Somaliland Cyberspace

Food Import Dependence in Somalia Magnitude, Causes, and Policy Options

Y. Hossein Farzin

The World Bank, Washington, D.C.
World Bank Discussion Papers
First printing January 1988

ABSTRACT

This paper has three main objectives (a) to examine in detail the trends in food import, its composition and the degree of food import dependency in Somalia, (b) to analyze various micro and macro economic channels through which both foreign food aid as well as imprudent domestic economic policies have resulted in Somalia's excessive reliance on imported food, and (c) to highlight main policy options that both Somali decisionmakers and her foreign aid donors should consider in order to arrest Somalia's accelerating trend in food import dependency.

I. Introduction

During the past ten years, Somalia has been experiencing a drastically rising food deficit that reflects a rapidly increasing per capita food consumption on the one hand, and a declining per capita food production on the other. Consequently, to meet its domestic food demand, Somalia has become alarmingly, and more than any other country in Sub- Saharan Africa, dependent on imported food. This is particularly striking when one notes that Somali economy has been predominantly agricultural/ pastoral and that, up until early seventies, it used to be self-sufficient in food grains. Questions then naturally arise as to (i) what have been the principal factors responsible for such an adverse development, (ii) what have been the main micro and macroeconomic channels through which these factors have operated, and (iii) what are the policy options for Somalia to achieve a reasonable degree of self-sufficiency in food.

While in recent years a large number of valuable studies have emerged that deal with the food problems of developing countries,l they have not always adequately addressed these questions at the individual country level. The studies have often been more concerned with such aspects of the food problem as poverty, hunger, malnutrition, food insecurity of vulnerable groups, role of food aid in emergency shortages, etc., than with economic analyses of food import dependency.2

Furthermore, even if concerned with economic aspects of the food problem, most studies have usually addressed issues at highly aggregated (LDC or regional) levels.3 As such, they have often lacked adequate country specificity to allow a more detailed study of the extent and root causes of food import dependency and to identify corrective policy options. By focusing on economic analysis of food import dependency in Somalia, the present study attempts to somewhat narrow that gap.

The remainder of the paper is organized as follows: Section II describes the trade data situation in Somalia and the data used in the paper. Section III analyzes in detail the trends in food import and its composition and provides estimates of the magnitude of food import dependency in Somalia. Section IV concentrates on unwanted effects of food aid and shows how in variety of ways it has acted as one of the principal factors underlying Somalia's excessive dependency on imported food. In Section V, imprudent domestic policies are found to have been another root cause of Somalia's present food problem. Section VI highlights major policy actions that both the Government of Somalia and its donors need to consider in order to reverse Somalia's worsening trend in food import dependency. Section VII concludes the paper.

II. Data

In Somalia, the state of national economic statistics is very poor and badly in need of an infrastructural development. In most areas where serious economic analyses are to guide policy decisions on key issues, either data are lacking or, if they exist, they suffer from (a) lack of consistency in definition, (b) incomplete coverage of what has been recorded, (c) delays, irregularities and discontinuities in reporting data.

This latter situation is particularly the case with Somali trade data. For instance, import data reported by Somali authorities based on foreign trade returns suffer from the following shortcomings:
(i) they are reported according to SITC categorization only up to 1981; thereafter, they are scatteredly reported and on an ad hoc classification of commodities; (ii) there are no systematic data on imports by end-use sectors/subsectors (agriculture, industry, services, etc.). This makes it inter alia very difficult to properly assess the minimum import requirements of different sectors of the economy, the efficiency with which imported commodities have been used, and the prospects for import-substitution in the economy; (iii) there are serious underestimation of imports due to incomplete coverage of all imported goods, smuggling activities and/or underinvoicing of imports as a way to evade import duties; (iv) there are substantial delays in preparing, processing and reporting foreign trade returns; and (v) local- currency valuation of imports have frequently failed to reflect current or appropriate exchange rates.

Because of these deficiencies in Somali import data, in the analysis that follows we have taken the alternative approach of using the United Nation's "partners trade data", known also as "D-Series". "D-Series" provide annual time-series data on trade flows between a reporting country and each of its trading partners, and are classified according to the Standard International Trade Classification (SITC) of the United Nations.

For most countries, these data are available at the SITC "five digit" level of disaggregation; for a few (including Somalia), they are available only to the "three digit" level. By aggregating export data reported by all countries exporting to Somalia, one obtains exports of "the rest of the world" to Somalia, which can then be used as the best proxy for Somalia's imports.

III. Food Import Dependency: the Magnitude

As shown in Table 1, Somalia's share of food import 4 in total import has been steadily and significantly rising, reaching an average of nearly 29 percent for the 1980-84 period compared with 19 percent during the decade of the seventies. While it may be tempting to attribute this to the adverse effects of events such as 1978-79 border conflict with Ethiopia and the droughts in 1974 and 1983, it is nevertheless more reflective of an underlying rising trend in food import. Indeed, even with the exclusion of the years of war and drought from the sample period, the value of food import shows a trend that has been growing at an annual rate of more than 21 percent.5 The extent of Somalia's dependency on food import is better understood when it is noted that the share of food import in total volume of food consumption rose from less than 33 percent on average for 1970-79 period to over 63 percent during the 1980-84 period, implying a trend growth in Somalia's food import dependency at the average rate of 8.3 percent per annum during the 1970-84 period.6

TABLE 1: SOMALIA - Food Imports: Trends and Composition, 1970-84 a/ 1970-1979 1980-1984 1970-1984


Food Imports Share (as percentage of total imports) 19.2 28.9 22.4
Composition: (as percentago share of food imports)
-  Cereal                               60.0   68.0  62.7
-  Dairy products                       8.6    18.8  12.0
Trends: (annual percentage growth rates)/Food Imports  19.9 
-  Cereal                               40.0
-  Dairy products                       21.6

pj All figures refer to import values in current US dollars.

All trend growth rates are OLS estimatos of the relationship log Mit const. + log(l+g).t + et for the 1970-84 period; where Mit denotes import category i (or its components), t is the time, g is the trend growth rate to be estimated, and et is the error term. Wherever necessary, the estimates have been corrected for autocorrelation by using Cochrane-Orcutt procedure.

The steeply rising trend of food import in Somalia is also reflected in a comparatively high elasticity of food imports with respect to GDP. As the following estimation results indicate, employing a "constant elasticity" (or "loglinear") specification7 one obtains an estimated food import elasticity of 1.42 which is statistically highly significant. In other words, for every one percent growth in GDP the value of food import has in all probability been increasing by 1.42 percent.

COEFFICIENT STANDARD ERROR T-STATISTIC

C-5.9964619  0.9492587  -6.3169941 
LGDP----1.4220984 -- 0.1397704 -- 10.174533
R^-squared 0.888432 --  Mean of dependent var         3.624359
Adjusted R-squared  0.879850   S.D. of dependent var  0.933043
S.E. of regression  0.323417   Sum of squared resid   1.359784
Durbin-Watson stat  1.517851   F-statistic            103.5211
Log likelihood      3.278646
 
In fact, this estimated elasticity refers only to a narrow definition of food import, namely food and live animals categorized under 71 Formally, the specification used is Log Mit Const. +6 Log GDP + et where Mit is the value (in US dollar) of the import category i, GDP is gross domestic product (in US dollar), E is the constant elasticity of Mi with respect to GDP, and et is the error term.

SITC 0. Once the definition is broaden to also include imports of animal fat and vegetable oil (SITC 4), beverages (SITC 11), and oil seeds, nuts and kernels (SITC 22), then a higher estimate of nearly 1.50 (with an estimated standard error of 0.132) is obtained for the food import elasticity.8

Furthermore, by examining the composition of food import at the 'two-digit level" classification, it becomes evident that cereals (wheat, rice, maize, barley in forms of unmilled flour or meal) account for a lion share of total food import (63 percent on average) followed by dairy products (milk, butter, cheese and eggs) with an average share of 12 percent of food imports. In other words, cereals and dairy products together have been accounting for 75 percent of the value of food import.

Furthermore, as Table 1 indicates, both of these categories have not only been following steeply rising trends (21 percent and 40 percent annual growth rates respectively) but also their average shares in total food import have been rising substantially during 1980-84 period compared with those in the seventies.

It is important to note that the sharp rise in Somalia's food import dependency, specially after 1975, has not been associated with any significant increase in exports. Indeed, as Table 2 indicates, during 1975-84 major export categories declined both in value and volume terms.

Once allowance is made for the war years of 1979-80 and the drought years of 1974 and 1983, the food import elasticity estimates will be even higher than those presented above; 1.475 and 1.542 for the narrow and broad definitions of food import, respectively.

Livestock export accounting on average for nearly 80 percent of total export earnings,9 declined at the annual rate of -4.4 percent in real terms over the 1975-84 period, with some notable fluctuations around the trend.10 Following a period of stagnation, livestock output showed a cyclical recovery in the last part of the seventies and early eighties. This, combined with price incentives provided under the "franco valuta" system, boosted livestock export to a peak of US$106 million (94 m. tons in wight) in 1982. However, in 1983 export fell dramatically (to about US$72 million, or 43.5 m. tons in weight) as Saudi Arabia (the principal market) banned livestock imports from Somalia, and in 1984 it reached a record low level of US$33 million (or, 21 m. tons in weight) as the ban continued.

Crop export has been limited almost entirely to banana export, with an average share of about 9 percent of total export value.11

Banana export fell sharply from an average of 105 thousand tons during 1970-75 to 42 thousand tons during 1975-84, with a low record level of 20 thousand 91 This excludes hides and skins and mirrah with an average share of 7 percent in total exports.

Export earnings from banana declined to an average of US$11.7 million (in real terms), or nearly one half of its average level during 1970-74. This poor export performance was partly due to increased competition in export markets, but principally reflected the dramatic decline in the area planted and hence in domestic output as a consequence of inadequate domestic price incentives and shortages of inputs, fuel, and tractor services during 1975-81.

Export of fish and fish products, accounting for nearly 2 percent of export earnings, increased from a low average of 2.1 thousand tons during 1970-74 to 3.9 over 1975-78 mainly due to offshore projects supported by the USSR. However, with the withdrawal of Russian trawlers in 1978 and lapse of joint Somali-USSR fishing projects in the early eighties on the one hand, and Government's strict control on pricing and marketing, on the other, output and export of fish declined notably during 1980-84.12

Table 2. Somalia: Exports of Various Food Categories

(Period Averages in Millions of U.S. Dollars)Percentage Rate of Change
75-79 80-84    Over 1976-84 Period
1. Live Animals.    87.0.83.0 (660)a   (64.0)     - 4.4
2. Bananas          14.9.11.7 (59.0)    (42.0)    - 3.0
3. Fish A Fish Products3.01.4 (3.9)     (3.2)     - 3.8
Notes: a - Figures in brackets present period averages in thousands of metric tons. b - OLS estimates corrected for autocorrelation.

Source: Export values and volumes are calculated on the basis of data from "Somalia: Economic Development Review", IMF (several issues) and from "Agricultural Statistical Handbook for Somalia", World Bank, 1985.

Given Somalia's declining food exports during the period under study and the fact that the economy of Somalia is overwhelmingly dependent on livestock and agriculture, (jointly generating about 60 percent of GDP and over 90 percent of exports earnings), the question then naturally arises as to what have been the causes of Somalia's excessive food import dependency. The subsequent sections of the paper attempt to shed light on this question.

IV. Food Aid

To meet its domestic demand for food, Somalia has become increasingly dependent on food aid from its donors.13 This is seen from Table 2 which, among other things, shows a trend growth rate in volume of food aid (81.4 percent per annum) that is dramatically (14 times) higher than the trend growth in food consumption (5.8 percent per annum); implying therefore a rapidly rising share of food aid import in satisfying domestic food consumption.

One cannot dispute the beneficial role of food aid in assisting low-income countries to sustain domestic food consumption of vulnerable groups at times of transitory shortages.14 Nor one can deny the partial (albeit temporary) relief that food aid can provide for the balance of payments in countries facing acute foreign exchange shortages. However, these should not mask several potentially distortive effects that are associated with increased and prolonged food aid.

TABLE 3: SOMALIA - Food Crops Aid. Import. Production and Consumption. 1970-84 a/ Shares 1970-79 1980-84 1970-84

Food Import (as percentage of domestic food consumption) 23.0 38.0 28.0
Food Aid (as percentage of Food Import) 20.3 51.7 30.8 1970-74 1975-79
Food Gap (production-consumption, as percentage of food production)+6.0        -11.1    -30.9    -12.3
Trend Annual Growth Rates c/ (Percent)
Food Consumption             6.8
Food Production              1.B
Food Deficit J               22.7
Food Import                  13.4
Food Aid Import              81.4
a/ All figures in this Table refer to volumes (and not values) of food crops.
b/ Averages over specified periods.
c/ OLS estimates of the logarithmic trend over 1970-84 period, except for Food Deficit.
d/ Covers the subperiod 1975-84 during which food deficit (domestic consumption exceeding domestic production) was experienced, but as an loutlayer excludes 1983 in which food deficit was exceptionally large due to crop failure caused by drought.

Source: Computed on the basis of data provided in Agricultural Sector Review, Task Force No. 5, Agricultural Statistics Handbook, World Bank, July 1986.

First, by increasing the supply (or sometimes oversupplying the market), food aid often acts to dampen domestic food prices or, at least, prevent them from rising, thereby reducing incentives to domestic producers of food crops and exacerbating the national food deficit.15 This disincentive effect has been strong in the case of Somalia because food aid (a) has substantially augmented total food imports rather than just replacing commercial imports, and (b) has been sold in the domestic market at prices significantly below import parity levels evaluated at a realistic market exchange rate. For instance, at an exchange rate of 35 shillings to the US dollar,16 it is estimated that in late 1983 the import parity prices for maize and sorghum exceeded their open market wholesale prices by nearly 10 percent and 16 percent respectively. And, at the then prevailing free market exchange rate of 51 shillings to the dollar, the excesses were about 55 percent and 67 percent respectively.17

Second, the donors of food aid have rarely requested the Government to use the counterpart funds (local currency) generated from the sales of food aid to provide incentives to domestic food grain producers. Where there has been any conditionality on the use of such funds, it has often been limited to payment to donors' project contractors and personnel to cover their local costs. So that, by and large, the Government has had a free hand in using such funds to finance its current expenditures which, by fueling inflation, have further reduced the real producer prices (see also Section VI).

Third, food aid, when provided for an extended period of time, can change the food habit of consumers (particularly that of urban groups) in the recipient country through replacing traditional food grains by imported substitutes. This has certainly been the case in Somalia where there has been a notable shift in consumption out of sorghum and maize to rice and wheat products. In fact, the average combined share of rice and wheat products in per capita grain consumption has increased (on volume basis) from around 15 percent during 1970-75 to over 40 percent in 1984-85.

Concommitant with this, the per capita consumption of rice and wheat products has been increasing from an average of less than 15 kilograms in 1970-74 to 29 kilograms in 1975-79 and to over 38 kilograms in 1980-84, reflecting, in turn, the continued overwhelming share of these items in total grain import (over 77 percent on average during 1970-85).18 In a rational response to such a change in food consumption habit, domestic grain producers alter their crop mix by switching to substitute grains, in production of which they may have no comparative advantage.19 Thus, by distorting both food consumption habit and the composition of crop production, food aid can cause severe costs to the recipient economy. Fourth, and more importantly, food aid increases the uncertainties facing domestic grain producers. This is because donors often fail to coordinate their food aid programs with respect to the type and size of food aid supplies, the timing of their releases, their selling prices to the domestic market, and the conditions associated with aid. Lack of a close coordination in these areas gives rise to additional risks and possible disincentives to domestic farmers.

In this regard, it is important to note that the donors' food aid budgets are primarily influenced by prospects for commercial exports of their food surpluses rather than being determined in accordance with the needs and objectives of recipient countries to reduce their food import dependency. Accordingly, donors usually reduce their food aid budgets when prospects for commercial exports of surpluses are good and increase them when those prospects are poor. As a result, significant price fluctuations are likely to occur in the domestic food market, particularly when the former decision of donors happen to coincide with a poor harvest in the recipient country and the latter with a good one.20 To this should also be added the fact that, depending on the donor country/organization, there have been considerable variations in prices at which food aid has been sold to the domestic market. Some donors (e.g. Italy) have provided food aid for free distribution, or have left the decision about domestic selling prices entirely to the Government; some (e.g. EEC) have preferred government's concessional pricing to the domestic market; and yet some (e.g. the United States) have requested the government to adhere to auction-market pricing. Part of this diversity in domestic pricing of food aid stems from differences in the exchange rates that various donors specify for local-currency evaluation of their contributions.21 However, it principally reflects the differences among donors in their objectives of food aid programs.

Since political and economic interests of donors (particularly bilateral donors) have played an important part in deciding about the size, timing, pricing and conditions of the food aid, it has been difficult for them to reach a coordinated decision on these aspects. On the other hand, since domestic producers act as residual suppliers, in the sense that their supplies should meet domestic food demand less of food imports (including food aid), their price expectations and production decisions as to when, by how much, and what type of food crops to produce are naturally affected by the donors' decisions. Consequently, the uncertainties generated by the donors about the quantity and composition of food aid, the timing of its release, the pricing to domestic market, and the conditionalities associated with aid are directly or indirectly transmitted to domestic food producers.

V. Domestic Policies

Another important cause of Somalia's increased dependency on food import can be attributed to the Government's failure to provide adequate price incentives to producers of food crops, on the one hand, and its policy of subsidization of food consumption, on the other. Government's food pricing policy has been suffering from a strong bias to favor urban consumers at the expense of farmers producing food crops. The Government has implemented this policy bias through two main channels: explicitly, by controlling producers prices for most agricultural crops at considerably below market levels, and implicitly by maintaining an effective exchange rate that has been highly overvalued.22 Both of these policies have the effect of taxing rural grain producers in order to ensure cheap food (whether domestically produced or imported) to urban consumers. And in doing so, they both have contributed to increased food import dependency by discouraging domestic food production and boosting domestic food consumption. That these policies have long been in effect in Somalia is transparent from figures presented in Table 3. They show that over the period 1975-84, the real producer prices for maize, sorghum and rice declined continuously and sharply at annual average rates of about 6.25 percent, 8.30 percent and 20.5 percent, respectively.23 Moreover, during the same period the real effective exchange rate appreciated rapidly (at an annual average rate of about 9.5 percent) thereby lowering the cost of imported ccreals (consumed largely by urban consumers) and reducing the price incentives to local producers. These distortionary policies (and particularly the maintenance of an overvalued exchange rate) strongly appeal to the government because they relieve the Government budget from part of the burden of providing direct food subsidy by shifting the burden to producers.

Ratio of Market to Official Price   1981    1982     1983

Whit, maze                          1.65   2.14      3.22
Red sorghum                         2.09   1.70      2.34
White rice                          3.03   1.27      1.68
Domestic spaghetti                  1.60   1.86-2.67 2.67

Despite the information on market and official selling prices, the precise extent of food subsidy is hard to quantify because some of the officially priced grains have always passed to the open market through wholesalers and retailers and that the data on volumes sold at each price tier is unavailable. However, according to the officials in the Ministry of Finance, in 1985 the cost of providing foodstuffs in kind or at highly subsidized prices to hospitals, schools, government employees, police and the army amounted to So.Sh. 1.2 billion, or 13.3 percent of the Government's ordinary expenditures. These transfers/subsidies were granted as a means to supplement salaries and/or to compensate for the impact of increased consumer prices of food.

TABLE 4: SOMALIA - Indices of Exchange Rate and Real Producer
Prices for Selected Food Crops (1974100)

Real Producer Prices / Percentage Annual Rate of Change Over the
      1976  1978  1977  1978  1979 1980 1981  1982 1983  1984 
Maize 92.1  88.2  99.8  90.5  73.2 73.4 67.1  62.1 82.3  42.5  -6.26
Sorghum 92.1 88.2 99.8  90.5  73.2 73.4 60.7  61.8 87.2  36.8  -8.31
Rice   83.8  73.4  88.4  60.4 48.8 30.6 19.4  16.8 19.0  10.1 -20.67
Sesame 83.8  73.5  78.9  72.6 68.5 45.8 49.5  62.6 67.0  31.0  -8.08

Exchange Rate Real Effective Rate b(pd. averages)    
103.4  122.7  123.4  117.6  125.9  172.4  208.9  186.7 183.1 293.4 +9.S2
p/ Lacking a consumer price index for the rural areas, official producer prices have been deflated by Mogadishu Consumer Price Index.
I/ Increase in index denotes real effective appreciation.
c/ OLS estimates.

Sources: Real Producer Prices calculated from IMF, 'Recent Economic Developments, July 1981.
World Bank, "Agricultural Sector Review', Task Force No. 5, Agricultural Statistics Handbook, July 1985.
Exchange Rate calculated from IMF Data Fund.

Apart from these imprudent policies, the Government has also followed the unwise policy of depending excessively on local-currency funds generated from sales of food aid to finance its budget. In recent years, these funds have on average constituted over 10 percent of the total revenues and grants in the government budget, compared to approximately 7.5 percent in Sudan, 5.1 percent in Madagascar, and 3.0 in Zaire.24 Since in financing its expenditures it is politically more convenient for the Government to rely on revenues from sales of food aid than to resort to often unpopular alternative of raising tax revenue, the Government has developed a preference for the former at the expense of severely undermining the fiscal base of the economy and the need for domestic resource mobilization.25 This presents yet another potentially distortive effect that can be attributed to food aid programs.

Furthermore, it is likely that the negative impacts of food aid and government domestic food policies have significantly reduced farmers'real incomes and their incentives to plant, produce, or invest in food crops - thereby causing a shift away from production of traditional food grains to non-food crops and/or to non-tradeable commodities.

Unfortunately, at the present time, the lack of data does not allow one to provide quantitative evidence of such a shift. However, the economic implications of such a shift are important enough to be pointed out here. Among other things, a shift away from production of traditional food grains means increased dependency on food imports, and therefore increased pressure on the balance of payments. The question of national food security aside, this in turn raises important questions for the efficiency of resource allocation and agricultural development strategy in Somalia.

For, unless the resources shifted out of food crops production can generate larger value added in non-food agricultural production or elsewhere in the economy, or can generate larger net foreign exchange earnings (compared to foreign exchange costs of food imports that the shift necessitates), such a shift would inflict a net loss on the economy both in terms of domestic as well as foreign exchange resources. It is therefore important that the potential of foreign exchange savings through expansion of food crops production (and hence reduced food imports) are carefully weighed against potential of foreign exchange earnings through allocation of resources to export crops.

VI. Policy Options

The preceding analysis points to a number of important policy actions that both the Somali Government and its aid donors should consider seriously if they are committed to arrest Somalia's accelerating trend in food import dependency. On the Government side, to encourage domestic food production, immediate decontrol of foodgrain prices to producers and establishment of food crops floor prices are essential first steps.

This should be accompanied by non-price supply incentives such as provision of agricultural credits to small farmers, granting the peasants with the right of access to the vast, unused rainfed lands; provision of essential inputs such as water, seeds, fuel, fertilizer and insecticides; provision of research and extension services; improvement in grains marketing; provision of storage facilities and infrastructures. These actions should constitute core elements of a sound national food strategy for Somalia whose central objective ought to be gradual substitution of domestic food production for food imports over the medium to long-term horizon.

Complementary to this strategy is an industrial development strategy of which an integral element should be import-substitution in the areas of dairy products, meat processing and preparation, animal fat and vegetable oil production, and grain food processing and preparation. As was indicated earlier, together these account for a notable share of the total value of food imports to Somalia. So that, given the potentials in the agriculture and livestock sectors, a policy of import-substitution in these areas should in principle succeed to generate sizable value added and foreign exchange savings. The success of such a policy will also hinge, to a significant degree, upon the government's efforts to adjust the real exchange rate from its presently overvalued level to a realistic level which would enhance import-substitution possibilities.

The foregoing policy measures, if undertaken fully and on a sustained basis, will encourage growth of domestic food production. But, unless these supply-side policies are adopted simultaneously with measures to check the accelerating trend in domestic food demand, the achievement of the overall goal of reducing food import dependency will prove a particularly formidable task. To better appreciate this point, it should be noted that if no measures are taken to curb the growth of food consumption from its trend rate of 5.8 percent per annum, then even if one were to assume that the supply-side measures would result in a 6 percent annual growth of food crop production; that there would be no adverse external events (drought, war, refugee immigration); and no additions to the existing grain stock; still it would take some 20 years before the food crop import-consumption ratio can be reduced to half of its current level of nearly 40 percent.26 On the other hand, this same objective can be achieved in less than five years if, in addition to the supply assumptions mentioned above, one assumes that the growth of food crop consumption would be curbed only by 1 percentage point per annum from a trend rate of 5.8 percent to 4.8 percent. Of course, the assumption of 6 percent annual growth in food crop production may be too optimistic, compared to perhaps a more realistic growth rate of 4.5 percent. In this case, the objective of lowering the import-consumption ratio to half its current level in five years time would require that the growth rate of food crop consumption be contained to 3.45 percent per annum.27 The message from these numerical stipulations is plain: in order for Somalia to attain the goal of reduced food import dependency it is essential that in addition to undertaking supply-side measures the Government abandons its present policy bias of favoring urban consumer groups through direct or indirect subsidization of their food consumption.28

While the removal of food subsidies and adjustment of the exchange rate will help curtail food imports, a change in the present food aid policies of donors is a prerequisite for reducing Somalia's food import dependency. In this regard, it is important to recognize that the best way donors can assist in alleviating Somalia's food problem is to redesign their food aid programs to make them compatible with, and promotive to, development of agriculture and related industries which should be the ultimate source of food for Somali people. Accordingly, donors, as well as the Government, should consider food aid as a transitory program that will help to sustain food consumption of vulnerable groups and fill a narrowing gap between food production and consumption until domestic production has grown enough to meet demand. In order for food aid programs to fulfill such a role, a close coordination among donors is imperative. Donors' orchestrating actions are required in several crucial respects:

a) There is an urgent need for a donor-coordinated plan to reduce over time the overall volume of food aid to levels that are consistent with adequate price incentives to domestic farmers and with the objective of gradual substitution of domestic production for imported food.

b) To prevent the bias towards consumption of imported rice and wheat products and to facilitate achieving the goal of reduced food import, the composition of food aid programs needs to be altered to encourage production and consumption of sorghum and maize and to promote substitution for imported dairy products. and vegetable oils.

c) To minimize the adverse effects of uncertainties about volume, composition, and timing of food aid inflows, donors should restructure their aid programs from a year-to-year basis (which is often influenced by commercial export prospects for their surpluses) to multi-year commitments that are based on Somalia's future food import needs consistent with planned targets to reduce its food import dependency.

d) To make food aid an effective development instrument, it is essential that donors abandon the present inefficient marketing practices whereby each donor uses a different government agency to channel its food aid. Instead, donors and the Government should jointly establish a special unit that will be responsible for a coordinated management of all food aid programs.

e) To eliminate the present large variations in pricing of food aid to the domestic market, there is a need for donors to adopt a uniform pricing policy that reflects market realities and avoids price disincentives to local producers and price subsidies to consumers. As part of this policy, it is important that donors uniformly use the free market exchange rate for local-currency evaluation of their food assistance.

f) Donors should end their present individualistic approaches to specifying uses of counterpart funds generated from sales of food aid. Such an approach has not only distorted the focus of food aid programs but also has been a source of Somalia's weak fiscal and monetary discipline. To enhance the effectiveness of food aid programs, a close coordination among donors is needed to ensure that these counterpart funds are used in support of policy measures (discussed above) necessary for increasing domestic food production.

VII. Summary and Conclusion

The main findings and arguments of this paper can be summarized as:
a) Over the past decade, Somalia has been experiencing a rapidly growing gap between its food consumption and production which has resulted in a change of her position from a virtually self- sufficient country in food until 1975 to a major food- deficit country since then. To fill this gap, Somalia has become increasingly, and to an alarming degree, dependent on imported food (particularly food aid).
b) The chief factors responsible for this development have been:
(i) ill-coordinated and ill-planned food aid programs of Somalia's donors which have often had the effect of undermining the incentives to local farmers, augmenting risks and uncertainties faced by them, intensifying import bias in food consumption and production patterns, and contributing to Somalia's weak financial discipline by allowing the public budget to become heavily dependent on counterpart funds generated from sales of food aid; and
(ii) imprudent government policy of favoring the welfare of urban consumers at the expense of food growing peasants through controlling food crop prices, subsidizing consumer prices of food, and maintaining an overvalued real exchange rate.

c) The paper has highlighted the major policy actions that both Somali decisionmakers, as well as donors, ought to seriously consider if they are to arrest the accelerating trend in Somalia's food import dependency. Indispensable to that task will be a well-planned national food strategy that aims to gradually substitute domestic production for imported food. The success of such a strategy will crucially depend on two fundamental reforms. First, a change in the Government's policy that would end the present bias of protecting the living standard of the urban groups at the expense of those in the agricultural sector, and would, instead, emphasize growth of food production through providing adequate incentives to farmers. Second, a change in donors' food aid policies that would replace their presently self-interested, bilateral approach by a coordinated, multilateral approach in order to make food aid programs not only compatible with, but also conducive to, the goal of reducing Somalia's food import overdependency.

Obviously, since certain urban groups are bound to lose from these reforms (even if only temporarily), the effective implementation of the reforms can face serious political economy constraints. Indeed, under pressure from such groups to maintain their standard of living, the Government may resort to countervailing economic measures that can mitigate or even nullify the impact of the reforms. This is particularly likely in the case of agricultural price reform, whether the reform takes the form of direct price increases or is done indirectly through exchange rate devaluations. That whether in such a situation it would be wiser if policy reform emphasizes expansion of domestic food supply through direct, non- price measures remains an open and interesting area for further research. In any event, it is important that in formulating agricultural price reforms and assessing their effectiveness, political economy constraints are explicitly taken into consideration. While general work in this area has already attracted researchers' interests (see, e.g., Sen (1972), Lindbeck (1973), Lipton (1977), and Braverman and Kanbur (1986)), further research in the context of individual country cases can be highly beneficial to the design of policy reforms in countries, like Somalia, where concerns for the well-being of certain urban groups are especially strong.

NOTES

1/ See, for example, Ghai and Radwan (1983), Huddleston (1984), Mellor (1984), Taylor and Mervin (1985), and FAQ (1985).
2/ See, for example, Lane (1980), Reutlinger and Knapp (1980), Green and Kirkpatrick (1982), and Reutlinger and van Holst Pellekaan (1986).
3/ See, for example, Abalu (1982), US Department of Agriculture (1981)(1983), Timmer, Falcon and Pearson (1983), Delgado and Mellor (1984), and de Wilde (1984).
4/ Food import is defined as the aggregate of SITC 0 (food and live animal) + SITC 4 (animal, vegetable, oil, fat) + SITC 11 (beverages) + SITC 44 (oil seeds, nuts, kernels).
5/ The trend growth rate over the whole 1970-84 period is estimated at a slightly lower rate of 19.9 percent per annum.
6/ Food import dependency is defined here as the ratio of food import to available food for consumption, where the latter is equal to food production - waste and seed + imports - exports - stock change. The figures are computed on the basis of data in Shapouri, Dommen and Rosen (1986), Appendix Table 8. The trend growth rate is OLS estimate corrected for autocorrelation.
8/ These estimates are based on data for the whole sample period 1970-84.
10/ These fluctuations partly reflected the nomads' decisions to sell fewer animals to reconstitute their herds following the drought years.
11/ A small amount of rice, averaging 4000 metric tons a year, was exported during 1970-74, but it ceased completely after 1974 as domestic rice consumption rose sharply.
12/ It is worth mentioning that between 1970 and 1979 Somalia exported some meat and meat products, averaging US$5 million a year, to USSR and some East European countries. However, meat export completely ceased after 1979 with Somalia's political reorientation.
13/ Somalia's principal food donors are the United States (with a dominant share of total food aid to Somalia), World Food Program (which utilizes the contributions of various donors), EEC, and Italy.
14/ The transitory shortages may be caused by crop failure, drought, flood, war and similar factors.
15/ For two very interesting studies of the possible disincentive effects of food aid in India and Colombia, see Isenman and Singer (1977) and Dudley and Sandilands (1975), respectively. For a useful summary list of studies on the price effects of food aid in other countries, see Schneider (1975).
16/ This would be approximately the exchange rate in 1983 if the purchasing power parity were to be maintained at 1974 level.
17/ It should be noted that the late 1983 wholesale prices were judged to be extraordinarily high by historical comparisons. Note also that the excess of import parities over domestic selling prices would be drastically larger if comparisons were made with official selling prices, which in 1983 were around half the open market prices.
18/ According to a family expenditure survey for Mogadishu undertaken in 1977 by Central Statistical Department of the then State Planning Commission, food accounted for 60 percent of family expenditures and, of the total expenditure, 19.4 percent was spent mostly on imported grains (in the proportions of 6.3 percent rice, 4.3 percent wheat flour, 4.8 percent wheat bread, and 4.0 percent spaghetti) compared with 9.7 percent spent on domestically produced grains (in the proportions of 0.5 percent sorghum, 7.7 percent maize, and 1.5 percent ground maize). Further, as a joint report by Somali Government and World Bank (1984) indicates, once it is taken into account that during 1974-1981 about 30 percent of maize consumption was imported, the ratio of imported to domestic grain in Mogadishu diets becomes three to one.
19/ The observation in recent years that Somali farmers in the Shebelli Valley and the Juba Valley have been increasingly growing rice and wheat (in addition to maize) indicates such a change.
20/ For an analysis of possible adverse effects of food crop price instability on farmers' output, income, and welfare, see Newberry and Stiglitz (1981) and particularly Bigman (1985).
21/ The exchange rates used by donors have ranged between the official and the market rates for which the averages in 1985 were 40.17 So.Sh/US$ and 101.60 So.Sh/US$ respectively.
22/ Of course by keeping official consumer prices of food below free/parallel market levels, the Government has also been providing direct food subsidies to various urban consumer groups. The joint report by Somali Government and World Bank (1984) gives the following estimates of the excess of market retail prices over official selling prices for various cereals during 1981-83 (August of each year).
23/ The declining trends in real producer prices for grains in Somalia should not be misinterpreted to reflect the declining world grain prices after 1974. Such an association should be viewed more as apparent than real for since early 1970's agricultural producer prices in Somalia, instead of being influenced by world market prices, have been set by Governmental Agency (Agricultural Development Corporation (ADC)) at artificiary low levels to fulfill the Government's special concern for the well-being of selected urban consumer groups. Indeed, in contrast to the world nominal prices of cereals that showed consinderable fluctuations during the period 1966-1984 - with two peak levels in 1974 and 1981 - Somalia producer prices, being administratively fixed, have followed step-wise rising trends since 1973. An apparent association (after 1974) between real producer prices in Somalia and world real prices is mainly due to very high inflation rates in Somalia since early 1970's.
24/ For Somalia, the ratio is calculated on the basis of data from Table 24 in National Development Strategy and Program, Ministry of National Planning, Somali Democratic Republic, September 1985. For other countries, the ratios are based on World Bank and IMF data.
25/ The deteriorating performance of the Government in raising tax revenue is well reflected in a tax revenue-GDP ratio that has been rapidly declining from an already low average of 11.7 percent during 1979-81 to 8.3 percent in 1982-83, down to an average of 3.9 percent for 1984- 85. (Source: IMF)
26/ The calculation is based on the formula (M/C)T 1 - (Po/Co)[(l+Pg)/(l+Cg)]T where (M/C)T is the targeted import-consumption ratio (0.20 here), T is the number of years before the targeted import-consumption ratio is achieved, Pg and Cg are respectively the assumed annual growth rates of food crop production and consumption, and POICO is the base year food crop production-consumption ratio (currently around 0.76). Note that no change in the food stock is assumed.
27/ With a population growth rate of nearly 3 percent, this would still imply a 0.45 percent annual growth of per capita food grain consumption.
28/ This does not mean that the Government should not be concerned about the food security of the urban poor. However, such concerns would best be served if the government sharply targets its food assistance (in cash or in kind) to the very poor and tries to raise the income of this group through public employment programs. For specific intervention policies that can be used to alleviate the food insecurity of the poorest groups, see Timmer, Falcon, and Pearson (1983) and particularly Poverty and Hunger, World Bank (1986).

BIBLIOGRAPHY

1. Abalu, G.O.I., "Solving Africa's Food Problem", Food Policy, Vol. 7, No. 3, August 1982.
2. Bigman, D., "Food Policies and Food Security under Instability: Modelling and Analysis", Lexington, Massachusettes, Lexington Books, 1985.
3. Braverman, A. and Kanbur, S.M.R., "Urban Baias and the Political Economy of Agricultural Price Reform", mimeo, World Bank, Washington, D.C., June 1986.
4. De Wilde, J.C., "Agricultural Marketing and Pricing in Sub-Saharan Africa", African Studies Center, University of California, 1984.
5. Delgado, C.L. and Mellor J.W., "A Structural View of Policy Issues in African Agricultural Development", American Journal of Agricultural Economics, Vol. 66, No. 5, December 1984, pp. 665-70.
6. Dudley, L. and Sandilands, R.J., "The Side Effects of Foreign Aid: The Case of PL480 Wheat in Colombia", Economic Development and Cultural Change, Vol. 23, January 1975, pp. 325-36.
7. Food and Agriculture Organization of the United Nations, Food Situation in African Countries Affected by Emergencies: Special Report, Rome, December 1985.
8. Ghai, D. and Radwan, S., eds., Agrarian Policies and Rural Poverty in Africa, Geneva: International Labor Organization, 1983.
9. Green, C. and Kirkpatrick, C., "A Cross-Sectional Analysis of Food Insecurity in Developing Countries: Its Magnitude and Sources", The Journal of Development Studies, Vol. 18, No. 2, January 1982, pp. 185-204.
10. Huddleston, B., Closing the Cereals Gaps with Trade and Food Aid, IFPRI Research Report No. 43, Washington, D.C.: International Food Policy Research Institute, 1984.
11. Isenman, P.J. and Singer, H.W., "Food Aid: Disincentive Effects and their Policy Implications", Economic Development and Cultural Change, Vol. 25, No. 2, January 1977, pp. 205-37.
12. Lane, S., "The Contribution of Food Aid to Nutrition", American Journal of Agricultural Economics, Vol. 62, No. 5, December 1980, pp. 984-87.
13. Lindbeck, A., "Endogenous Politicians and the Theory of Economic Policy", Seminar Paper No. 35, Institute for International Economic Studies, Stockholm, 1973.
14. Lipton, M., Why Poor People Stay Poor: A Study of Urban Bias in World Development, Temple Smith, London, 1977.
15. Mellor, J.W., "Food Aid, Equity, and Agricultural Growth", IFPRI Report 1983, Washington, D.C.: International Food Policy Research Institute, 1984, pp. 7-11.
16. Reutlinger, S. and Knapp, K., "Food Security in Food Deficit Countries", World Bank Staff Working Paper No. 393, Washington, D.C., 1980.
17. Reutlinger, S. and van Holst Pellekaan, J., Poverty and Hunger: Issues and Options for Food Security in Developing Countries, Washington, D.C.: World Bank, 1986.
18. Schneider, H., The Effects of Food Aid on Agricultural Production in Recipient Countries, Paris: OECD Development Center, 1975.
19. Sen, A.K., "Central Areas and Accounting Prices: An Approach to Economic Evaluation, Economic Journal, 1972.
20. Shapouri, S., Dommen, A.J., and Rosen, S., Food Aid and the African Food Crisis, Economic Research Service, FAER-221, U.S. Department of Agriculture, June 1986.
21. Somali Government and World Bank, Report of the Study of Agricultural Incentives and Grain Marketing in Somalia, mimeo, Washington, D.C., January 1984.
22. Taylor, C. and Yetley, M.J., Food Gaps and Surpluses in Developing Countries: Status, Trends, and Implications, 1967-88, ERS Staff Report No. AGES840912, U.S. Department of Agriculture, Economic Research Service, September 1985.
23. Timmer, C.P., Falcon, W. and Pearson, S., Food Policy Analysis, Baltimore, MD., John Hopkins University Press, 1983.
24. U.S. Department of Agriculture, Economic Research Service, Food Policies in Developing Countries, FAER-194, December 1983.
25. U.S. Department of Agriculture, Economic Research Service, Food Problems and Prospects in Sub-Saharan Africa: The Decade of the 1980s, FAER-166, 1981.