Abstract
Rapid export growth has been a key characteristic of fast-growing economies. The recent southeast Asian experience suggests that, along with improved international competitiveness, export growth plays a major role in fostering economic growth in developing countries. The importance of export promotion, as a means of achieving rapid economic growth, has gained wide acceptance in the literature. Outward-oriented trade strategies lead to faster growth because they improve resource allocation by promoting competition and encouraging learning-by-doing (Lucas (1988)). Moreover, improved access to trade opportunities enhances the positive externalities resulting from increased availability of technology (Romer (1986)). Krueger (1995) has argued that export promotion policies are superior to import substitution policies since they allow a country to extract efficiency gains from the scale of the world economy.
A. Introduction
Rapid export growth has been a key characteristic of fast-growing economies. The recent southeast Asian experience suggests that, along with improved international competitiveness, export growth plays a major role in fostering economic growth in developing countries. The importance of export promotion, as a means of achieving rapid economic growth, has gained wide acceptance in the literature. Outward-oriented trade strategies lead to faster growth because they improve resource allocation by promoting competition and encouraging learning-by-doing (Lucas (1988)). Moreover, improved access to trade opportunities enhances the positive externalities resulting from increased availability of technology (Romer (1986)). Krueger (1995) has argued that export promotion policies are superior to import substitution policies since they allow a country to extract efficiency gains from the scale of the world economy.
More and more policymakers in developing countries are emphasizing export diversification and rapid export growth as a means of creating employment and averting foreign exchange scarcity. In this regard, understanding the determinants of export performance is important in the formulation of export-oriented policies. Several studies have investigated the determinants of exports with pooled data (e.g., Arize (1988), Bond (1987), Ghura and Grennes (1995), Islam and Subramanian (1989), Khan and Knight (1988), and Pick and Vollrath (1994)). Both domestic and external factors have been found to be important determinants of exports. Among the domestic factors are the productive capacity of the economy, the domestic demand pressure, and the misalignment of the real exchange rate, while external factors include real world income and international prices for exports.
This study reviews the performance of Cameroon’s main export commodities during the period 1970–95 and empirically investigates the factors that influenced this performance. Aside from being a net oil exporter, Cameroon has a relatively diversified export base. In 1995, primary commodities (cocoa, coffee, and cotton) accounted for about 50 percent of Cameroon’s total export earnings, while oil exports accounted for 35 percent. Cameroon also exports agro-manufactured and semimanufactured goods such as sawnwood, aluminum sheets, and cocoa butter. Industrial exports, albeit still small, are growing. Undoubtedly, with the appropriate set of economic policies, Cameroon could take advantage of its diversified export base to encourage rapid economic growth.
The results of the empirical investigation can be summarized as follows. First, export commodities react positively and significantly to changes in their respective international prices. Second, exports are boosted by gains in international competitiveness. Finally, improvements in the domestic productive capacity have positive effects on non-oil exports, whereas domestic demand pressures exert adverse effects. The organization of the paper is as follows: Section B reviews the developments in the structure and composition of Cameroon’s major exports commodities between 1970 and 1995. Section C resents the elements of the theoretical model used as a basis for empirical investigation. Section D reports the results of the econometric analysis. The last section summaries the main findings and discusses a few policy implications.
B. Structure and Composition of Cameroon’s Exports
In discussing Cameroon’s export performance, the period 1970–95 is divided into four main subperiods for ease of presentation.
Pre-oil boom era (1970–77)
Prior to 1978, the year in which oil exports started, Cameroon’s agricultural sector contributed some 80 percent of total export earnings and some 30 percent of the gross domestic product (GDP). Agriculture also generated employment for about three fourths of the labor force. Among agricultural commodities, cocoa and coffee largely dominated the composition of Cameroon’s exports, constituting more than half of the total export revenue and 80 percent of agricultural export revenue.
This period was characterized by an increase in the level of taxation on the major commodity exports in order to finance the modernization of the agricultural sector. This was accompanied by the expansion of the modern plantation sector and the creation of managed agricultural systems under which crop producers had no decision-making autonomy (Tshibaka (1996)).
Between 1970 and 1975, world prices for most of Cameroon’s commodity exports showed decreasing trends. Then, in 1976 and 1977, world prices (measured in CFA francs) for coffee and cocoa more than doubled, leading to a 30 percent improvement in the terms of trade (Charts 1 and 2). This resulted in a large jump in the real export values for these commodities (Charts 3 and 4). Total export revenue increased from CFAF 64 billion in 1970 to CFAF 179 billion in 1977 with an average annual growth of about 18 percent in nominal terms.
CAMEROON: Exports, Terms of Trade, and Real Effective Exchange Rate, 1970–95
(In percent)
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Sources: Data provided by the Cameroonian authorities; and IMF, Information Notice System.1/ Left scale; in billions of 1980 constant CFA francs. The GDP deflator was used for converting the nominal series into constant ones.2/ Right scale; index, 1980=100.CAMEROON: Exports, Terms of Trade, and Real Effective Exchange Rate, 1970–95
(In percent)
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Sources: Data provided by the Cameroonian authorities; and IMF, Information Notice System.1/ Left scale; in billions of 1980 constant CFA francs. The GDP deflator was used for converting the nominal series into constant ones.2/ Right scale; index, 1980=100.CAMEROON: Exports, Terms of Trade, and Real Effective Exchange Rate, 1970–95
(In percent)
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Sources: Data provided by the Cameroonian authorities; and IMF, Information Notice System.1/ Left scale; in billions of 1980 constant CFA francs. The GDP deflator was used for converting the nominal series into constant ones.2/ Right scale; index, 1980=100.
CAMEROON: World Market Prices for Export Commodities, 1970–95
(In 1980 constant CFA francs; indices, 1980=100)
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Source: IMF, Commodity Price System.CAMEROON: World Market Prices for Export Commodities, 1970–95
(In 1980 constant CFA francs; indices, 1980=100)
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Source: IMF, Commodity Price System.CAMEROON: World Market Prices for Export Commodities, 1970–95
(In 1980 constant CFA francs; indices, 1980=100)
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Source: IMF, Commodity Price System.
CAMEROON: Major Agricultural Commodity Exports, 1970–95
(In billions of 1980 constant CFA francs) 1/
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Source: IMF, Commodity Price System.1/ The GDP deflator was used for converting the nominal series into constant ones.CAMEROON: Major Agricultural Commodity Exports, 1970–95
(In billions of 1980 constant CFA francs) 1/
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Source: IMF, Commodity Price System.1/ The GDP deflator was used for converting the nominal series into constant ones.CAMEROON: Major Agricultural Commodity Exports, 1970–95
(In billions of 1980 constant CFA francs) 1/
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Source: IMF, Commodity Price System.1/ The GDP deflator was used for converting the nominal series into constant ones.
CAMEROON: Major Nonagricultural Commodity Exports, 1970–95
(In billions of 1980 constant CFA francs) 1/
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Source: IMF, Commodity Price System.1/ The GDP deflator was used for converting the nominal series into constant ones.CAMEROON: Major Nonagricultural Commodity Exports, 1970–95
(In billions of 1980 constant CFA francs) 1/
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Source: IMF, Commodity Price System.1/ The GDP deflator was used for converting the nominal series into constant ones.CAMEROON: Major Nonagricultural Commodity Exports, 1970–95
(In billions of 1980 constant CFA francs) 1/
Citation: IMF Staff Country Reports 1996, 125; 10.5089/9781451807998.002.A003
Source: IMF, Commodity Price System.1/ The GDP deflator was used for converting the nominal series into constant ones.Oil boom era (1978–86)
With the start of oil exports in 1978, Cameroon’s structure of exports changed drastically. Oil became rapidly the main source of export revenue, accounting for 43 percent of total export earnings in 1982, and international oil prices increased sharply between 1978 and 1985.
In principle, the oil boom would have created disincentives for agricultural exports by (i) raising relative factor prices, (ii) enhancing profitability of investment in the nontradable goods sector, and (iii) raising wages in the public sector, thus putting upward pressures on rural wages (Benjamin and others (1989)).58 Nevertheless, this appreciation was avoided in Cameroon as the real exchange rate actually depreciated during this period, despite the foreign exchange inflows associated with the oil boom. This occurred largely on account of two factors. First, a sizable portion of the windfall oil income was saved as the oil boom was perceived by the authorities as a transient rather than a permanent phenomenon (Benjamin and others (1989)). Second, the French franc depreciated vis-ὰ-vis the U.S. dollar between 1980 and 1985. These factors helped prevent the real exchange rate appreciation usually associated with a substantial inflow of foreign exchange. Tshibaka (1996) reports that the real exchange rate was in fact undervalued between 1980 and 1984. In this regard, no major loss in competitiveness was associated with the oil boom.
Nevertheless, the real value of Cameroon’s main agricultural exports declined between 1978 and 1981. At the same time, the real world prices for cocoa, coffee, and banana fell. The decline in export values was aggravated by the severe drought that affected the agricultural sector in 1982 and 1983. The output of these export commodities recovered quickly, however, as the windfall income from the oil boom made it possible for the government to raise the producer price of cash crops, lowering the effective export tax on these commodities (Benjamin and others (1989)). Concurrently, the marketing costs of agricultural products were substantially reduced as a result of the improvement of physical infrastructure (for example, the Yaoundé-Douala and Bafia-Bafoussam roads), which was partly financed by oil revenue (Tshibaka (1996)).
Although Cameroon’s terms of trade deteriorated during the 1978–86 oil boom, they remained broadly favorable to the country. In 1983 and 1984, the world prices for coffee and cocoa rose in response to the weather induced shortages in the world market. The favorable terms of trade, combined with the boom in the oil industry, resulted in an acceleration of the growth of total export revenue to an average of 22.8 percent a year during this period. Another major development during this period was the establishment of the Office National de Commercialisation des Produits de Base (ONCPB) to replace the existing commodity stabilization funds. The ONCPB was an autonomous public institution that was given the monopoly for the internal and external marketing of the major cash crops (cocoa, coffee, cotton, groundnuts and palm oil). Its main objectives were to maintain a link between the producer price and the world price of the major exports, and to protect producers from large fluctuations in the world prices. At the beginning of each crop year, the government would set the producer prices of export crops on the basis of proposals by the ONCPB. In determining the level of producer prices, a number of factors were taken into account, including international market prices, the liquidity position of ONCPB, public sector salary increases, producer prices paid in neighboring countries, and the strength of attractions for migration to the urban centers.
Post-oil boom era (1987–93)
In the second half of the 1980s, Cameroon experienced a steady decline in its economic activity, as well as a significant deterioration in its terms of trade resulting from the fall in the world prices of crude oil, cotton, coffee, and cocoa.59 Between 1985 and 1988, the world oil price (expressed in CFA francs) fell by two thirds, while the prices of coffee and cocoa dropped by one half and one third, respectively. The producer prices for these commodities were kept unchanged until 1989, however, which led to heavy losses for the ONCPB. Because of widespread inefficiencies and mounting financial losses, the government liquidated the ONCPB in 1992 and opened exports to private traders.
Between 1987 and 1993, both the nominal and the real exchange rates appreciated as a result of the marked appreciation of the French franc against the U.S. dollar, as well as the authorities’ attempt to boost aggregate demand through an increase in government expenditure. Most observers agree that the real exchange rate was overvalued during this period. According to Devarajan and others (1993), a real depreciation of the currency of 36 percent (or a 46 percent decline in the domestic price level) was needed in order to achieve equilibrium in the real exchange rate. Tshibaka (1996) found that the average rate of overvaluation of the real exchange rate was about 40 percent between 1985 and 1993. This overvaluation led to a major loss of international competitiveness, and as a result, total export earnings expressed in nominal CFA francs dropped by more than half between 1985 and 1993.
Meanwhile, the export structure continued to undergo major transformations. The real value of cocoa and coffee exports declined substantially and continuously during the post oil boom period. Concurrently, their shares in total export revenue declined considerably as world prices for these commodities continued to fall in both nominal and real terms. From 1990 onward, timber exports surpassed both coffee and cocoa exports, becoming the main non-oil export commodity and accounting for more than 6 percent of total export earnings and 12 percent of non-oil export revenue.
In the face of declining international competitiveness and deteriorating external and internal conditions, Cameroon started in 1988 a series of internal adjustment programs with a view to restructuring the productive sector and to restoring competitiveness. The adjustment efforts emphasized the reduction of government expenditures, restructuring of the public enterprises, and liberalization of input and product prices. Some progress was made toward increasing productivity in the agricultural sector by improving pricing arrangements, marketing, extension services, and support for infrastructure. However, the progress made in achieving international competitiveness was jeopardized in 1992, owing to difficult social and political situations and a public disobedience campaign.
Despite a sharp contraction in oil output, total exports expanded in 1993 as the terms of trade improved. Non-oil exports grew, reflecting the incipient recovery in world commodity prices and favorable weather conditions. In particular, production of cocoa, robusta coffee and arabica coffee rose by 7 percent, 18 percent and 50 percent, respectively, between 1992 and 1993. Nevertheless, Cameroon’s overall economic performance continued to deteriorate.
Recent post-devaluation performance (1994–95)
The large deterioration in the terms of trade, sharp real appreciation of the currency, and continuous decline in oil production were the main reasons for Cameroon’s difficulties during the 1986–93 period. Despite some improvements in the terms of trade, Cameroon’s economic situation continued to deteriorate in 1993, indicating that the internal adjustment strategy in the late 1980s had failed to bring positive results in terms of competitiveness or the overall economic situation. It became clear that a more comprehensive approach was required, and that an adjustment of the exchange rate was necessary to restore external competitiveness, thus laying the foundations for the resumption of economic growth.
The 50 percent nominal devaluation of the CFA franc in January 1994 led to a reallocation of resources toward tradable goods.60 The devaluation was followed by a comprehensive adjustment program. Concurrently, the terms of trade improved sharply. Expressed in CFA francs, total export revenue increased by 130 percent between 1993 and 1994. Cameroon regained its competitiveness almost instantly, as the real effective exchange rate depreciated by 28 percent on a cumulative basis between 1993 and 1994. The improvement in competitiveness was especially pronounced in the sectors that were labor intensive and less intensive in imported inputs. Immediately after the devaluation, exports of timber, sawnwood, rubber and bananas rose sharply, in both nominal and real terms. Cocoa and coffee exports also expanded, but to a lesser extent. In 1995, timber continued to be Cameroon’s first non-oil commodity exports, accounting for 18 percent of non-oil export earnings, with cocoa, robusta coffee, and arabica coffee, respectively accounting for 10 percent, 11 percent, and 2 percent. Available data indicate a continued improvement in Cameroon’s export earnings in 1995 and 1996.
C. Determinants of Commodity Exports: Some Theoretical Considerations
The supply function used in this study is specified independently of an export demand function, based on the assumptions that Cameroon is a price taker in all its export commodities and that it faces a perfectly competitive world market. It is assumed that foreign demand is already reflected in the world prices. The following equation is used to discuss the theoretical considerations underlying Cameroon’s export supply:
where f1 > 0, f2 < 0, f3 < 0, f4 < 0, f5 > 0, f6 < 0; X denotes quantity exported, P represents export price; UC represents per unit cost of production; REER is real effective exchange rate, OILSHR is the share of oil value added in total GDP; CAPACITY represents the domestic productive capacity; and DEMAND is an indicator of the domestic demand pressure.61 The rationale underlying equation (1) is straightforward and can be explained as follows.
First, producers and exporters respond to market price (P) signals. Export supply increases with the profitability of producing or selling exports. Recent evidence suggests positive responses of primary exports to changes in export prices in developing countries (e.g., Arize (1988), Binswanger (1989), Bond (1983 and 1987) and Jaeger (1991)). Second, production of export commodities would be expected to fall with arise in the unit cost of production (UC). Since these factor costs are likely to move with the general level of domestic prices, the GDP deflator is used as a proxy for them (Goldstein and Khan (1985)).
Third, low levels of external competitiveness, as reflected in an appreciated real effective exchange rate (REER), lower production of export commodities by reducing profitability. An appreciated real exchange rate could be associated with an overvaluation, which would be expected to induce gaps between the marginal costs and prices, and thus to act as an implicit tax on the production of exportable goods and reduce the supply of exports. On the other hand, an appreciated real exchange rate could result from a change in the fundamentals. An appreciation of the real exchange rate reduces the profitability of producing exports. In both cases, international competitiveness weakens as resources are shifted away from the tradable goods sector toward the nontradable goods sector. The deleterious effects of overvalued real exchange rates on agricultural and overall exports, as well as on macroeconomic performance in developing economies, are well established in the literature (e.g., Agarwala (1983), Cottani and others (1990), Dollar (1992), Edwards (1988), Elbadawi (1992), and Ghura and Grennes (1993, 1995)).
The fourth variable (OILSHR) is used to capture the effects of the oil boom on the non-oil export commodities. This variable is used to test the Dutch disease hypothesis, characterized by a rise in the relative price of nontraded to traded goods triggered by a large capital inflow.62 It seems that the Dutch disease was largely averted in Cameroon, as the real exchange rate depreciated by about 50 percent on a cumulative basis between 1979 and 1984, reflecting largely the depreciation of the French franc. In addition, Benjamin and others (1989) have noted that the government saved a large portion of the windfall income from oil since it perceived the oil boom as a temporary phenomenon, thus avoiding a spending boom.
The fifth variable (CAPACITY) is an indicator of domestic capacity. A country’s ability to produce export commodities is enhanced by improvements in its production capacity (Goldstein and Khan (1985)). The beneficial impact of improvements in capacity on exports has been confirmed by Faini (1988) for Turkey and Morocco. The last variable (DEMAND) captures the effects of domestic demand pressure on exports. When domestic demand pressure rises, the profitability of selling in the domestic market is enhanced, an effect that is particularly pronounced if the export market carries the perception of higher risks in the minds of traders (Islam and Subramanian (1989)). Empirical evidence demonstrating the adverse effects of domestic demand pressure on commodity exports in developing countries has been provided by Artus (1973), Islam and Subramanian (1989), and Zilberfarb (1980).
D. Econometric Analysis and Results
To estimate the effects of the factors discussed in the previous section on Cameroon’s exports, the following empirical equation is used:
where In denotes the natural logarithm, and et denotes an error term. The estimated coefficients can be interpreted as elasticities with respect to X. The GDP deflator, which serves as a proxy of the unit cost of production (UC), is used to deflate Xt and Pt in order to obtain the real value of exports.63 In order to avoid the problem of “spurious regressions,” the Engel-Granger (1987) two-step procedure is used.64 The rest of this section summarizes the results of the unit root tests and discusses the results from the long-run estimated equation for total non-oil exports and the exports of major non-oil individual commodities (cocoa, robusta coffee, arabica coffee, bananas, timber, aluminum, and sawnwood).65
Time-series properties of the data
The augmented Dickey-Fuller statistics (ADF) were used to test for stationarity of the variables used in the study (X, P, REER, OILSHR, CAPACITY and DEMAND). The results of the unit root tests are reported in Table 1. The analysis suggests that, at the 5 percent level, all the relevant variables are integrated of order one, I(1), with the exception of OILSHR, which is integrated of order two, I(2). Accordingly, the first difference, and not the level, of OILSHR was used in the analysis so that all the variables at hand are I(1). This property is necessary for the cointegration analysis conducted below.66
Unit Root Test Results
(Augmented Dickey-Fuller test statistics) 1/
Critical values: −3.63 for the 5 percent level (*), −4.44 for the 1 percent level (**); a constant term and a time trend are included.
All variables are expressed in natural logarithm.
Unit Root Test Results
(Augmented Dickey-Fuller test statistics) 1/
| Variables 2/ | Level | First difference | Second difference | |
|---|---|---|---|---|
| Real exports (X) | ||||
| Total non-oil exports | −3.12 | −4.66 ** | — | |
| Cocoa | −2.42 | −4.14 * | — | |
| Robusta coffee | −2.31 | −4.52 ** | — | |
| Arabica coffee | −3.42 | −6.96 ** | — | |
| Timber | −2.29 | −6.13 ** | — | |
| Sawnwood | −2.09 | −6.26 ** | — | |
| Aluminum | −3.16 | −5.46 ** | — | |
| Bananas | −3.03 | −9.89 ** | — | |
| Real prices (P) | ||||
| Total non-oil exports | −3.61 | −4.72 ** | — | |
| Cocoa | −2.53 | −4.30 * | — | |
| Robusta coffee | −2.09 | −4.16 * | — | |
| Arabica coffee | −2.32 | −4.48 ** | — | |
| Timber | −2.84 | −5.24 ** | — | |
| Sawnwood | −2.75 | −6.43 ** | — | |
| Aluminum | −3.03 | −4.62 ** | — | |
| Bananas | −3.49 | −5.43 ** | — | |
| Real effective exchange rate (REER) | −1.22 | −4.87 ** | — | |
| Investment/GDP (CAPACITY) | −0.94 | −3.74 * | — | |
| Consumption/GDP (DEMAND) | −3.19 | −5.56 ** | — | |
| Value added of oil/GDP (OILSHR) | −2.25 | −3.27 | −3.96 * | |
Critical values: −3.63 for the 5 percent level (*), −4.44 for the 1 percent level (**); a constant term and a time trend are included.
All variables are expressed in natural logarithm.
Unit Root Test Results
(Augmented Dickey-Fuller test statistics) 1/
| Variables 2/ | Level | First difference | Second difference | |
|---|---|---|---|---|
| Real exports (X) | ||||
| Total non-oil exports | −3.12 | −4.66 ** | — | |
| Cocoa | −2.42 | −4.14 * | — | |
| Robusta coffee | −2.31 | −4.52 ** | — | |
| Arabica coffee | −3.42 | −6.96 ** | — | |
| Timber | −2.29 | −6.13 ** | — | |
| Sawnwood | −2.09 | −6.26 ** | — | |
| Aluminum | −3.16 | −5.46 ** | — | |
| Bananas | −3.03 | −9.89 ** | — | |
| Real prices (P) | ||||
| Total non-oil exports | −3.61 | −4.72 ** | — | |
| Cocoa | −2.53 | −4.30 * | — | |
| Robusta coffee | −2.09 | −4.16 * | — | |
| Arabica coffee | −2.32 | −4.48 ** | — | |
| Timber | −2.84 | −5.24 ** | — | |
| Sawnwood | −2.75 | −6.43 ** | — | |
| Aluminum | −3.03 | −4.62 ** | — | |
| Bananas | −3.49 | −5.43 ** | — | |
| Real effective exchange rate (REER) | −1.22 | −4.87 ** | — | |
| Investment/GDP (CAPACITY) | −0.94 | −3.74 * | — | |
| Consumption/GDP (DEMAND) | −3.19 | −5.56 ** | — | |
| Value added of oil/GDP (OILSHR) | −2.25 | −3.27 | −3.96 * | |
Critical values: −3.63 for the 5 percent level (*), −4.44 for the 1 percent level (**); a constant term and a time trend are included.
All variables are expressed in natural logarithm.
Long-run relationships
The possibility of the variables (X, P, REER, OILSHR, CAPACITY, and DEMAND) being linked by some long-run equilibrium relationship is investigated. The first step of the Engel-Granger method is to identify a “possible” long-run relationship among the variables in equation (2) by using Ordinary Least Squares (OLS).67
The expected signs of the estimated coefficients in equation (2) are as follows. As the relative export price of a commodity rises, production and export of that commodity become more profitable and therefore they would be expected to increase. Thus, the expected sign of the price elasticity β is positive. The sign of γ is expected to be negative because an appreciated and/or overvalued real exchange rate is harmful to exports. The Dutch disease effect is captured by the share of the oil sector value added in GDP. If the boom in the oil industry has been harmful to the non-oil export sector, the coefficient on OILSHR, δ, is expected to be negative. The effect of the domestic productive capacity is captured by the ratio of total investment to GDP; the sign of λ is expected to be positive since the effect of an increase in the domestic productive capacity is to raise export volume. Finally, a rise in domestic demand would tend to divert export products, as well as resources, away from the export market, and thus the sign on θ is expected to be negative. The effect of this domestic demand pressure is captured by the ratio to GDP of private consumption.
In general, the postulated model performs well in explaining the variations in exports. The regression results are presented in Table 2. In order to determine if the variables are actually cointegrated, the residual series from the OLS regression for each regression is tested for a unit root using the ADF test statistics (reported in Table 2). If the residual series from a regression is stationary, then the variables in the estimated equation are cointegrated. In other words, if all the series included in the regression analysis are I(1) but there is a linear combination of them which is I(0), then, the series are said to be cointegrated. The results show that the error terms are stationary for the individual commodities considered, as well as for the total non-oil exports at the 5 percent significance level. Thus, the variables used in each equation are cointegrated and the regression estimates provide long-run relationships.68
Estimates of the Long-Run Commodity Export Equations
(Dependent Variables: Real Value of Exports) 1/
Expressed in billions of 1980 constant CFA francs.
Various lag structures were used for the explanatory variables in the estimated equations; for ease of presentation, only the sum of coefficients for each variable is given. All the variables used in the regressions are I(1). Standard errors are given below the parameter estimates. The symbols ***,**,* denote significance at the 1 percent, 5 percent, and 10 percent levels, respectively.
R2 is the coefficient of determination.
DW is the Durbin Watson test statistic.
ADF is the augmented Dickey-Fuller (ADF) test statistic in a regression with the residuals, without a drift term (critical values: −2.67 at the 1 percent level; and −1.95 at the 5 percent level).
Equation corrected for autocorrelation, using the Cochrane-Orcutt iterative technique.
Estimates of the Long-Run Commodity Export Equations
(Dependent Variables: Real Value of Exports) 1/
| Explanatory variables 2/ | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Dependent variables | PRICE | REER | OILSHR | CAPACITY | DEMAND | Intercept | R2 3/ | DW 4/ | ADF 5/ | |||
| Total non-oil exports | 0.979 *** | −0.469 | −0.455 ** | 0.719 ** | −0.640 | 3.986 | 0.67 | 2.04 | −3.82 | |||
| (0.217) | (0.391) | (0.175) | (0.265) | (0.578) | (3.164) | |||||||
| Cocoa | 0.695 *** | −1.451 *** | 0.345 * | 0.581 ** | −0.560 | 6.539 ** | 0.91 | 2.20 | −2.82 | |||
| (0.112) | (0.504) | (0.197) | (0.247) | (0.456) | (2.589) | |||||||
| Robusta coffee | 1.021 *** | 1.775 *** | −0.385 *** | −0.023 | −0.841 ** | −8.862 *** | 0.96 | 1.77 | −3.55 | |||
| (0.097) | (0.421) | (0.134) | (0.279) | (0.299) | (2.361) | |||||||
| Arabica coffee 6/ | 1.485 *** | −2.491 *** | −0.811 *** | 0.481 | 4.309 *** | 36.500 *** | 0.95 | — | −2.18 | |||
| (0.232) | (0.729) | (0.240) | (0.463) | (0.667) | (5.424) | |||||||
| Bananas 6/ | 0.330 | −3.110 * | −0.116 | −1.672 | −1.339 | 15.830 | 0.48 | — | −5.05 | |||
| (1.098) | (1.729) | (0.642) | (1.058) | (2.361) | (12.510) | |||||||
| Timber | 0.464 | 0.873 | 0.185 | −1.540 *** | −0.396 | 1.003 | 0.84 | 2.16 | −3.31 | |||
| (0.306) | (0.753) | (0.184) | (0.351) | (0.952) | (3.305) | |||||||
| Aluminum | 0.974 ** | −3.293 *** | −1.727 *** | 2.060 *** | −5.206 *** | 17.030 ** | 0.69 | 1.90 | −3.40 | |||
| (0.484) | (1.005) | (0.383) | (0.664) | (1.331) | (7.370) | |||||||
| Sawnwood 6/ | 0.222 | 0.096 | 0.785 *** | −1.233 *** | −1.536 *** | −2.496 | 0.92 | — | −4.82 | |||
| (0.143) | (0.241) | (0.122) | (0.197) | (0.493) | (2.206) | |||||||
Expressed in billions of 1980 constant CFA francs.
Various lag structures were used for the explanatory variables in the estimated equations; for ease of presentation, only the sum of coefficients for each variable is given. All the variables used in the regressions are I(1). Standard errors are given below the parameter estimates. The symbols ***,**,* denote significance at the 1 percent, 5 percent, and 10 percent levels, respectively.
R2 is the coefficient of determination.
DW is the Durbin Watson test statistic.
ADF is the augmented Dickey-Fuller (ADF) test statistic in a regression with the residuals, without a drift term (critical values: −2.67 at the 1 percent level; and −1.95 at the 5 percent level).
Equation corrected for autocorrelation, using the Cochrane-Orcutt iterative technique.
Estimates of the Long-Run Commodity Export Equations
(Dependent Variables: Real Value of Exports) 1/
| Explanatory variables 2/ | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Dependent variables | PRICE | REER | OILSHR | CAPACITY | DEMAND | Intercept | R2 3/ | DW 4/ | ADF 5/ | |||
| Total non-oil exports | 0.979 *** | −0.469 | −0.455 ** | 0.719 ** | −0.640 | 3.986 | 0.67 | 2.04 | −3.82 | |||
| (0.217) | (0.391) | (0.175) | (0.265) | (0.578) | (3.164) | |||||||
| Cocoa | 0.695 *** | −1.451 *** | 0.345 * | 0.581 ** | −0.560 | 6.539 ** | 0.91 | 2.20 | −2.82 | |||
| (0.112) | (0.504) | (0.197) | (0.247) | (0.456) | (2.589) | |||||||
| Robusta coffee | 1.021 *** | 1.775 *** | −0.385 *** | −0.023 | −0.841 ** | −8.862 *** | 0.96 | 1.77 | −3.55 | |||
| (0.097) | (0.421) | (0.134) | (0.279) | (0.299) | (2.361) | |||||||
| Arabica coffee 6/ | 1.485 *** | −2.491 *** | −0.811 *** | 0.481 | 4.309 *** | 36.500 *** | 0.95 | — | −2.18 | |||
| (0.232) | (0.729) | (0.240) | (0.463) | (0.667) | (5.424) | |||||||
| Bananas 6/ | 0.330 | −3.110 * | −0.116 | −1.672 | −1.339 | 15.830 | 0.48 | — | −5.05 | |||
| (1.098) | (1.729) | (0.642) | (1.058) | (2.361) | (12.510) | |||||||
| Timber | 0.464 | 0.873 | 0.185 | −1.540 *** | −0.396 | 1.003 | 0.84 | 2.16 | −3.31 | |||
| (0.306) | (0.753) | (0.184) | (0.351) | (0.952) | (3.305) | |||||||
| Aluminum | 0.974 ** | −3.293 *** | −1.727 *** | 2.060 *** | −5.206 *** | 17.030 ** | 0.69 | 1.90 | −3.40 | |||
| (0.484) | (1.005) | (0.383) | (0.664) | (1.331) | (7.370) | |||||||
| Sawnwood 6/ | 0.222 | 0.096 | 0.785 *** | −1.233 *** | −1.536 *** | −2.496 | 0.92 | — | −4.82 | |||
| (0.143) | (0.241) | (0.122) | (0.197) | (0.493) | (2.206) | |||||||
Expressed in billions of 1980 constant CFA francs.
Various lag structures were used for the explanatory variables in the estimated equations; for ease of presentation, only the sum of coefficients for each variable is given. All the variables used in the regressions are I(1). Standard errors are given below the parameter estimates. The symbols ***,**,* denote significance at the 1 percent, 5 percent, and 10 percent levels, respectively.
R2 is the coefficient of determination.
DW is the Durbin Watson test statistic.
ADF is the augmented Dickey-Fuller (ADF) test statistic in a regression with the residuals, without a drift term (critical values: −2.67 at the 1 percent level; and −1.95 at the 5 percent level).
Equation corrected for autocorrelation, using the Cochrane-Orcutt iterative technique.
In this long-run framework, the variables included in equation (2) yield their respective expected signs for the total non-oil exports, although their significance levels vary. An important result relates to the significant and positive response of the total non-oil export variable to its price with an elasticity of 0.97. The indicator of international competitiveness (real effective exchange rate) yields the expected negative sign, although it is not statistically significant. Confirming the Dutch disease hypothesis, the estimated coefficient for the oil share is significantly negative, suggesting that total non-oil exports were hurt by the oil boom. Improvements in the domestic capacity, as captured by the ratio of total investment to GDP, exert a positive and significant effect on exports. Finally, domestic demand pressures, albeit not statistically significant, appear to lower total non-oil exports.
The remainder of this section discusses the results at the individual commodity export level. With respect to price incentives, all the commodities considered yield positive price elasticities ranging form 0.22 for sawnwood to 1.48 for arabica coffee.69 However, although the estimated coefficients for bananas, timber, and sawnwood are positive, they are statistically insignificant. Most commodity exports exhibit inelastic responses to changes in their respective international price. It is also important to note that the price elasticities for the traditional export commodities (cocoa and coffee) are higher than those for the nontraditional ones (bananas, timber, sawnwood).
The effects of increases in the real effective exchange rate are negative and significant for cocoa, arabica coffee, bananas, and aluminum. This evidence supports the view that an appreciated or an overvalued real exchange rate hurts the export sector, confirming similar results by Jaeger (1991), Pick and Vollrath (1994), and Ghura and Grennes (1995).
The coefficient estimates for the oil share (OILSHR) are significantly negative for robusta coffee, arabica coffee, and aluminum. With the exception of bananas, all the remaining commodities yield positive estimates. This result supports the view that, during the period of the oil boom, Cameroon experienced the Dutch disease phenomenon only partially. The windfall income permitted the government to promote some of the non-oil sectors by reducing the taxation on those sectors and by improving the existing infrastructure. Both the resource movement effect and the spending effect were mild in Cameroon during the oil boom. The first effect was small because the oil sector’s capital and labor are primarily foreign. The second effect was limited because the windfall income from the oil boom was mostly saved in foreign assets and not immediately introduced in the domestic monetary system (Benjamin and others (1989)).
As far as productive capacity (CAPACITY) is concerned, the analysis yields the expected positive estimates for cocoa, aluminum, and arabica coffee. The estimated coefficients for timber and sawnwood are negative and significant. Finally, domestic demand pressure (DEMAND) appears to be an important explanatory variable, yielding the expected negative sign for all the commodities with the exception of arabica coffee.
E. Conclusions and Policy Implications
This paper has investigated the determinants of Cameroon’s main export commodities using annual data during 1970–95. The results and policy implications can be summarized as follows. First, the analysis indicates that in general exports are boosted by gains in international competitiveness. Thus, as indicated by the sharp increase in Cameroon’s main exports in the aftermath of the early 1994 devaluation of the CFA franc, exports can indeed be boosted by correcting an overvalued real exchange rate.
Second, export commodities react positively to changes in their respective international prices. Thus, the avoidance of distortions, such as export taxes, in the domestic price structure of export commodities would be expected to be beneficial to the export sector. It is also established that export price elasticities differ significantly in both magnitude and significance level for the different commodities considered. In general, traditional export commodities are more responsive than nontraditional ones to changes in prices.
Third, improvements in the domestic productive capacity have significant and positive effects on Cameroon’s total non-oil exports. This finding suggests that the implementation of institutional and structural policies that expand the domestic productive capacity and improve the efficiency of resource allocations is beneficial to exports. Finally, domestic demand pressures exert negative impact on export commodities, a result that suggests that tight government policies designed to curb domestic demand and improve external competitiveness are beneficial to the export sector.
Data Definition and Sources
(a) The sample.
Data used for this study consist of 26 annual observations pertaining to the years 1970–95. All the data are on a calendar-year basis.
(b) X: Real exports.
Obtained by deflating export revenue (in billion of CFA francs) with the GDP Deflator. Data for the export values (in CFA francs) for cocoa, robusta coffee, arabica coffee, bananas, timber, aluminum, sawnwood, were provided by the Cameroonian authorities.
(c) P: Relative prices.
The choice of an export price is not straightforward in Cameroon’s case because, until recently, producers of exports received the administratively determined “producer price” which diverged from the export price. To the extent that producer prices accrue to commodity producers and export (world) prices accrue to commodity exporters, the two types of prices constitute incentives to produce and to export, respectively. Hence, the export (world) price is more relevant in modeling exports. This series was obtained by deflating nominal prices (in CFA francs per unit) of each commodity with the GDP deflator. For the total non-oil export, an export price index is used for the P variable, and it is computed as a weighted average of the individual commodities export prices. Source: IMF, Commodity Prices System Database.
(d) UC: Unit cost of production.
Given that the unit cost of production tend to move with the domestic price level, this variable is proxied by the GDP deflator. Source: Cameroonian authorities.
(d) REER: Real effective exchange rate.
The REER is a weighted index of the nominal exchange rates adjusted for the differential between the domestic inflation and the rates of inflation in partner countries using a geometric weighing method (see Wickham (1987) for details). An upward movement in REER denotes an appreciation. Source: IMF, Information Notice System.
(e) OILSHR: Oil sector’s value added as a ratio to GDP.
Source: Cameroonian authorities.
(f) CAPACITY: Domestic productive capacity.
In order to capture capital and technological accumulations over the period considered, this variable is proxied by Cameroon’s gross domestic investment as a percentage of GDP. Source: Cameroonian authorities.
(g) DEMAND: Domestic demand pressure.
The ratio of private consumption to GDP is used to proxy the domestic demand pressure. Source: Cameroonian authorities.
References
Agarwala, Ramgopal (1983), “Price Distortions and Growth in Developing countries,” World Bank Staff Working Papers, No. 575 Management and Development Series No. 2 (Washington: World Bank July 1983).
Arize, Augustine (1988) “Modelling Export Prices and Quantities in Selected Developing Countries” Atlantic Economic Journal, Vol. 16, pp. 19–24.
Artus, J. R. (1973) “The Short-Run Effects of Domestic Demand Pressure on Export Delivery Delays for Machinery” Journal of International Economics, Vol. 3.
Benjamin, Nancy, C. (1990) “Investment, the Real Exchange Rate, and Dutch Disease: A Two-Period General Equilibrium Model of Cameroon” Journal of Policy Modeling, Vol. 12, pp. 77–92.
Benjamin, Nancy, C., Shantayanan Devarajan, and Robert Weiner (1989) “The ‘Dutch’ Disease in a Developing Country: Oil Reserves in Cameroon” Journal of Development Economics, Vol. 30, pp. 71–92.
Binswanger, Hans P. (1989) “The Policy Responses of Agriculture,” Proceedings of the World Bank Annual Conference on Development Economics, (Washington: World Bank).
Bond, Marian E. (1987) “An Econometric Study of Primary Commodity Exports from Developing Country Regions to the World” IMF Staff Papers, Vol. 34, pp. 191–227.
Bond, Marian E. (1985) “Export Demand and Supply for Groups of Non-Oil Developing Countries” IMF Staff Papers, Vol. 32, pp. 56–77.
Bond, Marian E. (1983) “Agricultural Responses to Prices in Sub-Saharan African Countries” IMF Staff Papers, Vol. 30, pp. 703–26.
Clement, Jean A.P., Johannes Mueller, Stephane Cosse, and Jean Le Dem (1986) “Aftermath of the CFA Franc Devaluation” IMF Occasional Paper No. 138 (Washington: International Monetary Fund).
Corden, W. M. (1984) “Booming Sector and Dutch Disease Economics: A Survey and Consolidation” Oxford Economic Papers, Vol. 36 pp. 359–80.
Cottani, Joaquin A., Domingo Cavallo, and M. Shahbaz Khan (1990) “Real Exchange Rate Behavior and Economic Performance in LDCs” Economic Development and Cultural Change, Vol. 39 pp. 61–76.
Devarajan, Shantayanan, Jeffrey D. Lewis, and Sherman Robinson (1993) “External Shocks, Purchasing Power Parity, and the Equilibrium Real Exchange Rate” World Bank Economic Review, Vol. 7 pp. 45–63.
Dollar, David, (1992) “Outward-Oriented Developing Countries Really Do Grow More Rapidly: Evidence From 95 LDCs, 1976–1985” Economic Development and Cultural Change, Vol. 40 pp. 523–44.
Edwards, Sebastian (1988) Exchange Rate Misalignment in Developing Countries, (Baltimore, Maryland: Johns Hopkins University Press).
Edwards, Sebastian (1989) Real Exchange Rates, Devaluation, and Adjustment: Exchange Rate Policy in Developing Countries, (Cambridge, Massachusetts: MIT Press).
Elbadawi, Ibrahim A. (1992) “Real Overvaluation, Terms of Trade Shocks, and the Cost to Agriculture in Sub-Saharan Africa: the Case of Sudan,” Journal of African Economies, Vol. 1 pp. 59–85.
Engel, Robert F. and C.W.J. Granger (1987) “Co-integration and Error Correction: Representation, Estimation and Testing” Econometrica, Vol. 55 pp. 251–76.
Faini, Ricardo (1988) “Export Supply, Capacity and Relative Prices” World Bank Policy, Planning and Research Working Paper, WPS 123 (Washington: World Bank).
Ghura, Dhaneshwar and Thomas J. Grennes (1995) “Aggregate Trade Response to Economy-Wide Distortions in Sub-Saharan Africa” Journal of African Economies, Vol. 3 pp. 359–86.
Ghura, Dhaneshwar and Thomas J. Grennes (1993) “The Real Exchange Rate and Macroeconomic Performance in Sub-Saharan Africa” Journal of Development Economics, Vol. 42 pp. 155–74.
Goldstein, Morris and Moshin Khan (1978) “The Supply and Demand for Exports: A Simultaneous Approach” Review of Economics and Statistics, Vol. 60.
Goldstein, Morris and Moshin Khan (1985) “Income and Price Effects in Foreign Trade,” in Handbook of International Economics, edited by R.W. Jones and P.B. Kenen. (Amsterdam: North-Holland, 1985, Vol. 2)
Islam, N. and A. Subramanian (1989) “Agricultural Exports of Developing Countries: Estimates of Income and Price Elasticities of Demand and Supply” Journal of Agricultural Economics, Vol. 40 pp. 221–31.
Jaeger, William (1991) “Impact of Policy in African Agriculture: an Empirical Investigation,” World Bank Policy, WPS 640 (Washington: World Bank).
Johansen, Sores and Katarina Juselius (1990) “Maximum Likelihood Estimation and Iinference on Cointegration With Applications to the Demand for Money” Oxford Bulletin of Economics and Statistics, Vol. 52 pp. 169–210.
Khan, Moshin and Malcolm D. Knight (1988) “Import Compression and Export Performance in Developing Countries” Review of Economics and Statistics, Vol. 70 pp. 315–21.
Krueger, O. Anne (1995) Trade Policies and Developing Nations, (Washington: Brookings Institution).
Lucas, Robert (1985) “On the Mechanics of Economic Development: W.A. Mackintosh Lecture 1985” Journal of Monetary Economics, Vol. 22 PP. 3–42.
Lukonga, Inutu (1994) “Nigeria’s Non-oil Exports: Determinants of Supply and Demand. 1970–90” IMF Working Paper 94/59 (Washington: International Monetary Fund May 1994).
Pick, Daniel H. and Thomas L. Vollrath (1994) “Real exchange Rate Misalignment and Agricultural Exports Performance in Developing Countries” Economic Development and Cultural Change, Vol. 42 pp. 555–71.
Romer, Paul (1986) “Increasing Returns and Long-Run Growth” Journal of Political Economy Vol. 4 pp. 1002–1037.
Tshibaka, Tshikala B. (1996) “Domestic Policies, External Shocks, Export Prices, Output and Income in Cameroon,” Interim Report (Nairobi, Kenya: African Economic Research Consortium Kenya, May 1996).
Zilberfarb, Ben-Zion (1980) “Domestic Demand Pressure, Relative Prices and the Export Supply Equation: More Empirical Evidence” Economica, Vol. 47.
Prepared by Dhaneshwar Ghura and Soamiely Andriamananjara (summer intern).
The Dutch disease hypothesis postulates that, during an oil boom, traded goods sectors (e.g., agriculture) contract as the real exchange rate appreciates.
In late 1980s, the entry of new southeast Asian suppliers into the world market drove the world prices of cocoa and coffee down. Cameroon was faced with tough competition in the world market since these suppliers used more advanced technology than that used in Cameroon.
The parity of the CFA franc was changed from CFAF 50 to CFAF 100 per French franc.
See Appendix I for the definitions and sources of all variables used.
This problem arises both from a resource-movement effect and a spending effect (Corden (1984)). The spending effect is attributed to the increase in purchasing power, reflected in an increased demand for both traded and nontraded goods. As noted by Benjamin (1990, p. 78), the “new demand for traded goods is met by imports at constant world prices, but the excess demand for nontraded goods causes their price to rise relative to those of traded goods.” The resource-movement effect stems from the upward pressure on the domestic resource costs associated with the production of nontraded goods. In the case of Cameroon, Benjamin and others (1989) note that the oil sector is an enclave with respect to the rest of the economy, as it uses mainly imported factors of production, including labor. Thus, movements in the relative price of nontraded good would have been most likely dominated by the spending effect.
The definitions and sources of the variables are presented in Appendix I.
It should be noted that the Johansen-Juselius (1990) procedure is generally superior to the Engel-Granger two-step method since it does not impose a single linear cointegrating vector. However, the Engel-Granger method can be used in the context of a single equation model if it produces theoretically sound and statistically significant results that can be confirmed by the Johansen-Juselius procedure.
The discussion of the short-run dynamics is excluded from this study. Nevertheless, the analysis of the short-run properties of the model using the Error Correction Mechanism (ECM) provides evidence in favor of convergence toward the long-run equilibrium for each commodity.
The first step of the Engel-Granger procedure, as well as the Johansen methodology, suggests that all the variables in the long-run relationship have to be I(1).
Where there was evidence of autocorrelation, the Cochrane Orcutt iterative technique was used to correct for it. In order to allow for the possibility of delayed export responses, different combinations of the lagged values of the explanatory variables were included in the estimating equations.
The framework developed by Johansen is used in order to ensure that the cointegration vector obtained from the Engel-Granger procedure for each commodity actually lies in the cointegrating space. In this regard, the vector is introduced as a restriction in the Johansen analysis, and the validity of this cointegration restriction is tested using the log-likelihood ratio test. This exercise is conducted for each export commodity as well as for the total non-oil exports. The analysis largely accepts the restriction for total non-oil exports, robusta coffee, arabica coffee, bananas, timber, aluminum, and sawnwood. The cointegrating restriction for cocoa is marginally accepted.
Bond (1983) reported an average long-run elasticity of 0.2 for the major export crops in Africa. Jaeger (1991) obtained price elasticities ranging from 0.1 to 0.3 for aggregate agricultural export supply. In a survey of price responsiveness of agricultural commodities, Binswanger (1989) reported that aggregate agriculture is price inelastic in sub-Saharan Africa. In addition, Ghura and Grennes (1995) reported price elasticities of 0.65 and 0.68 for total and primary exports, respectively.
