The effects of government policies on cereal consumption pattern change in the Gambia

Introduction

Market liberalisation has been the preferred route of major development agencies such as the World Bank and the International Monetary Fund (IMF) for increasing agricultural productivity in sub-Saharan Africa over the past couple of decades. The liberalisation usually includes the break-up of government monopolies (through parastatals) in agricultural marketing. The concern about government involvement in agricultural marketing centres around the possibility that price guarantees to farmers could lead to excessive draining of treasuries (because of subsidies), discourage private-sector investment in the agricultural sectors and therefore inhibit higher agricultural productivity. It was therefore believed that with government withdrawal and the resultant participation of many competing agro-dealers in the markets, farmers would respond to this competition by increasing output. Many of these policies were strongly pushed when the Structural Adjustment Programmes (SAPs) were implemented in sub-Saharan Africa in the 1980s. As a consequence of this policy view, the sub-Saharan Africa region has witnessed the privatisation of many government agencies and the concomitant reduction or removal of agricultural subsidies beginning in the 1980s (Gibbon et al. 1993).

Contrary to predictions, the reduction of the state's role in agricultural marketing did not lead to agricultural productivity increases in most African countries. This is because the debate about the merits of agricultural market liberalisation sometimes occurs with little reference to the specific political economy contexts of individual countries. The impediments to agricultural productivity growth in many sub-Saharan African countries are due more to structural problems (for example, high transportation and input costs) and tend to be country-specific rather than price distortions caused by government interventions (Delgado and Mellor 1984). In fact, the involvement of government agencies was most likely premised on these structural problems since their policies in agriculture were characterised mainly by the provision of credit and input subsidies and reduction of transportation costs. As a result, even after the privatisation of government parastatals, there is little evidence that any significant agricultural productivity growth resulted (Gabre-Madhin and Haggblade 2004).

The persistent lack of agricultural productivity growth in sub-Saharan Africa and frequent food crises in the region are consistent with the view that structural impediments are indeed quite important, and likely more so than any possible distortionary effects of government involvement. For example, the 1994 devaluation of the CFA franc (a currency used mostly by former French colonies in West and Central Africa) has led many to hope that the production of local cereals such as millet would increase and possibly displace (mostly imported) rice since the devaluation would lead to relative price increase of the latter (Singare et al. 1999). However, there has been no evidence of local production response to this favourable change induced by the CFA franc devaluation (Diagana 1999).

This paper provides evidence from the Gambia on the effect of government policy changes in the agricultural sector. We focus specifically on the changing cereal consumption pattern in the country and provide evidence that its cause lies in major PPP government policy changes starting in the 1980s. Instead of leading to productivity increase in the relevant crop (groundnuts) as hoped for by proponents of reforms, it actually led to a production decline in that crop and an increase in the production and consumption of millet in the country, and had no corresponding agricultural productivity effect. In actuality, what the agricultural policy changes achieved was the compounding of a problem that had its basis in the absence of sound and long-term agricultural policies since independence in 1965, and earlier during the colonial period. Our paper builds on and extends Sallah (1990), who not only noted many of the policy problems with regards to the agricultural sector in the Gambia between independence in 1965 and 1990, but was able to highlight numerous instances of misplaced priorities such as higher government expenditures for defence and external affairs over agriculture. The paper also questions implicitly the efficacy of the World Bank–IMF imposed SAP austerity measures, which provide the backdrop for policies of both the Jawara and Jammeh governments.

Like most countries in West Africa, (mostly imported) rice has been the staple food in the Gambia despite the presence of many locally grown cereals such as millet and sorghum. Beginning in the 1990s, however, the production and consumption of millet in the country increased and, by 1999, total millet consumption had overtaken rice consumption, despite the continuing importance of factors that have traditionally favoured rice consumption such as increased urbanisation and government subsidies. We provide evidence that this shift in consumption from rice to millet is caused by abrupt changes in PPP and APRC government policies, such as government withdrawal from the marketing of groundnuts and the reduction of state agricultural subsidies, which were driven by SAP conditionalities. Thus, beginning in the 1980s, the PPP government privatised parastatals that provided credit and input subsidies to groundnut producers with disastrous consequences, given that earnings from the sale of groundnuts, the major cash crop, enables farmers to purchase imported rice. Difficulties faced by farmers in producing and selling groundnuts left them with no option but to turn to locally grown cereal to ensure food security.

In the absence of structural problems in the agricultural sector, changes in government policy, such as the privatisation of agricultural marketing and elimination of subsidies, would be welcome developments. However, in the context of the Gambian economy, such policy changes were ill-advised, and even though the adoption of reforms during structural adjustment led to macroeconomic stabilisation, governments failed to implement long-term agricultural policies. Thus, policy instruments meant to stimulate agricultural production exclusively through market liberalisation only exacerbated the hardships faced by farm households, even though the country was often touted as a model of success and vindication of SAP austerity measures.

Agricultural policy has always been a sensitive political area for Gambian governments. Since the agricultural sector is the main source of livelihood for most Gambian households, and periodic elections have been held since independence, elected governments have always found it politically expedient to respond with short-term measures rather than long-term agricultural policies. Therefore, political factors play a crucial role in explaining the failures of Gambian regimes in implementing long-term agricultural policies.

The importance of agriculture in the Gambian economy

A poor agricultural country (Sallah 1990) with a GDP per capita of US$430 in 2009,¹ the Gambia is characterised by Low Human Development, being currently ranked 160 (out of 179 countries) in the UNDP Human Development Index (UNDP 2009).

Agriculture employs 68% of the labour force (World Bank 2010) and accounted for 33% of GDP in 2007 (IMF 2007). Beyond this, productivity growth in the sector is essential for economic growth and development (Timmer 1988). Like most economic processes, however, sectoral growth has numerous determinants, some of which depend on government policies. Understanding these determinants will help provide a context for appreciating the effects of government policy changes.

The key policy determinants of agricultural productivity include good infrastructure, access to modern inputs, agricultural research and favourable policies on trade and human capital investments (Bravo-Ortega and Lederman 2004). Good and reliable infrastructure lowers the cost of inputs by reducing transportation cost and reduces the cost of marketing for farmers, an essential requirement as farm households produce increasingly for the market. Modern agricultural inputs such as improved seeds and fertiliser which increase crop productivity (both in output per land and per worker) are essential to the growth of agriculture. Consequently, policies that do not actively facilitate farmer access to these inputs also forego considerable productivity gains in the agricultural sector. Investment in human capital such as education also facilitates agricultural growth because the cost of public awareness campaigns in agriculture (and other sectors) is reduced. For example, more educated farmers are less likely to make mistakes in the precise application of modern inputs, some of which have little margin for error. The presence of local agricultural research institutes is also vital, not only to develop new technologies, but also to ensure that modern innovations developed elsewhere are customised to local conditions. These local institutions usually work in tandem with extension agents who facilitate the transmission of information to individual farm households.

Before addressing government policies directly in what follows, we provide a brief background note on agricultural crops grown and consumed in-country, the most important of which are groundnuts, coarse cereals (millet and, to a much smaller extent, sorghum) and rice.

Major Gambian crops

Groundnuts, grown by at least 80% of all agricultural households, are the main cash crop in the country, contributing approximately 6% of the country's GDP (Ministry of Agriculture 2002). About 45% of cultivated land has been devoted to this crop, on average, since independence; approximately 60% of the crop is exported (as both raw or shelled nuts and groundnut oil), with the rest being marketed locally. Although 34,000 metric tonnes (MT) of groundnut (shelled) were exported in 1965, exports were down to 16,000 MT by 2006 (FAOSTAT 2009).

Although Figure 1 shows national groundnut production was roughly the same in 2007 (100,000 MT with shell) as in 1965 (128,000 MT with shell),² annual production fluctuated widely over this period, with the great Sahelian drought of the 1970s, for example, alongside low fertiliser use during these and other years, impacting negatively on production totals, which come in their entirety from rain-fed farms (Ministry of Agriculture 2002; FAOSTAT 2009). Farmers preserve a portion of their harvests for use as seed the following rainy season which results in the use of low quality seeds since groundnut traders insist on the highest quality seeds right after harvest. And, while the National Agricultural Research Institute (NARI) has been promoting the use of new and improved groundnut varieties, the take-up rate among farmers has been low, with the result that little growth in productivity or yield has been recorded (Figure 2).

Figure 1.

Production trends of groundnuts, millet and rice, the Gambia, 1965–2007.

Source: Graphed from data in FAOSTAT (2009).

Figure 2.

Yields of groundnuts, millet and rice (metric tonnes per hectare – MT/ha), the Gambia, 1965–2007.

Source: Graphed from data in FAOSTAT (2009).

Rice has been the staple food in the Gambia for over a century, although more so in urban than in rural areas. At independence in 1965, rice (milled) consumption was 28,400 MT, since when the average annual rate of growth in consumption has been almost identical to the rate of population growth, at some 3% per annum. However, beginning in the mid 1990s, the level of consumption of rice started to fall, and by 2003, total national consumption was back to late-1970s levels (FAOSTAT 2009).

Most of the rice consumed in the country is imported since local production falls far short of total demand. As a result, the level of local production in Figure 1 (37,000 MT of paddy rice in 1965 and 40,000 MT in 2007) is a poor indicator of the extent of domestic consumption. As with groundnut production, output totals and average yield are both low as virtually all the area cultivated is rain-fed, levels of fertiliser application are low and use of new seed varieties is limited. Average yield in the 1960s was 1.3 metric tonnes/hectare (MT/ha), climbing to 1.9 MT/ha between 2000 and 2007.³ Currently, only about 2,300 ha of land planted to rice is irrigated, with all irrigated rice fields being situated in Central River Region.⁴ Although average smallholdings are as low as 0.5 ha in irrigated perimeters, for example, this is due only partly to a high national population density of 154/km², as only about 40% of the potential area that can be dedicated to rice production is being utilised (African Development Bank 2005).

Because of the higher level of consumption relative to local production, rice imports (mainly from Asia) are vital for food security, growing from 6,800 MT in 1965 to about 92,000 MT in 2006. Not all rice imports are consumed within the Gambia, however, as a significant proportion leaves the country through the re-export trade, mainly to Senegal and Guinea. Prior to market liberalisation in 1985, the government agency, the Gambia Produce and Marketing Board (GPMB), had the sole licence to import rice into what was a tightly regulated market. An average post-independence per capita rice consumption rate of 67kg is one of the highest in Africa, with only Madagascar (102kg/year), Guinea Bissau (92kg/person), Sierra Leone (89kg/person) and Liberia (86kg/person) recording higher average per capita rice consumption between 1980 and 2003 (FAOSTAT 2009).

Millet is currently the most important cereal grown locally. At independence in 1965, 35,400 MT was produced in the country, a figure which had reached 160,000 MT by 2007 (Figure 1), reflecting an average annual increase in production, between 1992 and 2007, of 7% (FAOSTAT 2009). Most of this growth was the result of an expansion in the area under cultivation rather than an increase in yield (Figure 2). For example, the land devoted to millet cultivation in 1965 (33,000 ha) was just one-third that devoted to groundnut cultivation (100,000 ha), but by 2005, the area under millet was 21% higher than that planted to groundnuts (105,000 ha versus 127,000 ha).

The flat trends in the productivity of these major crops provide strong evidence that growth in agricultural productivity has been largely absent over the past four decades. The underlying reasons for this general low productivity can be traced to policy decisions.

Historical reasons for high rice consumption in the Gambia

Although the key motivation of this paper is to explain the shifting consumption pattern in which millet is competing with rice as the main staple, it also raises the question of why the Gambia has rice, a cereal it is not self-sufficient in, as its staple food. But while the origins go back centuries, the reality of rice's continued importance can be attributed to path dependence and the privileging of groundnuts as the colony's cash crop to satisfy European industrial demand (Swindell and Jeng 2006).

Following the end of the transatlantic slave trade in the late 1870s, the Gambia became enmeshed in the new but growing legitimate trade in primary commodities, principally groundnuts. The consequence of the shift in trade from enslaved Africans to primary production was to irreversibly restructure both economic and social relations of the Senegambia region and, specifically, in the Gambia. This gave rise to the intensified use of the ‘strange farmer’ system in the Gambia,⁵ which not only helped to increase the production of groundnuts for export, but also effectively incorporated Gambians and Gambian farmers into a global capitalist economy. In time, export crop production undercut production of food crops, partly because of demand and, partly too, because of support it received from the British colonial administration. The result was the reliance on rice imports which, when combined with seasonal famines and droughts, deepened the dependence of farmers on imported rice and European consumer goods. In time, a complex relationship between European subsidiaries and their Gambia intermediaries, Gambian farmers and the colonial state ensued (Saine 2008).

In the end, this economic relationship was inherently exploitative, with the major beneficiaries being the foreign business sector and colonial state which, through collusion, operated a de facto monopoly at the expense of farmers. This economic relationship also had the negative effect of stimulating groundnut exports while simultaneously dampening production of food crops (Saine 2008). This pattern of dependence and privileging of groundnuts did not change after political independence, continuing under both the Jawara-led and Jammeh-led regimes. Despite the increasing importance of Asian and hybrid rice varieties, the West Africa subregion has an indigenous variety, Oryza glaberrima. Archaeological and linguistic evidence suggests that its domestication occurred about 3,000 years ago around the Senegambia region (Porteres 1970). The Asian variety (Oryza sativa) is believed to have been introduced by Portuguese traders around the sixteenth century (Linares 2002). Groundnuts were introduced to the region around the same time (Brooks 1975), with its production increasing significantly in the early nineteenth century due to demands from Europe for raw materials in the manufacturing of soap and lubricants.

Most of the Gambia's early exports of groundnuts went to France because Britain depended more heavily on palm oil, which was exported from Ghana and Nigeria (Hughes and Perfect 2006). In response to an increase in Gambian groundnut exports from about 50 MT in 1835 to 11,000 MT in 1851 (Brooks 1975), local farmers shifted resources away from locally grown cereals such as millet and sorghum to groundnuts, generating concern over food security as local cereal production declined. Over time, this led, as in 1857, to serious periodic food shortage in the country (Carney and Watts 1991), to which the colonial administration's solution was to increase imports of rice from Asia. We have here then the main reason behind the eventual dominance of rice as the staple food in many Sahelian countries in West Africa, especially in the Senegambia subregion (Swindell and Jeng 2006).

As testament to the importance of this arrangement in maintaining reasonable food security, both the Senegalese and Gambian governments continued to be heavily involved in the marketing of groundnuts and the importation of rice. In the Gambia, the government agency, the GPMB,⁶ had the sole licence to market groundnut and import rice in the country until 1985. Similarly, in Senegal, a government agency, the Office National de Coopération et d'Assistance Pour le Développement (ONCAD), had the sole monopoly in marketing groundnut and importing rice up to 1981 (Ross 1982).

Over time, socio-economic changes have reinforced the importance of rice as a staple. Across all countries in Africa (Gajigo and Denning 2010), rice consumption is strongly correlated with urbanisation rates. The reason behind this correlation is that the preparation of milled rice is less time-consuming than that for local cereals such as millet and sorghum. For non-agricultural households in urban areas, the high opportunity cost of time leads to a preference for rice. Taking the Gambia, while only about 15% of the population lived in urban areas at independence in 1965, that percentage had increased to 57% by 2009, the result of an average annual urbanisation rate of 7% (United Nations Populations Division 2009). But there are also political reasons that favour high rice consumption in urban areas, with many African governments providing rice subsidies to both feed and appease an increasingly restless urban population.

Although rice is mainly imported and locally produced groundnut is mostly exported, the fate of the two crops has continued to be highly intertwined, as most smallholder farmers depend on the proceeds from the sale of their groundnut harvests in order to purchase imported rice. Consequently, any serious disruption to either groundnut production or marketing is likely to have an adverse effect on farmer ability to afford imported rice, the price of which is set internationally.

Government policies and the groundnut sector

While many of the factors that underlie low agricultural productivity are long established (for example, the exclusive dependence on rainfall and poor infrastructure), a more recent phenomenon, major changes in government policies, has compounded the difficulties faced by traditional smallholder farmers. Since before the colonial period, British and French traders have purchased groundnuts directly from farmers along the banks of the River Gambia (Brooks 1975). The long navigable river was instrumental in making the interior of the country accessible to vessels and significantly reduced the transportation cost faced by farmers. Given the small average farm size, transportation cost relative to the value of crop harvests has always been high. Following independence in 1965, the Jawara government retained the system of purchasing harvests from farmers close to their farms. The Gambia Cooperative Union (GCU), which had been established by the colonial government, continued to buy groundnuts directly from farmers and provided credit and subsidies for seeds and fertilisers.⁷ In 1973 the government created the Gambia Produce Marketing Board (GPMB),⁸ which added to and modernised groundnut collection points throughout the country. This new agency continued to be the governmental arm in providing subsidies to farmers, while farmer credit was financed through the government-owned Gambia Commercial and Development Bank (GCDB).

This arrangement worked reasonably well for the first 15 post-independence years, with groundnut (with shells) production averaging some 124,000 MT annually between 1965 and 1980. The problem with the groundnut sector started in the 1980s, a period which was preceded by a decade or so of rapid expansion in government expenditure and indebtedness, some of which could not be blamed on simple profligacy. For example, drought affected crop output during the 1970s. Nonetheless, as part of a Structural Adjustment Programme (SAP), known locally as the Economic Recovery Programme (ERP), government was forced to drastically reduce the size of the public sector by inter alia cutting back on subsidies for seeds and fertilisers, after the short-term Economic Stabilisation Programme (ESP) failed to stabilise the economy (Cooke and Hughes 1997; McPherson and Radelet 1996; Sallah 1990). Prior expansion in public expenditure and debt made it difficult for the government to guarantee the convertibility of the currency – the dalasi – and to meet its debt service obligations (Sallah 1990; Jabara 1994). The seriousness of the crisis left the government with little choice but to quickly implement IMF reforms to bring both expenditure and debt to sustainable levels. Within the agricultural sector, the most important reforms led to the privatisation of the GCDB.⁹ With much government support for the groundnut sector removed, the GPMB struggled to provide the same services that farmers had come to expect. The effect on farmers was acute enough for the Jawara government to promise more subsidies despite commitments to the contrary made under ERP (Jawara 2009), thereby highlighting the role of politics in frequent policy reversals and changes in agricultural policy.

The government tried several measures between the late 1980s and early 1990s to reform its most important parastatal in agriculture, the GPMB, without success. Finally, in 1993, the Jawara government sold it to Alimenta SA, which renamed it Gambia Groundnut Corporation (GGC). Needless to say, while this new privately owned enterprise continued to purchase groundnuts from farmers and licensed traders, it not only did so at fewer collection points but farmers no longer received seed and fertiliser subsidies. In 1994 the then government (still headed by Dawda Jawara) was overthrown in a military coup led by Yahya Jammeh (Ceesay 2006; Saine 2009). The new government faced serious funding problems since almost all development aid was suspended until after elections were held in 1996. As the difficulties faced by farmers in selling their groundnut harvest continued, the new Jammeh-led government nationalised the GGC in 1998 through a presidential decree, accusing the company of money laundering. This charge was never proven in a court of law, and the nationalisation was most likely a political response to farmer discontent about groundnut marketing. Alimenta successfully sued the government and received US$11.4 million in settlement (IMF 2001). The government is currently under pressure to privatise the GGC as it is obligated to do under the requirements of the Highly Indebted Poor Countries (HIPC) programme (IMF 2007).

The repossession of the GGC by the Jammeh government resulted in the restoration of some subsidies, although subsidy levels remained lower than they had been during the1980s, before ERP. This decision provided opportunities for the Jammeh government to portray itself as responsive to the concerns of farmers, a highly important voting bloc. However, after a brief surge in production in the late 1990s and 2000, groundnut production dropped again.

In 2004, the Jammeh government gave sole authorisation to the government-owned Gambia Agricultural Marketing Corporation (GAMCO) to market groundnuts, a decision which had a disastrous impact on the following year's harvest. In the event, the under-capitalisation of GAMCO meant that insufficient funds for the purchase of all of the 2004/05 harvest led to a significant fall in groundnut exports that year and, subsequently, untold hardship to poor farmers (IMF 2006).

As the preceding discussion suggests, government policy changes and reversals started with the abrupt privatisation of the GPMB and the GCDB by the Jawara government, with the resulting withdrawal of credit and subsidies previously provided by these agencies worsening the problem. Far more abrupt was the nationalisation of the GGC by the Jammeh government, which raised hopes that subsidies would return to their previous levels, although these were never fulfilled. The reduction of groundnut collection points further added to the cost of production for the mostly small-scale farmers. By 1998, many farmers close to the border sold their produce in Senegal (Samarasinghe and Gray 2006). Those without such an option had to pay to transport their harvest to points where it could be collected for export. Finally, the government's creation of GAMCO and the monopsonist status it was granted raised hopes that were later dashed when it could not purchase groundnuts due to limited funds. Taken together, these changes effectively led to the decimation of the returns to groundnut production in the country.

The World Bank–IMF SAP policies that the Gambian government was required to follow, and which led to the privatisation of the GPMB and the GCDB, and the subsequent removal of agricultural subsidies and credit, were based on the theory that the best incentive for agricultural productivity was through producer price incentives that government involvement in the sector was supposedly impeding. But as Delgado and Mellor (1984) had presciently argued, supply response to price changes was likely to be very limited as long as small-scale farmers continued to face high input prices, high transportation cost due to poor infrastructure and inadequate agricultural extension services.

Although the effects of SAP represented a proximate, rather than underlying cause of the problems faced by the Gambian agricultural sector, SAP policies made a poor economic environment all the more untenable. And, as a macroeconomic stabilisation programme, SAP in The Gambia (ERP) was successful but not successful enough to improve the lives of farmers. Specifically, not only did staff retrenchment in the public sector reduce the government budget deficit, but foreign reserves increased and debt obligations were brought to sustainable levels. Nonetheless, the price of this success included a significant cutback in government agricultural expenditure, including reduced subsidies, which exacerbated both the structural weaknesses of the agricultural sector and existing challenges to groundnut productivity. Furthermore, the austerity measures contributed directly to an increase in unemployment due to retrenchment of public employees and may have worsened income inequality among Gambians.

In hindsight, it is not difficult to understand how groundnut production in the country was adversely affected by PPP government policies, notably the removal of price guarantees by the GPMB, which saw small-scale farmers, whose small farm sizes and rural location placed them outside the reach of the few formal financial institutions that could provide services such as credit and savings instruments, facing wildly fluctuating and unpredictable prices. Predictably, average groundnut production fell by 29% from 1980s to 1990s (Figure 1). Indeed, the extent of the problem of limited access to essential modern agricultural inputs is illustrated in Figure 3, which is based on data from a representative national household survey carried out in 2003 by the Central Statistics Department of the Gambia. The vast majority of farm households do not have access to productivity-enhancing inputs such as fertilisers, seeds and extension services. Government services that provided these inputs were those that were removed as a result of policy changes that began under the Jawara government and continued under Jammeh.

Figure 3.

The problem of access to inputs in the Gambia in 2003.

Source: Drawn from data in the Household Surveys, Central Statistics Department (2003).

Significantly, political considerations played a key role in the persistence of such short-term and ineffective agricultural policies in the Gambia. The patronage system under Jawara's rule meant that government agencies such as the GPMB and the GCDB played key roles in funnelling money to politically connected individuals and groups (McPherson and Radelet 1996). Furthermore, the periodic announcements of agricultural subsidies provided opportunity for the government to present their implementation as largesse during election campaigns.

The removal of subsidised fertilisers and seeds under both Gambian governments has not been the only problem to have affected the groundnut sector. The quality of Gambian groundnuts is relatively low compared to other groundnut producing countries and this has lowered the price fetched by the country's exports (Samarasinghe and Gray 2006). For example, after the EU imposed standards for aflatoxin contamination for edible groundnuts, Gambian groundnut exports could no longer benefit from high prices that are available in the European market (Otsuki et al. 2001; Samarasinghe and Gray 2006). As a result, Gambia's groundnut exports target mainly the groundnut oil and European birdseed markets, and are completely absent from the lucrative edible groundnut market in Western Europe.

The link between groundnut and millet production

How did the change in policies under the PPP and APRC regimes lower rice consumption and increase millet production through their effects on groundnut production? Recall that rice consumption in the Gambia and in the subregion took off because of the rapid and sustained adoption of groundnut as a cash crop. Over most of the twentieth century, the production of local cereals such as millet and sorghum remained marginal as farmers increasingly devoted land to groundnut production in response to global demand. This arrangement, in which groundnuts were grown as the cash crop and rice was imported, worked well as long as government intervention reduced the transaction cost for farmers, and rice imports were not obstructed. However, with the disruption of groundnut production and its export in the wake of SAP, critical elements of this arrangement that had ensured reasonable food security for over a century were disrupted. Without guaranteed proceeds from groundnut exports, many farm households found it difficult to purchase imported rice. Near specialisation in groundnut production for most agricultural households became untenable in light of government policy changes with regards to the provision of credit and input subsidies. To ensure food security, farmers had little choice but to grow their own cereal.

The cereal that has benefited most directly from this is millet. The production and consumption of millet started to expand rapidly by the late 1990s (Figures 1, 4 and 5). By 2000, millet consumption had overtaken rice consumption. The average annual per capita consumption of millet went from 39 kg between 1990 and 1999 to 59 kg between 2000 and 2007, an increase of 50%. The situation for rice was almost the reverse, with average annual per capita rice consumption decreasing from 68 kg between 1990 and 1999 to 40 kg between 2000 and 2007.¹⁰ Figure 6, based on the findings of nationally representative household surveys carried out by the Gambia Central Statistics Department in 1992 and 2003, shows that household production of millet went up significantly in all regions over the period under review, thus providing corroboration of the aggregate increase in millet production mentioned above. A further way in which the disruption to groundnut production has affected rice imports is through the sharp reduction in foreign exchange reserves it has caused, given that groundnuts are the country's biggest export and its largest foreign currency earner. With groundnuts bringing in less in the way of foreign currency than previously, the government found it increasingly difficult to meet the import bill to satisfy the demand for rice of a growing and increasingly urbanising population.¹¹

Figure 4.

Trends in the production and consumption of the major crops in the Gambia.

Source: Graphed from data in FAOSTAT (2009).

Figure 5.

Per capita consumption of millet and rice in the Gambia.

Source: Graphed from data in FAOSTAT (2009).

Figure 6.

Percentage of households growing millet in the Gambia between 1992 and 2003.

Source: Graphed from data in Household Surveys, Central Statistics Department (2003).

Consumption pattern change

(1) Comparing evidence from the Gambia and Senegal

To the extent that the fall in groundnut production and the rise in the consumption and production of millet can be attributed to policy changes within the Gambia, it is important to show that the observed production and consumption trends in these crops are specific to the Gambia and not a local manifestation of a region-wide phenomenon. We would like to stress that we are not providing a comparison between Senegal and the Gambia in terms of agricultural policies in general. Rather, the goal here is to show that the observed consumption pattern change is limited to the Gambia.

Senegal provides a useful comparison when examining whether the consumption pattern change in the Gambia can be attributed to state policies in the latter. While it is a separate country with different policies, Senegal surrounds the Gambia on all sides except for a small strip of land on the Atlantic coast, and shares similar agroecological zones with the Gambia (Figure 7). More relevantly, the historical reasons behind the high consumption of rice and heavy export dependence on groundnuts are identical in both countries.

Figure 7.

The Senegambia Region.

Source: Geographic Guide. Available from: http://www.geographicguide.net/africa/gambia.htm

Figure 8 shows production trends in groundnuts and millet in Senegal. Groundnut production does not show a pronounced decline in production from the late 1990s. And millet production does not show an upward trend over the same time period. In Figure 9, the levels of consumption of rice and millet show patterns significantly different from, maybe even contrasting with, those observed in the Gambia over the period under review (see Figure 5): consumption of rice increasing in Senegal but falling in the Gambia and millet consumption in Senegal falling while rising in the Gambia over the past decade and half.

Figure 8.

Production of groundnuts and millet in Senegal.

Source: Graphed from data in FAOSTAT (2009).

Figure 9.

Consumption of rice and millet in Senegal.

Source: Graphed from data in FAOSTAT (2009).

Comparing consumption and production trends in millet and rice between Senegal and the Gambia helps to underscore the role of policy, notably the key similarities and differences in policy implementation and impact between the countries, both of which have experienced very little agricultural productivity growth since independence (Kelly et al. 1996). Like the Gambia, Senegal also underwent economic reforms as part of Structural Adjustment Programmes in the 1980s (Masters 2007) which, via a reduction in direct agricultural input subsidies, adversely affected agricultural productivity (Fall 2003; Kelly et al. 1996). However, there are some key policy differences that can account partly for differing patterns in rice and millet consumption. In 1994, Senegal (along with several other West African countries that use the CFA franc) underwent a 50% currency devaluation. Despite its adverse consequences (Creevey et al. 1995), the devaluation led to increased competitiveness of all export goods in the country, including groundnuts (Diagana 1995). Furthermore, Senegal has acted to subsidise the importation of rice on numerous occasions in response to economic shocks. For example, import duties on rice were reduced after the 1994 currency devaluation to offset rising import costs. Another example is the suspension of import duties on rice imports during the food crises of late 2007 and early 2008 (Gajigo and Denning 2010). Devaluation thus helped to improve groundnut production, while the import duty policies adopted prevented a significant reduction in rice imports. Consequently, Senegal did not experience the changing consumption patterns which were recorded in the Gambia.

(3) Cross price effects

As the production of crops is affected by both their own prices and that of relevant competitors, it is worth examining whether exogenous price movements in favour of millet have been responsible for the increase in millet production. Figure 10 shows that, overall, there has not been an increase in the per unit price of millet relative to groundnut. It is therefore likely that policy changes rather than increases in the relative price of millet have been driving its increasing production.

Figure 10.

Producer prices of millet and groundnuts between 1991 and 2008.

Source: Graphed from data in FAOSTAT (2009).

Implications of consumption pattern change for food security in the Gambia

Is the growing replacement of rice with millet as the staple food a favourable development for future food security in the Gambia? The answer depends on many factors. The West African subregion is increasingly reliant on rice imports. Only a small percentage of the world rice production is exported and sub-Saharan Africa (SSA) accounts for a third of global imports (Gajigo and Denning 2010). Thus the region is highly exposed to price or production shocks on the world rice market. This precarious position was highlighted during recent food crises when food prices (especially rice) increased precipitously over the course of a few months in 2007 and 2008. For many of SSA's poor, food security is unlikely to be achieved under the current reality. Although this situation can be ameliorated through increased local production stimulated by appropriate policies, it is unlikely that SSA can become self-sufficient in rice in the short- or medium-term if current consumption trends and government policy instruments persist. At projected levels of consumption, production growth would have to increase tremendously (and perhaps unrealistically) for local West African and SSA production to match consumption (Gajigo and Denning 2010). It is therefore possible that the Gambia could improve its food security by relying increasingly on millet, something which could also have long-term positive health implications, as it would begin to reverse consumption of highly refined (and less nutritious) rice in the diet of Gambians.

This potentially favourable effect on food security, however, is not guaranteed. As Figure 2 shows, virtually all the growth in the production of millet comes from expansion of the area cultivated rather than an increase in yield. This implies that structural problems that inhibit productivity growth, which are caused by the absence of effective long-term agricultural policies, persist. For example, the percentage of arable land in the Gambia that is equipped for irrigation has not increased since independence: 0.8% (1965–1964), 0.6% (1975–1984), 0.8% (1995–2005) and 0.8% (1995–2005) (FAOSTAT 2009).

In contrast, the situation with transportation cost has actually been getting worse. Between 1990 and 2000, the proportion of roads in the country which were paved remained constant at about a third, but has since dropped under the Jammeh-led APRC regime to just a fifth (World Bank 2010). This suggests, when taken along with the current low level of millet yield, that the overall level of millet production is far below its full potential. Figure 11 shows that the potential impact on total production of a 50% increase and doubling of the current average yield value of 1.03 metric tonnes per hectare (MT/ha) would be average annual increases in total production of, respectively, 47% and 96%. Additionally, the continuing lack of agricultural productivity under the current regime, of which low millet yield is just one specific example, is in part attributable to the frequent changes and lack of continuity at the topmost level of policy making in the Ministry of Agriculture.¹² In short, food security would be improved if farm households could achieve substantial increases in yield, but improved access to modern varieties and fertilisers and, therefore, government-funded subsidies, as well as improvements in infrastructure, would be prerequisites for such improvements in yield. Given the Gambia's high population density, increasing crop yield from the acreage already under cultivation, rather than expanding the cultivated area, is the only feasible long-term path to food security. The gender dimension of the increasing importance of millet consumption is also important. Traditionally, cash crops such as groundnuts are cultivated by men while women cultivate mainly food crops such as millet and rice (Carney and Watts 1991). This sexual division of labour implies that a lot of the recent government interventions in terms of providing inputs have largely benefited cash crop producers who are overwhelmingly male. The increasing importance of millet provides an opportunity for gender balance in agricultural intervention, although such an outcome should not be taken for granted, given that the current division of labour is not immutable. For example, government intervention in the 1980s to help commercialise rice cultivation in Central River Region (known at the time as MacCarthy Island Division) led to an influx of men in rice production, and the subsequent crowding out of some women from this traditional women's activity (von Braun and Webb 1989).

Figure 11.

Actual and potential levels of millet production in the Gambia.

Source: FAOSTAT (2009). Note: Actual yield is taken as an average value of 1.03MT/ha.

In addition to modern inputs, technology that prolongs the storage life of millet flour is also essential. Part of the reason why the consumption of rice increases with urbanisation is that rice is more easily processed, and milled rice can be stored for longer than millet flour without deteriorating. While non-milled millet can be stored for a long time without any damage, millet flour is susceptible to pest infestation and chemical reactions with atmospheric elements. Therefore, food processing and storage technologies that close the gap between millet and rice would also be required. Significantly, too, such technologies would have the added advantage of much-needed time- and labour-saving opportunities for women, given the prevailing intra-household gender division of labour in the Gambia.

Conclusion

This paper has looked at the nature, (presumed) causes and effects of frequent policy changes in relation to agricultural production and marketing, notably the provision of subsidies and credit, under the PPP and AFRC governments. Contrary to predictions, the withdrawal of government subsidies did not result in productivity increases in groundnut production, which had benefited most directly from government intervention. Rather, groundnut production stagnated and rice imports, historically funded from the proceeds of groundnut exports, declined. At the same time, millet production and consumption have both risen.

The lack of productivity response in the case of groundnuts is explained by the presence of structural factors such as high transportation costs resulting from poor infrastructure, high input prices and inadequate extension services, rather than limited competition caused by government involvement. These structural problems remained, even after government withdrew from agricultural marketing, thereby casting doubt on the potential attractiveness of agricultural marketing for private sector investors. In turn, they reflect the failures of successive post-independence governments which failed to implement meaningful long-term strategies to promote growth and development of the agricultural sector. Both regimes have found it politically expedient to implement short-term populist measures that could be unveiled during election campaigns in a bid to win votes. And yet, these policy failures have had adverse consequences for the economic development of the country as a whole, given the livelihood significance of agriculture for individuals and households, as well as the national economy. Most telling, however, is that real per capita GDP has grown at an average annual rate of only 0.3% between independence in 1965 and 2007 (Penn World Tables 2009).

We have also argued that the shift from rice to millet consumption is a direct outcome of government policy changes, rather than a simple local Gambian manifestation of a West Africa-wide phenomenon. The neighbouring country of Senegal does not show a similar shift in cereal consumption, despite similarities in agroecological conditions and a shared history. We show, too, that the fall in rice consumption cannot be explained by sudden world price surges, since none occurred around the time rice consumption started to fall in the Gambia. Furthermore, there has been no relative price increase in favour of millet to induce more farmers to start growing it, which further strengthens the view that policy change has been the main driving factor.

Although the switch from rice to millet consumption could improve food security in the Gambia, this would depend on productivity increases in millet production, something that has not happened so far. For the past five decades, average millet yields have been stagnant. The recent increase in production has come from expansion in the area cultivated, a response which is not sustainable because of the high population density of the country. Therefore, public–private partnerships and long-term agricultural development programmes that address structural barriers such as poor infrastructure and access to modern inputs are needed to raise productivity of millet, as well as other important crops, to ensure future food security.

Given the importance of gender in Gambian agriculture, it is not surprising that the observed shift in consumption has gender implications. Groundnuts are mostly cultivated by men while women are relatively more involved in food crop growing. So the increasing importance of millet has the potential to increase the relative income of women. However, this is only a possibility. As long as existing economic and cultural norms responsible for gender inequality persist, there is a real likelihood of the gains earmarked for women being captured by men. This therefore underscores the need for gender mainstreaming in agricultural policy in The Gambia.

Notes on contributors

Ousman Gajigo is Economist at the African Development Bank, Tunis, Tunisia.

Abdoulaye Saine is Professor of African Studies and International Political Economy in the Department of Political Science, Miami University, Oxford, Ohio, United States.