Why do farmers behave as they do? Understanding compliance with rural, agricultural, and food attribute standards

Herzfeld, Thomas; Jongeneel, Roel Article — Accepted Manuscript (Postprint) Why do farmers behave as they do? Understanding compliance with rural, agricultural, and food attribute standards Land Use Policy Provided in Cooperation with: Leibniz Institute of Agricultural Development in Transition Economies (IAMO), Halle (Saale) Suggested Citation: Herzfeld, Thomas; Jongeneel, Roel (2012) : Why do farmers behave as they do? Understanding compliance with rural, agricultural, and food attribute standards, Land Use Policy, ISSN 0264-8377, Elsevier, Amsterdam, Vol. 29, Iss. 1, pp. 250-260, https://doi.org/10.1016/j.landusepol.2011.06.014 , http://www.sciencedirect.com/science/article/pii/S0264837711000688 This Version is available at: https://hdl.handle.net/10419/149556 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by-nc-nd/4.0/ 1 Why do farmers behave as they do? Understanding compliance with rural, agricultural, and food attribute standards Thomas Herzfeldaand Roel Jongeneela aAgricultural Economics and Rural Policy Group, Wageningen UR NL-6706 KN Wageningen, The Netherlands [email protected] [email protected] corresponding author: Thomas Herzfeld Tel: +31-(0)317/484876, Fax: +31-(0)317/484736 [email protected] This is a “Post-print” accepted manuscript, which has been published in “Land Use Policy”. Please cite this publication as follows: Herzfeld, T., Jongeneel, R. (2012). Why do farmers behave as they do? Understanding compliance with rural, agricultural, and food attribute standards. Land Use Policy 29, 250-260. You can download the published version at: http://dx.doi.org/10.1016/j.landusepol.2011.06.014 © <2012>. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 2 Why do farmers behave as they do? Understanding compliance with rural, agricultural, and food attribute standards Abstract Agricultural production experiences a shift in underlying institutions during the last years. The importance of private stakeholders like retailers, processors, consumers as well as tax payers is emerging. Eligibility for single farm payments and marketing of products is linked to compliance with diverse codes of practice and standards. Voluntary certification schemes are not only relevant with respect to agriculture’s traditional activity (i.e. food production), but are also highly relevant in the case of rural policies. Examples are the EU’s agrienvironmental schemes, aimed at preserving biodiversity, landscape elements, etc., which are characterized by voluntary participation and the contractual commitment to apply certain standards. These relationships, however, are characterized by strong information asymmetries. Economic literature offers several approaches to analyse similar relations. The classical approach is based on the assumption of utility maximising agents that will comply with rules as long as the net utility of compliance will be higher than the net utility of an offence. Recent evidence challenges this result. Experiments show that people behave more honestly than the classical deterrence model predicts. The main objective of the paper is to compare the various approaches to explain compliant behaviour. Additionally, an outlook for empirical applications of the theoretical framework developed here is given to motivate further research. However, our outlook shows that the selection of appropriate variables to analyse compliance empirically is open to debate. Keywords: regulation, cross-compliance, quality standards, compliant behaviour 3 Theories of choice are at best approximate and incomplete. (Tversky and Kahneman, 1992; p. 317) Introduction Agricultural production currently faces a fundamental shift in underlying constraints. Farmers have to adapt themselves not only to changed price signals, but also to a new institutional arrangement giving agriculture its proper place in society. An example is the current debate on the reform of the Common Agricultural Policy, which emphasizes the need to make agriculture more sustainable and improve its role in the provision of public goods. Whereas this often implies making increased use of regulatory measures affecting land management, the literature is rather silent on the issue of compliance with such regulations. However, compliance might be crucial for the effectiveness of the regulations with respect to achieving the specified policy objectives. Another development is that judicial regulations are increasingly strengthened by privately agreed regulations or contractual arrangements. Even more important, those regulations and arrangements often exhibit a quasi-mandatory character. Examples are obligatory and voluntary cross-compliance, the EU’s agrienvironmental schemes--i.e. voluntary standards aimed at preserving biodiversity and landscape element--and privately initiated food quality standards or Quality Assurance Schemes (QAS). All these domains were previously exclusively governed by European commission’s Directives and Regulations as well as national laws on the environment, health and food safety. They should guarantee an equal level of environmental protection and consumer’s health within the European Union and at the same time give the producers economic freedom as how to satisfy the statutory targets. Two institutional developments over the recent decades introduced a greater variety of production conditions within the EU. Starting with the introduction of voluntary programs, like the agri-environmental schemes (AES) in the late 1980s, participating farmers face 4 additional constraints on production or changed incentives to comply to existing regulations from a policy perspective. With the introduction of compulsory cross-compliance regulations with the 2003 CAP reform (Regulation 1782/2003), agricultural policy provides an additional enforcement system beside existing legislation. More specifically, then about five million EU farmers have to comply with 19 pre-existing legislative acts, also called Statutory Management Requirements (SMR), and a number of standards ensuring good agricultural and environmental condition of agricultural land (GAEC) to be qualified for the Single Farm Payment (SFP). Beyond these minimum requirements each Member State has put in place several AES which entitle farmers to additional compensations. Main focus of this instrument mix is to reach environment, biodiversity, food safety, animal welfare and soil conservation related goals. Turning from policy to new institutions invented by trading partners, beside established processor-specific standards in the framework of contract agriculture, food retailers and processors set up standards of Good Agricultural Practices at the end of the 1990s. These standards define food safety, animal health and welfare, environmental and land use requirements (e.g. soil erosion). However, some of them operate to ‘baseline’ standards, i.e. close to minimum legal requirements, and others extend far beyond legal requirements. The majority of them have only limited regional or national relevance. However, it seems that a minority relate to a standard of global relevance (e.g. GlobalGAP, British Retail Consortium Global Standard Food). Farmers can, theoretically, freely decide to adopt those standards. But it is quite reasonable to assume that with an increasing range of standards, they will develop a quasi-mandatory nature, at least indirectly or for larger farmers. For instance, 90 per cent of the pig meat traded in the Netherlands satisfies the IKB-Pig standard. Meat not having this standard will face obstacles hindering smooth trade. For milk the Dutch dairy farmer’s participation in the KKM quality standard is even close to 100 per cent (Jongeneel, 2006). 5 Moreover in the UK, today, the most dominant of the quality assurance schemes (in terms of market share) are those which come under the umbrella of Assured Food Standards (AFS), well-known because of the ‘Little Red Tractor’ logo, a registered trademark (Farmer et al., 2007, p. 24). Economic analysis of these new institutions 1 regard usually the costs and benefits of regulations. A full evaluation and measurement of these costs and benefits (welfare impact assessment) involves great difficulties in practice and is beyond the scope of this paper (for a discussion, see Gardner (2003). But one central element of the effectiveness of regulations is afarmer’s behaviour. Farmers can (or have to) adopt the new requirements. However, they do not necessarily adjust their input and output mix to mitigate the impacts of the regulation. Thus, compliance with the imposed regulations/ standards is a critical issue. Non-compliance at a large scale might undermine the acceptance of regulations or even provoke their complete failure. Due to its uniform nature of the described regulations for every farmer, the optimal design of several types of regulations for heterogeneous set of farmers from a regulator’s perspective is neglected here (see for instance Moxey et al., 1999; Hart and Latacz-Lohmann, 2005). Furthermore, a rich body of literature analyses the voluntary adoption of agrienvironmental schemes (e.g. McEachern, 1992; Beedell and Rehman, 1999; Morris and Winter, 2002). Analysis of compliance leads to an inherent problem of these contractual arrangements. All of the above mentioned institutions are exposed to information asymmetry and therefore, provide examples of incomplete contracts (Brousseau and Farès, 2000). Whereas compliance with cross-compliance is motivated by two parallel enforcement mechanisms, pecuniary incentives for farmers are less clearly identifiable in the case of private food quality standards or QAS. Whether producers receive higher prices after adoption of QAS is less transparent, 1 Institutions are commonly defined as the written and unwritten ‘rules of the game’ or, in other words, arrangements which structure repeated human interactions (Ostrom and Basurto, 2011). 6 but certainly they face higher costs. Monetary benefits of additional incentives like improved market access, new access to certain supply chains or higher consumer trust are difficult to quantify. Incentives for moral hazard clearly exist. The economic literature provides various approaches to the analysis of compliant behaviour. 2 More recently, the economic literature has been influenced by sociological and psychological findings. Within this paper we review the various approaches to the economic analysis of compliance and apply them to the situation of an agricultural producer. For various reasons individual farmers and managers of enterprises, like cooperatives or processors, might behave differently. Slemrod (2007) highlights the different role of risk aversion and bearing of a possible penalty in both types of organisation. Farmers facing a potential loss of all Single Farm Payments might have to be treated distinctly from a food processor facing the potential loss of one client demanding some food standard certification. However, the overwhelming majority of European farms are run as a family business and some food quality standards, like GlobalGAP or SQF1000, are targeted towards farms and not processors. Finally, a continually increasing range of standards at the retail level reduces the market potential for non-certified produce, increasing the deterrence potential of the loss of certification. Therefore, differences in decision making at farm or firm level will be neglected here, and the analysis will concentrate on one individual decision maker. Our contribution to the literature is twofold. Firstly, the disparate literature is presented and applied to issues in agricultural production. Thereby, we go beyond the neoclassical theoretical framework mainly used in previous papers (Sutinen and Kuperan, 1999; Hart and Latacz-Lohmann, 2005). Secondly, a discussion of a possible empirical analysis is provided. The remainder of the paper is as follows: Section 2 starts with a model of compliance based on neoclassical utility maximising the theoretical framework. Underlying assumptions are 2 See for instance a review of the economic literature on tax compliance by Andreoni et al. (1998) or the literature on corruption reviewed by Jain (2001). 7 criticised in Section 3 and other motivations of compliance to regulations are presented. Section 4 aims at comparing the different theoretical approaches. An outlook on possible empirical applications concludes the paper. The standard neoclassical approach Starting with Becker’s (1968) seminal work on the economics of crime, a variety of theoretical attempts to analyse non-complaint behaviour emerged. Applications and further refinements of the Becker model is based on the classical utility maximizing theoretical framework, more specifically, a von Neumann-Morgenstern utility function. The central element of this neoclassical theory is the comparison of an agent’s utility in case of compliance in contrast to his/ her utility in the case of non-compliance, the latter being weighted with a probability of detection. The modelling framework described is applied to the case of an agricultural producer in the following discussion. According to the standard neoclassical model, farmers maximize their profits subject to a production technology constraint. The standard outcome of this optimization problem is a (short-run) system of variable input demand and variable output supply relationships, which are a function of input and output prices, quasi-fixed factors (capital, land, family labour), and dynamic shifters like technological change and genetic progress. If some inputs or outputs are restricted (e.g. the milk quota), than these restricted variables are also included in the set of explanatory variables. The outcome of the optimization reflects the farmers’decisions regarding input and output mix, where increasing costs of production (input price increases) lead generally to a decline in input demand as well as output supply. Increasing output prices, in contrast, show a reverse effect: they lead to an increased output (supply) as well as increasing demand for variable inputs. 8 Understanding the impact of regulation within the neoclassical economic framework can be obtained in two steps: Firstly, one could include regulation as a further constraint on production possibilities, and therewith as factors affecting the production technology. Secondly, one could allow for the possibility that farmers might violate the regulations or show non-compliant behaviour (Sutinen and Kuperan, 1999). To begin with the first approach, regulation is treated as further constraint on production possibilities, and therewith as factor affecting the (effective) production technology. Since imposed regulations reduce the production possibility set (excluding possibilities that were allowed without the regulations being imposed) in general they are expected to negatively affect production or to increase the costs of production. Defining output supply vectors as q(.) and input demand as x(.), a typical solution to this problem would be )};,,(|{Max tf rxqwxpq  with   tq ;,, rwpq  and   tx ;,, rwpx  (1) where pand wrepresent a vector of output prices and input prices, and trepresents technical change. Regulatory constraints are represented by the vector r, with the bar indicating that they are assumed to be binding (i.e. restricting a farmer’s behaviour). 3 All functions are assumed to be ‘well-behaved’, i.e. output supply as well as input demand increasing in pand decreasing in w. Substituting the supply and demand relationships into the profit condition would yield the dual (optimum value) profit function   t;,, rwp which gives the profits associated with the regulatory regime. As denoted before the profits including the regulatory constraint impact will be lower than without the regulation, i.e.     tt ;,;,, wprwp   . The drawback of this first approach is that it is implicitly assumed that the regulatory constraints 3 Here we discuss the case of restrictions on input use like manure or stocking density. In the Annex A we provide the opposite example if a certain regulation requires more of one factor, for instance more labour, or one-time investments. The general conclusions remain unaffected. 15 framework. Research over the last two decades has shown that people do not only care about economic incentives but also about the well-being of other humans; they also exhibit social preferences (Fehr and Falk, 2002). Therefore, drivers of human behaviour are discussed from a psychological perspective followed by determinants derived from the sociological literature. We use a very simple rule to discriminate between both strands: all factors that relate directly to an individual’s attitude are classified as psychological, and the determinants which shape honesty indirectly via the society around the agent are termed as sociological. Our final concern is the link between institutional quality and compliance. Criticisms of assumptions of expected utility theory The following three axioms of agent’s preferences related to certain prospects are underlying the expected utility theorem, more specifically the existence of a von Neumann-Morgenstern utility function: ordering, requiring itself completeness and transitivity; continuity; and independence. Whereas the first two axioms--ordering and continuity--are common for every economic model of choice, the independence axiom implies that any prospect that is weakly preferred to another will also be weakly preferred if weighted with the same probabilities (Starmer, 2000). Additionally, it is assumed within the expected utility framework that the agent knows all possible outcomes and related probabilities of his or her choices. Kahneman and Tversky (1979), Tversky and Kahneman (1986), Machina (1987) and Rabin and Thaler (2001), to name only some examples, provide an ample discussion of those axioms. In particular, experiments show that respondents show no consistent behaviour if faced with non-weighted and weighted prospects. Therefore, most alternatives of the expected utility theory relax the independence axiom. For an extensive review of the literature the interested reader is referred to Schoemaker (1982) and Starmer (2000). In the following, some selected critics which seem to be most relevant for our examples are discussed more extensively, but 16 an exhaustive survey of the literature would go beyond the scope of this paper. More specifically, we will discuss the relevance of losses and gains versus changes in wealth levels, the framing effect and problems of deriving objective probabilities of detection. One of the most prominent alternative developments of the expected utility theory is the prospect theory introduced by Kahneman and Tversky (1979). Their work is based on results from various simple choice experiments at three different universities and includes a procedural aspect, assuming that an agent’s decision follows some heuristics or rules (Starmer, 2000). One central element of prospect theory is the notion that arguments of the utility function are rather changes in wealth than the final asset position. Thus, in a first stage, respondents order the different prospects they face with respect to gains and losses related to a subjectively chosen reference point. They do not integrate previous entitlements and potential gains to determine a maximised outcome as a first choice. In a second stage, potential outcomes and related probabilities are evaluated and the alternative with the highest utility selected. Consequently, the first phase of decision in prospect theory violates the transitivity axiom of the expected utility theory. Kahneman and Tversky (1979) show that their proposed approach is capable to explain risk-averse behaviour in the case of potential gains but riskseeking behaviour in the case of potential losses. In a later paper they use the terms ‘diminishing sensitivity’ to explain the convex and concave shape of their value function, and ‘loss aversion’ to describe the steeper curve in the domain of losses compared to the domain of gains (Tversky and Kahneman, 1992). 8 Many authors usually assume the reference point equivalent to a status-quo wealth position. Therefore, it follows from the prospect theory that each farmer will evaluate a certain fixed penalty differently. Implementation of crosscompliance regulation takes account of this in assuming fines as a percentage of the farm specific Single Farm Payment. 8 Additional elements of the prospect theory like the certainty and the reflection effect seem to be of more theoretical relevance and will not be discussed here. See Kahneman and Tversky (1979) for a detailed description. 17 A formal exposition of the prospect theory to the question of compliance is developed below based on the work by Tversky and Kahneman (1992). 9 The major change with respect to the deterrence model above is the distinction between the profit πif a violation is detected and if not.          trwp ;,, if 1 0     A value function is specified depending on the changes in utility, where in turn utility is a function of the change in profits, or u(Δπ). As outlined above, the choice of a reference point is crucial. 10 Under the assumption of compliant behaviour as the reference, the two possible outcomes are )(r   in the case of non-detection of deviation and )()( rrr  g if a violation is detected and punished. It is assumed that the value function is concave for gains and convex for losses. The third prospect, no change in income, is valued zero and, therefore, drops out. A second important element of prospect theory is the underestimation of high probability events and overestimation of low probability events. To take this discrepancy between the perceived and actual probability of detection into account, the probability of detection θis replaced by a weighting function φ(θ), a necessary nonlinear transformation of the probability scale (Tversky and Kahneman, 1992). Following Al- Nowaihi et al., (2008), and if preference homogeneity holds, the value function could be written as a power function with equal exponents for gains and losses. Subsequently, the farmer’s new objective function becomes: ))(()())(()1( rr    uuV (4) The variable is called the coefficient of loss aversion, implying that > 1. Maximizing Vwith respect to riyields the following optimal solution (see Annex B for further details): 9 Yaniv (1999) as well as Dhami and al-Nowaihi (2007) provide a similar exposition in the case of tax evasion. 10 Most recent developments allow to model an uncertain point of reference (Schmidt et al., 2008). However, as the focus of this paper is to compare the theoretical approaches, a simplification is opted for. 18 ii r g uu u r                                          )( )( )( )( )1( )( )( r           for all i= 1, …, Nr(5) As (5) shows, marginal profits are related to marginal penalty, just like in equation (3), but now a ‘correction’ factor (see first right hand sight term in brackets) is added. This term depends on    )(u and    )(u which together capture the degree of loss aversion. Moreover, it depends on the weighting function reflecting the difference between the objective and perceived probabilities on detection and non-detection. As can be easily checked, if there would there be no loss aversion, i.e.    )(u =    )(u , and the perceived and objective probabilities would coincide, i.e. )1()1(    and   )( , then (5) would be exactly equivalent with (3). Nowadays the violation of the independence axiom within the expected utility framework is better known as framing effect (Tversky and Kahneman, 1986). Preferences over prospects are not independent on how they are presented. Starmer (2000) describes this phenomenon as violation of the description invariance underlying conventional theory. In order to give an example, consider two alternative descriptions of the same situation: Case A: A farmer’s violation of a regulation will be punished with a 1% of the SFP given a probability of 10 per cent of detection (inducing an expected loss of 0.1% in payments). Case B: Having violated a regulation and being detected, a farmer is only eligible to 99.9% of the full SFP 11 . A framing effect would exist if a farmer would behave differently facing one of the two descriptions. Similarly, people might react differently on the probability of a penalty and the potential benefit of a higher price. The example of framing in tax law by Schelling discussed in Tversky and Kahneman (1986) has also a high relevance for the case of cross-compliance and their acceptance among the general public. Obviously, payments if expressed as a reward for 11 Note that the expected value of the payments received for both cases is the same! 19 complying with the law will be judged much more sceptically by the non-farmer audience than payments in exchange for farmers’ efforts to reduce negative environmental effects of production activity. Finally, people may fail to estimate and revise subjective probabilities for themselves, resulting in systematic mistakes. Alm et al. (1992) show that people appear to overestimate the probability of an audit. Similarly, it has been shown that agents subjectively overestimate the expected penalty, as well (OECD, 2004). As pointed out by Starmer (2000), non-conventional approaches of utility theories, incorporating evidence from psychology and sociology, are not able to define one single utility function. Instead of defining one utility function over individual prospects, nonconventional theories assume that individuals base their choices on decision rules or choices. Therefore, we are not able to continue the formal presentation of a farmer’s compliance decision as presented above within the next paragraphs. 12 Psychological literature Part of the reason for a farmer’s compliance that is unexplained by the above utility maximisation framework might be influenced by external and internal factors. Whereas economic theory focuses almost exclusively on external rewards of human behaviour, psychology is mainly concerned with the internal motivation of certain behaviour. Intrinsic motives are qualified through the absence of any apparent incentive (Frey and Jegen, 2001), or “superego” in the terminology of Mazar and Ariely (2006). In this context, Sutinen and Kuperan (1999) distinguish between two groups of thoughts. 12 In this respect we deviate from the approaches developed by Sutinen and Kuperan (1999) as well as Sandmo (2006). The authors include a variable of an individual’s moral standing as a function of the similarity between regulation and individual’s values into their objective function. However, their solutions are still based on the expected utility framework. 20 The first, so-called instrumental perspective, assumes individuals as driven purely by selfinterest. They respond only to changes in the tangible, immediate incentives and penalties associated with non-compliance. Frey and Jegen (2001) summarise sources that explain the possible disincentive effect of external intervention in intrinsic motivation. This effect is known as cognitive evaluation theory in the psychological literature. A beneficial personal motivation, e.g. a positive attitude towards active environmental protection, might be corrupted through the introduction of explicit monetary rewards. Frey and Jegen (2001) term it as a crowding-out effect and present empirical relevance from several economic studies. The second perspective, called normative, emphasizes what individuals consider just and moral, instead of what is in their self-interest. Individuals tend to comply with the law to the extent that they perceive the law as appropriate and consistent with their internalized norms. The key variables determining compliance in the normative perspective are individuals’ perceptions of the fairness and appropriateness of the law and its institutions (Sutinen and Kuperan, 1999). Within this normative perspective, the cognitive theory explains the impact of personal motivation and the level of personal development on an individual’s behaviour. People show pro-social behaviour (Alm et al., 1992). Finally, psychological research has shown that people might behave completely differently in situations of emotional engagement then in situations without being emotionally aroused (“heat of the moment”) (Amir et al., 2005). Put differently, despite the cognitive awareness about how to treat animals or how to apply chemicals correctly, farmers can violate codes of practice in stressful situations or under similar circumstances. Contrary to the factors discussed above, this behaviour might increase non-compliance. Task complexity, an additional argument raised by Carroll (1978), implies that agents are behaving with a ‘limited’ rational at best. This could be especially relevant in the case of cross-compliance with 19 requirements. However, such a concept works better to explain 21 non-compliance instead of a higher compliance than predicted by the expected utility theory. 13 Sociological literature Following Fehr and Fischbacher (2002), the theory of (positive) reciprocity represents quantitatively the most important type of social preferences. The theory implies that people respond positively to some friendly actions without expecting further benefit from their doing so. Fehr and Gächter (2000) distinguish between reciprocity and cooperation as the latter requiring a future interaction of the involved agents and, subsequently, a positive expected value from today’s friendly response. Applying the expected utility theory, agents would expect to show no friendly behaviour if no further interactions are foreseen. Human behaviour develops over time, probably to a larger extent during adolescence, and is shaped by parents, peer groups or stigmas. Stigma might result as a reaction of the surrounding people to revealed non-compliant behaviour (Deffains and Fluet, 2007). However, different peer groups could have quite opposing influences depending on the involvement. For instance, a farmers’ association might have a negative attitude towards a stricter environmental regulation, whereas the surrounding non-farming village community might appreciate it. Similarly, social norms or society’s expectations with respect to animal welfare conditions, soil conservation, water use, application of chemicals or biodiversity differ in many instances from beliefs of a farming peer group. Reference to reciprocity might help to solve this discrepancy. However, reciprocity needs a positive incentive. For instance, Bardsley and Sausgruber (2005) find that a reciprocally motivated agent will contribute to a public good only if he or she receives a benefit from it. The non-binding chosen level of worker’s effort, as another example, is shown to be 13 Garoupa (2003) provides a critical review of behavioural approaches to the law and crime literature. However, his conclusion about the potential contribution of behavioural economics to this literature is rather sceptical. 22 positively related to the principal’s wage offer (Fehr and Fischbacher, 2002). Within the literature about reciprocity, several authors describe the orientation of own behaviour with regard to certain reference groups with the terms ‘conditional cooperation’,‘conformity’or the ‘desire of social approval’(Pretty and Ward, 2001; Fehr and Falk, 2002; Henrich, 2004). Results from various experiments support these concepts. Thus, an individual farmer is expected to be more willing to comply with a regulation if many farmers respect it (Fehr and Falk, 2002) or, similarly, a farmer has a motivation to fulfil social norms or expectations of the farmer’s peer group (Henrich, 2004). Bohnet and Zeckhauser (2004) and Bardsley and Sausgruber (2005) describe conformity as an incentive to adjust own behaviour according to “social comparisons” even without material consequences. 14 Bohnet and Zeckhauser (2004) show that people in ultimatum game experiments do not react to information about randomly chosen offers but revise their action if faced with average offers. That is, information about a certain type of consensus within a group or society acts as an incentive to adjust own behaviour. Generally, it is expected that above mentioned factors will be more influential in cooperative cultures with strong communal norms (OECD, 2004). The influence of a society on an individual’s behaviour is a central element of the field experiments by Henrich et al. (2001). The authors find in 15 small-scale societies involved in agricultural production clear evidence of group-level economic organisation and market integration as a driver of individual behaviour instead of self-interest. More specifically, two dimensions of societal interaction, payoffs to cooperation and market integration, explain the variation in the results of the field experiments to a large extent. Moreover, none of the individual level explanatory variables added a significant contribution to an actor’s behaviour in the games. Henrich et al. (2001) conclude from their results that the economic conditions and social interactions of everyday life shape an individual’s preferences, and that these are 14 Bardsley and Sausgruber (2005) highlight explicitly the difference between the economic concept of conformity and that here applied in calling it “normative conformity”. 23 not exogenous. Turning to empirical support of the theory of reciprocity from cross-country analysis, Frey and Torgler (2007) present several illustrative examples using data from the European Value Survey (EVS). Their econometric results point to a significant relation between the degree of tax evasion in a country perceived by respondents and the respondent’s own readiness to evade taxes. One illustrative example of social motives to comply with regulations requiring a greater quality or higher level of response is the case of the Canadian Best Management Practice (BMPs) programs, which have a similar aim as the EU’s Good Agricultural and Economic Conditions (GAEC) standards. Agriculture and Agri-Food Canada (AAFC) (2006) defines Beneficial Management Practices (BMPs) as “farm management practices that: minimize and mitigate impacts and risks to the environment, by maintaining or improving the quality of soil, water, air and biodiversity; ensure the long term health and sustainability of natural resources used for agricultural production; and, support the long-term economic and environmental viability of the agriculture industry.”As part of the program, farmers attend an Environmental Farm Plan workshop and complete a workbook designed to assess the current state of the farm and identify areas of concern. Then farmers develop an action plan for addressing the areas of concern. The action plan is confidentially reviewed by a group of locally appointed farmers. Once the Peer Review Committee approves the Action Plan, a farmer can participate in the EFP Cost-Share Program that helps cover a portion of the costs of implementing eligible projects from the action plan. Producers are eligible to apply for cost-share incentives through the Canada Farm Stewardship Program, Greencover Canada, and the Canada Water Supply Expansion Program. The federal government covers up to 60% of the cost of implementing eligible practices. Many practices covered through federal cost share programs are also eligible for funding under different provincial cost share programs. As a result up to 90% of the total 24 project cost can be covered by combining federal and provincial funds. However, the coverage varies depending on farmer eligibility, provinces and type of Best Management Practices. Despite the fact that the implementation of Best Management Practices is not mandatory, there has been a relatively high degree of participation. For example, between 2005 and 2007, more than 11,000 of 57,211 Ontario farmers implemented or were in the process of implementation of BMPs. Even though the implementation of Beneficial Management Practices is partly subsidized by the federal and provincial governments, it is not costless to the farmers. As an example, Ontario farmers bore about a third of the cost of implementation of the management practices eligible for funding. Net costs for participating farmers could amount to about €1000 per farm. The reason farmers are prepared to pay these costs is that they recognize the environmental problems and the societal demands to behave as good citizens that each contribute their ‘fair share’ (Fox and Rajsic, 2007). The study by Burton et al. (2008), which analyses the non-economic determinants of adoption of voluntary agri-environmental programmes, provides another empirical illustration of society’s influence on afarmer’s behaviour . The authors highlight the expression of cultural capital, that is to demonstrate oneself to be a ‘good farmer’ as an important motive for a farmer. Following the line of this argument implies to expect a higher degree of compliance if the result of such a behaviour is more visible to the farming community around the farm. Institutional quality and compliance Tyran and Feld, (2006) summarise the discussion about mild (without deterrent sanctions) and severe law (with deterrent sanctions). The above mentioned mechanism of conditional cooperation or the own commitment act as motivation for compliant behaviour in the case of endogenously enacted laws. This type of regulation applies more to quality standards than to 31 reforms aimed at simplification and improving effectiveness, consistency, accountability and transparency of public administration (e.g. the Commission’s White Paper on Governance (EC, 2001, pp. 18-20); all this is part of the better regulation movement (EC, 2005). One element of such assessments is always the analysis of compliance, although until now such types of analysis tend to be done badly. This paper contributes to a better understanding of compliance by exploring and comparatively analysing the different approaches to explain compliance that are found in the social sciences literature. It was argued that the classical economic approach, which interprets compliance as a decision based on a cost –(forgone) benefit-decision, is too limited for a full understanding of compliance behaviour. Moreover, the limits of the expected utility framework underlying this theory were briefly reviewed and contrasted with results from the prospect theory. It was further argued that alongside this utility framework the contributions from the psychological and sociological literature should also be acknowledged. This adds factors such as intrinsic motivations, moral convictions (e.g. obeying laws, stick to one’s given word), social preferences, reciprocity and impacts of peer groups. Moreover, it was argued that institutional quality matters as a factor to understand compliance. The general conclusion does not underline the rather pessimistic statement by Tversky and Kahneman (1992) cited above. We conclude that a theory of choice cannot consist of only one single approach. Whereas one coherent model integrating all these aspects is lacking, at least at this moment, we think that a first priority is to undertake more empirical analyses. These analyses should have a broader focus than deterrence issues. Some suggestions for this are given, although it is recognized that this kind of research will be quite data demanding. However, going along this route should in the end contribute to a better selection of appropriate variables relevant to analyse compliance issues. A better understanding of compliance behaviour is also a crucial element to not necessarily more but better regulation. 32 References Agra Focus, 2008. Cross compliance rules agreed (no. 144, February 2008); p. 8. Agriculture and Agri-Food Canada (AAFC), 2006. http://www.agr.gc.ca/index.html. Al-Nowaihi, A., Dhami, S., Bradley, I.G., 2008. A Note on the Utility Function Under Prospect Theory. Economics Letters 99, 337-339. Alm, J., McClelland, G.H., Schulze, W.D., 1992. Why do People Pay Taxes? Journal of Public Economics 48, 21-38. Amir, O., Ariely, D., Cooke, A., Dunning, D., Epley, N., Gneezy, U., Koszegi, B., Lichtenstein, D., Mazar, N., Mullainthan, S., Prelec, D., Shafir, E., Silva, J., 2005. Psychology, Behavioral Economics, and Public Policy. Marketing Letters 16, 443-454. Andreoni, J., Erard, B., Feinstein, J., 1998. Tax Compliance. Journal of Economic Literature 36, 818-860. Andvig, J.C., Moene, K.O., 1990. How Corruption May Corrupt. Journal of Economic Behavior and Organization 13, 63-76. Bardsley, N., Sausgruber, R., 2005. Conformity and Reciprocity in Public Goods Provision. Journal of Economic Psychology 26, 664-681. Becker, G.S., 1968. Crime and Punishment: An Economic Approach. Journal of Political Economy 76, 169-217. Beedell, J.D.C., Rehman, T., 1999. Explaining Farmers' Conservation Behaviour: Why Do Farmers Behave the Way They Do? Journal of Environmental Management 57, 165- 176. Bohnet, I., Zeckhauser, R.J., 2004. Social Comparisons in Ultimatum Bargaining. Scand. J. Econ. 106, 495-510. Brousseau, E., Farès, M.h., 2000. Incomplete Contracts and Governance Structures: Are Incomplete Contract Theory and New Institutional Economics Substitutes or Complements?, in: Ménard, C. (Ed.), Institutions, Contracts, Organizations, Perspectives from Now-Institutional Economics. Edward Elgar Publisher, Cheltenham, pp. 399-421. Buller, H., 2000. Regulation 2078: Patterns of Implementation, in: Buller, H., Wilson, G.A., Hoell, A. (Eds.), Agri-environmental Policy in the European Union. Ashgate Publishing Ltd., Aldershot, UK, pp. 219-253. Burton, R.J.F., Kuczera, C., Schwarz, G., 2008. Exploring Farmers' Cultural Resistance to Voluntary Agri-environmental Schemes. Sociologia Ruralis 48, 16-37. 33 Carroll, J.S., 1978. A Psychological Approach to Deterrence: The Evaluation of Crime Opportunities. Journal of Personality and Social Psychology 36, 1512-1520. Deffains, B., Fluet, C., 2007. Legal versus Normative Incentives under Judicial Error, in: Centre Interuniversitaire sur le Risque les Politiques Économique et l'Emploi (Ed.), CIRPEE Working Paper No. 07-18, Montréal, p. 33. Dhami, S., al-Nowaihi, A., 2007. Why do People pay Taxes? Prospect Theory versus Expected Utility Theory. Journal of Economic Behavior and Organization 64, 171-191. EC, 2009. Council Regulation (EC) No. 73/2009. Official Journal of the European Union L30/16, Brussels. EC, 2007. Report from the Commission to the Council on the Application of the System of Cross-compliance, in: Commission of the European Communities (Ed.). Commission of the European Communities, Brussels, p. 10. EC, 2005. Implementing the Community Lisbon Programme: A Strategy for the Simplification of the Regulatory Environment, in: Commission of the European Communities (Ed.). Communication of the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, Brussels. EC, 2001. Commission’s White Paper on Governance, in: Commission of the European Communities (Ed.). Commission of the European Communities, Brussels. ECA, 2008. Is Cross-compliance an effective policy?, in: European Court of Auditors (Ed.). Special Report No. 8/2008, Luxembourg. Farmer, M., Swales, V., Jongeneel, R., Karaczun, Z., Müssner, R., Leipprand, A., Schlegel, S., Poux, X., Ramain, B., De Roest, K., Varela Ortega, C., Simó, A., 2007. Exploring the Synergies between Cross Compliance and Certification Schemes, in: LEI (Ed.), Report. Wageningen UR, The Hague, p. 82. Fehr, E., Falk, A., 2002. Psychological Foundations of Incentives. European Economic Review 46, 687-724. Fehr, E., Fischbacher, U., 2002. Why Social Preferences Matter: The Impact of Non-Selfish Motives on Competition, Cooperation and Incentives. Economic Journal 112, C1-C33. Fehr, E., Gächter, S., 2000. Fairness and Retaliation: The Economics of Reciprocity. Journal of Economic Perspectives 14, 159-181. Fox, G., Rajsic, P., 2007. Estimates of Farmer Costs of Current Environmental Programs and Potential Abatement Costs for Canadian Cereals Growers, in: Food, Agriculture and Resource Economics (Ed.), mimeo. University of Guelph, Guelph, Ontario. 34 Frey, B.S., Benz, M., Stutzer, A., 2004. Introducing Procedural Utility: Not only What, but also How Matters. Journal of Institutional and Theoretical Economics 160, 377-401. Frey, B.S., Jegen, R., 2001. Motivation Crowding Theory. Journal of Economic Surveys 15, 589-611. Frey, B.S., Torgler, B., 2007. Tax Morale and Conditional Cooperation. J. Compar. Econ. 35, 136-159. Gardner, B., 2003. U.S. Food Quality Standards: Fix for Market Failures or Costly Anachronism? Amer. J. Agr. Econ. 85, 725-730. Garoupa, N., 2003. Behavioral Economic Analysis of Crime: A Critical Review. European Journal of Law and Economics 15, 5-15. Harrison, G.W., Rutström, E.E., 2009. Expected Utility Theory and Prospect Theory: One Wedding and a Decent Funeral. Experimental Economics 12, 133-158. Hart, R., Latacz-Lohmann, U., 2005. Combating Moral Hazard in Agri-environmental Schemes: A Multiple-agent Approach. Eur Rev Ag Econ 32, 75-91. Henrich, J., 2004. Cultural Group Selection, Coevolutionary Processes and Large-scale Cooperation. Journal of Economic Behavior and Organization 53, 3-35. Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., McElreath, R., 2001. In Search of Homo Economicus: Behavioral Experiments in 15 Small-Scale Societies. Amer. Econ. Rev. 91, 73-78. Jain, A.K., 2001. Corruption: A Review. Journal of Economic Surveys 15, 71-121. Jongeneel, R., 2006. Deliverable 6: Overview of Certification Schemes in 7 EU Countries and 3 non-EU Countries, Facilitating the CAP Reform: Compliance and Competitiveness of European Agriculture. LEI, The Hague. Jongeneel, R., Brouwer, F., Farmer, M., Müssner, R., de Roest, K., Poux, X., Fox, G., Meister, A., Karaczun, Z., Winsten, J., Ortéga, C., 2007. Compliance with Mandatory Standards in Agriculture: A Comparative Approach of the EU vis-à-vis the United States, Canada and New Zealand, in: LEI (Ed.), Report No. 6.07.21. Wageningen UR, LEI, The Hague. Kahneman, D., Tversky, A., 1979. Prospect Theory: An Analysis of Decision under Risk. Econometrica 47, 263-292. Kaufmann, D., Kraay, A., Zoido-Lobaton, P., 1999. Aggregating Governance Indicators, in: World Bank (Ed.), Policy Research Working Paper No. 2195. World Bank, Washington D.C., p. 39. 35 Levi, M. 1998. A State of Trust, in: Braithwaite, V., Levi, M. (Eds.), Trust and Governance. Russell Sage Foundation, New York, pp. 77-101. Machina, M.J., 1987. Choice Under Uncertainty: Problems Solved and Unsolved. Journal of Economic Perspectives 1, 121-154. May, P.J., 2005. Compliance Motivations: Perspectives of Farmers, Homebuilders, and Marine Facilities. Law and Policy 27, 317-348. Mazar, N., Ariely, D., 2006. Dishonesty in Everyday Life and Its Policy Implications. Journal of Public Policy and Marketing 25, 117-126. McEachern, C., 1992. Farmers and Conservation: Conflict and Accommodation in Farming Policies. Journal of Rural Studies 8, 159-171. Morris, C.T., Winter, M., 2002. Barn Owls, Bumble Bees and Beetles: UK Agriculture, Biodiversity and Biodiversity Action Planning. Journal of Environmental Planning and Management 45, 653-671. Moxey, A., White, B., Ozanne, A., 1999. Efficient Contract Design for Agri-Environmental Policy. J. Agr. Econ. 50, 187-202. OECD, 2004. Economic Aspects of Environmental Compliance Assurance, in: OECD (Ed.), OCED Global Forum on Sustainable Development, Paris, p. 179. Ostrom, E., Basurto, X., 2011. Crafting Analytical Tools to Study Institutional Change. Journal of Institutional Economics forthcoming (DOI: 10.1017/S1744137410000305). Pretty, J., Ward, H., 2001. Social Capital and the Environment. World Development 29, 209- 227. Rabin, M., Thaler, R.H., 2001. Anomalies: Risk Aversion. Journal of Economic Perspectives 15, 219-232. Sandmo, A., 2006. The Theory of Tax Evasion: A Retrospective View. National Tax Journal 58, 643-663. Schmidt, U., Starmer, C., Sugden, R., 2008. Third-generation Prospect Theory. Journal of Risk and Uncertainty 36, 203-233. Schoemaker, P.J.H., 1982. The Expected Utility Model: Its Variants, Purposes, Evidence and Limitations. Journal of Economic Literature 20, 529-563. Slemrod, J., 2007. Cheating Ourselves: The Economics of Tax Evasion. Journal of Economic Perspectives 21, 25-48. Starmer, C., 2000. Developments in Non-Expected Utility Theory: The Hunt for a Descriptive Theory of Choice under Risk. Journal of Economic Literature 38, 332-382. 36 Sutinen, J.G., Kuperan, K., 1999. A Socio-economic Theory of Regulatory Compliance. International Journal of Social Economics 26, 174-193. Tversky, A., Kahneman, D., 1986. Rational Choice and the Framing of Decisions. Journal of Business 59, S251-S278. Tversky, A., Kahneman, D., 1992. Advances in Prospect Theory: Cumulative Representation of Uncertainty. Journal of Risk and Uncertainty 5, 297-323. Tyran, J.-R., Feld, L.P., 2006. Achieving Compliance when Legal Sanctions are Nondeterrent. Scand. J. Econ. 108, 135-156. Williams, R.E., 1997. The Political Economy of the Common Market in Milk and Dairy Products in the European Union, in: Food and Agricultural Organisation (Ed.), FAO Social and Economic Paper No. 142, Rome Winter, S.C., May, P.J., 2001. Motivation for Compliance with Environmental Regulations. Journal of Policy Analysis and Management 20, 675-698. Yaniv, G., 1999. Tax Compliance and Advance Tax Payments: A Prospect Theory Analysis. National Tax Journal 52, 753-764. 37 Table 1: Best estimates of degrees of compliance for selected regulations and EU member states *) France Germany Italy Netherlands United Kingdom Spain Environment Bird and Habitat Directives n.a.; probably very high management plans not yet in place in most areas management plans not yet in place in most areas very high very high very high Protection of groundwater not very high for exhaustible oils very high very high high very high very high Sewage Sludge Directive very high very high very high very high very high very high Nitrate Directive dairy farmers low and beef farmers extremely low not high extremely low; national implementation tool place only recently low, (mainly due to recent change in the regulations) very high high Identification and registration Identification and Registration of bovine animals high, but not always within 7 days very low n.a.; databank working since 2005 very high low very high Identification and Registration of ovine and caprine animals extremely low; new regulation since 2005 very low n.a.; databank working since 2005 high very high very high Public, Animal and Plant Health Food Traceability and Food Safety n.a. n.a. n.a. high n.a. n.a. 38 Notification of diseases high, no precise estimate available n.a n.a. high n.a.; since 1 January 2006 imposed n.a. Animal welfare expected to be high expected to be high expected to be high expected to be high expected to be high n.a. Good Agricultural and Environmental Condition n.a. very high n.a. not high very high very high Source: (Jongeneel et al., 2007)), p. 57. *) Because of uncertainties and measurement problems only general characterizations are given. The following legend was used: Compliance is considered ‘very high’ if the degree of compliance is greater than 95% (95% of the farmers or more are fully compliant). Compliance is labelled as ‘high’ in case the degree of compliance is in the interval 90%-95%. Compliance is labelled as ‘not high’ if compliance rates were in interval 80%-90%. Compliance was labelled ‘low’ when the degree of compliance was in the interval 70%-80%. It was labelled as ‘very low’ when the degree of compliance was in the interval 40%-70%. Finally, it was labelled to be ‘extremely low’ in case of compliance rates below 40%. n.a. –not available. 39 Table 2: Determinants of compliant behaviour according to different strands of the literature Theory/ Literature strand Determinants Expected utility theory Chance of getting caught/ penalised Size of penalty Individual’s degree of risk aversion Prospect theory Gains and losses with respect to reference point Diminishing sensitivity Loss aversion Psychological literature Internalised attitudes “superego”/ personal moral beliefs Pro-social behaviour Sociological literature Reciprocity with principal’s action Conformity with peer group Institutions Trust in institutions in general Satisfaction with institutional framework – “against all” Source: Own presentation 40 Annex A: Case of higher input requirements and non-continuous requirements Assuming a regulation which requires a farmer to use more labour than would be supplied otherwise, denoted r )};,(|{ trfMax xq,wxpq  with   trq ;,wp,q  and   trx ;w,p,x  (A-1) Examples from the cross-compliance catalogue are animal registration, especially double ear tags for bovine, ovine and caprine animals, and all documentation tasks. It still holds that     ttr ;;, wp,wp,   . If the farmer violates the regulation and will be detected subsequently, a penalty g might be imposed. In contrast to the previous, the penalty is not a direct function of the degree of violation. So if detected    otherwise0 if0 )( rr rg and rr r g r g     0,0 2 2 where it is implicitly assumed that g(.) is a continuous function and differentiable for all rr  . The decision under expected utility theory becomes:           trrgtr ;1;, w,p,wp,   (A-2) Subsequently, the first order condition for the optimal r level is     r rg r tr      ;,wp, (A-3) where the inequality is due to the discontinuity allowed for in the punishment function for the case rr  .