2012). What is important is that it should be
formed on agricultural producers’ own initiative, and its aim should be to primarily improve economic
effectiveness of its member farms, mainly by adapting production and sales to the market requirements
(Falkowski & Ciaian, 2016). In this regard, POs can reduce transaction costs and strengthen the col-
lective bargaining power of farmers by, for example, concentrating supply, providing technical and
logistical assistance to their members, helping with quality management and transferring knowledge.
Empirical evidence shows that the presence of POs has a positive impact on market performance, as
it supports the ‘yardstick effect’ due to the presence of these hybrid forms of governance (Hanisch,
Rommel, & Müller, 2013; Wills, 1985).
In the European Union (EU) the special role played by POs is acknowledged and, as a result, they
can ask for recognition from each Member States they are based in. The Common Agricultural Policy
(CAP) has recognized POs and legally supported farmers’ cooperatives since 2001 in the fruit and
vegetable sector, and since 2011 in the milk sector. The 2013 CAP reform extended this form of farming
cooperative to all agricultural sectors. This can be seen as a direct consequence of the progressive
liberalization of the European agricultural market, which has increasingly exposed farmers to price
volatility and greater market risks in a general framework of declining public resources for European
agriculture (Bijman & Iliopoulos, 2014). As a consequence, in Italy the 583 recognized POs as of 2017
(that is 13.5% of 3,400 POs recognized in the EU) can benefit from access to EU funding as well as
from exemptions from EU competition rules for certain activities (such as collective negotiations on
S. CILIBERTI, A. FRASCARELLI AND G. MARTINO 3
behalf of their members, planning of production or for certain supply management measures), since
negotiating with contractors is a basic function of any PO (Van Herck, 2014). Overall, what emerges is
that both policymakers and stakeholders are becoming increasingly aware of the problems relating to
the functioning of agri-food supply chains. However, the central question concerning the factors that
stimulate participation in collective forms of organization in the agro/food supply chain remains almost
unexplored, at least empirically.
Economists have long been aware of the importance of cooperatives and collective arrangements
in agriculture and significant contributions have been published on the changes in their status and
the challenges they represent (Cook, 1995; Nuhanovic-Ribic, Tortia, & Valentinov, 2017). The
rationale for analysing the determining factors of participation in cooperatives and POs is based on
the assumption that the diffusion of contracts, the nature of the ex-post negotiations between farmers
and their contractors and, consequently, the division of ex-post surplus between them are sensitive
to the organizational structure of the agents involved in a given transaction (Coase, 1937; Grossman
& Hart, 1986; Williamson, 1975; 1985). According to Coase (1998), transactions matter, because
their organization under different types of arrangements and under the umbrella of institutions that
make them happen more or less easily, determines the capacity of economic activities to develop
and take advantage of the division of labour and of specialization. In that sense, the choice of a form
of organization for arranging transactions, i.e. the transfer of rights among parties to an activity of
production or exchange, is crucial. Although transaction cost economics (TCE) has been extensively
criticized from various angles, due to the very concepts of this analysis and, more specifically, to
the notion of transaction costs itself (Daidj, 2017), cooperatives can be considered as a substitute for
the market. However, since they are not fully, vertically integrated firms that replace a market, where
transaction costs are too high, it follows that cooperatives are typical, hybrid, governance structures
(Bijman et al., 2012; Ménard, 2007). The concept of ‘hybrids’ has been proposed by economists
in particular. Their analyses are grounded in the TCE theory to encapsulate the properties of the
family of arrangements with characteristics, which differ significantly from those underlying market
exchanges. On the other hand, they also differ substantially from those presiding over the organization
of transactions within integrated firms (Ménard, 2007). This means that cooperatives combine market
elements (e.g. the price continues to play a motivational role in the transaction between farmer and
cooperative) and hierarchy (e.g. the ownership relation between members and the cooperative).
The TCE perspective is therefore adopted, in order to investigate the factors that determine farmers’
cooperation, paying particular attention to those forms of collaboration that are institutionalized in the
current policy framework. TCE predicts that specific governance structures depend above all on the
asset specificity and uncertainty of the transaction. These attributes influence the costs of transactions
and, in turn, affect the adoption of the optimal forms of governance that minimize such costs. Further-
more, scholars have increasingly paid attention to the interaction between organizational forms and
the surrounding institutional environment, which establishes the rule of the game. To this regard, the
agri-food system represents an interesting case study, since public authorities are traditionally influ-
ential in this sector, both in terms of financial incentives and in terms of specific regulations (Ménard,
2017).
Against this background, the paper aims to investigate the factors which determine producers’ partic-
ipation in collective forms of organization in the Italian agri-food system, in the light of the transaction
cost approach. Article 45 of the Constitution of the Italian Republic, dating back to 1948, explicitly
promotes cooperatives. Agriculture and the food industry can count over 10,000 cooperatives, approx-
imately 15 per cent of the Italian total. Cooperatives are mainly important for wine, fruit & vegetables
(with a combined market share of approximately 50%) and dairy products (42%); the market share for
olives and olive oil is low and on the decline (5% in 2010). Interestingly, most agri-food cooperatives
4S. CILIBERTI, A. FRASCARELLI AND G. MARTINO
are associated with one of the five recognized, national associations. The choice to join a cooperative
or PO is conceptualized as a choice of governance structure, since it is also crucial to raise awareness
of farmers’ collective actions (Bijman, 2016). This obviously poses the research question of how and
why cooperative behaviour takes place, i.e. the factors motivating the set-up of cooperatives and POs
and/or the barriers discouraging it, as well as of what reading can be done of the role of collective
arrangements and cooperative enterprises in an agri-food system characterized by a high institutional
instability. In this regard, this paper aims to quantitatively analyse the factors determining cooperation
in Italy based on an extensive dataset. This country represents an interesting case study within the
European framework since, despite a high number of family-managed farms (over 1 million in 2010)
with a very small, average farm size (less than 10 hectares), cooperation assumes a fundamental role
and such types of collective arrangements are widely established, as shown by Pascucci, Gardebroek,
and Dries (2012).
The paper is organized as follows. Section 2 describes the theoretical framework dealing with the
determining factors of various forms of governance in the agri-food system according to the TCE.
Section 3 defines the dataset and the specification used for the quantitative analysis and discusses the
econometric models adopted, in order to test the research hypotheses. Section 4 reports the findings
and discusses both the theoretical and practical implications in the light of TCE assumptions. Lastly,
conclusions are drawn in Section 5.
2CONCEPTUAL FRAMEWORK AND RESEARCH
HYPOTHESES
The study of institutions in the agri-food systems is gaining momentum, since it represents an intricate
and undoubtedly relevant case study as regards intermediate-product markets (Ariyaratne, Feather-
stone, Langemeier, & Barton, 2000; Hubbard, 1997; Ito, Bao, & Su, 2012). Due to their perishable
nature, the impact of quality on consumers, their fragmented supply, their vulnerability to climatic
variations and to epidemics of various nature, agricultural products have always raised problems of
coordination and control across the different stages of the supply chain, with high transaction costs
as a consequence (Royer, Mènard, & Gouin, 2015). In this regard, TCE makes the assumptions that
exchanges are not cost free, since transaction costs are those resulting from a transfer of property rights
between agents (Royer, 2011). More specifically, transaction costs are defined as ‘costs for negotiating,
enforcing and monitoring a contract’ (Hubbard, 1997; Matthews, 1986; Royer, 2011). TCE assumes
that the transaction is the basic unit of analysis (Commons, 1924; 1934; Williamson, 1991), although
it is assumed that governance of the transaction does not operate in isolation. Indeed, the comparative
efficacy of alternative modes of governance varies with the attributes of economic players on the one
hand, and the institutional environment on the other (Williamson, 1993).
2.1 The transaction cost attributes
Agricultural cooperatives have the potential for high transaction costs (Nilsson, 1996). TCE is based
on the discriminating (i.e. transaction-cost economizing) alignment hypothesis (David & Han, 2004;
Williamson, 1979). It implies that the comparative, economic organization never examines organiza-
tional forms separately, but always in relation to alternatives (Williamson, 1991). The heuristic model
captures the impact of transaction costs on organizational arrangements. It states that, depending on
the dimensions of the transactions (asset specificity, uncertainty and frequency) and the behavioural
assumptions (bounded rationality and opportunism), economic agents will choose institutions and
S. CILIBERTI, A. FRASCARELLI AND G. MARTINO 5
Asset specificity, Uncertaint
y
Transaction costs
markets hybrids hierarchies
FIGURE 1 Organizational form responses to changes in asset specificity and uncertainty (adopted from Ménard,
2004 and Royer et al., 2015)
organizational forms that minimize the cost of exchange (Williamson, 1985). The following equation
summarizes the relations between transaction costs (TC) and these variables, with Ffor frequency, U
for uncertainty and AS for asset specificity, where the signs indicate their effects on transaction costs
(Ménard, 2007):
𝑇𝐶 =(𝐹−,𝑈+,𝐴𝑆+).
These attributes (or dimensions) of the transaction are better described below, based on TCE liter-
ature. The frequency refers strictly to buyer activity in the market. Although discrete transactions are
intriguing, few transactions feature this very isolated character. For those that do not, the difference
between a one-off and an occasional transaction is not apparent. Accordingly, a distinction is made only
between an occasional and a recurrent frequency (Williamson, 1979). However, this dimension loses
importance in the TCE narration, since it is only of significance in the presence of asset specificity,
pushing transactions away from a spot market towards a hierarchical arrangement (David & Han, 2004).
As for the second attribute, it must be noted that transactions conducted under certainty are relatively
uninteresting. Uncertainty is widely conceived as a critical attribute (Williamson, 1979), even though
such a dimension is considered as conditional (Figure 1). It implies that when asset specificity is low,
market governance should be preferred whatever the degree of uncertainty and, only in presence of
such specificity, does uncertainty increase the relative attractiveness of hierarchies and hybrids (David
& Han, 2004; Williamson, 1985). Uncertainty surely aggravates the costs of market exchange more
than those of internal organization, so it is likely to favour integration (Masten, Meehan, & Snyder,
1991). Recently, uncertainty is gaining importance as the leading force pushing towards alternative,
organizational solutions to market and/or hierarchy, the so-called hybrid forms of governance. Like-
wise, environmental uncertainty is important for agricultural cooperatives, as agricultural production is
strongly influenced by nature, affecting the quantity, quality and timing of the harvest (Bijman, 2012).
To this regard, the parties have a strong incentive to change the arrangement that minimizes their cost of
governance conditioned by the degree of uncertainty, which determines these costs. Moreover, Royer
et al. (2015) suggest that the density of shared rights depends on the intensity of uncertainty the parties
to the arrangement face. It follows that when uncertainty threatening producers increases, the parties
6S. CILIBERTI, A. FRASCARELLI AND G. MARTINO
have an incentive to share more rights, in order to reduce contractual hazards and increase coordination
and control over the strategic rights at stake. However, what is the main source of uncertainty in agri-
culture? Moschini and Henessy (2001) single out different sources: (i) production uncertainty, caused
by weather conditions or pests, (ii) price uncertainty, due not only to an inelastic demand for food, but
also to market shocks, (iii) policy uncertainty, since agriculture is typically characterized by an intricate
system of government interventions, the changes in which may also create a risk for agricultural invest-
ment and profitability. Other scholars have addressed the consequences of considerable institutional
uncertainty, closely linked to a rapid, unpredictable change in the company’s regulatory framework,
which makes agreeing on price, quality and volume more complex (Hoffman, 2007; Jolink & Nesten,
2012; Mayer & Teece, 2008; Ménard, 2017). Such an aspect has gained particular importance in the
agri-food sector over recent decades. This has been due to the drastic changes in the economic and
institutional environment, triggered by the affirmation of a market-oriented paradigm, which led to the
inclusion of the agricultural sector in the GATT negotiations, subsequently promoted and fostered by
the WTO (Royer et al., 2015). OECD (2009) also recognizes financial uncertainty related to changes in
income and/or in interest value, which somehow hampers access to credit. As a result, this dimension
is indirectly connected to the above-mentioned price uncertainty. Other types of uncertainty are those
related to counterpart behaviour and information asymmetry, as recognized by Royer (2011), or those
concerning the quality and quantity of deliverables (Royer et al., 2015): such types of uncertainty are
clearly related to production, so they can be included in that dimension.
With regard to asset specificity, this has always represented the main factor determining transac-
tion cost and, as a consequence, a huge amount of literature has analysed its main facets (Figure 1).
Basically, it arises when the specific identity of the parties has major cost-bearing consequences
(Williamson, 1979). Transactions of this kind are referred to as idiosyncratic. What happens is that
the supplier is effectively ‘locked into’ the transaction to a significant degree: such a situation is well
known in economic literature as ‘the hold-up problem’ (Ménard & Klein, 2004). As a result, when
asset specificity increases, the transaction costs associated with market governance increase accord-
ingly and more vertically integrated solutions are necessary (Pascucci et al., 2012). This study aims
to adopt Williamson’s definitions (1979; 1991) and descriptions of asset specificity. From this view-
point, such a concept refers to the degree to which an asset can be redeployed for alternative uses
and by alternative users without sacrificing productive values. Furthermore, it must be noted that asset
specificity entails several dimensions (Ménard, 2004). The specialization process entails physical asset
specificity, due to the fact that purchases of a specialized component are required for production. More
specifically, special purpose equipment or dedicated assets are often needed to produce a specific com-
ponent, implying discrete investments made at the request of a particular customer/client. Moreover,
investments related to the production of high quality end products with a high reputation generate
brand-name capital specificity. Lastly, as for human and relational capital, idiosyncratic investments
generate transaction-specific human assets and networks.
Table 1 summarizes the main attributes of transaction costs and their dimensions according to TCE.
It outlines the frame of the attributes of transactions and their dimensions that inspire the empirical
analysis carried out in this paper.
However, it must be noted that there is also a large amount of criticism of the TCE. For instance,
critics have argued that TCE not only ignored the role of differential capabilities in structuring
economic organization (Richardson, 1972) and neglected power relations, trust, and other forms of
social embeddedness (Granovetter, 1985), but also overlooked evolutionary considerations, including
Knightian uncertainty and market processes (Langlois, 1984). These criticisms have been echoed and
refined in numerous, more contemporary contributions, mainly by non-mainstream economists (Dosi
& Marengo, 2000), although new entrants are increasingly recruited from the ranks of management
S. CILIBERTI, A. FRASCARELLI AND G. MARTINO 7
TABLE 1Main attributes of transaction costs and their dimensions according to TCE
Attributes of transaction Dimensions/source
Asset specificity Human capital
Network/relational capital
Specialization asset
Dimension/size
Uncertainty Market
Production
Institutional/ Policy context
scholars (Conner & Prahalad, 1996; Ghoshal & Moran, 1996). Nevertheless, TCE gives rise to two
types of institutional responses (Tortia, Valentinov, & Iliopoulos, 2013). First, since transaction costs
can be reduced, it causes the emergence of institutions facilitating market exchange, such as markets,
hybrids and integrated organizations (Williamson, 1975). Secondly, since transaction costs act as a
constraint on the division of labour, it causes the emergence of institutions of self-sufficiency. In recent
institutional economic literature, agricultural cooperatives have traditionally been shown to economize
on transaction costs by protecting their members from being exploited by opportunistic, contractual
partners, often using highly specific assets (Hansmann, 1996; Staatz, 1987). The theory of the social
division of labour has shown, however, that transaction costs affect a market in two fundamental ways,
only one of which has been reflected in the explanations of traditional transaction costs of agricultural
cooperatives and in recent institutional economic literature more generally. First, in line with this
literature, transaction costs determine the optimal institutional choice by helping to sort out those
governance mechanisms, which economize on these costs most efficiently. Secondly, transaction costs
act as a constraint on the social division of labour and thus circumscribe the boundaries of market
economy as a whole (Valentinov, 2009) and the process of integration of independent farmers into
investor-owned firms.
Based on this elaboration on the main attributes of transaction costs, it follows that the aim of this
research is to understand how TCE premises applied to the relationship between cooperatives and farms
regarding transaction attributes can explain the decision to participate in a cooperative. For this reason,
this paper contributes to the ongoing debate on hybrid collective arrangements in the agri-food supply
chain, by investigating the determining factors of co-operation in Italian agriculture. The following
hypothesis was, therefore, tested:
H1: sources of asset specificity and uncertainty are able to affect participation in hybrid collective
arrangements.
2.2 The role of institutional environment in shaping hybrids
Bearing in mind the TCE rules of thumb, a new awareness is emerging in the field of TCE: the
institutional embeddedness of the various forms of governance (Ménard, 2014b; Tortia et al., 2013;
Williamson, 2000). Such a topic deserves special attention, where governance forms are strongly
affected by the institutional environment, as in the case of the agri-food sectors.
According to Davis and North (1971), institutions set the fundamental, political, social and legal
rules, which establish the basis for production, exchange and distribution. Deeply rooted in the
Coasian tradition, Williamson’s approach is aware that organizational arrangements are embedded
in their institutional environment. Williamson (1993) certainly recognized that, since there is a
8S. CILIBERTI, A. FRASCARELLI AND G. MARTINO
strategic feedback mechanism at stake, the importance of the institutional environment influences
the governance of contractual relations. Consequently, the set of rules, laws, policies, customs and
norms that determine the rules of the game has to be taken into consideration, since organizational
arrangements are embedded and enforced in this institutional environment (Ménard & Valceschini,
2005). To this regard, the rationale for initiating policy instruments, which aim to provide incentives to
promote cooperative behaviour, has often been based on the assumption that acting collectively should
allow farmers to cope more effectively with the various challenges, which agricultural producers face
and struggle to address individually (Ostrom, 1990).
Moreover, it must be considered that for an arrangement to be implemented and to remain sustain-
able, it needs to gain institutional legitimacy, on which the capacity to enforce the rules of the game
also depends (Royer et al., 2015). In this sense, an organization is very often the way to implement
and operationalize the ‘rules of the game’, as they are defined by the institutional environment and this
process gives birth to ‘hybrid forms’ (Ménard, 1995). Recent decades have seen an increasing interest
in the development of these non-standard modes of organization in agri-food networks, particularly in
Europe where, not by chance, agricultural production is embedded in various, changing, institutional
environments, yet producers compete in an increasingly global market (Mènard & Klein, 2004).
Hybrids are a class of arrangement, which Williamson places between market and hierarchies. Such
a mode is characterized by semi-strong incentives and an intermediate degree of administrative appa-
ratus (Williamson, 1991). Indeed, modes of hybrid collective organizations have spread everywhere
in the food industry. Despite the apparent heterogeneity of hybrids, three main characteristics allow
such governance forms to be identified: (i) parties pool part of their resources, while keeping property
rights and associated decision rights distinct, (ii) the main mechanism implemented for coordinating
is contractual and (iii) competition persists among partners, making the issue of rent-sharing partic-
ularly acute (Ménard & Valceschini, 2005). More specifically, the research agenda also focuses on
the relevant understanding of the ‘jungle’ of types of organizations (or ‘strange animals’ as stated by
Ménard, 2012) that coexist in the agricultural sector, e.g. cooperatives, networks, chain systems. Some
of those collective arrangements (e.g. POs) also share a distinctive property, since they are ‘institutional
hybrids’, combining a self-regulation mechanism operated by private partners along the supply chain
with a legal framework, which determines the terms and conditions under which these mechanisms
operate (Royer et al., 2015). The investigation is, therefore, on whether hybrid forms of governance are
mainly the result of government policies, rather than an efficient means of reducing transaction costs.
Previous evidence demonstrates that rules and regulations in the institutional environment influence
the formation of hybrids (Jolink & Nielsen, 2012). Such a process mostly concerns European efforts to
improve the functioning of the agri-food supply chain and address the increasing environmental uncer-
tainties, which make agreeing on price, quality and volume more complex. On the other hand, the
establishment of such hybrids greatly depends not only on the ability to gain institutional legitimacy,
but also on the capacity to find support (Royer et al., 2015). However, more effort is needed in order not
only to shed light on the bidirectional relationship between the institutional environment and hybrid
arrangements in the agri-food supply chain, but also to better understand, from the TCE perspective,
the importance and the role of the embedment of such forms of organizational arrangements, when
faced with a scenario of increasingly widespread environmental uncertainty (Ménard, 2004; 2014a;
Nuhanovic-Ribic et al., 2017).
Previous attempts investigated what drives participation in institutional hybrids, such as POs.
Henceforth, the main contributions are reviewed. Most of the scholars who analysed the factors
determining participation in the agricultural sector dealt with the bigger cooperatives (Karli, Bilgiç, &
Çelik, 2006) and the marketing cooperatives in developing countries (Fischer & Quaim, 2012; Wollni
& Zeller, 2007). This literature identified a set of drivers in the decision to join cooperatives or related
S. CILIBERTI, A. FRASCARELLI AND G. MARTINO 9
organizations. Important insights into factors affecting the emergence of POs are also highlighted by
Pascucci et al. (2012), who clearly showed that the patterns of horizontal integration between farmers
are likely to be very sector and region-specific. Wollni and Fischer (2015) studied the intensity of
participation in producer organizations of coffee farmers in Costa Rica. They highlighted that the share
delivered to cooperatives decreases with farm size. Fischer andQaim (2014) studied the case of Kenyan
banana producers, providing evidence that low participation can mostly be attributed to structural and
institutional conditions, such as group size and the timing of payments for collective product sales. The
same authors in another work on the same sector (2012) reported that, although farmer organizations
are generally inclusive of poor farmers, ownership of land and other agricultural assets and access
to credit significantly increased the probability of joining a group. Several studies suggested that
producers with larger farms are more likely to belong to cooperatives (Bernard & Spielman, 2009; Ito
et al., 2012; Ma & Abdulai, 2016). Hellin, Lundy, and Meijer (2009) showed the importance of market
access for the emergence of producer organizations in Mexico and several Central American countries.
They argued that the benefits of farmer organizations are closely linked to transaction costs associated
with market access: the higher these costs, the stronger the incentives for farmers to engage in
collective action and farmer organizations. Lastly, various studies have shown that there are important,
complementary roles for the government and the private sector in enabling producer organizations to
deal with the constraints they face in marketing their products (Markelova, Meinzen-Dick, Hellin, &
Dohrn, 2009). Zheng, Wang, and Awokuse (2012) suggested that educational attainment, risk comfort
level, farm expansion, operational costs, geographic location and crop types are significant factors that
influence producers’ participation behaviour. As regards the role of social and human capital, numer-
ous studies provide evidence that the age of the household head, along with educational attainment,
farming experience and access to social networks and information have a positive effect on the likeli-
hood of cooperative membership (Bernard & Spielman, 2009; Fischer & Qaim, 2012; Francesconi &
Heerinck, 2010; Markelova & Mwangi, 2010; Okello & Swinton, 2007; Zheng et al., 2012). Overall,
human and social resources can be said to be commonly found to affect the development of producer
organizations.
In this regard, the paper contributes to the existing literature, by providing quantitative evidence of
the factors affecting participation in collective forms of organizations in the Italian agri-food supply
chain. The main purpose is to explore the similarities and differences between membership in (sponta-
neously established) cooperatives and institutional hybrids (e.g. POs). Such differences/analogies may
be explained in the light of the TCE, since transaction attributes (asset specificities and uncertainty)
may once again play a key role in fostering the allocation of decisional rights by means of specific
collective arrangements promoted and enforced by the institutional environment. As a consequence, a
second research hypothesis is elaborated:
H2: sources of asset specificity and uncertainty are able to affect participation in institutional hybrids
(POs).
3METHODOLOGY
Bearing in mind the purposes of the paper and based on the availability and nature of data pro-
vided by the Italian version of the Farm Accountancy Data Network (FADN), reliable and robust
empirical models were built, in order to test the research hypotheses stemming from the conceptual
framework.
The dataset and empirical models are subsequently described in detail.
10 S. CILIBERTI, A. FRASCARELLI AND G. MARTINO
3.1 Dataset and specification of variables
The FADN is a European system of sample surveys, conducted every year to collect accountancy
data from farms. Derived from national surveys, the FADN is the only source of microeconomic data
based on harmonized bookkeeping principles. The survey does not cover all agricultural holdings, but
only those that could be considered commercial, due to their size. Moreover, it is an important source
of information to understand the impact of the measures taken under the CAP on different types of
agricultural holdings.
The Italian version of the FADN (named RICA, acronym of ‘Rete di informazione contabile
agraria’) is a dataset that provides several pieces of information on economic and organizational
aspects and aims to monitor the business activities of EU agricultural holdings. Therefore, a list of
variables related to the transaction cost attributes was identified in the RICA dataset for 2015 (n=8536
after processing data for outliers), based on the TCE conceptual framework. More specifically, due to
the fact that FADN represents an annual, repeated, cross-sectional survey that does not guarantee a fully
balanced panel dataset, a cross-sectional analysis was preferred instead of a longitudinal approach, in
order not only to ensure and maintain the representativeness of the sample, but also to avoid misinter-
pretations of the results.
Above all, it must be noted that the RICA uses categorical variables to establish when farms are
members of cooperatives (COOP) and when they are members of producers’ organizations (PO). What
emerges is that 17.7 per cent and 11.4 per cent respectively of farms in the sample were members of a
cooperative or of a PO in 2015. Furthermore, a small percentage of farms in the sample (2.16%) opted
for joint memberships of a COOP and a PO. Moreover, the dataset contains several variables that can
be adopted as proxies for the main attributes of the transaction cost attributes and their dimensions
(reported in Table 1). Two main groups of co-variates are identified that refer to asset specificity and
uncertainty.
As regards asset specificity, five sub-groups of variables were created. They concerned:
(i) according to Williamson’s (1993) definition of human asset specificity, which refers to some-
thing that ‘arises in learning by doing’, and based on the availability of variables in the FADN
dataset, we selected some proxies referring to the farmer’s age (Manager_age), self-employment
(Manager_empl) and to the presence of parents as predecessors (Parents), as a whole representing
strong sources of transaction costs for farms;
(ii) networks (relational capital), with reference to membership of associations (OPPAA) or other
networks (oth_NETW),
(iii) specialization assets, which regroup the variables that identify the dominant farm activity
(TF_COP, TF_orch&wine, TF_hort, TF_mixed, TF_cattle&milk, TF_graniv), variables mainly
referring to specific high-value production (such as Organic,PDO and PGI) and covariates that
reveal the presence of diversification activities (GO_processact_share) or high quality production
(GO_qualact_share),
(iv) dimensional specificity, by reference to a control group of variables referring directly or indirectly
to the farm size, e.g. farm size (UAA), rented (Rented_UAA) and irrigated land (Irrigated_UAA),
intensity of family labour (FWU_UAA), mechanical force (Machpow_UAA) and capital intensity
(Fixasset_UAA).
With regard to uncertainty, two sub-groups were established. They regarded:
5(a) production and market uncertainty, with specific attention to gross output and farm income
volatility (GO_sd and FI_sd, expressed as standard deviation) and production uncertainty,
S. CILIBERTI, A. FRASCARELLI AND G. MARTINO 11
expressed as the intensity of perceived risk (Insurance_UAA, i.e. the amount of the insurance
premium per hectare),
5(b) institutional/policy uncertainty, represented by a set of variables referring to the concentration
of cooperatives (Index_agr_coop), the number of regional processing companies (Proc_comp),
the rate of municipal agricultural employment (Employment) and the variation of the CAP aids
(CAPaids_sd, expressed as a standard variation of public support).
Table 2 shows the entire list of variables used in the quantitative analysis and their descriptive
statistics.
3.2 Empirical approach
As both dependent variables are dichotomous in nature, a probit and logit model seemed appropriate.
However, since the error terms of the two models are likely to be correlated, an extension of the probit
model, known as the bivariate probit model, is usually more appropriate (Greene, 2012; Pascucci et al.,
2012). More specifically, taking into account that participation in a cooperative (and/or an institutional
hybrid) depends on several dimensions of the transaction cost attributes, the bivariate probit model has
the following specification:
𝑍𝑖1=𝛽′
1𝑋𝑖1+𝜀𝑖1;𝑦𝑖1=1if𝑧𝑖1>0,𝑦
𝑖1=0ifz
𝑖1≤0,
𝑍𝑖2=𝛽′
2𝑋𝑖2+𝜀𝑖2;𝑦𝑖2=1if𝑧𝑖2>0,𝑦
𝑖2=0if𝑧𝑖2≤0,
where 𝛽′1is the coefficient of the covariates (Xi1) related to asset specificity (AS) and uncertainty
(U), when COOP is the dependent variable and 𝛽′2is the coefficient of the covariates (Xi2) to asset
specificity (AS) and uncertainty (U), when PO is the dependent variable.
In order to test both first and second hypotheses, this model is able to provide (consistent) estimates
of the coefficient vectors 𝛽1and 𝛽2for the two equations, of the correlation between the error terms 𝜀ij
of the equations (𝜌), and of the standard errors for these parameters. Moreover, this approach makes it
possible to test whether the correlation between the equations is statistically significant. Thus, we can
determine whether the bivariate probit model provides the best fit. If this correlation is not significantly
different from zero, however, the separate (univariate) probit estimation of the equations is preferable,
as the bivariate specification is less efficient (Greene, 2012).
By estimating probabilities, the probit model is actually used to refer specifically to the problem in
which the dependent variable is binary – i.e. the number of available categories is two. Probit regression
measures the relationship between the categorical, dependent variable and one or more independent
variables, which need not necessarily be continuous. More specifically, the probit models adopted have
the following specification:
𝑍𝑖1=Φ
(𝛽0+𝛽′
𝑙𝑋𝑖+𝛽′
2𝑋𝑖+𝛽′
𝑛𝑋𝑖)
Both bivariate probit and probit models are estimated with Stata 12, using the maximum likelihood
procedure with the robust estimator for variance.
4FINDINGS AND DISCUSSION
First of all, since the correlation coefficient (𝜌) of the bivariate model does not differ significantly
from zero the error structures of the two equations are not correlated and, therefore, two separate
12 S. CILIBERTI, A. FRASCARELLI AND G. MARTINO
TABLE 2Conceptual approach and descriptive statistics of variables
Class of variable Variable Label Obs. Mean SD
Dependent
variables
1 if a farm is member of a cooperative COOP 8208 0.177 0.382
1 a farm is member of a Producer
Organizations
PO 8208 0.114 0.318
Human capital Farmer’s age Manager_age 8535 58.030 13.559
1 if the manager is also employee in the
farm
Manager_ empl 8509 0.377 0.485
1 if a predecessor is present Parents 8535 0.094 0.292
Network 1 if it is a member of a farmer association OPPAA 8208 0.746 0.435
1 if it is a member of other networks,
associations, etc.
oth_NETW 8208 0.206 0.404
Specialization
asset
1 if organic production is present Organic 8536 0.054 0.225
1 if PDO production is present PDO 8536 0.053 0.225
1 if PGI production is present PGI 8536 0.030 0.169
Share of gross output from processing
activities
GO_processact_
share
8531 0.108 0.230
Share of gross output from high quality
products
GO_qualact_
share
8531 0.021 0.128
1 if a farm is specialized in arable crops
(benchmark)
TF_COP 8536 0.265 0.441
1 if a farm is specialized in permanent
crops
TF_orch&wine 8536 0.316 0.465
1 if a farm is specialized in horticulture TF_hort 8536 0.022 0.147
1 if a farm is specialized in mixed
production
TF_mixed 8536 0.129 0.335
1 if a farm is specialized in animal
breeding
TF_cattle&milk 8536 0.223 0.416
1 if a farm is specialized in breeding of
granivores
TF_graniv 8536 0.045 0.207
Dimensional
specificity
Farm size (expressed in hectares) UAA 8536 34.860 59.322
Rented land (expressed as % of the total
UAA)
Rented_UAA 8536 0.374 0.412
Irrigated land (expressed as % of the total
UAA)
Irrigated_UAA 8536 0.294 0.398
Intensity of mechanical force (expressed
as machine power – in kW per hectare)
Machpow_UAA 8536 12.704 22.059
Intensity of family labour (expressed as
Family Work Unit per hectare UAA)
FWU_UAA 9120 0.129 0.249
Intensity of capital (expressed as fixed
asset per UAA)
Fixasset_UAA 9120 26242.600 49520.160
(Continues)
S. CILIBERTI, A. FRASCARELLI AND G. MARTINO 13
TABLE 2Continued
Class of variable Variable Label Obs. Mean SD
Production and
market
uncertainty
Turnover volatility (expressed as standard
deviation of the gross output)
GO_sd 7080 22518.360 76144.560
Income volatility (expressed as standard
deviation of the farm income)
FI_sd 7080 22267.280 79990.750
Intensity of perceived risk (amount of the
insurance premium per hectare)
Insurance_UAA 8536 103.790 374.966
Institutional /
policy
uncertainty
CAP aids variation (expressed as standard
deviation of the CAP support between
2013 and 2015)
CAP aids_ sd 7080 2251.999 8188.251
Index of concentration of regional
cooperatives
Index_agr_coop 8536 0.029 0.016
Number of regional processing companies Proc_comp 8536 3290.167 2055.733
Index of municipal agricultural
employment
Employment 8532 0.117 0.087
probit models enable a more appropriate solution to be obtained. As a consequence, these models
yield more efficient parameters than a single, bivariate probit estimate. Furthermore, before revealing
and discussing the results of the probit models, some relevant specification issues are discussed. Above
all, due to the fact that the models contain over 30 covariates, a test for multicollinearity is required. A
calculation of the variance inflation factors (VIFs) reveals not only an average value of 1.45, but also
the absence of variables with a VIF of over 2.6. As a result, the models do not have multicollinearity
problems. Moreover, a Smith–Blundell test (1986) implemented by Baum (2003), is used to identify
potential instrumental variables able to solve potential issues for some instrumented variables whose
exogeneity is under suspicion (mainly those referred to the dimensional specificity) due to reversal
causality. Then, a Wald test of exogeneity (Wooldridge, 2002) is implemented using efficient two-step
estimator according to Newey (1987). Tests on both probit models are not significant implying that the
null hypothesis of no endogeneity cannot be rejected and probit equations perform well with no need
for instrumental in absence of endogeneity.
As a consequence, the empirical results of the two probit models reported in Table 3 are discussed
below, according to the conceptual framework presented in Section 2.
As regards asset specificity, the results show that the vast majority of the dimensions analysed sig-
nificantly influence participation in cooperatives and/or POs in the Italian agricultural sector.
With regard to human capital and relational network, despite the absence of any ‘age effect’ on the
adoption of a collective organizational solution, empirical evidence points out that the willingness to
join a hybrid form of arrangement in agriculture is positively affected by the presence of a predecessor
in the farm (+0.065) for cooperatives and by the presence of a manager, who is also employed in
farm activities for POs (+0.025). Such results, however, confirm the importance of the ‘human’
or ‘people’ factors, since members’ characteristics have an important influence on their behaviour
towards a cooperative (Bhuyan, 2007). Very interestingly, cooperation or participation in POs is
negatively affected by membership of farmers’ associations (coefficients for the two collective
arrangements are −0.103 and −0.244, respectively) or other types of networks (−0.104 and −0.121).
Counter-intuitively, empirical evidence does not confirm what Pascucci et al. (2012) showed for Italy.
A possible explanation could be that since participation in several organizations is time consuming,
cooperatives and POs compete with other association forms, networks etc. Moreover, it must be
14 S. CILIBERTI, A. FRASCARELLI AND G. MARTINO
TABLE 3Results of probit models: marginal effects
COOP PO
Coeff. (SE) P>zCoeff. (SE) P>z
Human capital Manager_age 2.96 *10−4(3.49 *10−4)4.82
*10−4(2.99 *10−4)
Manager_ empl −0.004 (0.009) 0.025 (0.008) **
Parents 0.065 (0.017) *** 0.012 (0.013)
Network OPPAA −0.103 (0.013) *** −0.244 (0.015) ***
oth_NETW −0.104 (0.009) *** −0.121 (0.006) ***
Specialization asset Organic 0.003 (0.020) −0.008 (0.016)
PDO 0.090 (0.022) *** 0.016 (0.018)
PGI 0.099 (0.028) *** 0.018 (0.023)
GO_processact_share −0.187 (0.023) *** 0.037 (0.017) **
GO_qualact_share 0.195 (0.033) *** 0.048 (0.027) *
TF_cattle&milk 0.083 (0.015) *** 0.052 (0.013) ***
TF_graniv −0.038 (0.022) *0.043 (0.023) *
TF_hort 0.182 (0.042) *** −0.025 (0.025)
TF_orch&wine 0.142 (0.016) *** 0.044 (0.013) **
TF_mixed 0.072 (0.017) *** 0.052 (0.016) **
Dimensional
specificity
UAA −1.71 *10−4(1.03 *10 −4)*1.20 *10−4(7.30 *10 −5)
Rented_UAA −0.012 (0.012) 0.007 (0.010)
Irrigated_UAA 0.039 (0.012) ** 0.023 (0.010) **
Machpow_UAA −4.23 *10−4(3.11 *10−4)6.90 *10−4(2.42 *10−4)**
FWU_UAA −0.054 (0.032) *−0.082 (0.028) **
Fixasset_UAA 9.51 *10−7(1.29 *10−7)*** −1.24 *10−7(1.07 *10−7)
Production and
market
uncertainty
GO_sd −2.91 *10−7(1.14 *10−7)1.51
*10−7(8.31 *10−8)*
FI_sd −1.10 *10−7(1.09 *10−7)−2.80 *10−8(6.01 *10−8)
Insurance_UAA 8.40 *10−5(1.48 *10−5)*** 2.00 *10−5(1.29 *10−5)
Institutional / policy
uncertainty
CAP aids_ sd 8.35 *10−7(6.60 *10−7)1.22 *10−6(4.98 *10−7)**
Index_agr_coop 2.583 (0.290) *** −0.878 (0.271) **
Proc_comp −2.94 *10−5(2.33 *10−6)*** −1.08 *10−5(2.05 *10−6)***
Employment 0.211 (0.052) *** −0.220 (0.048) ***
N6779 6779
LR 𝜒2(28) 1415.87 *** 623.31 ***
Log likelihood −2667.872 −2100.056
Pseudo R20.209 0.129
*<0.100, **<0.050, *** <0.001.
Source: Author’s calculation based on RICA 2015.
noted that sometimes the functions of farmers’ associations and cooperatives/POs could overlap.
Therefore, since these organizations are somehow substitute, farmers opt for just one membership,
in order to avoid duplication of procedures, possible procedural conflict and, above all, an increase
of transaction costs. Another possible interpretation could be due to the influence played by specific,
embedded meso-institutions, e.g. farmers’ organizations (Monderlaers, Baecke, & Lauwers, 2014). As
these interest groups were responsible for interpreting and implementing public, agricultural policies
S. CILIBERTI, A. FRASCARELLI AND G. MARTINO 15
according to Ménard (2014a; 2017), these circumstances historically enabled them to interfere in the
cooperation process in the Italian agri-food system. Sometimes a top-down approach was used and
POs or cooperatives were established merely to take advantage of the available financial support.
Results confirm that specialization triggers cooperation for all the types of farming under investiga-
tion. Assuming arable crops as the benchmark, the models reveal that farms specializing in horticulture
(+0.182), orchards and vineyards (+0.142), cattle and milk (+0.083) and mixed production (+0.072)
are more likely to join cooperatives; the opposite applies to granivores (−0.038). The findings reveal
similar, but weaker effects for POs. More specifically, farms specialized in cattle and milk and mixed
production (both +0.052), orchards and wine (+0.044) and granivores breeding (+0.043) are more will-
ing to participate in POs than farms cultivating arable crops. The findings are consistent with existing
studies showing that the development of POs is a sector-specific phenomenon, with cooperation being
present more often in sectors characterized by higher heterogeneity in terms of quality (e.g. wine, fruit
and vegetables, milk) and of the number of products offered (such as mixed production) (Guzmán,
Arcas, Ghelfi, & Rivaroli, 2009; Pascucci et al., 2012).
The findings highlight that brand-name capital also plays a key role in triggering cooperation. This
is clearly indicated by the fact that high quality and high value production with a great reputation
and high brand values, such as PGI (+0.099) and PDO (+0.099), positively affect cooperation in
Italian agriculture. This evidence is confirmed by the fact that farms producing high-quality food
are highly motivated to participate in cooperatives (+0.195) and, to a lesser extent, in POs (+0.048).
According to Raynaud, Sauvee, and Valceschini (2005), this finding shows that quality labelling and
governance of the vertical chains are related. To this regard, Ménard and Valceschini (2005) recognize
that the promotion of food quality, together with the enforcement of quality certification, requires
highly specific investments and tight coordination among transacting parties. Therefore, hybrid forms
of governance can enable farmers to coordinate on quality control and comply with stringent food
standards, as recognized by Narrod et al. (2009). No significant effects were observed for POs, a reason
for which could be that this type of institutional, collective arrangements does not aim specifically at
enhancing and promoting specific (high quality) productions. Lastly, apart from a slight, but positive
effect on participation in POs (+0.037), farms carrying out processing activities are not likely to join
cooperatives (−0.187). A typical explanation of such a result is the ‘make or buy’ decision (Klein,
2005; Walker & Weber, 1984). Indeed, when a farm decides to make a product (i.e. to process a food),
it clearly tends to compete with processing cooperatives. Therefore, when farms cannot gain significant
advantages from collaboration, the reasons for participating in collective arrangements disappear.
As regards specificity affected by farm size, some control variables were tested. The findings high-
lighted that size very slightly affects participation in cooperatives (−1.71*10−4). More relevant is the
impact due to the presence of family workers (expressed by means of FWU/UAA), that may represent
an obstacle for cooperation (−0.054) and, to a greater extent, POs (−0.082). A reasonable interpre-
tation is that there can be some conflict between family work and participation in cooperatives and
POs. Based on different cultural conceptions, farms in which family members keep decision-making
may be more reluctant to join cooperatives than farmers more open to membership based organization
such as cooperatives and POs (Birchall, 2010). Findings show that in Italy the former case prevails
on the latter one. In this regard, Ménard and Klein (2004) also noted that while a farm remains a
family-owned business, it is difficult to design and enforce effective, collective arrangements. How-
ever, this evidence does not necessarily contrast with the fact that there is large consensus in literature
(Valentinov, 2007) on the fact that family farming has been an instrumental ingredient of successful
agricultural cooperation in most, if not all, agriculturally advanced countries, in order to take advan-
tage of the transaction-cost economizing properties of family farms. At most, this finding only adds
information related to the attitude of family farms in Italy to cooperation.
16 S. CILIBERTI, A. FRASCARELLI AND G. MARTINO
Turning to indicators of specific investment, the results revealed that the intensity of assets related to
farm size (slightly) stimulates cooperative membership (+9.51*10−7) as well as specific investments for
irrigation positively influence participation in COOPs (+0.039) and POs (+0.023). Likewise, dedicated
investments for mechanization slightly stimulate the participation in POs as well (+6.90*10−4).
The second transaction attribute analysed in the model was uncertainty. Empirical evidence showed
that such a dimension influences the decision to participate in both cooperatives and POs. More specif-
ically, the findings concerning production uncertainty revealed that the use of agricultural insurance
slightly (+8.40 +10−5) triggers producers’ participation in cooperatives. This result has interesting
implications, since it highlights that farms which perceive risks and use risk management tools are
more likely to recognize the important role of cooperatives as an institutional means to collectively
address and manage uncertainty. As regards the uncertainty that stems from the market and directly
affects the level of revenues, the findings revealed a very weak, positive effect on stimulating partic-
ipation in POs (+1.51*10−7). This institutional hybrid arrangement could indeed be approached as a
means to reduce the impact of price volatility on revenues.
Furthermore, some very interesting findings related to the institutional environment emerged.
Indeed, findings highlighted that the greater the relevance of cooperation, the more farmers are stim-
ulated to join them (+2.583). Conversely, there is a sort of competition with POs at stake, since a
widespread diffusion of cooperation decreases participation in institutional hybrids (−0.878). By tak-
ing into account the concept of mesoinstitutions according to Ménard (2014b; 2017), a possible inter-
pretation could be that these devices play a key role, since they are in charge of actually implementing
the general rules of the game by converting them into sector-specific rules and/or geographic areas,
thus framing and delineating the domain of the players’ activities. More specifically, when mesoinstitu-
tions are able to implement and enforce rules and property and decision rights, and guarantee a reliable
institutional environment, certain types of collective arrangements (e.g. cooperatives) can spread and,
as a consequence, companies are encouraged to participate. This phenomenon causes a sort of compe-
tition with other types of arrangements that, according to Pascucci et al. (2012), grows as the number
of collective arrangements (e.g. POs) increases.
Furthermore, the results revealed that the higher the number of processing companies in the agri-
food system, the more competition there will be among processors and, as a consequence, the lesser the
need for farmers to join a cooperative (−2.94*10−5)oraPO(−1.08*10−5). This evidence points out
the general abilities of cooperatives to economize on transaction costs and to develop ‘countervailing
power’ in the presence of downstream monopsony in output markets (Tortia et al., 2013). Bijman et al.
(2012) also found evidence of the fact that cooperatives provide their members with market access and
bargaining power, which is especially important in markets characterized by monopsonistic structures
where, by bulking the produce of many (small) producers into one offer and negotiating on price and
delivery conditions, cooperatives are able to establish countervailing power. Indeed, it is not uncom-
mon to find cases where the local monopsony processors went out of business and farmers’ associ-
ation members, or some of them, decided to buy the processing plant and operate it as a processing
cooperative.
The employment rate, which identifies more agricultural-oriented areas, also plays a role in fostering
participation in cooperatives (+0.211) and, at the same time, has a negative effect on participation in
POs (−0.220). According to Pascucci (2012), a possible interpretation is based on the socio-economic
differences at stake between geographical areas in Italy. More specifically, where the agri-food sector
plays a key role in terms of employment due to the smaller size of farms – mainly in southern Italy –
farms rely more on cooperatives, rather than on POs, which are actually far more widespread in central
and northern Italy.
S. CILIBERTI, A. FRASCARELLI AND G. MARTINO 17
Lastly, a slight, but very interesting effect is due to the influence of policy uncertainties (related to
the evolution of CAP aids over time) on participation in “institutional” collective arrangements. The
findings show that PO membership was somehow slightly influenced (+1.22*10−6) by this type of
uncertainty, due to the considerable variation in public support in the period 2013–15 as a result of
several reforms of the CAP, which affected many strategic, agricultural sectors in Italy. Many authors
have indeed recognized the increasing role of uncertainty in shaping institutional arrangements
and modifying the collaborative attitudes of farmers, faced with the increasing liberalization of the
agricultural market, a decreasing dependency on public aid and the increasing regulatory support to
collective arrangements (Ciliberti & Frascarelli, 2018; Ménard & Valceschini, 2005; Royer et al.,
2015).
To sum up, these findings enabled us to confirm both H1 (the sources of asset specificity and
uncertainty are able to affect participation in hybrid collective arrangements) and H2 (the sources of
asset specificity and uncertainty are able to affect participation in institutional hybrids (POs)). Indeed,
according to TCE literature, the findings reveal that, despite some differences, both asset specificity
and uncertainty play a major role in stimulating/hindering participation in cooperatives or institutional
hybrids, such as POs, in the Italian agri-food sector. The role of asset specificity is relevant for both
types of collective arrangements under investigation, since the higher the level of specialization and
specific fixed investments made by farmers, the more participation rises. Evidence tends to suggest that
human capital also matters, even though what emerges is that the family farming model could somehow
not be able to perceive the importance of collective actions as a means to enhance market access for
small scale farms. Moreover, very interestingly, a contrast between participation in cooperatives and
in farmers’ organizations or other networks is at stake. Apart from the reasonable competition among
these organizations, this evidence could confirm the scarce reputation of farmers’ organizations in
leading and fostering the process of participation in collective arrangements in Italy. Indeed, the main
players in the process of PO formation are often not farmers, but rather farmers’ organizations, for the
sole purpose of benefitting from public incentives offered by the CAP. This, in turn, poses the question
as to what extent these organizations can be effective in promoting farmers’ interests, since POs set up
using a top-down approach lack economic justification, which is an essential condition for the success
of these collective forms of arrangements.
The other transaction attribute according to the TCE, i.e. uncertainty, also plays a relevant role in
fostering access to cooperatives and POs. Farmers respond to the risks linked to prices and weather
instability that negatively affect farmers ‘incomes by means of insurances or diversification strate-
gies. In addition, also the participation in collective arrangements is seen with increasing interest as a
solution to share these risks as well as to jointly address the institutional uncertainty surrounding trans-
actions. In this regard, it must be noted that these latter phenomena have been considerably affected by
several CAP reforms, which have reduced the dependence of farm incomes on public aid. As a con-
sequence, what emerges is also the significant influence of the mesoinstitutional context on fostering
or hampering participation in cooperatives or POs, since it is largely responsible for governing the
increasing uncertainty in the agri-food supply chain.
5CONCLUSIONS
The ongoing liberalization of the European agri-food market adds complexity and uncertainty to trans-
actions between farmers and their counterparts, due not only to the intrinsic characteristics of products
(e.g. credence attributes, quality related to location and methods), but also to external factors (e.g.
role of institutions, price volatility). Although Olson (1965) reported that due to their concentrated
18 S. CILIBERTI, A. FRASCARELLI AND G. MARTINO
nature, minority interests will be over-represented and widespread majority interests trumped, what
has emerged in the last decade is a growing interest in collective forms of organization in the agri-
food sector. These organizational solutions have been gaining momentum since they are able to play
a role of socio-economic protection to vulnerable farms thanks to their attitude to be resilient in time
of global turbulence. Their renewed importance stems from the need of rearranging the coordination
of decisions among economic actors consequent to the process of progressive dismantlement of the
protectionist architecture of the CAP started in the early 1990s.
This paper has focused its attention on the determining factors of hybrid forms of governance in the
Italian agri-food sector, such as cooperatives and POs, from the TCE perspective. It offers an innova-
tive, original contribution to scholars and policymakers, corroborated by a quantitative analysis, based
on an extensive dataset, of the factors that foster/hinder the establishment and participation in collective
arrangements in Italy. Moreover, in addition to asset specificity, the empirical model carefully takes
into account the uncertainty surrounding transactions, paying specific attention to the influence of the
institutional environment.
The main contribution of the work is represented by the fact that it has shed light on the circum-
stances in which successful POs and cooperatives can be established and what policy signals may be
used to support this process. Under the TCE lens, the paper has shown that farmers’ relations with
agricultural cooperatives depend to a large extent not only on specialization, human capital and the
network, but also on the institutional context, in which farmers operate and face uncertainty. It has,
therefore, confirmed that whether farmers are better off being members of a cooperative or acting on
their own depends on the particular characteristics of the product and market structure. In general,
collective arrangements appear to be more beneficial not only in high-value supply chains (e.g. wine
and cheese), but also in the presence of cash crops, such as fruit and milk. What emerges is a rele-
vant role of cooperatives where there is high quality, local production (PDO and PGI), as they may
help farmers comply with the stringent food standards. At the same time, participation in hybrid forms
of governance in the agri-food system is somehow highly site-specific, since it is more widespread
in areas where agriculture plays a key role in terms of employment. However, participation in insti-
tutional hybrids (e.g. POs) is common in more developed areas, where the primary sector absorbs a
smaller share of workers. Lastly and very interestingly, collective arrangements in the agri-food sectors
are increasingly approached by farmers as a solution to jointly address the increasing uncertainty sur-
rounding transactions. In more details, as a result of a free-rider problem, collective actions are often
approached as a means to share uncertainty and the costs of a hold up, stemming from opportunistic
behaviours of their counterparts, are split over the entire group. In this regard, they are perceived as
resilient organizational solutions able to keep their hallmark features (aimed to foster social trust and
sustain equitable rural growth) as well as to ensure flexibility vis-à-vis an unusual, difficult and hostile
business context. Such an innovative function opens room for a new role of cooperatives and POs to
the benefit of the adaptability of their members facing disturbances in the institutional environment
within which the agricultural sector functions. Moreover, where farms ‘co-operate’ (Dagnino, 2007)
by pooling decisional rights on output, these hybrid solutions provide a stronger negotiating position
towards potential contractors.
Nevertheless, this work shows some limitations, which must be carefully taken into account. How-
ever, it may represent the starting point for further investigation on this topic. First, due to the limited
availability of the variables of the FADN, the paper does not take into account whether the sample
cooperatives are marketing, processing, supply, or multi-purpose cooperatives, or whether the sample
cooperatives are centralised or federated, which would arguably modify the incentives for individual
farmers to join a PO. As a consequence, the paper may not provide any implications of the nature of the
cooperatives. Secondly, time dimensions are excluded, and this constitutes another barrier to a correct
S. CILIBERTI, A. FRASCARELLI AND G. MARTINO 19
assessment of whether the motivation for joining cooperatives/POs changes over time. Thirdly, due to
the nature of the FADN dataset, the paper could not accurately take into account the social, institu-
tional and cultural contexts, which together contribute to shaping collective arrangements and the way
they function. Lastly, attention is exclusively focused on Italy, so any generalisation for other European
countries is not possible.
Future research may aim to include the above-mentioned relevant factors in order to pave the way
for future work in this field, aiming to fill the ‘knowledge vacuum’ of managers, stakeholders and
policymakers, who often lack a common understanding of organizational problems along the agri-food
supply chain.
ACKNOWLEDGEMENTS
The authors acknowledge the Council for Agricultural Research and Analysis of Agricultural Eco-
nomics (CREA) for providing access to the Italian FADN.
ORCID
Stefano Ciliberti https://orcid.org/0000-0001-7833-9547
REFERENCES
Ariyaratne, C. B., Featherstone, A. M., Langemeier, M. R., & Barton, D. G. (2000). Measuring X-efficiency and scale
efficiency for a sample of agricultural cooperatives. Agricultural and Resource Economics Review,29, 198–207.
Baum, C. F., Schaffer, M. E., & Stillman, S. (2003). Instrumental variables and GMM: Estimation and testing. The Stata
Journal,3, 1–31.
Bernard, T., & Spielman, D. J. (2009). Reaching the rural poor through rural producer organizations? A study of agri-
cultural marketing cooperatives in Ethiopia. Food Policy,34, 60–69.
Bhuyan, S. (2007). The people factor in cooperatives: An analysis of members’ attitudes and behaviour. Canadian Jour-
nal of Agricultural Economics,55, 275–298.
Bijman, J. (2016). The changing nature of farmer collective action: Introduction to the book. In J. Bijman, J. Schuurman,
& R. Muradian (Eds.), Cooperatives, Economic Democratization and Rural Development. Cheltenham, UK: Edward
Elgar.
Bijman, J., & Iliopoulos, C. (2014). Farmers’ cooperatives in the EU: Policies, strategies and organization. Annals of
Public and Cooperative Economics,85, 497–508.
Bijman, J., Iliopoulos, C., Poppe, K. J., Gijselinckx, C., Hagedorn, K., Hanisch, M., …van der Sangen, G.
(2012). Support for Farmers’ Cooperatives – Final Report. Retrieved from https://ec.europa.eu/agriculture/sites/
agriculture/files/external-studies/2012/support-farmers-coop/fulltext_en.pdf, last accessed 30.04.2019.
Birchall J. (ed.), (2010), People-Centred Businesses: Co-operatives, Mutuals and the Idea of Membership, London:
Palgrave.
Bouamra-Mechemache, Z., & Zago, A. (2015). Introduction: Collective action in agriculture. European Review of Agri-
cultural Economics,42, 707–711.
Budiguel, L. et al. (2016). Study on Agricultural Interbranch Organisations in the EU, Brussels.:
https://doi.org/10.2762/901778.
Ciliberti, S., & Frascarelli, A. Collective arrangements in the agro-food supply chain: The case of the interbranch organi-
zation Tabacco Italia, In G. Martino et al. (Ed.), (2018). It’s a Jungle Out There – The Strange Animals of Economic
Organization in Agri-Food Value Chains, Wageningen: Wageningen Academic Publishers.
Coase, R. H. (1937). The Nature of the Firm. Economica,4, 386–405.
Coase, R. H. (1998). New Institutional Economics. American Economic Review,88, 72–74.
Commons, J. R. (1924). Law and economics. Yale Law Journal,34, 371–382.
Commons J. R. (ed.), (1934). Institutional Economics, Madison, WI: University of Wisconsin Press.
20 S. CILIBERTI, A. FRASCARELLI AND G. MARTINO
Conner, K. R., & Prahalad, C. K. (1996). A resource-based theory of the firm: Knowledge vs. Opportunism. Organization
Science,7, 477–501.
Cook, M. L. (1995). The Future of US Agricultural Cooperatives: A Neo-Institutional Approach. American Journal of
Agricultural Economics,77, 1153–1159.
Dagnino, G. (2007). Preface: Coopetition strategy – toward a new kind of inter-firm dynamics? International Studies of
Management & Organization,37(2), 3–10.
Daidj N. (ed.), (2017), Cooperation, Coopetition and Innovation (Vol. 3). Wiley Online Library. (also ISTE).
David, R. J., & Han, S.-K. (2004). A systematic assessment of the empirical support for transaction cost economics.
Strategic Management Journal,25, 39–58.
Davis L. E., & North D. C. (eds.), (1971). Institutional Change and American Economic Growth, Cambridge: Cambridge
University Press.
Dosi, G., & Marengo, L. (2000). On the tangled discourse between transaction cost economics and competence-based
views of the firm: Some comments, In N. J. Foss & V. Mahnke (Eds.), Competence, Governance, and Entrepreneur-
ship, Oxford: Oxford University Press.
Falkowski J., & Ciaian P. (eds.), (2016), ‘Factors supporting the development of producer organizations and their impacts
in the light of ongoing changes in food supply chains’, EU JRC Technical Report No. 27929.
Fischer, E., & Qaim, M. (2012). Linking smallholders to markets: Determinants and impacts of farmer collective action
in Kenya. World Development,40, 1255–1268.
Fischer, E., & Qaim, M. (2014). Smallholder farmers and collective action: What determines the intensity to participate.
Journal of Agricultural Economics,65, 683–702.
Francesconi, G. N., & Heerink, N. (2010). Ethiopian agricultural cooperatives in an era of global commodity exchange:
Does organizational form matter? Journal of African Economies,20, 153–177.
Ghoshal, S., & Moran, P. (1996). Bad for practice: A critique of the transaction cost theory. Academy of Management
Review,21, 13–47.
Gow, H. R., Streeter, D. H., & Swinnen, J. F. M. (2000). How private contract enforcement mechanisms can succeed
where public institutions fail: The case of Juhocukor A.S.. Agricultural Economics,23, 253–265.
Granovetter, M. S. (1985). Economic action and social structure: The problem of embeddedness. American Journal of
Sociology,91, 481–510.
Greene W. H. (ed.), (2012), Econometric Analysis (7th edn.). Upper Saddle River, NJ: Prentice-Hall.
Grossman, S. J., & Hart, O. D. (1986). The costs and benefits of ownership: A theory of vertical and lateral integration.
Journal of Political Economy,94, 691–719.
Guzmán, I., Arcas, N., Ghelfi, R., & Rivaroli, S. (2009). Technical efficiency in the fresh fruit and vegetable sector: A
comparison study of Italian and Spanish firms. Fruit s,64, 243–252.
Hanisch, M., Rommel, J., & Müller, M. (2013). The cooperative yardstick revisited: Panel evidence from the European
dairy sectors. Journal of Agricultural & Food Industrial Organization,11, 151–162.
Hansmann, H. (1996). The Ownership of Enterprise, Cambridge, MA: The Belknap Press of Harvard University Press.
Harris, A., Stefanson, B., & Fulton, M. (1996). New generation cooperatives and cooperative theory. Journal of Coop-
eratives,11, 15–28.
Hellin, J., Lundy, M., & Meijer, M. (2009). Farmer organization, collective action and market access in Meso-America.
Food Policy,34, 16–22.
Hoffman, W. (2007). Strategies for managing a portfolio of alliances. Strategic Management Journal,28, 827–856.
Hubbard, M. (1997). The “new institutional economics” in agricultural development: Insights and challenges. Journal
of Agricultural Economics,48, 239–249.
Ito, J., Bao, Z., & Su, Q. (2012). Distributional effects of agricultural cooperatives in China: Exclusion of smallholders
and potential gains on participation. Food Policy,37, 700–709.
Jolink, A., & Nesten, E. (2012). Recent qualitative advances on hybrid organizations: Taking stock, looking ahead.
Scandinavian Journal of Management,28, 149–161.
Karli, B., Bilgiç, A., & Çelik, Y. (2006). Factors affecting farmers’ decision to enter agricultural cooperatives using
random utility model in the South Eastern Anatolian region of Turkey. Journal of Agriculture and Rural Development
in the Tropics and Subtropics (JARTS),107, 115–127.
Klein, P. G. (2005). The make-or-buy decision: Lessons from empirical studies, In C. Menard & M. M. Shirley (Eds.),
Handbook of New Institutional Economics, Boston, MA: Springer.
S. CILIBERTI, A. FRASCARELLI AND G. MARTINO 21
Langlois, R. N. (1984). Internal organization in a dynamic context: Some theoretical considerations, In M. Jussawalla &
H. Ebenfield (Eds.), Communication and Information Economics: New Perspectives, Amsterdam, North-Holland.
Ma, W., & Abdulai, A. (2016). Does cooperative membership improve household welfare? Evidence from apple farmers
in China. Food Policy,58, 94–102.
Markelova, H., Meinzen-Dick, R., Hellin, J., & Dohrn, S. (2009). Collective action for smallholder market access. Food
Policy,34, 1–7.
Markelova, H., & Mwangi, E. (2010). Collective action for smallholder market access: Evidence and implications for
Africa. Review of Policy Research,27, 621–640.
Masten, S. E., Meehan, J. W., & Snyder, E. A. (1991). The costs of organization. Journal of Law, Economics & Organi-
zation,7, 1–25.
Matthews, R. C. O. (1986). The economics of institutions and the sources of growth. Economic Journal,96, 903–918.
Mayer, K., & Teece, D. (2008). Unpacking strategic alliances: The structure and purpose of alliance versus supplier
relationships. Journal of Economic Behaviour and Organization,66, 106–127.
Mccorriston, S. (2002). Why should imperfect competition matter to agricultural economists? European Review of Agri-
cultural Economics,29, 349–371.
Ménard, C. (1995). Market as institutions versus organizations as markets? Disentangling some fundamental concepts.
Journal of Economic Behaviour and Organization,28, 161–182.
Ménard, C. (2004). ‘The economics of hybrid organizations’, Journal of Institutional and Theoretical Economics,160,
345–376.
Ménard, C. (2012). Hybrid modes of organization. Alliances, joint ventures, networks, and other “strange” animals, In
R. Gibbons & J. Roberts (Eds.), The Handbook of Organizational Economics, Princeton, NJ: Princeton University
Press.
Ménard, C. (2014a). Plural forms of organization: Where do we stand? Managerial and Decision Economics,34, 124–
139.
Ménard, C. (2014b). Embedding organizational arrangements: Towards a general model. Journal of Institutional Eco-
nomics,10, 567–589.
Ménard, C. (2007). Cooperatives: Hierarchies or hybrids? Vertical markets and cooperative hierarchies, In K. Karantini-
nis & J. Nilsson (Eds.), The Role of Cooperatives in the Agri-Food Industry, Berlin: Springer.
Ménard, C. (2017). Meso-institutions: The variety of regulatory arrangements in the water sector. Utilities Policy,49,
6–19.
Ménard, C., & Klein, P. G. (2004). Organizational issues in the agrifood sector: Toward a comparative approach. Amer-
ican Journal of Agricultural Economics,86, 750–755.
Ménard, C., & Valceschini, E. (2005). New institutions for governing the agri-food industry. European Review of Agri-
cultural Economics,32, 421–440.
Monderlaers, K., Baecke, M., & Lauwers, L. (2014). ‘Coping with market power: Dairy farmers’ interest in producers’
organizations and contracts’, Paper presented at the EAAEA 2014 Congress in Ljubljana (26–29 August 2014).
Moschini, G., & Hennessy, D. A. (2001). Uncertainty, risk aversion, and risk management for agricultural producers, In
B. Gardner & G. Rausser (Eds.), Handbook of Agricultural Economics, Amsterdam: Elsevier.
Narrod, C., Roy, D., Okello, J., Avendano, B., Rich, K., & Thorat, A. (2009). Public–private partnerships and collective
action in high value fruit and vegetable supply chains. Food Policy,34, 8–15.
Newey, W. K. (1987). ‘Efficient estimation of limited dependent variable models with endogenous explanatory variables’,
Journal of Econometrics,36, 231–250.
Nilsson, J. (1996). The nature of cooperative values and principles. Transaction cost theoretical explanations. Annals of
Public and Cooperatives Economics,67, 633–653.
Nuhanovic-Ribic, S., Tortia, E. C., & Valentinov, V. (2017). Agricultural co-operatives: A struggle for identity, In J.
Michie, J. R. Blasi, & C. Borzaga (Eds.), The Oxford Handbook of Mutual, Co-Operative, and Co-Owned Business,
Oxford: Oxford University Press.
OECD, (2009), An overview of risk-related policy measures, Managing Risk in Agriculture: A Holistic Approach,Paris:
OECD Publishing.
Okello, J. J., & Swinton, S. M. (2007). Compliance with international food safety standards in Kenya’s green bean
industry: Comparison of a small- and a large-scale farm producing for export. Applied Economic Perspectives and
Policy,29, 269–285.
Olson M. Jr (ed.), (1965). The Logic of Collective Action, Harvard: Harvard University Press.
22 S. CILIBERTI, A. FRASCARELLI AND G. MARTINO
Ostrom E. (ed.), (1990), Governing the Commons. The Evolution of Institutions for Collective Action, Cambridge: Cam-
bridge University Press.
Pascucci, S., Gardebroek, C., & Dries, L. (2012). Some like to join, others to deliver: An econometric analysis of farmers’
relationships with agricultural cooperatives. European Review of Agricultural Economics,39, 51–74.
Penrose-Buckley C. (ed.), (2007), Producer Organisations. A Guide to Developing Collective Rural Enterprises, Oxford:
Oxfam GB.
Raynaud, E., Sauvee, L., & Valceschini, E. (2005). Alignment between quality enforcement devices and governance
structures in the agro-food vertical chains. Journal of Management and Governance,9, 47.
Renda A., Cafaggi F., Pelkmans J., Iamiceli P., de Brito A. C., Mustilli F., & Bebber L. (eds.), (2014), Study on the
Legal Framework Covering Business-to-Business Unfair Trading Practices in the Retail Supply Chain.Bruxelles,
DG: Internal Market, European Commission.
Richardson, G. B. (1972). The organization of industry. Economic Journal,82, 883–896.
Royer, A. (2011). Transaction costs in milk marketing: A comparison between Canada and Great Britain. Agricultural
Economics,42, 171–182.
Royer, A., Mènard, C., & Gouin, D.-M. (2015). Reassessing marketing boards as hybrid arrangements: Evidence from
Canadian experiences. Agricultural Economics,47, 1–12.
Sexton, R. J., Sheldon, I., Mccorriston, S., & Wang, H. (2007). Agricultural trade liberalization and economic develop-
ment: The role of downstream market power. Agricultural Economics,36, 253–270.
Smith, R. J., & Blundell, R. W. (1986). An exogeneity test for a simultaneous equation Tobit model with an application
to labor supply. Econometrica,54, 679–686.
Staatz, J. (1987). Farmers’ incentives to take collective action via cooperatives: A transaction cost approach, In J. Royer
(Ed.), Cooperative Theory: New Approaches, Agricultural Cooperative Service,. Service Report No. 18, Washington,
DC: USDA.
Tortia, E. C., Valentinov, V., & Iliopoulos, C. (2013). Agricultural cooperatives. Journal of Entrepreneurial and Orga-
nizational Diversity,2, 23–36.
Valentinov, V. (2007). Why are cooperatives important in agriculture? An organizational economics perspective. Journal
of Institutional Economics,3, 55–69.
Valentinov, V. (2009). Third sector organizations in rural development: A transaction cost perspective. Agricultural and
Food Science,18, 3–15.
van Herck K. (ed.), (2014), ‘Assessing efficiencies generated by agricultural producer organisations’, Report for the EU
Directorate-General for Competition, https://doi.org/10.2763/76733.
Walker, G., & Weber, D. (1984). A transaction cost approach to make-or-buy decisions. Administrative. Science Quar-
terly,29, 373–391.
Williamson, O. E. (1975). Markets and Hierarchies: Analysis and Antitrust Implications, New York: The Free Press.
Williamson, O. E. (1979). Transaction-cost: The governance of contractual relations. Journal of Law and Economics,
22, 223–261.
Williamson O. E. (ed.), (1985)., The Economic Institutions of Capitalism, New York: The Free Press.
Williamson, O. E. (1991). Comparative economic organization: The analysis of discrete structural alternatives. Admin-
istrative Science Quarterly,36, 269–296.
Williamson, O. E. (1993). Transaction cost economics and organization theory. Industrial and Corporate Change,2,
107–156.
Williamson, O. E. (2000). The new institutional economics: Takingstock, looking ahead. Journal of Economic Literature,
38, 595–613.
Wills, R. L. (1985). Evaluating price enhancement by processing cooperatives. American Journal of Agricultural Eco-
nomics,67, 183–192.
Wollni, M., & Fischer, E. (2015). ‘Member deliveries in collective marketing relationships: Evidence from coffee coop-
eratives in Costa Rica. European Review of Agricultural Economics,42, 287–314.
Wollni, M., & Zeller, M. (2007). Do farmers benefit from participating in specialty markets and cooperatives? The case
of coffee marketing in Costa Rica. Agricultural Economics,37, 243–248.
Woolridge J. M. (ed.), (2002), Econometric Analysis of Cross Section and Panel Data,Cambridge,MA:TheMITPress.
Zheng, S., Wang, Z., & Awokuse, T. (2012). Determinants of producers’ participation in agricultural cooperatives: Evi-
dence from northern China. Applied Economic Perspectives and Policy,34, 167–186.
S. CILIBERTI, A. FRASCARELLI AND G. MARTINO 23
How to cite this article: Ciliberti S, Frascarelli A, Martino G. Drivers of participation in
collective arrangements in the agri-food supply chain. Evidence from Italy using a transac-
tion costs economics perspective. Annals of Public and Cooperative Economics. 2020;1–23.
https://doi.org/10.1111/apce.12263
