Abstract
The introduction of partial liberalization into Ghana’s cocoa market in 1992/1993 encouraged competition among cocoa buyers. However, since the government determines prices of cocoa bean and thus prohibits price-based competition among Licensed Buying Companies (LBCs), LBCs rely on building relationships with cocoa farmers by providing incentives like credit and loans to cocoa farmers to encourage them to sell to them preferentially. These LBCs employ non-price strategies including the provision of loans and credit to cocoa farmers. This study analyzed the effects of non-price competition on loan and credit provision by cocoa LBCs. Descriptive analysis revealed that LBCs offer credit and loans to cocoa farmers, with 37.4 and 54.7% of cocoa farmers who received loan and credit respectively citing LBCs or purchasing clerks as their main source. Our regression analyses, however, did not fully support the standard intuition that competition promotes credit and loan provisions. This indicates the limited effectiveness of the use of credit and loan provisions as a non-price competition strategy among LBCs in Ghana’s cocoa market. Additional analyses offer valuable insights, suggesting that loans and credits may serve different functions when considered as means of competition for LBCs despite being essential forms of advanced payments.
1 Introduction
Historically, commodity marketing boards were established to address issues of market sharing and price fluctuations among others. These boards also allowed the government to maintain control over the marketing of strategic commodities and collect export taxes [1]. Similarly, the cocoa marketing board (COCOBOD) in Ghana was responsible for all operations in the cocoa sector including internal and external marketing. Before each cocoa season, COCOBOD announced the official producer price of cocoa which was accepted by all stakeholders [2]. The state-owned company, Produce Buying Company (PBC), was the sole buyer of cocoa beans, making the cocoa market strictly monopsony. Nonetheless, as part of the Economic Reform Program to reinstate economic growth and stability in the mid-1980s to early 1990s, the World Bank and International Monetary Fund mounted pressure on commodity export markets to fully liberalize [2,3]. Subsequently, the government of Ghana 1992–1993 introduced a partially liberalized cocoa marketing system, where COCOBOD grants licenses to private companies known as licensed buying companies (LBCs) to participate in the cocoa market by purchasing cocoa beans from farmers at no less than the set producer prices and delivered back to COCOBOD for further distribution and export. In return, COCOBOD pays LBCs a fixed allowance to cover procurement and transportation costs and pays profit margins that vary with cocoa producer prices.
The involvement of LBCs in Ghana’s cocoa marketing has introduced competition in the cocoa market with competition increasing with the number of LBCs. Currently, there are 55 registered LBCs to participate in the purchases of cocoa beans in Ghana as compared to only the state-owned PBC before 1992 [4]. Studies on Ghana’s cocoa market after the introduction of partial liberalization concluded that despite the domination by a few large firms in the cocoa market, it remains competitive [5,6] but since LBCs do not have control over cocoa prices, they have adopted non-price competition strategies which include the provision of loans and credit to increase their market share [6,7]. However, how these non-price competition strategies affect the nature of competition in the cocoa market is still unresolved. To fill the knowledge gap and contribute to the existing literature on the subject area, this study seeks to analyze the effect of non-price competition on the provision of loans and credit by LBCs to cocoa farmers in Ghana. The study will serve as a source of policy guide to both COCOBOD and LBCs in the Ghanaian cocoa sector. Also, it will be a major source for comparison between markets where prices are controlled and markets without price control.
Most cocoa farmers in Ghana face budget constraints mostly because of limited access to the financial market [7], and this makes it difficult for them to buy inputs even to supplement the inputs freely supplied or subsidized by COCOBOD. For similar reasons, some farmers lack money for their daily needs, especially in the pre-harvest season [8,9], and these circumstances may encourage LBCs to use credit/loan provisions as their strategic tools for competition [10]. Vigneri and Santos [7] and Bannor et al. [11] found a positive relationship between competition among LBCs and farmers’ production, especially for cash-constrained farmers. They argue that a possible pathway of this linkage is that increased competition allows farmers to select among LBCs who provide cash by prompt payment as well as loan provision. Therefore, our study contributes to understanding the pathway from competition in the local market to production by cocoa farmers.
A study done on cocoa in Sierra Leone by Casaburi and Reed [12] supports the assertion that purchasing prices and loan and credit provisions thus advance payment are both strategic tools for buyers to attract farmers, especially farmers who preferred advance payment to after-trade payment. These two strategic tools (price, and loan and credit provision) can be mutual substitutes. On the contrary, Macchiavello and Morjaria [13] in a study on coffee chain in Rwanda argued that competition may reduce the supply of credit by diminishing the market power required to sustain it. They explained that if a monopolistic buyer uses loan/credit provisions to keep the relationship with farmers, the introduction of competition would weaken the function that the loan/credit provisions have served. Therefore, the relationship between competition and loan and credit provisions is context-based and should be empirically tested. These relevant studies were conducted in market settings that are different from the cocoa market in Ghana, such that governments had no control over prices, market prices are determined by the forces in the market, and competition could be on price bases. Therefore, our study provides important empirical investigations on the relationship between restricted competition and loan and credit provisions.
2 Methodology
2.1 Conceptual framework
This section discusses in detail the relationship between market competition and loan and credit provision in the case of Ghana’s cocoa market, where cocoa bean prices are determined by COCOBOD, and LBCs are price takers. Entry of more LBCs in the cocoa market increases the competition in the market and pressurizes LBCs to expand their loan and credit provision to farmers to secure supply [12]. This act of loan and credit provision intensifies the farmer–buyer relationship between cocoa farmers and cocoa LBCs. Also, a good farmer–buyer relationship encourages the provision of loans and credit to cocoa farmers. On the other hand, competition among LBCs can weaken the effectiveness of relational contracts to keep the farmer-buyer relationship where cocoa farmers fail to remain loyal to one LBC because they have many other options to sell to [13].
2.2 Data and data source
The study used household-level data collected in Ghana and Cote d’Ivoire by the Royal Tropical Institute (KIT) between November 2016 and March 2017 [14]. A mixed-method approach was used to sample 1,560 households in Ghana and 1,485 in Cote d’Ivoire making a total sample size of 3,045. However, only data collected on cocoa households in Ghana were used for the analysis as this study seeks to assess the influence of competition on loan and credit provision to cocoa farmers in Ghana. A sample of 1,318 cocoa households was used for the analysis. The data covered five of the six cocoa-growing regions in Ghana as shown in Table 1.
Sampled Regions from Ghana
| Region | Number of households | Percent (%) households |
|---|---|---|
| Ashanti | 266 | 20.2 |
| Central | 68 | 5.2 |
| Brong-Ahafo | 203 | 15.4 |
| Western | 532 | 40.4 |
| Eastern | 249 | 18.9 |
| Total | 1,318 | 100 |
Source: Authors’ construct from KIT data [14].
2.3 Method of analysis
In this study, we adopted a structural measure of competition using the Herfindahl–Hirschman index (HHI) approach to measure competition at the village level.
where
Competition was computed by summing the squares of the market share of the LBCs at the village level; lower values of HHI indicate a higher degree of competition among LBCs. The number of cocoa farmers that sold their cocoa beans to each LBC at the village level was used to calculate the market shares of the LBCs because data were not available for the quantity of cocoa beans sold to each LBC. The Ordinary Least Square analysis was used to ascertain the effects of competition on the provision of credit and loans by LBCs. With competition as the main explanatory variable and four dependent variables as a proxy for the provision of credit and loans together with other control variables, we generated four regression models. Model specification is as follows:
Here, the subscript i indexes farmer and v indexes village.
To circumvent the problem of endogeneity of competition which may result from unobservable factors that affect both competition and provision of credit and loans, such as cocoa production at village levels and the activeness of LBCs in the villages, we controlled for cocoa household characteristics (z) in model (2), cocoa agro-ecological zones (AEZ) for cocoa production in model (3), and contracting LBCs (lbc) fixed effects in model (4). The “credit” in this study is defined as the provision of farming inputs on credit, and “loan” is defined as the money LBCs lend to farmers for all purposes, including the expense of daily lives. The AEZ dummies indicate which of the cocoa AEZs the sampled villages fall within as classified by Bunn et al. [15]. The LBC dummies indicate which LBCs cocoa farmers sold their cocoa beans to within the village. Table 2 specifies all variables used in the regression models.
Description of variables used in the regression analyses
| Variable | Description | Measurement |
|---|---|---|
| Credit source_LBC | Received inputs on credit from LBC | 1 = if yes, 0 = otherwise |
| Loan source_LBC | Received Loan from LBC | 1 = if yes, 0 = otherwise |
| Loan or credit source_LBC | Received either loan or inputs on credit from LBC | 1 = if yes, 0 = otherwise |
| Price difference | Whether there is difference in prices paid for cocoa beans | 1 = if yes, 0 = otherwise |
| Competition (HHI) | Competition among LBCs at village level | Herfindahl-Hirschman index |
| Education | Respondent’s highest formal education attained | 0 = no formal education, 1 = basic education, 2 = Junior high/middle school, 3 = senior high school/A/O level, 4 = University 5 = technical/ vocational college, 6 = other |
| Gender | Sex of respondent | 1 = male 0 = female |
| Age | Age of respondent | Years |
| Cocoa training within 5 years | Whether respondent received training on cocoa within the previous 5 years | 1 = if yes, 0 = otherwise |
| Leadership position | Whether respondent has a leadership position | 1 = if yes, 0 = otherwise |
| Cocoa production group | Whether respondent is a member of a cocoa production group | 1 = if yes, 0 = otherwise |
| Cocoa certification | Whether cocoa farm is certified | 1 = if yes, 0 = otherwise |
| Household own land | Whether household owned land | 1 = if yes, 0 = otherwise |
| Household income from remittances | Whether household received income from remittance | 1 = if yes, 0 = otherwise |
| Cocoa farm as main income source | Whether cocoa farm is the main source of household income | 1 = if yes, 0 = otherwise |
| Agro_Eco_Type 1 | Agro-ecological zone with low precipitation and strong dry season | 1 = if yes, 0 = otherwise |
| Agro_Eco_Type 2 | Agro-ecological zone with cooler temperatures and long dry season | 1 = if yes, 0 = otherwise |
| Agro_Eco_Type 3 | Agro-ecological zone with higher temperatures and reliable precipitation | 1 = if yes, 0 = otherwise |
| Agro_Eco_Type 4 | Agro-ecological zone with high annual average temperature and weak dry season with high precipitation | 1 = if yes, 0 = otherwise |
| PBC | 1 = if yes, 0 = otherwise | |
| Olam | 1 = if yes, 0 = otherwise | |
| Kuapa cocoa | 1 = if yes, 0 = otherwise | |
| LBC_category_1 | LBCs who bought cocoa beans from 60 to 80 farmers | 1 = if yes, 0 = otherwise |
| LBC_category_2 | LBCs who bought cocoa beans from 10 to 50 farmers | 1 = if yes, 0 = otherwise |
| LBC_category_3 | LBCs marked as “others” and other non-licensed buyers | 1 = if yes, 0 = otherwise |
Source: Authors’ construct from KIT data [14].
Casaburi and Reed [12] argued that advanced payment is a strategic tool for traders under competition to attract farmers and a substitute for output prices. LBCs cannot change the fixed producer price set by COCOBOD but can offer different prices to farmers depending on the quality of the beans. For example, in October 2022, slated cocoa beans locally called “abinchi” were sold at 3 cedis per kilogram compared to 10.56 cedis per kilogram for grades 1 and 2 cocoa beans. LBCs may maintain the substitutability between the price and loan provisions by intentionally using price differentials by quality. In such a case, farmers who receive credits or loans are more likely to face differential prices because of the intentional use of quality checks by LBCs. The promoting effects of competition on loan provisions would also increase the probability of price differentials in the villages with higher competition. Therefore, the objective of introducing the fourth dependent variable, the dummy for the price difference between the highest and lowest prices, is to test this strategic relationship.
According to Bunn et al. [15], the suitability of a site for cocoa production is defined by both climatic and soil attributes. Figure 1 presents the AEZs for cocoa production and the sampled villages used in this study.
Map of Ghana showing AEZs for cocoa production and sampled villages with the number of cocoa buyers (LBCs) present. Source: Authors’ illustration with reference to the study by Bunn et al. [15].
They identified the AEZs in a North–South order from 1 to 4 after extrapolation: Type 1 has low annual precipitation and a strong dry season, typical of the Brong-Ahafo Region. Type 2 has cooler temperatures and a rather long dry season, typical of the central part of the Ashanti Region. Type 3 has higher temperatures and reliable precipitation throughout the year, typical of the Eastern Region, and lastly, Type 4 has high annual average temperatures and a weak dry season with high annual precipitation which is typical of the Western and Central regions.
In this study, we used dummies for AEZ as a control variable to cater to the difference in regional cocoa production, as cocoa productivity depends on the suitability of the soil and climatic conditions [15]. Therefore, villages with favorable climatic and soil conditions are more likely to have higher cocoa production which turns to attract more LBCs, hence increasing competition. We assumed that, within the same AEZ, the variation in competition is randomly assigned, such that competition might be the same if the degree of competition depends only on cocoa production. In that case, AEZ-fixed effects may fully control for the endogeneity of competition.
Similarly, we used dummies of LBCs the individual farmers sold to as fixed effects to control for the endogeneity of competition arising from selection problems. As the LBCs are not the same across all the villages, some LBCs might prefer to operate in more productive villages where competition is high and so might need to use strategies like the provision of credit and loans to farmers to compete with other LBCs in the same village. Other LBCs may prefer to operate in niche areas where competition is low and may not need to provide credit and loans to farmers. These types of selection strategies used by LBCs affect the supply of credit and loans to farmers. Furthermore, the LBCs may have their own preferred strategies to attract farmers regardless of the competition in the village. Well-capitalized LBCs may be more aggressive in credit and loan provisions to farmers, while less-capitalized LBCs may emphasize community support and other activities. In model (4), we introduced the dummy variables representing the LBCs to which the farmers sell cocoa beans. Most LBCs work in several villages, and controlling LBCs' fixed effects allows us to reduce the effects of differences in LBC characteristics on credit and loan provisions. In creating the LBC dummies used in the regression analyses, LBCs to which more than 80 farmers sold their cocoa beans were made to stand alone while LBCs to which less than 80 farmers sold their cocoa beans were grouped. This is because some LBCs had a small number of farmers selling their cocoa beans to them. Table 2 presents the details of the LBCs categorization.
3 Results and discussion
3.1 Descriptive analysis of loans and inputs received on credit
It was found that of the total respondents, cocoa farmers who received loans and credit accounted for 26.8 and 12.1%, respectively (Table 3, Panel A). There are two possible reasons for the unexpected lower rate of loan and credit. First, under the government support programs through COCOBOD, subsidized fertilizers, pesticides, and free seedlings were provided for cocoa farmers to a larger extent [16]. Second, many cocoa farmers lacked collateral, making it difficult to establish trust with lenders. Nonetheless, LBCs/purchasing clerks were the highest source of loans and credit to the farmers, accounting for 37.4 and 54.7%, respectively (Table 3, Panel B). A study by Bannor et al. [11] indicated that LBCs provide soft loans to cocoa farmers to balance pertinent livelihood issues, especially during off-seasons. These observations assure that LBCs are the main suppliers of loans and credit for cocoa farmers in Ghana.
Share of cocoa farmers who received loan and credit and their sources
| Loan | Credit | |||
|---|---|---|---|---|
| Frequency | Percentage | Frequency | Percentage | |
| Panel A: Share of cocoa farmers who received Loan and Credit | ||||
| Yes | 353 | 26.8 | 159 | 12.1 |
| No | 965 | 73.2 | 1,159 | 87.9 |
| Total | 1,318 | 100 | 1,318 | 100 |
| Panel B: Sources of loan and credit | ||||
| Family and Friends | 105 | 29.7 | — | — |
| Village money lender | 25 | 0.3 | — | — |
| Village Saving and Loans | 1 | 2.0 | — | — |
| Savings and credit cooperative/credit union | 7 | 2.0 | — | — |
| Microfinance institution | 9 | 2.5 | — | — |
| Bank | 68 | 19.3 | — | — |
| Trader | 3 | 0.8 | — | — |
| Farmer organization | — | — | 3 | 1.9 |
| Input supplier or COCOBOD | — | — | 54 | 34.0 |
| NGO | — | — | 1 | 0.6 |
| LBC or Purchasing clerk | 132 | 37.4 | 87 | 54.7 |
| Other companies | 3 | 0.8 | 11 | 6.9 |
| Other sources | 5 | 1.4 | 3 | 1.9 |
Some respondents had more than one source for loan. Source: Authors’ calculations from KIT data [14].
3.2 Regression analyses
Table 4 presents summary statistics for the variables used in the regression analysis, and Table 5 summarizes the estimation results for the effect of competition on inputs received on credit, loan, loan or credit, and cocoa price differential. The HHI has an inverse relationship with competition, this implies that the higher the HHI value, the lower the competition in the market. Therefore, results show a positive relationship between competition and provision of credit, loan, and loan or credit in all specifications, however not statistically significant in all. The coefficients for credit (Panel A) and for loan (Panel B) show weak statistical significance only in models (3) and (1), respectively. The coefficients for the loan or credit (panel C) show weak statistical significance across model (1) to (3).
Summary statistics of variables used in the regression model
| Variable | Observation | Mean | Std. Dev. | Min | Max |
|---|---|---|---|---|---|
| Dependent Variables | |||||
| LBC as a source of loan dummy | 1,318 | 0.10 | 0.30 | 0.00 | 1.00 |
| LBC as a source of credit dummy | 1,318 | 0.07 | 0.25 | 0.00 | 1.00 |
| LBC as a source of credit or loan dummy | 1,318 | 0.15 | 0.36 | 0.00 | 1.00 |
| Price difference dummy | 1,318 | 0.19 | 0.40 | 0.00 | 1.00 |
| Explanatory Variables | |||||
| Competition (HHI) | 1,318 | 0.55 | 0.18 | 0.28 | 1.00 |
| Controlled Variables | |||||
| Respondent Education | 1,318 | 1.55 | 1.18 | 0.00 | 6.00 |
| Respondent Gender | 1,318 | 1.33 | 0.47 | 1.00 | 2.00 |
| Respondent Age | 1,316 | 50.59 | 13.34 | 22.00 | 97.00 |
| Farmer received cocoa training within past 5years | 1,318 | 0.49 | 0.50 | 0.00 | 1.00 |
| Leadership position | 1,318 | 0.29 | 0.45 | 0.00 | 1.00 |
| Respondent is member of cocoa production group | 1,318 | 0.10 | 0.29 | 0.00 | 1.00 |
| Cocoa farm is certified | 1,318 | 0.35 | 0.48 | 0.00 | 1.00 |
| Household own land | 1,318 | 0.78 | 0.41 | 0.00 | 1.00 |
| Household income from remittances | 1,318 | 0.16 | 0.37 | 0.00 | 1.00 |
| Cocoa farm as main income source | 1,317 | 0.85 | 0.36 | 0.00 | 1.00 |
| Agro_Eco_Type 1 | 1,318 | 0.15 | 0.36 | 0.00 | 1.00 |
| Agro_Eco_Type 2 | 1,318 | 0.20 | 0.40 | 0.00 | 1.00 |
| Agro_Eco_Type 3 | 1,318 | 0.19 | 0.39 | 0.00 | 1.00 |
| Agro_Eco_Type 4 | 1,318 | 0.46 | 0.50 | 0.00 | 1.00 |
| PBC | 1,200 | 0.57 | 0.50 | 0.00 | 1.00 |
| Olam | 1,200 | 0.12 | 0.32 | 0.00 | 1.00 |
| Kuapa cocoa | 1,200 | 0.18 | 0.39 | 0.00 | 1.00 |
| LBC_category_1 | 1,200 | 0.11 | 0.32 | 0.00 | 1.00 |
| LBC_category_2 | 1,200 | 0.08 | 0.28 | 0.00 | 1.00 |
| LBC_category_3 | 1,200 | 0.11 | 0.31 | 0.00 | 1.00 |
Note: For the LBC dummies, only 1,200 cocoa farmers responded to have sold their cocoa beans to LBCs.
Source: Authors’ calculations from KIT data [14].
Regression estimates of competition on dependent variables
| (1) | (2) | (3) | (4) | |
|---|---|---|---|---|
| Panel A: Input received on credit from LBCs | ||||
| Competition (HHI) | −0.06 | −0.07 | −0.08* | −0.07 |
| (0.05) | (0.05) | (0.05) | (0.05) | |
| Household characteristics | No | Yes | Yes | Yes |
| Agro-ecological zones Fixed Effect | No | No | Yes | Yes |
| LBCs Fixed Effect | No | No | No | Yes |
| Constant | 0.10*** | 0.18*** | 0.15*** | 0.17*** |
| (0.03) | (0.05) | (0.05) | (0.05) | |
| R 2 | 0.002 | 0.023 | 0.030 | 0.057 |
| Panel B: Loan received from LBCs | ||||
| Competition (HHI) | −0.11* | −0.09 | −0.08 | −0.07 |
| (0.06) | (0.05) | (0.06) | (0.07) | |
| Household characteristics | No | Yes | Yes | Yes |
| Agro-ecological Zones Fixed Effect | No | No | Yes | Yes |
| LBCs Fixed Effect | No | No | No | Yes |
| Constant | 0.16*** | 0.12** | 0.10* | 0.10* |
| (0.04) | (0.05) | (0.05) | (0.06) | |
| R 2 | 0.004 | 0.017 | 0.018 | 0.024 |
| Panel C: Loan or credit received from LBCs | ||||
| Competition (HHI) | −0.14* | −0.14* | −0.14* | −0.13 |
| (0.08) | (0.08) | (0.08) | (0.08) | |
| Household characteristics | No | Yes | Yes | Yes |
| Agro-ecological zones Fixed Effect | No | No | Yes | Yes |
| LBCs Fixed Effect3 | No | No | No | Yes |
| Constant | 0.23*** | 0.25*** | 0.21*** | 0.22*** |
| (0.05) | (0.07) | (0.07) | (0.07) | |
| R 2 | 0.005 | 0.019 | 0.023 | 0.033 |
| Panel D: Price difference | ||||
| Competition (HHI) | 0.03 | 0.05 | 0.06 | 0.04 |
| (0.08) | (0.08) | (0.08) | (0.08) | |
| Household characteristics | No | Yes | Yes | Yes |
| Agro-ecological zones Fixed Effect | No | No | Yes | Yes |
| LBCs Fixed Effect | No | No | No | Yes |
| Constant | 0.18*** | 0.24** | 0.16* | 0.26** |
| (0.05) | (0.10) | (0.10) | (0.10) | |
| R 2 | 0.000 | 0.077 | 0.085 | 0.138 |
| N | 1,318 | 1,315 | 1,315 | 1,198 |
Source: Authors’ calculations from KIT data [14]. Cluster robust standard errors in parentheses; * p < 0.10, ** p < 0.05, *** p < 0.01.
As discussed in subsection 2.3, we expect that the promoting effects of competition on the provision of credit and loans (which is a form of advance payment in the case of Ghana’s cocoa market) would increase the probability of price differentials in the villages with higher competition. The estimated coefficients for “price difference” are positive in all specifications but statistically insignificant. Therefore, our analysis does not support the substitutive relationship between farmgate price and advanced payments as proposed by Casaburi and Reed [12].
The results of our basic regression models do not fully support the standard intuition that competition promotes credit and loan provisions. To be sure that we have adequately captured the characteristics of the Ghanaian cocoa market, we further estimated some alternative models that allow us to examine the competition structure in more detail.
3.2.1 Accounting for non-linear relationship between competition and loan and credit provision
The two conjectures on the effects of competition on advanced payments including credit and loan provisions in the literature may suggest a non-linear relationship between competition and advanced payments. If a monopolistic buyer offers inputs on credit or loan before harvest to keep ties with sellers, the entry of a few competitive buyers will break that relational contract between the buyers and sellers leaving the sellers no better off [9]. In contrast, Casaburi and Reed [12] argue that farm-gate prices and advance payments are substitute strategies for each other to attract farmers. They found that the provision of financial services (credit and loans) to cocoa farmers does not coincide with the market power which on the contrary is believed to sustain relational contracts between buyers and sellers. Following this route, the entry of more competitive buyers would increase loan and credit provisions, especially in an environment of limited price competition such as Ghana’s cocoa market. By integrating these two conjectures, competition may inhibit loan and credit provisions while the degree of competition is low. However, as competition increases, it may promote loan and credit provisions.
Therefore, we add the quadratic term of HHI to the regression models (1)–(4) to account for the non-linear relationships between competition and loan and credit provisions. As shown in Table A1, in the models for the loan (Panel B) and the loan or credit provisions (Panel C), the coefficients of HHI are negative, and the coefficients of the quadratic term are positive. The magnitudes of the latter coefficients are generally larger than half of the former, suggesting that the value of HHI that gives the minimum value of the dependent variable lies between zero and one. The findings are consistent with the above conjectures; however, the estimated coefficients are not statistically significant.
3.2.2 PBC as the main competitor for private LBCs after partial liberalization
Before the partial liberalization of Ghana’s cocoa market in 1992, PBC was the sole buyer of cocoa beans from farmers. After the market liberalization when competition was introduced, the main competitors for private LBCs seeking to enter the local cocoa markets might have been the PBC rather than other private LBCs. In such case, the HHI calculated using the share of all active LBCs in the village, including the PBC, may not reflect the competitive structure in the cocoa market, and if private LBCs provide loans and credit to farmers to compete with PBC, the share of PBC in the village would be a more accurate proxy for the competition structure. Therefore, we estimate the models with the share of PBC in the village instead of the HHI as a proxy for competition. The estimation results are summarized in Table A2. These estimation results are almost similar to those of the basic models. As with the basic models, the coefficients of the share of PBC show negative signs for the dependent variables “credit,” “loan,” and “loan/credit.” The coefficients are significant for “loan or credit” in the models without AEZ and LBC fixed effects. However, when we controlled for these fixed effects, the coefficients lost statistical significance.
The negative coefficients of the share of PBC for the provision of credit and loans suggest that LBCs may offer more loans and credit in the villages where the share of PBC is low. This result is consistent with the findings of [10] and [17] who indicated that Private LBCs may use loans and credit to compete among themselves but not with the government-owned buyer, PBC. Also, Bannor et al. [11] indicated that access to credit decreased the likelihood of cocoa farmers choosing PBC as a marketing outlet by 23.04%. To account for this possibility, we estimate the basic models using the sub-sample, which excludes the farmers who sold their beans to only PBC. The estimation results are summarized in Table A3.
We found some differences between the estimated results using full sample and sub-sample. When we exclude the sample of cocoa farmers who sold to only the PBC, the estimated coefficients of competition for “credit” tend to increase in size regardless of the estimation models. The coefficients are significant at 10% in models (2) and (3), and this is partly due to the efficiency loss of estimation with the reduced sample size. No such consistent tendency is found in the estimates for loans. In highly competitive villages, private LBCs who are highly conscious of collecting high-quality beans may have the incentive to differentiate prices to increase the volume of high-quality beans they buy from cocoa farmers, yet the cocoa marketing rules in Ghana do not give much room for the LBCs to implement such a price policy [18,19]; hence, they may compete with other private LBCs on the basis of offering more inputs on credit. On the other hand, since loan is used equally for competition between private LBCs and the PBC, excluding farmers who sell only to the PBC may not have affected the estimation results.
4 Conclusion
The study assessed the relationship between competition among cocoa bean buyers (LBCs) in Ghana and the provision of loans and credit to cocoa farmers. The results of this study indicated that majority (54.7%) of cocoa farmers who received input on credit got it from LBC.s. Likewise, 37.4% of cocoa farmers who received loan got it from LBCs. However, regardless of the estimation models used our results did not demonstrate a strong effect of competition on the credit and loan provision. This indicates the limited effectiveness of the use of credit and loan provisions to compete among LBCs in Ghana’s cocoa market.
The significance of the results of this study lies in the fact that LBCs’ competition under the partial liberalization of the cocoa bean market in Ghana is limited by the conditions of the government-intervened production factor market. This study initially assumed that under the government's fixed cocoa prices, LBCs would enhance non-price competition behavior in purchasing cocoa beans from cocoa farmers in the form of provision of loans and credit, but such LBC behavior was not statistically confirmed. A possible reason for this could be that although loans and credit are generally required to purchase production inputs, in Ghana, some production inputs (seedlings, fertilizer, and pesticides) are provided by the government at free or subsidized prices. By doing so, the government might aim to avoid the drastic change in the structure of cocoa production, presumably reducing the effectiveness of the LBCs' competition strategy for cocoa farmers and helping decelerate restructuring or polarization of the cocoa production structure that would be accompanied by increased income disparities among cocoa farmers. This suggests that the relationship between product and factor markets needs to be carefully considered when the government considers a change in policy that would enhance a particular output production or foster competition among private agents in a particular product market, not just the case of cocoa in Ghana.
It is therefore recommended that before governments decide to change policy on input subsidy, they must consider the effect of such policies on the relationship between LBCs and farmers. Specifically, a reduction in input subsidies may increase the provision of credit by LBCs, because that can increase the farmers’ demand for credit. As the theory of relational contract argues; the strong tie between LBCs and farmers may enhance the bargaining power of LBCs and this may weaken the feasibility of a fixed cocoa price policy.
This study is limited to the fact that data on the number of farmers which sold to individual buyers was used to calculate the market share of the cocoa buyers instead of data on the quantity of cocoa beans sold to individual buyers. This is due to the unavailability of such data in the dataset used. This may not give the best representation of buyers’ market share; it is therefore advised that future studies use the most appropriate measure of buyers' market share.
This study provides some insights for future studies. First, although credit and loan provision are forms of advanced payments, they may have different functions when considered as a means of competition for LBCs; therefore, there is a need to explore pathways other than credit and loan provision as a form of competition by LBCs. Second, it would be necessary to examine in detail the differences in competition among private LBCs and competition between private LBCs and the PBCs.
Acknowledgments
We are grateful to The Royal Tropical Institute (KIT) for availing their data online for research use. We also appreciate Professor Christian Bunn of International Center for Tropical Agriculture (CIAT) and his team for the permission to use their high-resolution map showing cocoa AEZs in Ghana [15].
-
Funding information: This work was supported by JSPS KAKENHI Grant Numbers 22K05846 and 19H03057.
-
Author contributions: SK: conceptualization, data preparation, methodology, formal analysis, writing- original draft preparation, writing-review and editing. AC: conceptualization, methodology, writing-review and editing, funding acquisition. MK: conceptualization, methodology, resources, writing-review and editing, funding acquisition. All authors have read and agreed to the published version of the manuscript.
-
Conflict of interest: The authors state no conflicts of interest.
-
Data availability statement: The datasets generated during and/or analyzed during the current study are available from the Harvard Dataverse repository, https://doi.org/10.7910/DVN/82TWZJ.
Appendixes
Regression estimates of competition on dependent variables with the quadratic term of HHI
| (1) | (1) | (2) | (2) | (3) | (3) | (4) | (4) | |
|---|---|---|---|---|---|---|---|---|
| Panel A: Inputs received on credit from LBCs | ||||||||
| Competition (HHI) | −0.06 | −0.22 | −0.08 | −0.12 | −0.08* | −0.04 | −0.07 | 0.03 |
| (0.05) | (0.36) | (0.05) | (0.34) | (0.05) | (0.32) | (0.05) | (0.31) | |
| Competition squared | 0.12 | 0.03 | −0.04 | −0.08 | ||||
| (0.26) | (0.24) | (0.23) | (0.23) | |||||
| Household x’tics | No | Yes | Yes | Yes | ||||
| Agro-ecological Zones Fixed Effect | No | No | Yes | Yes | ||||
| LBCs Fixed Effect | No | No | No | Yes | ||||
| Constant | 0.10*** | 0.14 | 0.18*** | 0.19 | 0.15*** | 0.14 | 0.17*** | 0.15 |
| (0.03) | (0.12) | (0.05) | (0.12) | (0.05) | (0.11) | (0.05) | (0.12) | |
| R 2 | 0.002 | 0.002 | 0.023 | 0.023 | 0.030 | 0.030 | 0.057 | 0.057 |
| Panel B: Loan received from LBCs | ||||||||
| Competition (HHI) | −0.11* | −0.42 | −0.09 | −0.47 | −0.08 | −0.48 | −0.08 | −0.55 |
| (0.06) | (0.42) | (0.06) | (0.44) | (0.06) | (0.47) | (0.07) | (0.52) | |
| Competition squared | 0.25 | 0.30 | 0.32 | 0.37 | ||||
| (0.30) | (0.31) | (0.34) | (0.37) | |||||
| Household x’tics | No | Yes | Yes | Yes | ||||
| Agro-ecological Zones Fixed Effect | No | No | Yes | Yes | ||||
| LBCs Fixed Effect | No | No | No | Yes | ||||
| Constant | 0.16*** | 0.25* | 0.12** | 0.22* | 0.10* | 0.21 | 0.10* | 0.23 |
| (0.04) | (0.14) | (0.05) | (0.13) | (0.05) | (0.14) | (0.06) | (0.16) | |
| R 2 | 0.004 | 0.005 | 0.017 | 0.018 | 0.018 | 0.019 | 0.024 | 0.026 |
| Panel C: Loan or credit received from LBCs | ||||||||
| Competition (HHI) | −0.14* | −0.54 | −0.14* | −0.51 | −0.14* | −0.47 | −0.13 | −0.49 |
| (0.08) | (0.53) | (0.08) | (0.52) | (0.08) | (0.54) | (0.08) | (0.54) | |
| Competition squared | 0.316 | 0.291 | 0.263 | 0.290 | ||||
| (0.377) | (0.370) | (0.377) | (0.379) | |||||
| Household x’tics | No | Yes | Yes | Yes | ||||
| Agro-ecological zones Fixed Effect | No | No | Yes | Yes | ||||
| LBCs Fixed Effect | No | No | No | Yes | ||||
| Constant | 0.23*** | 0.35* | 0.25*** | 0.35* | 0.21*** | 0.30* | 0.22*** | 0.32* |
| (0.05) | (0.18) | (0.07) | (0.18) | (0.07) | (0.17) | (0.10) | (0.18) | |
| R 2 | 0.005 | 0.006 | 0.019 | 0.020 | 0.023 | 0.023 | 0.033 | 0.034 |
| Panel D: Price difference | ||||||||
| Competition (HHI) | 0.03 | 0.14 | 0.05 | 0.16 | 0.07 | 0.03 | 0.04 | −0.33 |
| (0.08) | (0.68) | (0.08) | (0.66) | (0.08) | (0.65) | (0.08) | (0.65) | |
| Competition squared | −0.09 | −0.09 | 0.03 | 0.29 | ||||
| (0.50) | (0.48) | (0.49) | (0.51) | |||||
| Household x’tics | No | Yes | Yes | Yes | ||||
| Agro-ecological zones Fixed Effect | No | No | Yes | Yes | ||||
| LBCs Fixed Effect | No | No | No | Yes | ||||
| Constant | 0.18*** | 0.15 | 0.24** | 0.21 | 0.16* | 0.17 | 0.26** | 0.36* |
| (0.05) | (0.20) | (0.10) | (0.20) | (0.09) | (0.20) | (0.10) | (0.20) | |
| R 2 | 0.000 | 0.000 | 0.077 | 0.077 | 0.085 | 0.085 | 0.138 | 0.138 |
| N | 1,318 | 1,318 | 1,315 | 1,315 | 1,315 | 1,315 | 1,198 | 1,198 |
Notes: Household x’tics; household characteristics, cluster robust standard errors in parentheses; * p < 0.10, ** p < 0.05, *** p < 0.01.
Source: Authors’ calculations from KIT data [14].
Regression estimates of competition on dependent variables with SHARE of PBC in the village as a proxy for competition
| (1) | (2) | (3) | (4) | |
|---|---|---|---|---|
| Panel A: Input received on credit from LBCs | ||||
| Competition (share of PBC) | −0.0521 | −0.0588 | −0.0768 | −0.0604 |
| (0.0442) | (0.0448) | (0.0455) | (0.0489) | |
| Household characteristics | No | Yes | Yes | Yes |
| Agro-ecological Zones Fixed Effect | No | No | Yes | Yes |
| LBCs Fixed Effect | No | No | No | Yes |
| Constant | 0.0955*** | 0.169*** | 0.153*** | 0.172*** |
| (0.0317) | (0.0516) | (0.0472) | (0.0597) | |
| R 2 | 0.002 | 0.023 | 0.030 | 0.057 |
| Panel B: Loan received from LBCs | ||||
| Competition (share of PBC) | −0.104* | −0.0863 | −0.0853 | −0.0937 |
| (0.0605) | (0.0600) | (0.0640) | (0.0729) | |
| Household characteristics | No | Yes | Yes | Yes |
| Agro-ecological Zones Fixed Effect | No | No | Yes | Yes |
| LBCs Fixed Effect | No | No | No | Yes |
| Constant | 0.159*** | 0.124** | 0.104* | 0.114* |
| (0.0419) | (0.0505) | (0.0527) | (0.0579) | |
| R 2 | 0.005 | 0.018 | 0.019 | 0.026 |
| Panel C: Loan or credit received from LBCs | ||||
| Competition (share of PBC) | −0.137* | −0.129* | −0.138 | −0.253 |
| (0.0717) | (0.0734) | (0.0833) | (0.499) | |
| Household characteristics | No | Yes | Yes | Yes |
| Agro-ecological Zones Fixed Effect | No | No | Yes | Yes |
| LBCs Fixed Effect3 | No | No | No | Yes |
| Constant | 0.231*** | 0.246*** | 0.213*** | 0.261 |
| (0.0509) | (0.0708) | (0.0702) | (0.164) | |
| R 2 | 0.006 | 0.020 | 0.024 | 0.034 |
| Panel D: Price difference | ||||
| Competition (share of PBC) | 0.0107 | 0.0274 | 0.0603 | 0.0753 |
| (0.0786) | (0.0723) | (0.0788) | (0.0776) | |
| Household characteristics | No | Yes | Yes | Yes |
| Agro-ecological Zones Fixed Effect | No | No | Yes | Yes |
| LBCs Fixed Effect | No | No | No | Yes |
| Constant | 0.189*** | 0.247*** | 0.158* | 0.227** |
| (0.0472) | (0.0858) | (0.0870) | (0.0933) | |
| R 2 | 0.000 | 0.077 | 0.085 | 0.139 |
| N | 1,318 | 1,315 | 1,315 | 1,198 |
Notes: Cluster robust standard errors in parentheses; * p < 0.10, ** p < 0.05, *** p < 0.01.
Source: Authors’ calculations from KIT data [14].
Regression estimates of competition on-dependent variables excluding farmers who sold their cocoa beans to ONLY PBC
| (1) | (1) | (2) | (2) | (3) | (3) | (4) | (4) | |
|---|---|---|---|---|---|---|---|---|
| Full Sample | Only PBC excluded | Full Sample | Only PBC excluded | Full Sample | Only PBC excluded | Full Sample | Only PBC excluded | |
| Panel A: Credit from LBCs | ||||||||
| Competition (HHI) | −0.06 | −0.10 | −0.08 | −0.11* | −0.08* | −0.11* | −0.07 | −0.08 |
| (0.05) | (0.06) | (0.05) | (0.06) | (0.05) | (0.07) | (0.05) | (0.06) | |
| Household x’tics | No | Yes | Yes | Yes | ||||
| Agro-ecological Zones Fixed Effect | No | No | Yes | Yes | ||||
| LBCs Fixed Effect | No | No | No | Yes | ||||
| Constant | 0.10*** | 0.12*** | 0.18*** | 0.18*** | 0.15*** | 0.13** | 0.17*** | 0.16** |
| (0.03) | (0.04) | (0.05) | (0.06) | (0.05) | (0.05) | (0.05) | (0.07) | |
| R 2 | 0.002 | 0.004 | 0.023 | 0.022 | 0.030 | 0.033 | 0.057 | 0.069 |
| Panel B: Loan received from LBCs | ||||||||
| Competition (HHI) | −0.11* | −0.10 | −0.10 | −0.10 | −0.08 | −0.09 | −0.08 | −0.08 |
| (0.06) | (0.07) | (0.05) | (0.07) | (0.06) | (0.08) | (0.07) | (0.08) | |
| Household x’tics | No | Yes | Yes | Yes | ||||
| Agro-ecological Zones Fixed Effect | No | No | Yes | Yes | ||||
| LBCs Fixed Effect | No | No | No | Yes | ||||
| Constant | 0.16*** | 0.16*** | 0.12** | 0.15** | 0.10* | 0.15** | 0.10* | 0.19** |
| (0.04) | (0.04) | (0.05) | (0.06) | (0.05) | (0.06) | (0.06) | (0.07) | |
| R 2 | 0.004 | 0.002 | 0.017 | 0.013 | 0.018 | 0.013 | 0.024 | 0.022 |
| Panel C: Loan or credit received from LBCs | ||||||||
| Competition (HHI) | −0.14* | −0.15 | −0.14* | −0.16* | −0.14* | −0.16 | −0.12 | −0.13 |
| (0.08) | (0.09) | (0.08) | (0.09) | (0.08) | (0.10) | (0.08) | (0.10) | |
| Household x’tics | No | Yes | Yes | Yes | ||||
| Agro-ecological Zones Fixed Effect | No | No | Yes | Yes | ||||
| LBCs Fixed Effect | No | No | No | Yes | ||||
| Constant | 0.23*** | 0.24*** | 0.25*** | 0.27*** | 0.21*** | 0.23** | 0.22*** | 0.26*** |
| (0.05) | (0.06) | (0.07) | (0.09) | (0.07) | (0.09) | (0.07) | (0.09) | |
| R 2 | 0.005 | 0.005 | 0.019 | 0.017 | 0.023 | 0.019 | 0.033 | 0.033 |
| Panel D: Price difference | ||||||||
| Competition (HHI) | 0.03 | 0.08 | 0.05 | 0.07 | 0.07 | 0.07 | 0.04 | 0.05 |
| (0.08) | (0.10) | (0.08) | (0.08) | (0.08) | (0.08) | (0.08) | (0.11) | |
| Household x’tics | No | Yes | Yes | Yes | ||||
| Agro-ecological Zones Fixed Effect | No | No | Yes | Yes | ||||
| LBCs Fixed Effect | No | No | No | Yes | ||||
| Constant | 0.178*** | 0.146 | 0.237** | 0.206 | 0.159* | 0.167 | 0.258** | 0.359* |
| (0.05) | (0.20) | (0.10) | (0.20) | (0.09) | (0.20) | (0.10) | (0.20) | |
| R 2 | 0.000 | 0.001 | 0.077 | 0.112 | 0.085 | 0.116 | 0.138 | 0.187 |
| N | 1,318 | 908 | 1,315 | 906 | 1,315 | 906 | 1,198 | 789 |
Notes: Household x’tics; household characteristics cluster robust standard errors in parentheses; *p < 0.10, **p < 0.05, ***p < 0.01.
Source: Authors’ calculations using KIT data [14].
References
[1] Akiyama T, Baffes J, Larson D, Varangis P, Akiyama T, Baffes J, et al. Commodity market reform in Africa: some recent experience. Econ Syst. 2003;27(1):83–115. https://econpapers.repec.org/RePEc:eee:ecosys:v:27:y:2003:i:1:p:83-115.10.1016/S0939-3625(03)00018-9Search in Google Scholar
[2] Kolavalli S, Vigneri M, Maamah H, Poku J. The partially liberalized cocoa sector in Ghana: Producer price determination, quality control, and service provision. SSRN Electron J. 2013. http://dx.doi.org/10.2139/ssrn.2198609.10.2139/ssrn.2198609Search in Google Scholar
[3] Brooks J, Croppenstedt A, Aggrey-Fynn E. Distortions to agricultural incentives in Ghana. In: Anderson K, Masters W, editors. Distortions to Agricultural Incentives in Africa. World Bank Washington, DC: World Bank’s Development Research Group; 2007. p. 413–440. https://ageconsearch.umn.edu/record/48523.Search in Google Scholar
[4] COCOBOD. 51st Annual reports and consolidated financial statements-2020. Accra; 2020. https://cocobod.gh/resource_files/51st-annual-report-and-financial-statements-2019-2020.Search in Google Scholar
[5] Anang BT. Market structure and competition in the Ghanaian cocoa sector after partial liberalization. Curr Res J Soc Sci. 2011;3(6):465–70.Search in Google Scholar
[6] Laven A. Marketing reforms in Ghana’s cocoa sector. Partial liberalization, partial benefits? Background Note, Overseas Development Institute, London; 2007. http://www.odi.org.uk/resources/download/420.pdf.Search in Google Scholar
[7] Vigneri M, Santos P. What Does Liberalization without Price Competition Achieve? Proceeding of the International Association of Agricultural Economists’ Conference. Beijing, China; 2009.Search in Google Scholar
[8] Kyomugisha H, Mugisha J, Sebatta C. Potential determinants of profits and market efficiency of potato market chains in Uganda. J Agribus Dev Emerg Econ. 2017;7(1):52–68.10.1108/JADEE-06-2015-0031Search in Google Scholar
[9] Shiferaw BA, Obare GA, Geoffrey M, Silim S. Leveraging institutions for collective action to improve markets for smallholder producers in less-favored areas. Afr J Agric Resour Econ. 2009;3(1):1–8.Search in Google Scholar
[10] Varangis P, Schreiber G. Cocoa market reforms in West Africa. In: Akiyama T, Baffes J, Larson D, Varangis P, editors. Commodity market reforms: Lessons of two decades. Washington D.C.: The World Bank; 2001. p. 35–82.Search in Google Scholar
[11] Bannor RK, Oppong-Kyeremeh H, Atewene S, Wongnaa CA. Influence of non-price incentives on the choice of cocoa licensed buying companies by farmers in the Western North of Ghana. J Agribus Dev Emerg Econ. 2019;9(4):402–18.10.1108/JADEE-11-2018-0151Search in Google Scholar
[12] Casaburi L, Reed T. Using individual-level randomized treatment to learn about market structure. Am Econ J Appl Econ. 2022;14(4):58–90.10.1257/app.20200306Search in Google Scholar
[13] Macchiavello R, Morjaria A. Competition and relational contracts in the rwanda coffee chain. Q J Econ. 2021;136(2):1089–143.10.1093/qje/qjaa048Search in Google Scholar
[14] Tyszler M, Bymolt R, Laven A. Demystifying the cocoa sector in Ghana and Côte d’Ivoire. Havard Dataverse, V2, UNF:6:bKsOmmVySdVxa8Ksq2ajHQ== [fileUNF]; 2018. https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/82TWZJ.Search in Google Scholar
[15] Bunn C, Läderach P, Quaye A, Muilerman S, Noponen MRA, Lundy M. Recommendation domains to scale out climate change adaptation in cocoa production in Ghana. Clim Serv. 2019;16:1000123.10.1016/j.cliser.2019.100123Search in Google Scholar
[16] Takyi G. Cocoa farming and management. assessed on 2023 Sep 1; 2017. https://www.researchgate.net/publication/343431937_cocoa_farming_and_management_prepared_for_management.Search in Google Scholar
[17] Foundjem-Tita D, Donovan J, Stoian D, Degrande A. Baseline for assessing the impact of Fairtrade certification on cocoa farmers and cooperatives in Ghana. Nairobi, Kenya: World Agroforestry Centre; 2016. p. 1–104.Search in Google Scholar
[18] Quarmine W, Haagsma R, Sakyi-Dawson O, Asante F. Huis A Van, Obeng-Ofori D. Incentives for cocoa bean production in Ghana: Does quality matter? NJAS – Wagening J Life Sci. 2012;60–63(60):7–14.10.1016/j.njas.2012.06.009Search in Google Scholar
[19] United Nations Conference on Trade and Development. Agricultural commodity value chains: The effects of market concentration on farmers and producing countries – the case of cocoa. Geneva, Switzerland; 2016.Search in Google Scholar
© 2024 the author(s), published by De Gruyter
This work is licensed under the Creative Commons Attribution 4.0 International License.