2.1. Structure of Cocoa Markets and Trade Reforms

The domestic markets of most cocoa producing countries have been historically characterized by government interventions through national trade policies and direct control of the marketing channels. These interventions were materialized, for example, in the forms of prohibitively high levels of export tariffs and quotas, and contributed to isolating national markets from price signals in world markets (Rapsomanikis et al., 2006; Sharma, 2003). However, in the wake of trade liberalization reforms, governments retreated progressively (Table 1). For example, prior to 1995, the domestic cocoa market in Côte d’Ivoire was managed by a government-controlled cocoa marketing board. Thereafter, the sector was liberalized with private buyers sourcing cocoa beans directly from farmers from the 1995–1996 crop season. In Cameroon, liberalization of domestic cocoa markets started from the 1989–1990 crop season. In the case of Ghana, although the country’s marketing board is still responsible for cocoa beans export, the internal marketing of cocoa was liberalized during 1992–1993 with private buyers starting to source beans from farmers (UNCTAD, 2008). In Ecuador, market-oriented reforms started in 1992 (Vos, 2000), whereas the corresponding year for Indonesia is 1985 (Baffes and Gardner, 2003). However, in the latter two countries, cocoa markets have been historically characterized by limited government intervention (Panlibuton and Lusby, 2006; Ton et al., 2008).

The reforms were expected to result in improved transmission of price signals from international to domestic cocoa markets, thereby contributing to a better integration of cocoa farmers into global markets (Rapsomanikis et al., 2006; Shepherd, 2004). This improved relationship may be evidenced in Figure 1 which shows the evolution of the world price of cocoa and cocoa producer prices in the five countries considered in this paper. The trends in producer prices compared with world price of cocoa tend to demonstrate a similar pattern. More interestingly, producer prices for all the five countries seem to track more closely the world price of cocoa from mid-1990s compared with the 1970s and the 1980s. These improvements in price relationships may have been partly due to trade liberalization reforms in the various producing countries—it is the objective of this paper to empirically test whether this is true.

Figure 1

Producer prices in selected cocoa producing countries in relation to world price of cocoa. Source: Authors’, based on data from UNCTADStat, World Bank-WDI, FAOStat and ICCO. Note: The time period covered by price series depends on data availability (see Section 3, The Modelling Framework and Variable Description). It is important to note the “abnormal” value of the cocoa producer price in Ghana in 1982, which exceeded the world price significantly. This was primarily due to a highly overvalued exchange rate, combined with an increase in producer prices in Ghana in the early 1980s (Kolavalli and Vigneri, 2011). In this study, the effect of this abnormal value is captured by a dummy variable.

The outcomes of trade reforms depend on the effectiveness of their implementation, strong policy and institutional environment, which are key enablers.

The trade reforms have also unexpectedly resulted in a high degree of concentration in cocoa markets. At the upstream segment of the cocoa GVC, the main driving force is complexity of cocoa markets characterized by Transnational Corporations (TNCs)’ ease of access to resources, such as finance, risk management instruments, as well as technologies, along with their objective of achieving scale economies. These characteristics have led most small players to leave cocoa marketing channels or merge with multinationals who took control over their activities (ACET, 2014; Gilbert, 2009). The resulting market structure could contribute to improving cost efficiency of the cocoa GVC, with potential benefits passed onto smallholders, including small farmers. However, high concentration, especially in the domestic cocoa markets, may have offset the benefits for farmers if it led to oligopsonic or monopsonic structure, atomizing further the latter who as a result would be left with little, if any, room for negotiating remunerative prices for their output.

2.2. Price Linkages between Domestic and Global Cocoa Markets and Trade Reforms: Theoretical Underpinnings

Trade has arguably been considered the key to explaining the linkage between domestic and world markets of agricultural products. If a specific commodity is exported and traded in world markets, the prices received by domestic producers equal the international price minus transaction costs (Rapsomanikis et al., 2006). The assumption of price transmission from world to domestic markets for agricultural products is common in the literature (Baffes and Gardner, 2003; Sharma, 2003). The main argument is that the atomized nature of farmers compared with global agricultural commodity markets reduces them to price takers. In other words, these farmers are almost entirely at the mercy of price movements in international markets. This is particularly true for cocoa markets where buyers usually use prices in world markets as reference when sourcing cocoa beans from farmers, who usually after harvesting the pods, ferment and dry the beans on their farms or in their villages.

Historically, the marketing of cocoa beans from farm gates to export markets was controlled by national commodity boards. The boards would purchase the beans from farmers through collection points near villages, and act as principal sellers and exporters. Today, wherever the domestic cocoa market is liberalized, farmers directly sell their crops at buying stations to export agents or brokers. In all the cases, including in deregulated or non-deregulated markets, the buyers offer farmers prices that reflect the prevailing prices of cocoa beans in international cocoa markets (ITC, 2001). These markets include the London International Financial Futures and Options Exchange (where prices are denominated in British pound) and New York-based Intercontinental Exchange (where prices are set in United States dollars). These markets offer standardized futures contracts for cocoa beans with commitments to trade given quantities of cocoa at a certain price at the expiry of the contracts. The delivery ports of these contracts are located in Europe or the United States.

The business profitability of cocoa trading implies that farm gate prices of cocoa remain below world prices. The spread between the two prices contributes to covering the expenses incurred by traders from moving the beans from farms to delivery ports in importing countries. These expenses include inland and international transportation fees, taxes and other duties, storage costs at the ports of producing countries and the profit margins of traders (Cocoa Barometer, 2015) as illustrated in Figure 2. Moreover, farm gate prices of cocoa are denominated in the national currency of production countries, whereas the prices in world markets are set in international currencies. Therefore, volatility in the exchange rates in producing countries affect the prices paid to cocoa farmers.

Figure 2

Transmission channels between farm gate and international prices of cocoa.

The transmission channels depicted in Figure 2 also imply that improvement in domestic marketing channels (lower transactions costs including transport fees, reduced taxes and duties, higher bargaining power of farmers) would contribute to increasing the prices paid to cocoa farmers. These improvements were partly expected to be achieved through trade liberalization reforms under structural adjustment programmes undertaken by most developing countries in the 1980s and the 1990s.

Some key objectives of the trade reforms were the elimination of trade barriers, increased competition in the cocoa marketing channels and reduction in regulatory burdens that would result in higher prices paid to local producers including farmers (Ajetomobi, 2014; Gilbert, 2009). Before these reforms, parastatal entities were controlling national cocoa markets in producing countries. These entities set farm gate prices and buyers margins while collecting export taxes (Fold, 2002). Although the raison d’être of the parastatal agencies was to isolate farmers from volatility in global cocoa markets, an unintended consequence of their operations was that prices paid to farmers over time diverged greatly from price trends in international markets as a result of high taxes, and mismanagement in several cases (Rapsomanikis et al., 2006; Sharma, 2003; Williams, 2009). Trade liberalization entailed, therefore, the elimination of the parastatal entities and promotion of private institutions and market agents that were to make the marketing channel more efficient, with induced gains passed on to farmers. Yet, the extent to which the trade reforms, some three decades later, have achieved their expected outcomes is still debatable. This study contributes to this debate.

3.1. Price Transmission and Market Integration for Agricultural Products: Literature Overview

A number of studies have analysed price transmission dynamics in agricultural commodity markets. Quiroz and Soto (1995); Mundlak and Larson (1992) assessed transmission of price signals from world markets to producers using the Food and Agricultural Organization (FAO) of the United Nations annual price data. The former covered 78 countries over the 1966–1991 period, with between four and 15 products per country, and found non-existent or extremely low price transmission in 60 countries. By contrast, the latter used data on 58 countries over the 1968–1978 period covering 60 agricultural products and concluded that the variations in world price constitute the dominant component in the domestic price changes.

More recently, commodity price transmission analysis based on time series techniques has been carried out within the context of trade policy reforms with mixed results so far. Baffes and Gardner (2003) examined the degree to which world prices have been transmitted to domestic prices for eight countries and 10 commodities under trade policy reforms. They found that the effect of the reforms on world and domestic markets integration varies by countries and commodities. Markets integration was found to be significant in some cases and was either low or non-existent in others due primarily to national trade policies. Shepherd (2004) analysed price transmission through the coffee processing chain at a global level. He concluded that trade liberalization reforms have not significantly improved price transmission but rather have worsened it in some respects. The author attributed the cause of this to market power being potentially exerted by buyers in a context of high concentration of actors in the coffee GVC. Worako et al. (2008) investigated producer and world price transmission mechanism in Ethiopian coffee market. They concluded that domestic prices of coffee adjusted more rapidly to world price changes after the country had implemented trade policy reforms. However, the authors stressed that impacts of the reforms on the levels of prices received by Ethiopian coffee farmers have been limited by a number of market failures including lack of adequate market information system, high marketing and transaction costs, collusion among buyers and high concentration of exporters. Ajetomobi (2014) investigated the issue of market integration through price transmission in Nigerian cocoa markets over the postliberalization period, and concluded that market liberalization improved the efficiency of the country’s cocoa marketing channels.

Against this backdrop, this paper specifically investigates price transmission from world cocoa markets to farmers in the following key producing countries: Cameroon, Côte d’Ivoire, Ecuador, Ghana and Indonesia. As discussed earlier, the choice of the countries is guided by availability of quality data, the importance of the country in cocoa production at continental levels and the significance of the cocoa sector in their economies.

3.2. Empirical Model

The empirical framework is based on the ECM described later.⁴ Consider y_t and x_t as the annual producer and world prices of cocoa, respectively, expressed in logarithmic value. The relationship between the two prices may then be written as follows [Equation (1)]:

where u_t is the error term and β represents the long-term elasticity between the prices. A necessary condition to conclude that a producer price is strongly related to the world price of cocoa in the long run requires that y_t and x_t be co-integrated,⁵ meaning that [Equation (2)] is a stationary process. If the two price series are co-integrated, the producer price is said to be in a stable and long-run equilibrium with the world price of cocoa. The relationship may then be estimated through the following [Equation (3)]: where θ₁ gives the short-run relationship between the variation of y_t and that of x_t, and the coefficient γ measures the long-run adjustment of y_t to x_t. Indeed, γ is a measure of the speed of convergence, that is, the extent to which a disequilibrium in the two series in the previous period (previous crop year, for example, in the case of cocoa annual prices) is corrected in the current period (current crop year).

The existence of a short-run relationship implies that producer prices adjust immediately, that is, during the same period, to signals of world price. A short-run coefficient of one means that shocks in the world cocoa market are completely transmitted to national market during the same crop year. A significant long-run adjustment coefficient would suggest existence of a long-term equilibrium relationship between producer and world cocoa prices. In other words, the two prices track closely with each other in the long run, although they can drift apart in the short run. The overall adjustment proportion (k) of producer price to a given change in world price after n periods may be obtained through the following formula: k = 1 – (1 – θ)(1 – γ)ⁿ (see Baffes and Gardner, 2003).

To consider the impact of possible structural breaks such as the introduction of trade reforms that might have affected the relationship between the two prices, a dummy variable is introduced in a multiplicative form in the Equations (1) and (3) as in Worako et al. (2008). This multiplicative dummy also permits to estimate a pooled regression with one equation instead of two distinct equations. The main advantage is that a pooled regression increases observations, yielding greater degree of freedom.

The dummy variable is also added to Equation (1) in an additive form to capture possible changes in the constant term before and after the reform periods. Denoting the dummy as d_t = 1 for prices obtained during the period after the introduction of reforms and zero otherwise, a pooled regression form of Equation (1) becomes Equation (4)

With this formulation, the constant terms and coefficients α_i (i = 1, 2) differ from the pre- and postreform periods. The coefficients α_i (i = 1, 2) represent, respectively, the long-term elasticity between producer and international prices before and after reform periods.

With the introduction of the dummy variables, the equivalent of the ECM in Equation (3) is Equation (5)

Coefficients θ_i (i = 1, 2) describe the short-run effect of the variation in producer and world price, respectively, before and after the reforms were introduced, whereas γ_i (i = 1, 2) capture the speed of adjustment to equilibrium prices before and after reform periods, respectively. Equation (5) may be extended to capture any asymmetry in the short-run relation between producer and world cocoa prices.

Assessing potential asymmetry in the price transmission implies investigating whether producer prices react in the same manner for both increases and decreases in world price of cocoa. The investigation of the prevalence of asymmetric price transmission from world to domestic markets in the short-run is underscored by several potential arguments. One argument relates to market concentration and noncompetitive price behaviour. This argument posits that in the presence of market power derived from market concentration and imperfect competition, as is the case in cocoa markets (Gayi and Tsowou, 2016), the transmission of an increase in world prices of tradable products to producers is weaker than that of a decrease in prices (Scherer and Ross, 1990). A second argument for asymmetric price transmission relates to national trade policies such as floor price policy (Rapsomanikis et al., 2006). In the case of cocoa, this may be derived from the minimum guaranteed price set for cocoa farmers in producing countries such as Cote d’Ivoire and Ghana, which may insulate producer prices from downward swings in world price. Asymmetric price transmission may also be partly driven by inventory holding behaviour (Maccini, 1978). Indeed, in domestic cocoa markets, expectations of high prices in world markets or high indicative prices set by government in domestic markets may lead to stock accumulation locally. When high prices are recorded as expected, release of stocks may exert downward pressure on local markets and cause producer prices not to rise as much as it would in the absence of inventories.

Technically, the asymmetric effect may be captured by introducing a dummy that describes a price increase or decrease. Denoting the dummy D_t = 1 for an increase in world price (i.e., Δx_t ≥ 0) and D_t = 0 for a decrease in world price (i.e., Δx_t < 0), Equation (5) becomes Equation (6)

Coefficients θ_i (i = 1 – 4) describe the short-run responses of producer prices to increase and decrease in world price, respectively, before and after reforms.

3.3. Data and Sources

The study is mainly based on statistics from database and publications of the FAO, the International Cocoa Organization (ICCO) and the World Bank. Although in some cases, where more information was needed, or gaps in the information flow were found, interviews were conducted with key informants by telephone or e-mails.

The ICCO annual cocoa beans price, expressed in United States Dollar per ton, is used as a proxy of world cocoa price. The producers’ prices came from (FAO, 2016). The period covered by data, depending on their availability, varies with countries, although not much: Cameroon (1966–2010), Côte d’Ivoire (1966–2009), Ghana (1966–2011), Ecuador (1966–2012) and Indonesia (1967–2012). Nominal prices in national currencies are converted into the United States Dollar using annual average exchange rates from the World Bank (World Development Indicator) to facilitate comparison before they were transformed into natural logarithm.

4.1. Stationary Tests and Co-integration Relation

The results of stationary tests—Phillips–Perron and Augmented Dickey–Fuller—are provided in Table 2. The price variables are nonstationary in levels but become stationary after first difference. These imply that world and producer prices of cocoa could have a long-term relationship if their linear combination is stationary. This is formally assessed through co-integration tests with break dates relative to trade policy changes which might have affected internal cocoa marketing in producing countries.

The results of co-integration tests are provided in Table 3. They confirm the existence of a co-integration relation between the prices. The co-integration relation before trade liberalizing reforms may account for the fact that although producer prices were regulated in countries such as Cameroon, Côte d’Ivoire and Ghana before the reforms, producer prices were continuously adjusted to world price of cocoa by the commodity boards of each country. After the introduction of trade reforms, except for Indonesia, the relation between world and producer prices of cocoa improved as evidenced by the values of the long-term elasticity between world and producer prices for cocoa, which are higher during the period after the reforms compared with the period prior to the reforms. For example, a 1% increase in the world cocoa price resulted in a 0.80% increase in Cameroonian producer price before the reforms. After the reforms, the same proportion of increase in the price of world cocoa led to a higher (1.04%) increase in the price paid to the country’s cocoa producers. The equivalent values in Ghana are 0.90% for the pre-reform period and 1.05% for the postreform period. In Côte d’Ivoire and Ecuador, the increases are moderate from one period to other.

4.2. Results of Price Transmission

The results of ECM accounting for short- and long-run relationships in the context of structural breaks are reported in Table 4 with mixed outcomes.

Short-run transmission from world cocoa markets to producers is not statistically significant for Cameroon and Ghana over the pre- and postreform periods, although it seems to increase with the reforms. In Indonesia, the postreform period has been associated with a non-existent short-term relation between world and producer prices of cocoa. The situation in these countries could partly be attributed to market imperfections which delay the transfer of world market signals to domestic markets. These imperfections may include high transaction costs due to the poor state of rural infrastructure, market power of big players along the cocoa value chain and asymmetry in market information that usually prevails in developing countries, often to the detriment of small farmers. In the specific case of Ghana, the findings could also partly have resulted from high inflation and distortions in the country’s exchange rate (Kolavalli and Vigneri, 2011) as well as the fact that even after trade reforms, the country’s government has continued to play a significant role in the cocoa sector through a cocoa marketing board (Ghana Cocobod) which potentially acts as a buffer between farmers and global markets. Moreover, the cocoa producer price in Ghana is determined at the beginning of every crop year by the Producer Price Review Committee, a body consisting of Cocobod, government officials, representatives of private buyers and the national cocoa farmers’ organizations. The producer price setting mechanism does not allow for adjustments in line with the rate of domestic inflation thereafter, which almost certainly leads to a wedge between producer prices and the world cocoa price during the same crop season. For Côte d’Ivoire, the short-run transmission mechanism became effective after the implementation of trade reforms which might have therefore contributed to improved efficiency in the country’s domestic cocoa markets. The transmission effect in Ecuador is significant during the two periods, although it increased with the reforms. The results for Ecuador are not surprising per se due to historically weak government intervention in the country’s cocoa sector. For Côte d’Ivoire and Ecuador more specifically, the coefficient of short-run transmission is roughly 1 during the postreform period, meaning that shocks in world price of cocoa during a crop year are fully passed to domestic markets during the same crop year.

The speed of adjustment to long-run equilibrium between producer and world cocoa prices is significant for all the countries. Except for Indonesia, the speed of adjustment increased notably with trade policy reforms. For Côte d’Ivoire especially, results show a significant degree of over adjustment (i.e., adjustment coefficient higher than 1) of long-run price signals after the reform period. In the case of Indonesia, relative stability in the speed of adjustment during the periods prior to, and after, reforms could be attributed to the national marketing channels which were already characterized by liberalized trade practices even before the reforms.

These results imply that producer prices of cocoa in countries such as Cameroon, Côte d’Ivoire, Ecuador and Ghana adjusted more rapidly to shocks in world price after the reforms. An illustration is given for Cameroon and Ghana (Figure 2). For example, in the case of Cameroon, for a given change in the world price of cocoa, the 1-year price adjustment is estimated at 40%, during the period before trade liberalization reforms. After the reforms were introduced, the percentage of price adjustment increased to nearly 90%. This notwithstanding, heightened speed of adjustment between the prices also implies that exposure of the cocoa farmers to world markets has increased with the reforms.

4.3. Asymmetric Transmission between Cocoa Prices

The prevalence of asymmetry in short-term price transmission is estimated and tested for the countries; and the results are reported in Table 5. The modelling takes into account the effect of policy reforms as well as the transmission of increases and decreases in world cocoa prices to producers. During the prereform period, significant short-term price asymmetry is found for Cameroon, Côte d’Ivoire, Ghana and Ecuador. For Indonesia, the prereform period is associated with significant transmission of increases in the cocoa world price to farmers, whereas no evidence is found for price decreases.

During the postreform periods, transmission of upward swings in world cocoa prices to farmers is found to be significant for three countries, namely, Côte d’Ivoire, Ecuador and Indonesia. However, the transmission of price decrease is significant only for two of these countries, namely, Cote d’Ivoire and Ecuador. Where the transmission exists for both price increase and decrease, that is, in the cases of Côte d’Ivoire and Ecuador, we have not found any significant difference in the coefficients of asymmetry.⁷ These imply that transmission of an increase in the world price of cocoa to producers is not different from that of a decrease in these prices. A key implication is that farmers’ exposure to world markets has increased with the reforms which have therefore made them more vulnerable to volatile international cocoa markets (Maurice and Davis, 2011) at the time when they have limited, or no, access to price risk management instruments.

Therefore, this empirical investigation fails to support the existence of market power in the marketing channels of cocoa beans in Côte d’Ivoire and Ecuador which might have prevented farmers in these countries from benefitting from increases in the world cocoa price while bearing much of the cost of price decline. However, this conclusion must be interpreted with caution as many other causal factors may have affected the asymmetric price transmission, as discussed earlier. For example, Griffith and Piggot (1994) found asymmetries in the Australian lamb and beef markets but not in pork market although the former is less concentrated than the latter. Moreover, the frequency (annual) of our data and the size of the sample for each country have limited the number of observations considered in assessing price asymmetries. Higher frequency data (e.g., weekly and monthly), deeper analysis of national market structure and the behaviour of various market participants would be needed to improve the price asymmetry analysis. This is, certainly, an area for future research.

When comparing the short-term transmission in the prices over the pre with postreform periods, the incidence of price swings in the world market is greater during the latter period. For example, in the case of Cote d’Ivoire, a 1 unit increase in the world price of cocoa resulted in an increase of 1.17 units in the producer prices after the reforms. Over the same period, a 1 unit decrease in world cocoa price resulted in a 1 unit decrease in the country’s producer prices. Meanwhile, short-term price transmission was not found in Côte d’Ivoire before the reforms. These results are consistent with the previous findings: trade liberalizing reforms have increased cocoa producers’ exposure to both positive and negative shocks in world markets.

The results so far first suggest that while the transmission of world to producer prices of cocoa varies by countries, overall, it seems to have increased with trade reforms. Short-term transmission of the prices improved significantly with the reforms in countries such as Côte d’Ivoire and Ecuador. By contrast, in Ghana and Cameroon, although the short-term price transmission seems to have improved with the reforms, it is not statistically significant. In Indonesia, the short-term transmission exists before the reforms but seems to weaken thereafter. The long-run adjustment of producer prices to world cocoa prices has increased with trade liberalizing reforms in all the countries examined, except for Indonesia where the relation remains relatively unchanged. In other words, cocoa farmers in the producing countries such as Cameroon, Côte d’Ivoire, Ghana and Ecuador are integrated (i.e., are linked) with world markets in the long run, and this integration has increased with trade liberalizing reforms. The findings also imply that producers became more exposed to the volatile world cocoa prices under the reforms. The extent to which these developments have been associated with the evolution in the share of producer prices relative to world price of cocoa is explored in the next section.

4.4. Shares of Producer Prices Relative to World Price of Cocoa

The empirical analysis suggests that trade liberalization reforms undertaken by cocoa producing countries in the 1980s and 1990s have generally improved price adjustment between their domestic and world cocoa markets. The reforms might thus have been considered successful if their sole objective was to induce better response of cocoa producer prices to changes in world market prices. Have the changes in price transmission attributable to the reforms impacted on the shares of the world price accruing to farmers? This is a key question considering that a major argument in favour of the reforms was that they would contribute to generating a higher share of international prices to cocoa farmers by linking the latter directly with buyers (Gilbert, 2009; Shepherd, 2004). With regard to this latter objective, the outcomes have been mixed so far, as evidenced by the evolution in the shares of producer prices relative to world price of cocoa in our sample of producing countries (Figure 3).

Figure 3

Speed of price adjustment in Cameroon and Ghana over the pre-reform and postreform periods. Note: Vertical axis depicts the percentage of adjustment of domestic prices to a given change in world price after n periods (horizontal axis). The adjustment proportion (k) of producer price to a given change in international prices after n periods is calculated according to the following formula: k = 1 – (1 – θ)(1 – γ)ⁿ (Baffes and Gardner, 2003, p. 162).

Figure 4

Cocoa producer prices as a percentage of world price in selected countries. *Note: Data fully covering the sub-period 2006–2012 were not available for all the countries. Data coverage for this sub-period is as follows: Cameroon: 2006–2010, Côte d’Ivoire: 2006–2009; Ghana: 2006–2011; Ecuador and Indonesia: 2006–2012.

The shares of producer prices relative to the world price of cocoa vary within and across countries. In Côte d’Ivoire, for example, the postreform period has not been associated with a higher share of cocoa prices accruing to farmers. The average share of world price paid to Ivorian producers fell from roughly 70% during the period 1986–1990 to 54% over the first half of 1990s (i.e., before the reforms). Despite the reforms, it declined further and remained below 50% over the period 1991–2009. In Ghana, the share accruing to farmers increased in the 2000s compared with the 1990s, albeit it remained relatively low (between 50% and 55%) from 2001 to 2011. By contrast, in Cameroon, Ecuador and Indonesia, cocoa farmers received a relatively higher share of world price, especially after 2000. Ecuadorian cocoa farmers, for example, received more than 70% of world cocoa prices over the period 2001–2012.

The contrasting evolution in the shares of producer prices relative to the world price of cocoa confirms that the prices received by farmers are influenced by domestic factors. In Cameron, reduced taxes on cocoa beans exports might have contributed to higher prices paid to cocoa farmers (UNCTAD, 2008). In the case of Ecuador, higher prices paid to farmers have been driven by low export taxation, the establishment of minimum reference prices for cocoa beans and high quality of the beans produced in the country (Ton et al., 2008; WTO, 2005; WTO, 2011). Moreover, the fact that Ecuador is a dollarized economy has possibly contributed to keep the country’s producer prices higher compared with other cocoa producing countries where unfavourable exchange rates might have negatively affected the prices paid to farmers. In Côte d’Ivoire, domestic factors may explain why strong price transmission signals between domestic and world cocoa markets after trade reforms are associated with low share of producer prices relative to the world price of cocoa. In fact, the country’s domestic cocoa sector is still subject to high taxation, estimated, for example, at between 25% and 30% over the period 2002–2009, which has contributed to the downward pressure on prices paid to the country’s cocoa farmers (Kireyev, 2010). Other potential factors in the case of Côte d’Ivoire include political instability in the country from 2002 to 2010 (Gilbert, 2009) as well as market power exerted by multinational exporters or processors on cocoa farmers (Kireyev, 2010). In Ghana, low prices received by cocoa farmers have been partly attributed to high levels of inflation and overvalued exchange rate (Gilbert, 2009; Kolavalli and Vigneri, 2011). The case of Ghana is evidence of the extent to which the macroeconomic policy stance of a country can impact on prices paid to farmers. Indeed, the country’s Cocobod, under improved management, good governance and accountability after it was reformed, has enabled farmers’ to improve productivity and increase producer prices (Williams, 2009). Nevertheless, a major impediment to the country’s cocoa sector remains, that is, an unfavourable monetary policy environment with high inflation and overvalued exchange rates, which significantly erodes farmers’ incomes and as a result reduces the share of world cocoa prices accruing to them.

A number of other market imperfections could also explain the low prices received by cocoa farmers in the key producing countries at a time when trade reforms have improved their linkage with world markets. For example, the oligopsonic structure of national marketing channels in cocoa producing countries (Ecobank, 2012) could be detrimental for fragmented small-scale farmers, albeit empirical evidence of market power exerted by buyers has been inconclusive on this point (Anang, 2011; Traoré, 2009; Wilcox and Abbot, 2004). The main argument is, however, that this kind of market structure increases the risk of anticompetitive practices, creating a favourable environment for tacit or formal collusive behaviour by the big players (Losch, 2002, p. 225, Kaplinsky, 2004, p. 29) as has happened in other agro-food industries (Shepherd, 2004; Dobson et al., 2001; Goodwin, 1994). In the domestic markets of cocoa producing countries, the situation could have resulted in low prices paid to farmers who are entrenched in a weak bargaining position due to their scattered and unorganized nature. Low prices paid to cocoa farmers could also result from the number of intermediation levels between farm gate and the export point. In some cases, farmers have to deal with a large number of small- and medium-sized intermediaries, who often work for large-sized intermediaries in the domestic marketing channels. Subsequently, the margin available for farmers is very low as each intermediary tends to capture a share of profits that could have been passed onto farmers if the latter had direct access to exporters. This situation is almost certainly exacerbated by the poor state of transport infrastructure prevailing in the rural areas of many cocoa producing countries which increases transaction costs for traders and therefore discourages them from paying higher prices to farmers. Moreover, the poor access of farmers to financial and market information also affects their negotiation power, weighing down on the prices they receive from selling their products. For example, poor access to credit has led cocoa farmers in Cameroon to enter into partnerships with traders who provide them with the money they need during the planting season in exchange for their crops, often on very unfavourable terms for the farmers (Kamdem et al., 2010).

The discussion in this section reveals how the characteristics of domestic markets could affect cocoa farmers’ incomes. More detailed country case studies would, however, be required to identify specific determinants of domestic markets that are affecting cocoa farmers’ prices. This notwithstanding, the analysis carried out so far provides sufficient insights to discuss some policy options that could help in improving farmers’ situation so as to promote a sustainable cocoa farming.

5.1. Promoting Competitive Domestic Markets

As discussed earlier, the current oligopsonic structure of domestic cocoa markets results in power imbalances between highly integrated big players and other stakeholders, particularly farmers. In a major cocoa producing country such as Côte d’Ivoire, this situation might have been partly responsible for farmers not being able to capture a higher share of international prices although we found a complete transmission of price signals from world to the country’s domestic cocoa markets. If this market structure prevails unchecked, it will effectively undermine competition in the domestic cocoa market, adding further downward pressure on the price paid to farmers. The main arguments here are that a limited number of traders not only create a favourable environment for anticompetitive practices, but almost certainly reduce the bargaining room for fragmented farmers. As a result, the latter would not be able to bargain for higher and fairer prices for their output.

Creating a level playing field, by ensuring a competitive domestic cocoa market, is therefore critical to reversing the low prices received by cocoa farmers. Typically, this would entail a number of measures. First, it is imperative to enact, or reinforce, competition law in cocoa producing countries so as to prevent anticompetitive practices and limit market power of trading or processing companies which source beans from farmers. For this group of countries, the challenge with respect to competition laws is at two levels: how to enact and enforce these laws and how to address the challenge faced by legislators due to the ‘extraterritorial characteristic’ of national markets which stems from the fact that major TNCs active in cocoa trading or processing do not fall under the jurisdiction of cocoa producing countries. The former challenge could be addressed by improving institutional capacities at national levels with strong competition agencies. For the latter challenge, it would almost certainly require harmonization of rules dealing with anticompetitive practices, as well as cooperation among competition agencies at the global level with effective supervision by an international body.

Second, competitive domestic cocoa markets would require a supportive environment for local small participants, such as small traders and grinders, to remain in business. Indeed, key drivers of high concentration of buyers in the domestic markets of cocoa producing countries are the obstacles faced by local small participants to compete on a level playing field with TNCs, as the latter have better access to resources such as finance and technologies (ACET, 2014; Gilbert, 2009). Therefore, keeping local stakeholders including local small- and medium-sized enterprises involved in the national cocoa value chain would entail addressing the high costs of finance and high taxes on cocoa trading and processing activities in producing countries (ACET, 2014; Ecobank, 2014). In addition, keeping these local stakeholders in business would help to improve the value-added activities locally, which in turn could result in higher prices paid to cocoa farmers. For example, in Malaysia, thanks to a package of incentives, including an investment tax allowance, or partial tax exemption, for local players, several local entrepreneurs are directly involved in the cocoa sector; this contributes to improve the value-added activities retained by local cocoa communities (ACET, 2014). This is a good example of relevant policy that could be replicated elsewhere.

5.2. Ensuring Greater Transparency in Domestic Cocoa Markets

Agricultural markets in developing countries are often not transparent to farmers as they encounter difficulties in accessing reliable market information on demand, prices and quality, whereas buyers have usually better access to such information (Deardorff and Rajaraman, 2009). In the case of cocoa, this asymmetry which tends to favour traders is probably a major reason for the weak transmission of price signals from the world to domestic markets for some key cocoa producing countries, such as Cameroon and Ghana. Where high price transmission was found, as in the case of Côte d’Ivoire, weak access to market information that prevents farmers to enquire, for example, about market prices at national or international levels could also explain how the country’s cocoa farmers receive relatively low shares of world cocoa prices. A supporting argument is that a lack of market information may force small farmers to rely on intermediaries. The latter, which usually have better information on market prices compared with farmers, may use this information asymmetry to their advantage, offering producers low prices that might not reflect urban or international market prices. Thus, transparent markets for all the actors in the domestic cocoa sector would allow farmers to gain better access to information to make optimal marketing decisions. In Cameroon, for example, improved access to market information has contributed to enhancing the capacity of cocoa farmers to negotiate better prices for their beans (Wilcox and Abbott, 2006).

Governments in cocoa producing countries should therefore promote appropriate investments in new information and communications technologies through public and private partnership as a means of facilitating and enhancing farmers’ access to market information. Several successful examples exist in developing countries. The CocoaLink pilot project in Ghana involved at its pilot stage was a collaboration between the World Cocoa Foundation, the World Education, the Ghana Cocoa Board and the Hershey Company. This initiative helped to provide timely marketing, farming and social information (e.g., good farming practices, disease prevention, postharvest handling, crop marketing, farm safety and child labour issues) to Ghana’s cocoa farmers through SMS text and voice messages. Another successful agricultural market information system, which is not, however, a purely cocoa-linked initiative, is Esoko. Launched in 2005, with funding from international donors, Esoko is now a profit-making business entity with private investors providing agricultural information services to farmers in countries such as Ghana and Kenya. These success stories and others which have contributed to improving farmers’ access to information should be shared and replicated elsewhere, taking into consideration the specificities of each community.

5.3. Lowering Transactions Costs in Domestic Markets

High transactions costs resulting from impediments to trade such as the poor state of road transport, logistics and communication services in cocoa producing countries, particularly in those in Africa, could also contribute to explaining the relatively low levels of producer prices (Badiane and Shively, 1998; Njoh, 2008). These high costs erode the margins for buyers who in turn end up paying farmers a price which does not reflect world market price levels. The situation is often worse during the rainy season when roads in rural areas are hardly passable, making it more difficult for farmers and traders to get significant benefits from their trading activities. The problem becomes more acute when the distance from farm gates to markets is long, as is often the case in many developing areas.

These issues have endured in a number of cocoa producing countries, despite the recognition from governments that they impact negatively on the efficiency of domestic markets. The role of governments in providing public goods, such as roads, remains crucial, but given their limited financial resources, and in several cases, expertise, there is room for the engagement of the private sector through public and private partnerships.