Senegal: Research Findings

After several decades of heavy intervention in the groundnut marketing system, the Senegalese government initiated a reform process that was to allow participation by private traders and raise efficiency in the marketing system. A critical consideration behind public intervention in the groundnut sector has been the viability of the state-owned groundnut processing companies. This explains why the government chose a partial reform approach that consisted in:

1. keeping the processing sector outside of the reform process;

2. continuing its involvement in the procurement system of unshelled groundnuts through farmer cooperatives and a contract mechanism that allows its marketing board to control licensed traders;

3. further setting of pan-territorial and pan-temporal producer prices; and

4. not clearly removing extant, albeit not enforced, pre-reform restrictions on the movement of commodities.

• The Limits of Partial Liberalization
• Public Versus Private Marketing Sub-Systems
• Cost and Price Adjustment
• Liberalization, Profitability, and Competitiveness in the Processing Industry

The Limits of Partial Liberalization

The partial approach to reforms has led to a dualistic groundnut marketing system with, on the one hand, a state-controlled sub-system comprised of farm cooperatives and licensed official traders (OPS), and on the other hand, the large informal sub-sector of private traders. The official sub-system, cooperatives and OPS, dominates the procurement system of unshelled groundnuts, primarily destined to the oil mills. The private informal traders, while involved in the procurement of unshelled groundnuts, exclusively control the market for shelled groundnuts. The dichotomy of the marketing system, that to some extent is a legacy of groundnut marketing in Senegal, has been markedly reinforced by the partial approach to reforms. Partial liberalization has lead to a situation where two distinct but competing segments of the system are forced to carry out the bulk of their operations at different times of the year, handle different types of groundnut products, serve different geographic areas, procure from different sources, and sell to different clients. The reforms initiated more than ten years ago had, therefore, only limited, if any, positive impact on the performance of local groundnut markets.

Public Versus Private Marketing Sub-Systems

Senegal's groundnut marketing system has changed greatly during the period of liberalization. While allowing for greater participation from the private sector, the state, through the parastatals SONACOS and SONAGRAINES, still controls a large part of the groundnut marketing system. With the reforms initiated in the mid-eighties, however, the parastatals moved away from using their own agents and the official cooperative system (COOP) to contracting private licensed traders (OPS) to carry out procurement activities.

Even though the cooperatives are still part of the official marketing system, they have been losing market share rapidly to the licensed traders. Three years into the reform process, the OPS reached a 35 percent share of the procurement market. By the end of 1994, ten years after the introduction of the reforms, the share of the official farmers' cooperatives in the quantity of groundnuts marketed through the official channels fell from 65 to nearly 40 percent while that of the licensed private traders increased to 60 percent.

Perversely, the official groundnut marketing system allows the OPS and COOPs to operate only during the still officially determined marketing season. This reduces the competition between the two official channels to practically the 3 months during which the official marketing season is open. For the remaining 9 months of the year, marketing activities rest in the hands of the resurgent informal trading sector, which had been banned for the 30 years preceding the reforms. The majority of the informal traders, 56 percent of them, entered the market after the reforms. Moreover, 9 out of 10 traders feel that the effects of liberalization have been positive. More importantly, our survey data reveal that the informal sector has been strongly stimulated by the recent devaluation of the country's currency. The data indicate that investment by informal traders doubled in the first year after devaluation and that 83 percent of marketing equipment in use was purchased after that period. Despite the fact that agents in the official marketing systems (COOP and OPS) are active only during the official marketing season immediately after harvest, informal traders purchase between 60 and 70 percent of marketed quantities directly from farmers. The high percentage seems to indicate not only that informal traders may be competing with official traders during the official marketing season, when most of the harvest is sold, but it also suggests that some farmers may be withholding part of their production for sales outside of the official marketing season.

The described change in structure has also affected cost levels and losses in the groundnut marketing system. For the official marketing channels (COOP and OPS), the average cost per ton of marketed groundnuts, excluding transportation cost which is paid for by the two parastatals, amounts to CFA 930 and CFA 370, respectively. In the informal sector, average marketing costs including transportation amount to CFA 3,700 per ton. Deducting the cost of transport and taxes from marketing costs in the informal sector in order to make them comparable to that of the formal system, yields a ratio of 2 to 1 between the informal and OPS sectors and 5 to 1 between the informal sector and the COOPs. This seems to indicate that liberalization has been moving marketing out of the lower cost COOP channels to the higher cost OPS and informal channels. Some of this may be explained through the existence of diseconomies of scale in the informal sector. While the OPS operate through dispersed buying points with much lower procurement quantities per buying point, informal traders cover distances of up to 300 km with average loads of less than 6 tons. Moreover, over half of the informal traders frequent several weekly markets in different locations to buy and sell, which adds significantly to their operating costs.

Despite their higher operating costs, informal traders command an increasing albeit hitherto small share of the market because: i) they are free to cover wider geographic areas, whereas OPSs are assigned by the parastatals to designated areas, and COOPs can buy only from their members within a given area; ii) they sell to a bigger group of clients comprising private traders and consumers, whereas OPSs and COOPs sell only to the parastatals; iii) they are active on average 11 months a year, compared to 4 to 5 months for OPSs and COOPs; iv) they deal in various types of groundnuts, while OPSs and COOPs trade only in shelled groundnuts which are destined for the milling industry; and v) they pay in cash, while payment with COOPs and OPS often involves significant delays.

Cost and Price Adjustment

The impact of the partial and segmented approach to reforming groundnut markets is evident in the behavior of local prices. Tighter control of the market for unshelled, compared to shelled groundnuts has delinked the process of price formation between these two submarkets. Relative prices between these two types of groundnuts vary greatly between individual markets. The pattern of price variability also varies greatly, with unshelled groundnuts showing a considerably higher level of instability. Even though they have similar growing patterns, certain markets display quite dissimilar price seasonality patterns.

Price trends are also quite different across markets. During the 1990 - 1996 period covered by the study, prices for the same type of groundnuts have evolved quite differently across individual markets, rising significantly in some while falling in others. In general, however, farm prices and unshelled groundnut prices have risen more than retail prices and shelled groundnut prices, respectively, indicating that retailers have difficulty passing on price hikes to consumers. The major exception here is Dakar, the main consumption center, where retail prices for shelled groundnuts grew by more than twice that of unshelled groundnuts. It is most likely that the difference in price trends for the Dakar market is due to the many obstacles of transferring groundnuts between Dakar and the other markets, which are all located in producing areas.

As could already be expected from the significant divergences in relative price structures, spatial and temporal price spreads vary greatly between markets for similar distances and the same types of groundnuts. Price spreads also behave differently between the same pairs of markets for the two types of groundnuts. More interestingly, temporal (highest-to-lowest) price spreads seem to widen on average in all markets, which raises the question of increasing segmentation of local markets. In fact, the initial results of the analysis of market integration show no consistent patterns of market interdependence across local markets, particularly in the case of the controlled markets for unshelled groundnuts. No relationships could be detected between market interdependence and distance. The markets for shelled groundnuts on the other hand, exhibit a much higher level and more consistent pattern of integration across markets. The difference in behavior between the two types of markets, which itself seems to originate from the strong control of the markets for unshelled groundnuts by the state, may underlie the discrepancies in the patterns and spreads of local prices discussed earlier.

Liberalization, Profitability, and Competitiveness in the Processing Industry

The behavior of prices described above suggest that the reforms that were introduced in 1985 have not led to the establishment of an efficient marketing system. The pricing and marketing policies do not affect only the operation of markets but also the profitability of processing activities. Furthermore, these policies are associated with significant transfers of income between the sectors of production and processing. The effective liberalization of the groundnut sector would lead to a redistribution of the transfers between the two sectors. The success of further reforms in the sector would, consequently, depend on their redistributive effects and how these are managed. Hence, it is important to anticipate the transfer effect that would result from further liberalization of groundnut markets. The study has estimated the likely transfer effects using: a) actual producer prices, b) simulated mill-gate prices in case of liberalization based on a dynamic price time path model, c) the actual cost of processing in the milling industry, and d) the actual cost of marketing in the private trading sector.

In the context of liberalized and competitive markets, the simulated mill-gate price minus the cost of processing and the cost of marketing would determine the maximum producer price. The difference between that price and the actual producer price gives an idea of the volume of implicit transfer of income between the sectors of production and processing. In order to better illustrate the implications of policy changes, a retrospective analysis has been carried out, which simulates the effects that a liberalization of markets would have had on prices and procurement costs of the oil seed mills during the 1992-1997 period.

The simulation results indicate a significant level of transfer in favor of the producing sector from 1992 to the beginning of 1994, when the country's currency was devalued by 50 percent. The transfers resulted from the steep decline of world market prices combined with a gradual appreciation of the country's currency and relatively high but rigid producer prices. The average annual unit value of the transfers has been estimated at about 56-98 CFA francs per kilogram, compared to producer prices of around 60-70 CFA francs per kilogram. In other words, given the constellation of processing costs, world prices and the level of exchange rate during the years preceding the devaluation of the currency, the oil mills were losing for each kilogram of groundnuts that was processed more than the direct cost of these groundnuts.

These levels of transfers were possible only because the considerable amounts of subsidies by the government, which aimed at compensating the effects of marketing and pricing policies and of the CFA franc overvaluation on the procurement costs of the milling industry. It is interesting to note that the transfers and subsidies had not succeeded in stopping the erosion of incentives in the production sector nor did they reverse the chronic losses of the processing industry. The reason has been that the effects of the macroeconomic imbalances that had led to the sharp appreciation of the CFA franc and of the implicit taxation of producer prices through the prevailing marketing policies have more than compensated the transfers to producers and the subsidies to the processing industry over the years.

After the change in the exchange rate, the direction of transfers changed drastically in favor of the processing industries. The transfers in this case have resulted from the incomplete transmission of the change in the exchange rate to producer prices, while processing industries were exporting at higher world prices in terms of domestic currency. The average unit value of the transfers from producers to the processing industry since 1994 has been estimated at 10 to 30 percent of the producer price of each kilogram of groundnut that has been processed.

With the change in the direction of transfers that are associated with the country's pricing and marketing policies, the main issue in furthering the reform process that has been started in 1985 is with respect to the direct effect on the profitability and competitiveness of the processing industry. The simulation results indicate that the liberalization of producer prices would have raised the hypothetical price of groundnuts to the processing plants to levels above those of international prices. The simulation of the hypothetical price was carried out on the basis of the induced changes in producer prices that would have resulted from the liberalization of the marketing policies and the actual cost of operation within the private marketing sector between the producing zones and the regions where the processing plants are located. The ratio between the simulated price and the world market price during the period covered by the simulations indicates that the liberalization of markets without a substantial reduction in the costs of operations in the private marketing sector would have considerably reduced the competitiveness of the processing industries. The evolution of marketing costs in the context of further liberalization of domestic markets would, therefore, be of crucial importance as to the consequences of liberalization for the overall groundnut economy. The reduction of marketing costs would not only affect the procurement costs of processing plants, but would also create the scope for high producer prices.

The preceding discussion stresses the need to pay sufficient attention in the case of liberalization to the emergence of a competitive and lower-cost private marketing system. There are some indications of economies of scale in the marketing of groundnuts, which indicate that the unit costs in the sector would decrease with the increase of the average volume of market quantities by private traders. The private sector has a market share that is currently less than 15 percent of total marketed quantities. With increased participation, the average volume of their activities is expected to rise significantly and therefore lead to lower marketing costs overall. However, some of the main features of the reforms of 1985, such as the system of licensed traders (OPS) and the geographical quota and oligopolistic structure associated with it, are incompatible with the objective of lower unit cost of marketing. The same applies to the practices of wild taxation and the cumbersome and abusive controls which have been encouraged by the prevailing policy environment of the marketing sector and which unnecessarily raise the cost of operation by private traders.

Finally, the simulation results indicate that further liberalization and reduced marketing costs would not translate into significant gains in competitiveness for Senegal's oil exports, unless they are accompanied by improved efficiency and higher productivity in the milling sector. Unit costs in that sector seem to be substantially above that of international competitors. Without significant progress in the processing sector, the adjustment of domestic prices to their international levels would result in significant losses for the oil mills and a sharp decline in the competitiveness of Senegalese exports.

In contrast to the above findings, the recovery program that has been launched recently by the Senegalese government focuses exclusively on the production side. It is true that the problem of soil fertility and access to seed and fertilizer remains the main challenge in revitalizing the groundnut sector. One should, however, not lose sight of the fact that the problem of productivity cannot be resolved without the improvement of incentives which have to complement the technological efforts that so far have attracted the attention of the government. The improvement of incentives would necessitate the elimination of the implicit taxation resulting from the prevailing pricing and marketing policies and thus the liberalization of domestic markets. The liberalization in turn is conceivable only if adequate measures are taken in order to ensure a) a significant reduction of unit cost of operation in the private marketing sector, and b) a marked increase in the productivity and efficiency of processing mills.

• Collaboration
• Research Tasks
• Field Activities
• Research Findings

For further information please email ifpri-mti@cgiar.org or contact Markets, Trade and Institutions Division, IFPRI, 2033 K Street, N.W., Washington, D.C., 20006, U.S.A.