COOPERATIVES AND PRIVATE VENDORS IN THE
UGANDAN DAIRY INDUSTRY
What is the rationale of dairy farmers when selling to
cooperatives or private vendors?
Master Thesis Development Economics
Sophie van den Ende
Supervisor: Prof. M. Pradhan
University of Amsterdam
Student number: 10025707
24 -‐ 07 -‐ 2015
ABSTRACTThis paper researches the underlying motives of dairy smallholders when selling to cooperatives or private vendors. To introduce the topic, the theoretical background researches the suitability of contract farming in the dairy market in developing countries. It therefore looks at the theory of contract farming in general and the characteristics of milk and the market players. The paper shows that the dairy market fits the optimal environment for contract farming. The empirical section starts by conducting hypotheses on the underlying motives of farmers’ in choosing to sale their milk to cooperatives or private vendors. These hypotheses are based on the following features: the offered price of milk, the need for income risk
spreading, the volatility of milk production and the convenience of delivering milk to the cooperative. In order to obtain the results, and OLS and Tobit regression have been conducted on data from a household survey of Ugandan dairy farmers. As well the OLS as the Tobit regression disprove the hypotheses: only the estimator on the price of milk has a significant effect in the OLS regression. The result section clarifies this insignificance by elaborating on the statistical methods, the hypotheses and the used data. Although the first part of this paper shows that a coopertive system is appropriate in the dairy market; the emperical section can’t conclude on the underlying motives of farmers when selling to a coopertive.
TABLE OF CONTENTS
1. Introduction p.3
2. Theoretical background p.4
2.1. Contract farming p.4
2.1.1. Efficiency of contract farming p.5 2.1.2. Characteristics of the market and its’ players p.6 2.2. Dairy industry and contract farming p.7 2.2.1. Characteristics of milk p.7 2.2.2. Suitability of contract farming in the milk industry p.8 2.2.3. Dairy market in developing countries p.9 2.3. Effect of contract farming on farmers p.10 2.4. Dairy cooperatives in Uganda p.11
3. Methodology p.12
3.1 Data on Ugandan dairy farmers p.13 3.2 Hypotheses and descriptive statistics p.14 3.2.1. Price of milk p.14
3.2.2. Income risk p.15
3.2.3. Volatility of milk production p.15 3.2.4. Convenience of delivering milk p.16 3.3 Statistical methods p.17
4. Results p.19
4.1. Regression results p.19
4.1.1. OLS results p.19
4.1.2. Tobit results p.19
4.2 Concerns regarding the results p.21
4.2.1. Statistical methods p.21
4.2.2. Hypotheses p.22
4.2.3. Data on Ugandan dairy farmers p.22
5. Conclusion p.23
6. References p.25
Appendices I, II, III, IV and V p.27
1. INTRODUCTION
For decennia smallholders in rural Africa face problems with raising their income to a sufficient level. So far, the academic research on these issues focused mostly on increasing the production possibilities of farmers but this scope is shifting towards a focus on enlarging the market access (Shepherd, 2007). The smallholders in developing countries have to cope with several factors of market failure, which increase the costs of successfully participating in the market (Markelova, Meinzen-‐Dick, Hellin, & Dohrn, 2009). Contract farming and cooperatives are institutions that have shown to mitigate these costs for particular sectors and countries (Holloway, Nicholson, Delgado, Staal, & Ehui, 2000). They support as well farmers as buyers by being an intermediate form of industrial organization. They strengthen the relationship between farmer and buyer, a relative increase in comparison with a spot market but not to the level of tightness as in vertical integration (Grosh, 1994).
Cooperatives have been present in Uganda since the early 1900’s and by now there are about 10,000 cooperatives, divided over 40 cooperative unions. This research focuses on cooperatives in the dairy industry, a sector characterized by cooperatives, as there were 165 active dairy cooperatives in 2009 (Kyazze, 2010). The data supporting this research has been collected in context of a study of the Dutch Ministry of Foreign Affairs and contains information about dairy farmers in central and western Uganda (AIID, PwC, 2014).
A quick look at the data shows that the farmers who are a member of cooperative don’t
necessarily only sell to the cooperative but also to private vendors or direct consumers. This leads to the main research question of this study: what is the rationale of dairy farmers when selling to cooperatives or private vendors? This study will increase the understanding of farmers’ attitude towards cooperatives and private vendors. A further motivation is based on the problems that cooperatives face when trying engage in a long-‐term relationship with a farmer; this research shows methods to increase the loyalty of farmers (Vargas Hill, Temu, & Torero, 2012). To achieve answers to this research question, several hypotheses have been generated, based on the price of milk, the need for income risk spreading, the volatility of milk production and the convenience of delivering milk. These hypotheses have been tested on the data to research the underlying motives of smallholders when selling to cooperatives or private vendors.
The sequel of this thesis has the following structure. The theoretical background will focus on contract farming in general and its suitability in the dairy market in developing countries, based on the characteristics of milk and the market players. The methodology describes the data and the rationale of the hypotheses, accompanied with some descriptive statistics, to introduce the statistical model and methods that will be used. The results section shows the outcomes of the empirical research and discusses some concerns with the statistical methods, the validity of the hypotheses and the quality of the data. Lastly, the conclusion will give a brief summary of the paper and will then provide a concluding answer to the research question.
2. THEORETICAL BACKGROUND
All markets, whether in developed or developing countries, demand some kind of vertical coordination whereby the demand and supply can be matched. Vertical coordination can help a farmer, the supplier, in knowing what kind of harvest he can produce best at which price. On the demand side, vertical
coordination can help a factory in informing what products are available for which price. Hence vertical coordination deals with the relationship between demand and supply, which communicate with each other by carrying out transactions. The New Institutional Economics has analyzed the transactions and concluded that they are often complicated by the four following problems: imperfect information, limited ability to process information, dishonesty and transaction-‐specific investments (Williamson, 2000). It is hard for the market itself to deal with all these problems and to create an ideal environment. So informal and formal economic institutions have been established to address these problems by for example creating standards, trade associations and credit bureaus. But even with support of these institutions, markets face problems concerning the transactions between the demand and supply side. The markets in developing countries suffer relatively more from these problems due to poor economic institutions and governments. This is mainly applicable to the agricultural market as its characteristics increase the importance of good institutions: the agricultural products are often highly perishable, the production is geographically dispersed and farmers have limited access to resources and new information (Minot & Sawyer, 2014). Next to these, other characteristics that make it more sensitive to vertical coordination problems are: the fluctuation of supply due to the different seasons, the delayed response of supply to demand and a wide variation in quality (Minot N. , 1986).
The focus in this literature overview will be on the characteristics of the dairy market and its players that determine its suitability for contract farming. This section will start with an overview of the theory of contract farming in general. After an elaboration on the dairy market and contract farming, the research of effect of contract farming on farmers will be discussed. To conclude, and in anticipation of the empirical section, the history and current status of contract farming in Uganda will be described.
2.1 CONTRACT FARMING
Several levels of vertical coordination are embedded in today’s markets; from a low level of vertical coordination in spot markets to complete vertical coordination in vertical integration. The level of vertical integration in spot markets is minimal: the relationship between the buyer and seller only consists out of the transaction; there are no further commitments. The price is therefore the coordinating factor and is composed of the quality, quantity and timing. On the other hand, vertical integration has a maximal level of vertical coordination; production and processing are implemented within the same corporation (Minot & Sawyer, 2014). Contract farming lies between these two extremes: the production is carried out on basis of agreements between the producer and buyer (Grosh, 1994). Another way of describing contract farming is that it combines the advantages of large-‐scale production with the advantages of a small-‐scale production; there is improved access to financial resources, information about production methods and diminishing risk while the labor costs stay low and the incentives of the workers strong (Minot & Sawyer, 2014).
The agreements in contract farming can be settled on the characteristics of the product, on the methods of production, on delivery time or location. There is a wide variety in the contract made by buyers and producers; the contracts can differ by type of commitment, degree of formality and method of price determination and payment (Grosh, 1994). Within the different types of commitment, a contract can be based on market-‐specifications, resource provisions or production management techniques. On the price level a contract can have a fixed price, a formula price or a split payment system (Minot & Sawyer, 2014).
2.1.1. Efficiency of contract farming
Contract farming by itself is not the most efficient method of organizing transactions between the supply-‐ and demand-‐side, as both sides have to pay costs to participate in the commitments made. This makes that contract farming is only interesting and efficient if the increase in income exceeds the costs; this section lists the costs of the buyer and farmer to introduce the characteristics of the market which determine the efficiency of contract farming.
For buyers the costs are mostly the fixed costs of organizing the contract design: the negotiation process with the farmers, the distribution of inputs and the collection of the good. The negotiation process asks for field agents first to search for appropriate and willing farmers and secondly to discuss the
contracts content with these farmers. Some buyers offer to supply input to the contracted farmers to support a more stable production or higher quality of the goods, but the purchase and distribution of these inputs increase the costs for the buyer. The costs of contract farming have a more indirect effect on the finance of the farmers; they affect the autonomy and flexibility of the farmer. But it might also directly affect the income of the farmer: if the farmer agrees to sell a specified amount with a certain quality level to the buyer, he loses the liberty to change from buyer if the circumstances change, which might
occasionally lead to an increase in costs or decrease in income. These costs make that contract farming is not efficient for every producer, buyer or sector; the increase in profit has to exceed the increase in costs (Minot & Sawyer, 2014).
A simple scheme that determines the usefulness of contract farming is based on coordination requirements and scale complementarity. Scale complementarity measures the level of similarities between the production and processing procedure. If the two processes are very different and demand different techniques and equipment, then there is a low level of scale complementarity. In this case a low level of vertical integration is more appropriate, and the seller will conduct the production process and the buyer will carry out the processing process. The coordination requirements refer to the necessary level of organization and planning between the producing and processing party on for example the quality, methods of transport or state of delivered good. Vertical integration is more into place if production and processing procedure demand coordination to be able to adjoin one another. Thus, according to this theory, contract farming is most likely if a product demands a high level of coordination but works best on a low level of scale complementarity. Following this reasoning, products that are perishable, processed, exported, or labor intensive, or that need large inputs and careful husbandry, are more suitable for contract farming (Minot, 1986).
2.1.2. Characteristics of the market and its players
The characteristics of the product that increase the efficiency of contract farming, as described in the previous section, are indirectly included in the following more extensive theory on the effectiveness of contract farming. In this theory it depends on the efficiency of the market itself and on the characteristics of the product, of the buyer and of the demanding market whether it is worth to invest in contract farming.
Setting up contracts demands high fixed costs, for among others the negotiation process with the farmers, the establishment of collection centers and the distribution of inputs. The buyers in contract farming are therefore usually large-‐scale processors or exporters as they have the financial resources to afford these costs. In addition, these companies have better access to new market information, resources, credit and knowledge about production methods. The accessibility to these factors can increase the productivity and turnover of companies, but small farmers usually don’t have the connections or
overview. It is therefore beneficial for both parties if the buyer is a large-‐scale company who can support the producing smallholder by sharing knowledge or capital.
The characteristics of the destination market also influence the suitability of contract farming; a quality sensitive and food safety demanding final market increases the usefulness of commitments between producer and buyer. As well the quality sensitivity as the food safety benefit from a higher level of coordination between the producer and processor; more coordination leads to better agreements on the quality standards and can fasten the transportation process. A processor has better control over the production process when there is a contract with the farmer; this control could be necessary to provide the end consumers with a product that meets their demand.
The transaction costs of a product also influence the suitability of contract farming. If the transaction costs of a product are low, then the costs of contract farming will be likely to be higher than the increase in efficiency and so it is more appropriate to sell the product in a spot market. The
transaction costs are usually low if a good is homogenous and nonperishable, with an easily observable quality and a well-‐known production process.
The products that have higher transaction costs and call for a higher level of vertical integration have almost opposite characteristics. These goods are featured by a high perishability, a technical difficult production, a high value to bulk ratio and an economically important quality variation. If a good is highly perishable, then the timing of production, transport and processing is crucial for the quality and durability of the product. It therefore benefits from a high level of coordination between the seller and buyer. If the production process is technically difficult, a farmer can benefit from specialized inputs, technical expertise or other resources from a (large-‐scale) buyer. This buyer can also provide the smallholder with credit, allowing him to have the financial liquidity to buy agricultural inputs at planting time. Another feature of goods in contract farming is a high value to bulk ratio; more value squeezed into a less volume. High value-‐bulk ratio products are more likely to cover the additional costs of contract farming. The last feature of these goods is an economically important quality variation; goods can distinguish themselves by a higher quality or different production method (organic or more sustainable). If this is the case, then the higher asking price can cover for the added costs of the production and of contract farming. In this
situation, vertical coordination is necessary to incentivize the farmers to improve or change their production process (Minot & Sawyer, 2014).
In the next paragraph, the literature review will elaborate on the features of the milk industry and the consequential suitability for contract farming.
2.2. THE DAIRY INDUSTRY AND CONTRACT FARMING
The dairy market in developing countries focuses mostly on the production and processing of raw milk. This means that the players in the market have to deal with the features of milk such as its perishability, liquidity and visually elusive quality. This part of the paragraph will elaborate on the characteristics of milk to introduce the elaboration on why the dairy market is suitable for a contract farming system.
2.2.1. Characteristics of milk
One the most distinguishing features of milk is its perishability, it complicates the process and creates a high risk of spoilage. The components of milk, low acidity but a high level of nutrients, make it attractive to bacteria, which eventually could even cause food poisoning (Fellows & Hampton, 1992). The perishability influences all the different aspects of the daily affairs of the milk industry; it affects among others the transport, storage and industry risks. A consequence of the low durability of milk is that, quickly after milking, the milk has to be brought to a cooling storage, or directly to the market or
consumer. This has to happen fast to preserve food security and quality and to lower the post-‐harvest risk (Birthal, A., Tjongco, & Narrod, 2008). The transport is also complicated by another characteristic of milk; it consists for more than 85% out of water, which makes it heavy and increases the transportation costs. Next to that, it limits the quantity of milk that a farmer can transport if he doesn’t have vehicular transport (Staal, Delgado, & Nicholson, 1997). If the market or the consumer is too far away for daily transport, a farmer would need either need a storage location or a processing equipment at the farm to keep the milk fresh.
The quality of milk is another influencing feature in the dairy industry; milk quality is volatile and often hard to ascertain (Staal, Delgado, & Nicholson, 1997). But the quality of the raw milk is essential for the companies that process the milk into products as yoghurt, cheese and infant formula. As the quality is depending on as well exogenous factors, for example the quality and quantity of the feed for the cattle, as exogenous and environmental factors, farmers can only assure the specific quality to a certain point (Saenger, Torero, & Viceisza, 2013). This uncertainty leads to monitoring costs of the milk for the processing company and to potential losses for the farmer, trader and processor if the milk gets rejected (Staal, Delgado, & Nicholson, 1997).
As discussed before, raw milk can be processed into numerous different products such as butter and cheese. An advantage of the processed products is that they are less perishable, that they have a higher price per kilogram and that they have lower transport and storage costs (Staal, Delgado, & Nicholson, 1997). But the transformation of raw milk into these products asks for an investment in equipment.
A last feature of milk that influences the organizational structure of the dairy market is based on the milking process. The milking process in developing countries is mostly dependent on manual milking.
As most farmers are smallholders, they don’t have the necessary highly technical equipment for an automated milking process. This makes that not only the milking process but also the husbandry of the cows is very labor-‐intensive. As a consequence, labor accounts for a significant and important share of the total costs of the farm. In addition, labor has a great effect on the efficiency of the farm: a high motivation and productivity of the laborers will increase the output of the farm and the quality of the milk.
2.2.2. Suitability of contract farming in the milk industry
The above-‐described features of milk highly influence the characteristics of the dairy market and thereby the suitability of contract farming; this section clarifies the link between the two.
The perishability of milk increases the transportation and storage costs of milk significantly in comparison with non-‐perishable agricultural goods, such as grain. But as well the high transportation costs as the storage investment costs are subjected to economies of scale, thereby creating a situation in which a cooperative can act as a uniting organization. Cooperatives tend to invest in collection centers, by that decreasing the distance to the closest cooling point and the need for storage. This will lower the costs for all smallholders directly but also indirectly by lowering spoilage of milk.
Another consequence of the perishability of milk is that it exposes the farmers to higher market risks. As well the milk production as the consumption are subjected to significant seasonal variation (Holloway, Nicholson, Delgado, Staal, & Ehui, 2000). But due to the high perishability of milk, it is hard for farmers to flexibly respond to the milk demand. It is impossible to, unlike with for example grain, storage milk for a longer period to fill gaps between demand and supply later on. This makes farmers very sensitive for risks; they are susceptible for the whims of the market. This sensitivity also affects the attitude of farmers towards price offers; buyers have market power when determining the offering price, as it is hard for farmers to switch to a different purchaser on the short term (Minot & Sawyer, 2014). Farmers prefer to sell the milk as fast as possible to sustain a higher quality and lower level of spoilage, thereby giving power to the buyer as they are more willing to accept any price. A cooperative can help farmers by creating an environment in which the market risks are more shared between the different participating farmers and the buyer (Birthal, A., Tjongco, & Narrod, 2008). In the contract that the cooperative concludes with a producer, they settle on a certain amount of milk that will be bought by the cooperative, thus creating a reliable outlet for the farmer. Previous research has shown that farmers are willing to accept a lower price in exchange for a stable demand (Staal, Delgado, & Nicholson, 1997). As the milk quality is very important in the consecutive processing scheme, the farmers have to be able to offer a consistent quality. The advantage of a cooperative is that the transaction is based on a close relationship between the seller and the buyer. The relationship is backed up with a contract, of which part is focused on setting a milk quality standard. And as an extra nudge, sometimes the farmer is incentivized to reach this quality standard by the format of the payment schedule (Navarro, et al., 2015). But even without these incentivizing systems, the contract provides a foundation for a quality standard, whereas the buyer, in a common purchase, determines the quality of the milk after the transaction. So even though some issues with milk quality, such as the difficulty of measuring, are still present within a cooperative system; the contract ensures a basic standard quality level and therefor lowers the costs for as well the processor as the buyer.
Another advantage of contract farming in the dairy industry is due to the economies of scale of milk processing, giving a cooperative the opportunity to act as a combining institution. Just as with transportation costs, sharing the potential investment in processing equipment will lower the cost for farmers, while providing them with extra opportunities for their milk production.
These arguments show that the dairy market can highly benefit from a contract farming
organization, with a medium level of vertical integration, compared to complete vertical separation as in a spot market. Contract farming decreases the risks for as well the seller as buyer, makes use of the
economies of scale of transportation and storage and provides a better quality guarantee. But as the producing company is still completely separated from the processing company, in the contract they only define a few aspects as the price and quality; there is no question of complete vertical integration. As a consequence, the farmer and his family are the owner of the company and will benefit from any additional income. This provides a natural incentive for a larger and higher quality production and a more effective production process, an incentive that would not exist if the farm were part of a bigger company. The contract farming structure also economizes costs for the cooperative; they don’t have to take care of supervision costs or the procurement of land or herd (Birthal, A., Tjongco, & Narrod, 2008). This makes that the ideal vertical integration level is more located towards a cooperative structure, medium vertical integration, than towards complete vertical integration.
2.2.3. Dairy market in developing countries
The theory described above predicts that based on the characteristics of milk, the dairy market is very suitable for a contract farming structure. But that doesn’t have to be linked to the actual structure in de dairy market in developing countries. This part of the paragraph elaborates on the real market design, focusing on the farmers and cooperatives.
The features of the farmer are closely related to the characteristics of milk, as the farmer establishes his company to the needs of the production process of milk. Smallholder farms dominate the dairy industry in developing countries: they usually keep a small herd on a limited piece of land (most often less than 3 hectare), existing of a mixed crop-‐livestock (Wozemba & Rashid, 2008). The composition of a smallholder farm has great influence on the goings in the company: most often the whole family contributes to the production process. This makes not only that a farm has lower labor, supervision and search costs, but also that they are likely work more motivated than usual workers. But the smallholder organization also has its drawbacks: one of the main disadvantages of a smallholder company is a low productivity due to many challenges and constraints (Wozemba & Rashid, 2008). Most smallholders in developing countries don’t have the equipment to keep the milk save and cool at their farm and thus have to transport the milk on a daily base. The small size of the company and the low amount of laborers creates a challenge for the transportation of milk to a collection point or the market. Especially if the collection point is far away, the infrastructure bad and the availability of transportation vehicles low: farms face high transportation costs or a deficiency of laborers. Another limitation of a smallholder farm is the lack of access to the consumer, credit and knowledge market (Birthal, Joshi, & Gulati, 2005). It is hard for a smallholder in a developing country to get a clear and conclusive overview of the demand market, the market asks thereby for high marketing costs (Holloway, Nicholson, Delgado, Staal, & Ehui, 2000).
This makes that a producer doesn’t always manage to sell all his milk, let alone to receive a fair price. An additional deficiency of a smallholder is the credit constraint; this indirectly affects the poor production performance as it limits the initiatives of farmers to improve their company (Wozemba & Rashid, 2008). As the farmers have a relative small production and a wide variety of costs, they have to cope with small marketable surpluses (Holloway, Nicholson, Delgado, Staal, & Ehui, 2000). This influences them by increasing their risk adversity and by leaving them with only a very poor amount to invest.
Ill or non-‐functioning spot markets also characterize the dairy market in developing countries, resulting in high transaction and marketing costs for as well the farmer as the buyer (Moll, Staal, & Ibrahim, 2007). These weak spot markets also affect the level of competition between the dairy buyers. In a market without cooperatives, the private vendors have a low level of competition, which leads to unbalanced market power, prices that are well below the equilibrium price and in some cases to exploited dairy producers (Birthal, A., Tjongco, & Narrod, 2008).
The idea of contract farming can take different shapes in practice; this research focuses on contract farming with a cooperative. These farmer cooperatives are producer level institutes and act as the buyer in the settlement; a farmer then has to sell all his production to the cooperative (Eaton & Shepherd, 2001). Previous research has shown that cooperatives improve the effectiveness of the participation of farmers in the market by allowing them to compete with larger farmers (Markelova, Meinzen-‐Dick, Hellin, & Dohrn, 2009). One of the main benefits of a cooperative is the increase of bargaining power: by being in a group the smallholders have more influence on the price and market conditions (Wozemba & Rashid, 2008). This increase of bargaining power decrease the bargaining power of the buyer and by that the price risks for the farmer (Holloway, Nicholson, Delgado, Staal, & Ehui, 2000). The dairy producers in developing countries are vulnerable and dependent on their buyers: the
perishability of milk makes the farmers extra vulnerable. Once the milk has been milked; the bargaining power of the farmer is badly weakened when there is no contracted price or guarantee on a fair price (Minot & Sawyer, 2014). In this scenario the cooperative is extra appropriate, as the cooperative consists out of farmers, it stands up for the farmers and acts in line with the farmer’s wellbeing. This is the biggest distinction between a cooperative and other forms of contract farming: most buyers in other forms of contract farming are not connected to the producers and represent their own interests.
2.3 EFFECT OF CONTRACT FARMING ON FARMERS
The empirical part of this research focuses on small farmers in the dairy industry in Uganda. Small farmers face different problems and obstacles in their daily affairs than big farmers as they have less resources and possibilities. It is therefore interesting to study the effect of contract farming on the production, income and well being of small farmers.
Research has shown that small farmers have in general four types of constraints when accessing production resources or markets. Firstly, small farmers often lack information about production and marketing; information that is particularly useful if a farmer wants to expand or diversify his production. But secondly, even with this information, small farmers often still lack the financial assets, and a lack of collateral withholds them from applying for a loan. Thirdly, small farmers have shown to be more risk averse than larger farmers; they have to provide their family with a certain level of basic income before
they can use money for investments. Lastly, previous (public) interventions had no elevating effect on these constraints; governments’ actions generally favored the large farms and led to unreliable supplies and rationing (Minot N. , 1986). According to Roy (1972) and in line with the general theory: contract farming has several advantages that can help farmers in dealing with these problems. Theory
demonstrates that contract farming reduces risk, the deficit of financial liquidity and marketing problems, gives farmers a possibility of a more stable income and increases the access to inputs and technical assistance. But contract farming can also have its downsides: misperceptions or a lack of information might cause a negative balance in the end, contract terms might lock a farmer in a contract that is not profitable or the contract might have a negative effect on the other members of the household or community (Minot & Sawyer, 2014).
It therefore really depends on the product and environment of the market and farmer, whether contract farming has a positive or negative effect on the income and wellbeing of the farmer. Nonetheless, some empirical research has been conducted to test the effect of contract farming on the wellbeing of farmers. Little and Watts (1994) conclude that the income for a moderate to high part of the smallholder famers increases, but that non-‐farm income is still necessary to reach the basic needs. In the previous decennia, the research on contract farming focused on the possible social problems of the farmers and their environment such as an imbalance of bargaining power or a lack of control over the production (Porter & Phillips-‐Howard, 1997). In current research selection bias has been considered to play an important role in the effect on farmers; the unobservable characteristics of participating farmers could be significantly different from non-‐participating farmers. Although the researches conducted in the last years are analyzing specific sectors in specific countries, they still show a trend of positive outcomes; contract farming has an increasing effect on the income in labor. For example, Bolwig et al. (2009) conclude that participation in the contract increases the net revenue of coffee farmers with 75%, which is equal to 12,5% of the mean total household income. In this case this is mostly due to the higher offered prices because of an organic certification, but the increase of income can also be a result of lower transaction costs, as in the research of Birthal et al. (2008). This research focuses on the dairy sector in India and finds that participation in contract farming increases the net revenue with 80% compared to the average. Minot and Sawyer (2014) compare all the existing research on the effect on income of farmers and conclude that generally contract farming is beneficial for farmers and increases their income. In the few instances that contract farming doesn’t prove to be beneficial, the program didn’t exist long before it collapsed. So it can be assumed that in general, contract farming has a positive effect on the income of the participating farmers.
2.4 DAIRY COOPERATIVES IN UGANDA
The data used in the empirical part of this research focuses on farmers that are member of a dairy cooperative in Uganda. Also in Uganda, the smallholders dominate the dairy market: they produce about 80% of the total raw milk production of Uganda. They implement this by possessing over 90% of the total national herd (of 7.5 million cattle), and almost all of the small ruminants (Wozemba & Rashid, 2008). The milk of the Ugandan smallholders is obtained by milk a cow several times a day: on average this yields 3.6 liter milk per day per farm.
Even though theory would suggest that the dairy market in a developing country is highly suitable for medium vertical integration; research conducted on the number of contracts in this specific area has contrary results. A random survey in 2012 concluded that only 5% of the stratified sample of Ugandan dairy farmers had a contract with a buyer (Minot & Sawyer, 2014) but study of 2011 by the FAO showed that there were 241 primary cooperatives active in Uganda (Balikowa, 2011). Most of the active
cooperatives, 60%, were found in the South Western region and 25% in the Central region. This
distribution is line with the distribution of the number of cows and the milk production over the regions; the Western region is leading in milk production, followed by the Central region (Balikowa, 2011). The cooperatives in Uganda, existing since the 1960s, have had a troubled history; the former governments used the cooperatives for inappropriate political activities and extracted the rural surplus to provide it to urban population (Sinja, et al., 2006; Wozemba & Rashid, 2008). When in the 1990s the government no longer supported the cooperatives, they troubled with their financial self-‐reliance and slowly collapsed. But recently the cooperative organizations have redeveloped in various sectors, among which in the milk collection and marketing. They have even extended their activities to the further
processing of dairy products. The cooperatives have shown gain profits for the farmers as they offer stable prices and regular reliable payments. In, for example the region Mbarara, the cooperatives have decided on a fixed price for the milk over the year and guaranteed to buy all the collected milk from the farmers. This has lead to a more stable and reliable income for the farmers in the region (Wozemba & Rashid, 2008). Another example of the cooperative structure is situated in Western Uganda; the Uganda Crane Creameries Cooperative Union was established in 2005, born out the Western Uganda Dairies Association. The union consists out of 7 district cooperative unions, 76 milk collection centers and 10500 dairy farmers (Wozemba & Rashid, 2008). This illustrates a standard format of the cooperative unions in the Ugandan dairy industry.
3. METHODOLOGY
The empirical research of this thesis is focused on smallholder farmers in Western and Central Uganda. As the farmers in the data set are participants of a cooperative they officially have to sell all their dairy production to the cooperative, but some choose to also sell it to private vendors. The data, based on a household survey, on dairy farmers in western and central Uganda shows thereat that farmers do sell to both of the buyers. This raises the following question: what is the rationale of dairy farmers when selling to cooperatives or private vendors? Based on previous research and logical deduction, hypotheses have been composed, which will be tested by making use of the data on Ugandan dairy farmers. These hypotheses are based on characteristics of the producing farm, private vendor and cooperative, which influence the choice of utilizer of the milk production. The hypotheses are divided into the following themes: the price level of milk, the need for income risk spreading, the volatility of milk production and the convenience of delivering milk to the cooperative. So the following research question will be answered in the result section: what is the influence of price, production volatility, income risk and convenience of delivering milk on the farmers’ decision process of selling to a cooperative or private vendor?
As an preface of the results, this paragraph will elaborate on the studied data, the hypotheses, their underlying reasoning and some descriptive data, and will conclude with the exhaustive research model.
3.1 DATA ON UGANDAN DAIRY FARMERS
The empirical part of this research is based on data on Ugandan dairy farmers. A household survey has been conducted on selected villages in western and central Uganda. The data collection has been
organised by the Policy and Operations Evaluation Department of the Dutch Ministry of Foreign Affairs to evaluate the impact of the Dutch Food Security Programme. The aim of this programme was twofold: (1) to strengthen dairy market players by professionalizing the farms, support the dairy cooperatives and increasing the access to financial services and storage facilities and (2) to increase demand by improving the access to the market (AIID, PwC, 2014).
The data was gathered in 2014 and consists out of 840 observations adopted with a two-‐stage stratified sampling design, based on the dairy cooperatives (first stage) and dairy farmers (second stage). The questionnaire used in this research is the first of three surveys, and executed among households. 1 The subsequent surveys are executed among dairy cooperatives and milk vendors. Furthermore, this survey is a baseline survey; a distinction has been made between the control and treatment group but the groups are not yet treated differently. The household survey discusses several themes, such as housing, land use, dairy production and utilization, farm employment and other household income (AIID, PwC, 2014). A disadvantage of surveys is the high risk of measurement errors, subjective interpretations of the questions and socially responsible or overestimated answers. This could lead to all kinds of mistakes in the data, which will influence the final estimated effects of the independent variable. If the sample is big enough, the different errors are likely to cancel each other out, thereby creating a reliable regression. The farmers in the sample were clustered in cooperatives; in the treatment area they formed the Uganda Cranes Creameries Cooperative Union (UCCCU). The organisation of UCCCU is threefold: there is umbrellas headquarter, district level co-‐operative unions and primary cooperatives (AIID, PwC, 2014). Not all farmers in the sample state to be a member of a dairy cooperative, which gives the opportunity to compare member and non-‐members, but membership turns out to be an unreliable variable for selling milk to a cooperative. A quick look at the data shows that farmers that are member don’t necessarily sell to the cooperative, whereas farmers that aren’t a member do sell to a cooperative. The variable
membership has therefore not been used in the empirical research, instead the assumption has been made that farmers that sell to a cooperative have a contractual agreement with the cooperative. Appendix II contains graphs with more information about the cooperatives in the sample, including the milk collected and income generated.
As explained above, the data used for the empirical research consists out of a one-‐time shot of the production and status of the dairy farms. As a consequence, the data is not fully complete and contains missing values. To be able to work with and draw conclusions from the data, some statistical actions have been performed; these will be clarified in the following paragraph.
3.2 HYPOTHESES