Wednesday, 7 December 2016
It is time to move away from traditional subsidy based agricultural value chains and look for more efficient methods. Some organisations are already following these, yet to my surprise some of this information is new to exporters.
A couple of months back I had the privilege of addressing a forum at the National Chamber of Exporters (NCE) and my talk was on value chains, transaction costs and Smallholder Quality Penalty (SHQP). It made me realise, especially based on the discussions I had with exporters after the talk, that many of them do not have a clear idea on the format of their value chains.
They were making profits and they knew how to excel at it however, I saw a knowledge gap in terms of evaluating costs and benefits among different value chain models, where the opportunity cost is being generated and especially a theoretical view behind the implementation and failure of contracts. Therefore, my article on 27 October was focused on the transaction costs. However, with this article I want to develop a discussion around economic concept of ‘Principal-Agent Problem’.
I want to answer two main questions (1) Can the contracts in agricultural value chain be explained using the Principal-Agent Problem? and (2) if so, what can be done to come out of the problem. I want to start with some back ground to the contract farming, looking at the characteristics and economic theory behind. Later I will focus on the Principal-Agent Problem and remedies against it.
Contract farming: Some important concepts
Contract farming is a form of vertical integration within agricultural commodity chains, such that the firm has greater control over the production process, as well as the quantity, quality, other characteristics and the timing of what is produced. The conventional approach to vertical integration has been for firms to invest directly in production through large-scale estates or plantations (especially for commodities such as tea, coffee and rubber). Contract farming, in its various forms, allows a degree of control over the production process and the product without the firm directly entering into production.
Literature identifies several forms of contract farming:
(1) Centralised model: This is where a firm (often a large processor) contracts a large number of farmers, with strict quality requirements and quantity targets. Products suited to this contracting model require substantial processing prior to retail — for example, tea, coffee, cotton, sugarcane milk and poultry
(2) Nucleus-estate model: in here the firm (again, often a processor) enters the production node through an estate or plantation but also contracts with independent producers (for greater volumes, or for seed). This model is often used for perennial crops and is often the preferred model utilised with resettlement or transmigration programs (such as palm-oil production). Thus, this is the contract-farming model that utilises out- growers from a central estate.
(3) Tripartite model: This is a joint venture (between a public entity and a private firm) enters into a contract with farmers. This model can involve national and/or local government. Due to government involvement, contracting based on this model could potentially be politicised.
(4) Informal model: Here the smaller firms or traders enter into annual agreements, often on a verbal basis, with a limited number of farmers, frequently for fruit and vegetables that require minimal processing. As firm size is usually small, the success of such initiatives partly relies on the extent to which other providers (such as the state and/or NGOs) can offer inputs, such as extension and credit. Due to its non-formal nature, this model often suffers from extra- contractual side-marketing.
(5) Intermediary model: In here the firm sub-contracts interaction with the farmers to an intermediary, such as a farming committee or a trader. This model is popular in Thailand and Indonesia, and that the increased distance between firm and farmer decreases the degree of control that the firm has over the process and the product (one of the main reasons for contract farming).
Contract farming is something that has a deeply vested theoretical background. We generally attribute several economic theories when we talk about contract farming. First comes the life-cycle theory. Starting with Adam Smith’s dictum that “the division of labour is limited by the extent of the market”, Stigler’s life-cycle theory posits that industries tend to be more vertically integrated in the early stages of their development (since specialisation takes place when the size of the market supports economies of scale).
A more common approach to understanding contract-farming focuses on transaction costs. The starting point for this perspective is Coase’s (1937) simple question: why do firms exist? Coase’s answer is: to minimise the transaction costs of exchange. Convention theory focuses on the quality attributes that products exhibit. In well- established markets with perfect information, prices are assumed to reflect all relevant quality attributes. But if quality requirements are particularly exacting, or product quality is especially uncertain, certain quality conventions help to facilitate exchange. Therefore contract farming is based on the “Life-cycle theory” and the “transaction cost theory”.
What is a Principal-Agent Problem? What is the impact?
The crop contracting relationship can be modelled as a principal-agent problem where the agent (farmer) is growing a specialty crop that will be owned and exclusively used by the principal (processor). Asymmetric information reflects the processor’s uncertainty about the farmer’s efforts and performance under the contract. The “real” ex-post contract costs are the private information to the farmer who can truthfully disclose them to the principal or not.
The processor (principal) cannot completely observe the managerial effort spent by the farmer on the contracted activity and the care taken to reach the desired crop quality. The principal also cannot observe consistently whether the specific asset was used appropriately and exclusively to the contract (i.e., moral hazard problems).
Monitoring costs are assumed to be relatively high and quality measurement techniques are either costly or imperfect. As a result, the contract terms should be set so that the incentives and efficiency considerations are consistent with the risk-bearing capacities of the agent. The effects of these factors on contract choice, along with the farmer’s personal and business characteristics (e.g., risk attitudes, leverage, and farm size), can be approximated by his or her choice of cost and risk-sharing rates with contractors, pricing options, contract length, financing arrangements, and other contracting terms. I will talk about these points bit later when I propose ways to get around the Principal-Agent-Problem.
In simpler terms the Principal-Agent Problem occurs when the incentives between the principle and the agent are misaligned. As explained before, by definition the agent has been selected for his specialised knowledge and the principal can never hope to completely check the agent’s performance (for the most part). Delegation of a task to an agent who has different objectives than the principal is problematic when information about the agent is imperfect. This is the essence of “incentive questions”.
Let me spend little time in explaining this further. If the agent had a different objective function but no private information, the principal could propose a contract which perfectly controls the agent and induces the latter’s actions to be what he would like to do himself in a world without delegation. Again, incentives issues would disappear. The starting point of incentive theory corresponds therefore to the problem of delegation of a task to an agent with private information. This private information can be of two types:
(1) either the agent can take an action unobserved by the principal, the case of moral hazard or hidden action or
(2) the agent has some private information about its cost or valuation that is ignored by the principal, the case of adverse selection or hidden knowledge.
When this private information is a problem for the principal it is important to ask, what is the optimal way for the principal to cope with it? In addition another type of information problem has also been raised in the literature, the case of non-verifiability. This is where the principal and the agent share ex-post the same information but no third party and, in particular, no Court of Justice can observe this information.
The quality vs quantity is also an interesting issue that comes with the Principal-Agent-Problem. As mentioned earlier, the grower’s (agents) performance can often be measured fairly accurately in some efforts. In others, however, available performance measures that may be used to provide explicit incentives to the grower may not even exist.
For example, a grower may have to produce a certain amount of input that is easily measurable, but he may also have to make sure that the quality of input is high, which may be more difficult to measure. As a result, a grower that seeks to maximise his own income may not act in the best interest of the processor.
The multiple efforts performed by the grower in quantity and quality act jointly to determine the outcome. This is in contrast to, for example, a multi-task employee whose tasks are often modelled as additively separable in their contributions to the outcome. The processor has to consider the trade-off, often present, between the quality of a good and the quantity produced, when decides the compensation scheme.
In the standard vertical product differentiation model, it is usually implicitly assumed that quality and quantity are independent choices. That is, at any set level of quality, a grower is free to produce as much as he desires. However, in many goods, production can only be increased at the expense of lower quality. Indeed, the quality-quantity trade-off is an important research issue that has received substantial attention.
What would be the solutions?
In general, contracts should be contingent only on variables that the principal can more precisely observe and control the agent’s actions. Once such variables are identified, it is possible for the principle to come up with an easy incentive scheme to promote the good behaviour.
Let’s take a simple example. In environmental friendly/eco-friendly agriculture, especially the ones that are in the export value chain the biggest hurdle is the certification. Since most organisations that work with organic exports are earning a better premium for their products, there is an incentive for them to invest in quality checking. Therefore, a popular variable that is being looked at is the “maximum and minimum allowable limits of chemicals”. Such a variable allows the principal to design an effective incentive stricture.
Any farmer who is above the maximum allowable limits of chemical applications are rejected or banned from entering in to a contract. However, such a measurement will allow incentivising a farmer who would do extra to keep the allowable chemical levels down to the minimum. Therefore in such a situation closer to the minimum allowable chemical level, higher the incentive for the farmer.
Let’s take another example. For pepper producers, there is no incentive structure for using a thrashing machine as oppose to thrashing by foot. Obviously the ones that are being thrashed by a machine would have lesser contaminations and would have a higher quality/appearances. Therefore a company that is in contract with pepper farmers could monitor/measure the use of machinery in thrashing. The ones that are using a thrashing machine can be easily incentivised.
The point here is that an incentive would motivate the agent to leave out un-necessary risks of producing low quality produce. This is a suitable solution for the “moral hazard problem “associated.
As mentioned before the principal agent problem can result in adverse selection as well. Prior to a contract an agent can deceive the principal. To eliminate such an issue it is possible to implement a thorough screening process.
Let’s take an example. It is safe to assume that all organic farmers do not practice “pure organic methods”. Lack of enforcement on certifications has allowed farmers to put fertiliser and other chemicals and still come out as “organic farmers”. Though this might not be public information, it is possible to capture their actions through an easy farmer based questionnaire.
I have worked with organic paddy farmers in many districts of Sri Lanka and I have first-hand experience in seeing farmers put little amount of inorganic fertiliser and still call it “organic food”. The best way to go about this is to put in place a detection system. However it might be costly. It will be cheaper to implement a pre-screening process and identify the suitable farmers to enter in to a contract.
Value chain studies and information from the quarantine services show that, regardless of the number of enforcements put in place, still there are farmers who would try to deceive the agent. Needless to say almost all of them are being caught during quarantine inspections (I am not ruling out the possibility of bribing the way out of such strict inspections as well. I am sure one could find enough evidences all around the world on that). However, the literature suggests the importance of a monitoring and a reporting system in order to get around such matters.
A possible solution is an ICT based traceability system. An ICT based traceability system can establish an end-to-end communication system between the principal and the agent, and any wrongdoing can be traced down to a particular agent very easily.
Let’s take an example. Exporting to the European region with the Good Agricultural Practices (GAP) certification is on the rise. The program is being implemented with the heavy involvement and the leadership of the Department of Agriculture. However, there are still situations (though the incidences have reduced drastically over the years).
An ICT based traceability system would be an ideal solution for such a situation where each and every agricultural produce can be traced down to the individual farmer. This also is an opportunity for the principal to provide an incentive/disincentive for the farmer for maintaining/degrading the quality (whether it is to be an incentive or a disincentive is up to the principal).
As the final point, it is also possible to promote joint actions when enforcing contracts. Rather than entering into contracts with a single agent, the principal can work with a collective of agents. Such actions will benefit the principal since quality checks can be implemented at farmer organisation level. A farmer organisation has a reputation to maintain. Once the principal enters into a contract with a farmer organisation the stakes are high. It is not any more about an individual farmer. Therefore, a voluntary monitoring system will be implemented within the farmer organisation. In addition, working as a farmer organisation would bring in many benefits to the agents as well, especially achieving economies of scale and bargaining power over price.
(Dr. Chatura Rodrigo is an agriculture and environment economist. The author can be reached at chatura_rodrigo@yahoo.com and 94 77 986 7007.)
2016-12-06
2016-12-06
an elite that was consummately skilled and able in furthering its own agenda, its own interests, both social and economic, and safeguarding the privileges that were granted to them by the departing British Raj. An elite whose education was either in Oxford and Cambridge or Colombo schools, whose virtual first language was English, whose commuting was done in luxury cars, in other words, a class of people who were absolutely divorced from the common man in the country was charged with managing the socio-economic, cultural development of the nation. This elite clique was not limited to the leading community at the time, Sinhalese; it was the same for Tamils as well as Muslims and Burghers. As Kumari Jayewardene wrote in her celebrated exposition of Sri Lanka’s rise of Mercantilism and the elites in the twentieth century, ‘Nobodies to Somebodies’: ‘Members of another group of Sri Lankans, who were to form an important part of the emergent 19th century bourgeoisie, were landowners, whose holdings provided them with a means of accumulation and later, a basis for expanded growth in the plantation era. Just as the monopolistic policies of the Dutch and the British had located a stratum of officials in the cinnamon industry and endowed them with a basis for growth, their administrative policies also created a group of Sri Lankan officials, called Mudaliyars. Peebles (1973:1) has defined them as an economic and social status group “mediating between the alien rulers and the bulk of the indigenous population” performing functions that the foreign rulers were “unable or unwilling to perform”.’
