Risk Management in the Agri-Food Sector

This paper incorporates the interdisciplinary New Institutional Economics in a comprehensive framework for analyzing risk management in the agri-food sector. First, it specifies the diverse types of agri-food risks (natural, technical, behavioral, economic, policy, etc.) and the modes of their management (market, private, public, and hybrid). Second, it defines the efficiency of risk management and identifies the factors (personal, institutional, dimensional, technological, and natural) of governance choice. Next, it presents stages in the analysis of risk management and the improvement of public intervention in the governance of risk. Finally, it identifies the contemporary opportunities and challenges for risk governance in the agri-food chain.


Agri-food risks and modes of governance
Risk is any current or future hazard (event) with a significant negative impact(s). It is either idiosyncratic (accidental, low probability, unpredictable events) or systematic (high probability, "predictable" events). The risk/threat could be of natural (adverse weather, insectrelated, catastrophes), technological ("pure" technical failures), or human (individual or collective actions/ inactions), or a combination of these types. Individual behavior and actions that cause risks include agent's ignorance (lack of knowledge, information, training) and/or errors; risk-taking strategy (accepting above "normal" risk); mismanagement (inadequate planning, prevention, recovery); opportunistic behavior (precontractual cheating, post-contractual "moral hazard"); criminal acts (damaging/stealing or invading an individual's property); terrorist attacks (environmental contamination or other actions that aim to produce "mass terror"), etc. Collective actions are sources of risks that relate to: economic dynamics and uncertainty (changing demands, price volatility, competition, market "failures" and imbalances such as "lack" of labor, credit, inputs); collective orders (professional standards, strikes, trade and community rules and restrictions); or public orders (political instability and uncertainty, evolution in informal and formal norms, public "fail-ures" such as bad, delayed, or under/over intervention, law and contract enforcements, mismanagement, "inefficiency by design"), etc.
The agri-food sector can face risks associated with each component (dairy farms, food processors, trad-(dairy farms, food processors, trad-dairy farms, food processors, traders) or it could cause risks (risk from farming, from food processing, from food-distribution). Risk can be internal to agri-food chains such as hazards caused by one element affecting another within the sector. Risk can be external and associated with external factors (natural environment, government policy, international trade) and/or effecting external components (consumers, residents, industries, nature). Risk can be private, when it is assumed by individuals, collectives, entities, or industries, or it can be public when it affects large groups, communities, consumers, or future generations.
The magnitude of risk is large when an event has both a great likelihood of occurrence and substantial negative consequences -such as damage to humans and livestock (health and property), inferior yields and income, loss of market position, food and environmental contamination, etc. When risk is considerable, it would likely be associated with significant costs that often can hardly be expressed in monetary terms -human health hazards, degraded soil, and loss of biodiversity and eco-system services. Thus, "rational" agents that maximize their own welfare will be interested in investing in risk prevention and reduction. In a narrow (technical) sense, risk management comprises individual, collective and public action(s) for reducing/eliminating risk and its consequences. In a broader sense, risk management is the specific system of social order (governance) that is responsible Risk Management in the Agri-food Sector and organizational arrangements tailored to the features of risks and agents, such as codes of behavior, diverse (relational, security, future) contracts, cooperatives, associations, business ventures, etc. Market modes ("invisible hand of market") include various decentralized initiatives governed by free market price movements and competition, such as risk trading (selling/buying insurance), future contracts and options, production/ trade of special (organic, fair-trade, origins) products, etc. Public modes ("public order") include various forms of third-party public (government, international) intervention in market and private sectors, such as public information, regulations, bans, assistance, funding, assurance, taxation, provision, etc.
Sometimes, risk management can be conducted ef-be conducted ef-conducted effectively though "self-management" such as production management, adaptation to industry and formal standards, "self-insurance" though stock investments, financial reserves, etc. Primitive forms of on-farm risk management through production management improvement are widespread: control and security enhancement, use of appropriate crop varieties (resistance to pests, diseases, and weather), structure of technology and production, product diversification, dislocation, etc. Similarly, offfarm enterprise/income diversification is a major risk management strategy for most European farms (Bachev and Tanic, 2011). Very often, risk management requires an effective governance of relations with other agents such as the exchange and regulations of rights, the alignment of conflicts, the coalition of resources, and collective or public actions on regional, national, and transnational scales. Thus, a risk could be "managed" through market modes (purchase of insurance, hedging with future price contingency contracts), private modes (contractual or literal integration, cooperation), public forms (state regulation, guarantee, compensation), or hybrid combinations of other forms.

Risk management efficiency
Risk governance modes have different efficiencies be-ies be-be-because they have dissimilar potentials to reduce the like-the likelihood and impact of risks and involve different costs (Bachev, 2010). Market or collective governance has advantages over internal mode/"own protection" be-over internal mode/"own protection" be-/"own protection" be-own protection" be-" be-be-because market governance allows for the exploration of economies of scale/scope in preventing risks and man-/scope in preventing risks and man-scope in preventing risks and man-preventing risks and man-risks and man-s and man-and man-managing negative consequences. However, risk trading or sharing is associated frequently with significant transaction costs for finding the best partners and pric-the best partners and pric-best partners and pric-and pric-pric-prices, formulating/disputing terms of exchange, forming coalitions, and safeguarding against new risks from the opportunistic behavior of counterparts/partners. Con-behavior of counterparts/partners. Con-of counterparts/partners. Con-. Con-Consequently, markets and private sectors "fail" to effectively manage the existing/likely risks, and there is a need for "state intervention" in risk management (cooperation and assistance for farmers, public cost-sharing or provisions, mandatory assurance regulation). Thus, "governance matters" and applying proper risk man-applying proper risk man-proper risk man-proper risk man-roper risk man-management structure are important parts of the process of optimization (effective allocation) of resources.
Following Coase's logic (1960), if property rights were well-defined and transaction costs were zero,  erø, 2012;Williamson, 1996), and some transaction risk will always remain. Th erefore, eff ective risk man-Therefore, effective risk man-management is always connected with the needs for some trade-off between benefits from reducing particular risks (saved costs, minimized impacts) and related costs for risk governance. Thus, some "uncovered" risk will normally remain.
Individual modes can offer effective protection from different/multiple risks. Moreover, manage-manage-Risk Management in the Agri-food Sector from a partner's opportunism. The level of (overall) risk exposure is typically determined by the "critical" (most important) risk, and integral risk rarely equals the sum of the individual risks.
Frequently, there are a number of possible (alternative) forms of governance of a particular risk. For example, the "risk to the environment" might be managed by voluntary actions of farmers, environmental cooperation, private contracts with interested parties, assistance by third party organizations, public ecocontacts, and public regulations. Some forms of risk management are practically impossible or socially unacceptable, e.g., futures and insurance markets do not develop for many types of agro-food risks, and private management is the only option (Frank and Garcia, 2011); additionally, the management of many eco-risks require collective actions at local, eco-system, regional or transnational levels (Bachev, 2010;Sutherland et al., 2012). Many type of risk management are publicly imposed, e.g., there is harmonized public management of food safety risks in the EU, there are strict regulations of Genetically Modified Crops (GMC), the "precaution principle" is mandatory and carried out by states, "safety nets" are organized as public projects, etc.
Therefore, comparative analysis is to be employed to evaluate (technically, economically, socially) feasible alternatives and select the most efficient one that reduces the overall risk to an "acceptable" level and requires the minimum total (risk assurance and governance) costs.
The analysis must include all current and future costs associated with risk management, such as the current technological and management costs (for adaptation, compliance, information, certification), risk insurance premiums, contracting and coalition costs, and the longterm (future) costs for recovering damages, including the associated transaction costs (disputes, expertise, low suits) for claiming losses. Most risk management analyses ignore the current and likely long-term transaction costs associated with risk management.
In any case, the individual, group, community, sectoral, chain, national and international efficiency of risk management must be distinguished. Often, the elimination of risk for one agent induces (new) risk for another agent. For example, agri-food price fluctuation causes income risk for producers but benefits speculators, and application of chemicals reduces the risk for farmers but produces significant negative ef-fects (water, soil and air contamination) for residents, consumers, and affected industries. Furthermore, risk management is only a part of the overall governance of the diverse (production, consumption, transaction) activities of agents. For example, most managerial innovations in farming and the agri-food chain are driven by the motivation to economize and reduce transaction costs (Sporleder, 1992). Therefore, the total efficiency (benefits, disadvantages, cost saving and risk minimization potential) of various modes for individual agents and the public are to be taken into account.
It is frequently the case that the minimization of riskrelated costs is associated with an increase in production or transaction costs and vice versa. Similarly, the risk elimination costs of one agent frequently increase the security of another agent in the agri-food chain.
According to specific natural and socio-economic environments, personal characteristics of individuals, and social preferences, various structures of risk governance evolve in different sub-sectors, industries, supply chains, and societies. At one extreme, the system of risk management works well, and only "normal" (en-management works well, and only "normal" (en-works well, and only "normal" (en-s well, and only "normal" (en-well, and only "normal" (en-, and only "normal" (en-and only "normal" (en-"normal" (en-normal" (en-" (en-(en-(entrepreneurial) risk is left "ungoverned". In some cases, the market fails to provide adequate risk governance, but a variety of effective private modes emerge to fill the gap. Often, both market and private governance fail, but effective public involvement cures the problem.
Nevertheless, there are situations when specific institutional and risk management cost structures lead to failures of market and private modes, and needed public intervention in risk governance. Consequently, a complete range of risks is left unmanaged, which has adverse effects on the size and sustainability of agrifood enterprises, market development, the evolution of production and consumption, the state of the environment, and social welfare.
Depending on the costs and the efficiency of the specific system of governance in a particular (sub)sector, region, country, supply chain etc., there will be unlike outcomes in terms of "residual" risks and dissimilar states and costs of human, food, environmental, etc., security in different regions and periods of time ( Figure 1). For instance, when there is inefficient public enforcement of food, labor, and eco-safety standards (lack of political willingness, administrative capability), then enormous "gray" agrarian and food sectors develop with inferior, hazardous, and counterfeit components.

Factors of governance choice
The forms of risk management depend on the risk type, personal characteristics, institutional environment, progress in science and technology, culture, social education and preferences, evolution of the natural environment, etc. (Figure 1). Risk features such as the origin, probability of occurrence, likely damages, and scale are important factors for the choice of governance.
For instance, local risk could be managed through private modes, while most market and environmental risks require collective actions at the regional, national or transnational level. For high probability and harmful risks, agents prefer more secure/expensive modes such as security investments, purchase of insurance, maintaining reserves, taking hostages, and interlinked organizations. However, many smallholders cannot afford the related costs and practice no or primitive forms of risk management, such as cash-and-carry deals and product diversification. In that case, there is the need for third party (government, international assistance) intervention, though insurance, support, safety nets, etc., to decrease the vulnerability of farmers.

Personal and behavioral characteristics of agents
include interests, preferences, knowledge, capability, risk-aversion, reputation, trust, "contractual" power, and opportunism. For instance, some risks are not perceived (unknown) by private and public agents; therefore, no risk management is put in place. In some cultures, the cooperative is the preferred mode of organization; an experienced/trained farmer might design and manage a larger organization/hired labor and more outside (credit, insurance, inputs supply) contracts that are adapted to his needs. Additionally, a risk-taking entrepreneur prefers riskier, but more productive ventures.
Behavioral factors such as the bounded rationality and opportunisms of individuals are identified as being responsible for transaction costs and, thus, for organizational choice (Williamson, 1996). They are widely studied in insurance theory as sources of cheating on both sides of contracts (Derrig, 2002 (Olson, 1969). Bounded rationality often makes it costly/impossible to distinguish opportunistic from non-opportunistic behavior (e.g., a farmer finds that purchased seeds were of poor quality during the harvest), and agents protect their rights, investments, and transactions from the hazards of opportunism through means such as ex-ante efforts to find reliable counterparts and design efficient modes for partner commitments and ex-post investments for overcoming (monitoring, controlling, stimulating cooperating) possible opportunism during contract execution (Williamson, 1996).
In the agri-food sector, opportunism is widespread before signing insurance contracts (not disclosing information for possible risks) or during contract execution (not taking actions for reducing damages when events occur or consciously provoking damages in order to get insurance premiums). These events increase the insurance prices and limit the utilization of insurance contracts by small enterprises. The insuree also "discovers" the insurer's opportunism after the occurrence of a harmful event and finding that the assurance terms (protected risks, extent of coverage, dam-Electronic copy available at: https://ssrn.com/abstract=2253172 Risk Management in the Agri-food Sector ages assessments, costs) were not explained or adapted to the farmer's needs. For many farm-related risks, the markets evolve slowly or the insurance services are inaccessible to small operators. For some important risks, the insurance is not available "for purchase at all", e.g., the risk of lack of market demand, price fluctuations, opportunism of counterparts, etc. Farmers must develop other (private, collective) modes to safeguard investments and rights or lobby for public intervention in the assurance supply. The institutional environment ("rules of the game") includes formal and informal rights and rules and system(s) of the enforcement of these rights and rules (North, 1990 (theft) and organized ("security service") risks that devastated private businesses and household welfare.

Dimensional characteristics of activities/transac-
tions include uncertainty, frequency, and asset specificity, which were identified as critical factors by Williamson (1996), and appropriability was later added by Bachev and Labonne (2000).  Comparative analysis is to be employed to compare the feasible forms and select the most efficient alternative that reduces the overall risk to an "acceptable" level and minimizes the total (risk assurance and governance) costs. Most elements of risk governance ef-Most elements of risk governance ef-efficiency are difficult to quantify, such as individuals' characteristics, amounts of risk, levels of benefits and the costs associated with each mode. Th e "measure-costs associated with each mode. Th e "measure-ociated with each mode. Th e "measure-with each mode. Th e "measure-The "measurement problems" associated with the transaction benefits and costs are well-specified (Bachev, 2011b) and prevent the utilization of traditional (Neoclassical) models simply by adding a new "transaction" or risk management activity (Furuboth and Richter, 1998).

Extended Qualitative (Discrete structural) analysis
can be used as operationalized by Williamson (1981).
For assessing the efficiency of risk governing structures, the features of risk to be managed (probability, significance, acceptance level, collective action needs) and its critical (institutional, technological, behavioral, etc.) factors are to be matched with comparative advantages (potential) of alternative modes to inform, stimulate appropriate behavior, align interests, and overcome, reduce, control, share, dispute, and mini-and minimize overall costs of risk.
In a specific market, the institutional, technological and natural environment effective risk governance choice depends on a combination of risk features (probability of occurrence, likely magnitude of damages) and critical dimensions of activities/transactions (appropriability, assets specificity, and frequency), (Figure 2). High "standard" risks might be effectively managed through free market modes such as standard/classical insurance, inputs supply, marketing, etc.
contracts. Highly probable and damaging risks with a good appropriability and frequency of transactions between the same partners require a special (e.g. relational) contract. The later form is also appropriate for the risks surrounding with low uncertainty, high assets specificity and appropriability, and occasional character of the relations between the counterparts.
Principally, risks combined with high specificity, appropriability and frequency might be effectively managed though a vertical integration (internal risk man-

agement, contract forward or backward integration for
Electronic copy available at: https://ssrn.com/abstract=2253172 Risk Management in the Agri-food Sector risk sharing or mitigation). Highly likely and menacing risks combined with a high assets specificity and a good appropriability call for a collective organization (cooperation, collective action). Moreover, such risk/ costs sharing organization could be easily initiated and maintained since the condition of a high risk and assets dependency is in place.
However, serious transaction risk exists when assets specificity is combined with high uncertainty, low frequency, and good appropriability. In that case, the elab- there are increasing demands (premium) for organic, original and fair-trade products associated with some forms of (natural, poverty, labor, quality) risk management, but their supply could not be met unless effective trilateral governance, including independent certification and control, is put in place. Similarly, for risks with a low appropriability a third party (public) intervention is necessary to secure the effective risk management. Moreover, while a high probability low danger risks need a collective organization assisted by a third-party ("quasi" public organization for risk sharing and mitigation), the high damaging risks necessitate a public organization.

Stages in analysis and improvement of risk management
The analysis and improvement of risk governance in the agri-food chain is to include following steps: First, the existing and emerging threats and risks are identi-  and environmental changes and associated threats and risks. A holistic framework for assessing the efficiency and evolution of governing modes is suggested by OECD (2011) andBachev (2010). That stage identifies the deficiencies of dominating (market, private, public) modes to solve existing and emerging risks, and determines the needs for (new) public intervention. For instance, when appropriability associated with transaction/activity is low, there is no pure market or private mode to protect against the associated risks.
Respecting others' rights or "granting" risk protection rights to others could be governed by "good will" and charity actions (e.g., the eco-sustainability movement initially evolved as a voluntary activity). However, voluntary initiatives can hardly satisfy entire social demands, especially if they require significant costs.
The emergence of special large-member organizations for dealing with low appropriability to cover entire "social" risks is a very slow and expensive process, and it is unlikely to be sustainable in the long run (free riding). There is a strong need for third-party public intervention to make protection from risk possible or more effective, such as purely public organizations (public assurance for natural and economic disasters) or "quasi-public" modes (collective organizations assisted/ordered by third parties) for high probability, low damage risks (Figure 2). In other instances, it is more efficient to establish public regulations for risk minimization for the utilization of resources, products, and services (standards for labor, product, environmental safety); the introduction of foreign species, GMC, and (water, soil, air, comfort) contamination; bans on inputs, products, and technologies; regulations for trading ecosystem service protection; trade regimes; mandatory risk and eco-training, and licensing of operators, etc.
In other instances, using incentives and restrictions of the tax system is the most effective method. Different types of tax preferences are used widely to create favorable conditions for the development of (sub) sectors and regions, forms of organization, population segments, activity types. For example, environmental taxation on emissions or products (inputs, outputs) can be applied to reduce use or emissions of harmful substances; tax reduction can be used to overcome negative consequences of natural disasters, etc.
In some cases, public support of private organizations is the best mode for intervention. Programs for Electronic copy available at: https://ssrn.com/abstract=2253172 Risk Management in the Agri-food Sector The level of effective public governance depends on the type of risk and the scale of intervention. There are public involvements that are to be executed at the local (ecosystem, community, regional) level, while others require nationwide governance. There are other risk management activities that are to be initiated/ coordinated at the international (regional, European, worldwide) level due to the strong necessity for transborder actions or consistent (national, local) government failures. Often, the effective management of many risks requires multilevel governance with a system of combined actions at various levels that involve a diverse range of actors and geographical scales. If there is a strong need for third-party public involvement but effective (government, local authority, international assistance) intervention in risk management is not introduced in a timely manner, then significant risks to individuals and public persist while agrarian "development" is deformed.
Dealing with many risks in the agri-food sector/ chain requires multiform, hybrid, multilevel, and transnational intervention, and the appropriate governance mix is to be specified.
Comparative analysis can improve the design of public intervention for specific conditions of foodchain components in a particular country/region in terms of increasing security and decreasing costs. The suggested approach also can predict likely cases of (new) public failures due to the impossibility of mobilizing political support and resources or the ineffective implementation of otherwise "good" policies. Public failure is feasible, and its timely detection enables risk managers to foresee the persistence/increase of certain risks and inform local and international communities about the consequences.
Risk management analysis is to be made at different levels -individual components (input supply, farm, processing, transportation, distribution), regional, sub-sector, food-chain, national, and international, according to the types of risks and scales of collective actions necessary to mitigate risks. It is not a one-time exercise to complete the last stage with a perfect system of risk management. Rather, it is a permanent process to improve risk management along with the evolution of the socio-economic and natural environment, individual and community awareness, modernization of technologies, etc. Moreover, public (local, national, international) failure often prevails, which also brings us into the next cycle in the improvement of risk management.
For the application of this suggested approach, in addition to the traditional (statistical, industry, etc.) information, new types of data are necessary for diverse risks and forms of governance, critical factors for each agent, and levels of benefits and costs. Such data are collected through interviews with agri-food chain managers, stakeholders, and experts in these areas. Consumers' concerns about food safety risks have increased significantly after major food safety "events"/crises in recent years (avian flu; mad cow and foot-and-mouth diseases; poultry salmonella; contamination of dairies, berries, and olive oil; natural and industrial disasters, etc.). For example, since 2005 the number of respondents "worrying about food-safety problems" in the EU increased, and as many as 48%

Contemporary opportunities and challenges for risk governance in the agri-food chain
of European consumers (in Bulgaria, approximately 75%) indicated that consumed food was "very or fairly likely" to damage their health (Eurostat, 2010).
There are a number of (new) opportunities for risk governance in the agri-food chain ( Figure 3) Third, considerable development of specialization of activities (including risk taking, monitoring, management) and concentration of (integral) management in food-production, processing, servicing, and distribution allowing centralized innovation and enforcement; time, scale, and scope economies; and easy third-party control. For instance, the market share of the three largest food retailers comprise between 27 and 91% in the EU states (Eurostat, 2010), and food safety training, certification, inspection, and information are large international businesses (Humphrey and Memedovic, 2006). Fourth, quasi or complete integration of the consecutive or dependent stages in food chains are creating mutual interests and effective, long-term means for risk perception, communication, and management.
For example, in Bulgaria the (raw) milk supply is closely integrated by (dairy) processors through onfarm (collecting, testing) investments and interlinks (inputs, credit, and service supply against milk delivery) contracts with smallholders, while dairy marketing is managed by branding (standards, bio labels) and long-term contracts (Bachev, 2011a).
Fifth, the increasing consumer "willingness to pay" for food safety attributes such as chemical and hormone bans, safety and inspection labels, and original and special products (Trench at al., 2011). The latter justify and make economically possible the pay back of costs for special governance.  ii) Increasing new threats, risks, and uncertainty connected with inputs, technologies, and product differentiation and innovation. For example, the Fukushima nuclear accident severely affected the agri-food sector in Japan and beyond, and there are uncertainties associated with the growing application of nanotechnologies and GMCs, etc. (Eurostat, 2010).
iii) Increasing specialization and concentration of activity and organizations which separate "risk creation" (incidents, ignorance, opportunisms) and risk taking (unilateral dependencies, quasi-monopolies, spill-overs, externalities). Thus, risk assessment, pricing, communication, disputing, and liability through (pure) market and private modes are very difficult and costly. For instance, cheating, misleading, and pirating are common in food chain relations, including high information asymmetry, detection, disputing, and punishment costs. It is indicative that for food risk information, consumers in the EU place greater trust in "health professionals", "family and friends", "consumers associations", and "scientists" rather than "food producers" and "supermarkets and shops" (Eurostat, 2010). iv) Widespread mass production, distribution, and consumption increase the vulnerability of the agrifood chain expanding scope and severity of natural, incidental, opportunistic, criminal, and terrorist risks.
For instance, in Europe there is a growing number of official notifications based on market and non-member country controls, food poisoning, consumer complaints, company self-checks, border screenings, and rejections (Eurostat, 2010). v) Increasing adaptation and compliance costs (capital, training, certification, documentation, etc.) for rapidly evolving markets and institutional environments that delay or prevent the reformation of small farms and food chain enterprises (Trench et al., 2011;Bachev, 2010). For instance, in Bulgaria, the adaptation of dairy and meat processors to EU standards continued for 10 years, and two-thirds of them ceased to exist before the accession of their country (Bachev, 2011a).
vi) Public and private food quality and safety standards, and the efficiency of their enforcement, differ considerably between industries, countries, and regions (Humphrey and Memedovic, 2006). That is result of unequal norms (GAPs, formal and informal rules) and implementing and enforcing capability, or deliberate policies, or private strategies (e.g., multinationals that sell the "same" products with different quality in different countries). "Double/multiple standards" are responsible for exchange inequality and dissimilar threats and risks exposure of individual agri-food systems.
vii) Wide spread "public failures" in food chain (risk) management such as bad, inefficient, delayed, or under/over interventions; gaps, overlaps, infighting, and contradictions among different agencies and rules; high bureaucratic costs; and unsustainable costs and underfunding. For instance, the Bulgarian Food Agency was established after a 5-year delay; the Acquis Communautaire are still not completely implemented (capability deficiency, mismanagement, corruption); and trust of the EU rather than national institutions prevails (Bachev, 2010). There are also numerous instances of international assistance or governance failures when institutions are "imported" rather than adapted or designed for specific local conditions (Bachev, 2010).
viii) Production, marketing, and consumption traditions, high food or governance costs, and deficiencies of will and capacity are responsible for the persistence of the high risk informal/gray agri-food sector around the globe without effective control and substandard, fake, and illegitimate products and activities. For instance, only one-third of Bulgarian dairy farms comply with EU milk standards, only 0.1% possesses safe manure pile sites, and half of the milk production is consumed by the producer at home, exchanged or directly sold (Bachev, 2010).

ix) Multiplying new threats and risks associated
with adversary (competitor) and terrorist attacks, and emerging governing and exchange forms (street sales; internet, phone and mail order sales; shopping trips). These risks all require specific/non-traditional risk management methods and modes such as guards, policing, intelligence, multi-organizational and transnational cooperation, etc.

Conclusion
We have demonstrated that governance, along with technical, information and other issues, play a central role in risk management analysis and design. The entire spectrum of diverse types of threats and risks (natural, market, health, criminal, policy etc.), specific (natural, Electronic copy available at: https://ssrn.com/abstract=2253172 Risk Management in the Agri-food Sector technological, and often neglected behavioral, dimensional, and institutional) factors, and the comparative benefits and costs (including commonly ignored thirdparty costs, transaction costs, and time) are to be taken into account in assessing efficiencies, complementarities and prospects of all feasible and alternative (market, private, public, hybrid) modes. We suggest the use of discrete structural analysis to match the features of a risk to be managed (such as probability, significance, acceptance level, and collective action needs) and its critical (institutional, technological, behavioral, etc.) factors with the comparative advantages of the alternative modes to inform, stimulate appropriate behavior, align interests, and overcome, reduce, control, share, dispute, and minimize the overall costs of risk.
Moreover Finally, we argue that greater support must be given to multi and inter-disciplinary research on (factors, modes, impacts of) risk governance in the agri-food chain in order to provide effective support to national and international policies, the design of modes for public interventions, and individual, collective and business actions.