food aid, domestic policy and food security: contrasting ... · tries. this study reviews the...

Food Policy 32 (2007) 413–435

www.elsevier.com/locate/foodpol

Food aid, domestic policy and foodsecurity: Contrasting experiences from South

Asia and sub-Saharan Africa

Carlo del Ninno *, Paul A. Dorosh, Kalanidhi Subbarao

The World Bank, Safety Nets, Social Protection and Labor, 1818 H Street, NW,

Washington, DC 20433, United States

Received 25 October 2005; received in revised form 3 November 2006; accepted 15 November 2006

Abstract

Food aid, both for short-term emergency relief and as program food aid that helps address med-ium-term food ‘‘deficits’’, is often a major component of food security strategies in developing coun-tries. This study reviews the experience with food aid of four major recipients of food aid (India,Bangladesh, Ethiopia and Zambia) regarding food production, trade, markets, consumption andsafety nets, as well as the policy responses to food emergencies. The widely varying experiences ofthe study countries suggest that food aid that supports building of production and market enhancinginfrastructure, is timed to avoid adverse price effects on producers, and is targeted to food insecurehouseholds can play a positive role in enhancing food security. However, food aid is not the only, orin many cases, the most efficient means of addressing food insecurity. In many cases private marketscan more effectively address shortfalls in food availability and cash transfers may be a viable alter-native to food transfers in-kind.� 2006 Elsevier Ltd. All rights reserved.

JEL classification: Q18-Agricultural policy; Food policy; I38-Government policy; Provision and effects of wel-fare programs; F1-Trade

Keywords: Food aid; Food security; Trade; Bangladesh; Ethiopia; India; Zambia

0306-9192/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.foodpol.2006.11.007

* Corresponding author. Tel.: +1 202 4589579; fax: +1 202 5223252.E-mail address: [email protected] (C.del Ninno).

mailto:[email protected]

414 C. del Ninno et al. / Food Policy 32 (2007) 413–435

Introduction

Maintaining food security at the national and household level is a major priority formost developing countries, both for the welfare of the poor as well as for political stability.In order to help assure food security, developing country governments have adopted var-ious strategies including efforts to increase production (often with an explicit goal of foodself-sufficiency), government intervention in markets, public distribution of food andmaintenance of national food security stocks. Food aid, both for short-term emergencyrelief and program food aid that helps address medium-term food ‘‘deficits’’, is often amajor component of these food security strategies (Shaw, 2002; Haddad and Frankenber-ger, 2003).

Numerous concerns have been raised, however, about the efficiency of food aid – sup-ported programs in meeting their objectives.1 Problems identified include minimal impacton development (e.g., Pillai, 2000; Clay et al., 2005), high cost of procurement of tied aid,2

and poor targeting (Clay et al., 1999; Jayne et al., 2001). Moreover, along with food-assisted programs in general, food aid resourced programs often involve high administra-tive costs within-country (due in part to leakages; Subbarao and Smith, 2003a).Emergency food aid is also often criticized for lack of timeliness (late arrival or even can-cellation) and the high cost of delivery to the recipient country (Barrett and Maxwell,2005; Clay et al., 2005; Shaw, 2002). These limitations notwithstanding, emergency foodhas often been effective in addressing short-term relief needs.

Perhaps most important implication for long-term security, however, is the possibilitythat food aid may cause disincentives for domestic production through reductions indomestic prices and lead to reduced public and private investment in food production.To the extent that these disincentive effects are large, the benefits of food aid in addressingacute short-term food insecurity may be offset by the cost of reducing long-term food secu-rity. These possible medium-term disincentives of food aid are particularly importantbecause, although long-term program food aid is declining, emergency operations oftenextend for several years, and in some countries (e.g. Ethiopia), there has been almost acontinuous sequence of successive emergency operations. Several individual country stud-ies have examined these medium-term food aid issues in major food recipient countries(e.g. Shaw and Clay, 1994; WFP, 1994, 1995; Webb and Von Braun, 1994 and Von Braunet al., 1998; Ahmed et al., 2000). Clay and Stokke (2000) and Barrett (2006) also provide across-country analysis of medium-term issues and draw some useful lessons of experience.

The objective of this paper is to provide a systematic comparison of experiences ofselected countries of South Asia and sub Saharan Africa regarding impacts of food aid,not only in emergency situations, but with respect to medium-term production, tradeand consumption. The countries covered include four major recipients of food aid overthe past four decades, viz., India, Bangladesh, Ethiopia and Zambia. The two South Asiancountries have been the center of much of the debate on food aid and food security.3

1 Recent analyses include Shaw and Clay (1994, 1998); Hoddinot et al. (2004); Barrett and Maxwell (2005) andClay et al., 2005.

2 Cost efficiency estimates vary widely across donors, modes of supply, commodities and destinations, rangingfrom 10% below to more than 55% above the cost of alternative commercial imports (Clay et al., 2005, p.13).

3 India was among the leading recipients of food aid in the 1960s, as was Bangladesh (formerly East Pakistan)from the 1960s through 2000.

C. del Ninno et al. / Food Policy 32 (2007) 413–435 415

Among countries of sub-Saharan Africa, Ethiopia has been the leading recipient in mostyears since 1980; Zambia is chosen as a representative country of southern Africa.4

In the next section, we provide a brief description of the economic structure of the fourcase study countries, including an overview of their food economies and recent food aidflows. Section ‘staple food production, trade and availability’ examines aspects of theavailability of food: production, trade and markets. In Section ‘Food consumption andsafety nets’, we discuss consumption patterns and safety nets. Section ‘governmentresponse to crises covers’ policy responses to food emergencies, generally related to acuteproduction shortfalls. Finally, Section ‘conclusions and policy implications’ presents con-clusions and policy lessons.

Economic structure of the case study countries

India, Bangladesh, Ethiopia and Zambia differ substantially in terms of population,income levels and economic structure – differences which are crucial to understandingand interpreting their contrasting experiences with food aid and agricultural development.

India’s population in 2002 (1050 million) is about 50% greater than that of sub-SaharanAfrica. Bangladesh with a population of 144 million in 2002 and a population density ofnearly 1000 per square-kilometer has roughly twice the population of Ethiopia, in an areaonly 13% as large. Zambia’s population density of only 14 per square-kilometer, is onlyhalf the sub-Saharan African average, and less than 1/50th that of Bangladesh. Per capitaGDP (2002 data) is also significantly higher in India ($480/person) and Bangladesh ($360/person) than in Ethiopia (only $100/person), and per capita income of Zambia ($330/per-son) is close to that of Bangladesh.

Because of their very large populations, the total volume of food consumption require-ments in India and Bangladesh are substantially greater than in Ethiopia and Zambia andthose of most other countries in SSA. This suggests a greater possibility for gains fromeconomies of scale in marketing and storage. Moreover, lower per capita incomes, in Ethi-opia and Zambia suggest less market purchasing power among households in thosecountries.

Bangladesh and India also differ from Ethiopia is terms of growth and poverty out-comes over time. The agricultural growth in the two South Asian countries has beenlabor-demanding; real wages have risen, and the incidence of poverty has been reduced.For example, the head-count poverty in India prior to the Green Revolution was 55%;now it is about 30%, and in the agriculturally prosperous regions the incidence is now only20%. Both countries have also improved their human capital outcomes, Bangladesh lever-aging food-based safety net (food for education program) to promote both food securityand school enrollments. In Ethiopia, overall growth has been modest, and growth-inducedpoverty reduction has been minimal.

Nonetheless all four countries broadly share similar food security concerns. At varioustimes, all the countries considered here have had significant trade deficits in cereals, themajor staples for their households. However, India and Bangladesh have achieved large

4 Zambia has experienced several major droughts leading to large-scale inflows of food aid. UnlikeMozambique and Zimbabwe, but like Malawi, Zambia has enjoyed relative political stability and absence oflarge-scale armed conflict.


gains in per capita food production over time. None of the countries, though, has yetresolved problems of household-level food security, and overall per capita calorie con-sumption is still low in Bangladesh (2186), Ethiopia (2037) and Zambia (1885). Even inIndia, close to one-fifth of India’s population continue to be food insecure, and numerousprograms to promote access of the poor to food have met with only marginal success evenas food grain stocks mounted.

Consumption patterns differ in important ways as well and are affected by the vari-ability in domestic and international prices of staple foods. Cereals account for a highpercentage of calories in all countries, ranging from about 60% in India and Zambia,to 69% in Ethiopia to 81% in Bangladesh (Table 1). Moreover, a single major stapledominates cereal (and total) calorie consumption in Bangladesh (rice) and Zambia(maize), making poor consumers in these countries especially vulnerable to changes inthese prices. To the extent that imports are a major share of supply of staples or thatprices of these tradable staples are largely determined by import prices (as for maizeand wheat in Ethiopia, maize in Zambia and wheat and to a lesser extent, rice in Ban-gladesh), changes in international markets can also have a large impact on food con-sumption and food security. In principle, commercial imports of these major staples

Table 1Food economies of study countries, 2001

India Bangladesh Ethiopia Zambia

Production

Total Cereals (mn ton) 196.84 25.95 9.64 0.74Population (mns) 1024.9 140.4 64.5 10.6Total Cereals per capita (kg/year) 192.1 184.9 149.5 69.6

Shares of cereal production

Rice 47.3% 93.3% 0.0% 1.0%Wheat 35.4% 6.4% 16.6% 10.1%Maize 6.8% 0.3% 34.2% 81.2%Other cereals 10.5% 0.0% 49.2% 7.6%Total cereals 100.0% 100.0% 100.0% 100.0%Total cereal imports (mn ton) 0.06 2.06 1.11 0.13Of which food aid (mn ton) – 0.491 0.568 0.057Cereal consumption (mn ton) 166.35 25.39 10.04 1.49Per capita (kg/year) 162.3 180.9 155.8 139.8Cereal imports/total consumption 0.0% 8.1% 11.1% 8.7%Cereal food aid/total consumption 0.0% 1.9% 5.7% 3.8%Peak food aid year 1966 1988 2000 1992Food aid peak year (mn ton) 10.60 1.79 1.21 0.51Cereal food aid/total consumption 14.9% 12.1% 13.4% 27.7%

Per capita calorie consumption

Cereals (%) 59.8 80.9 69.2 62.3Starchy roots (%) 1.9 2.0 11.3 14.3Other vegetable products (%) 30.5 14.1 14.7 18.3Animal products (%) 7.8 3.0 4.7 5.0Total Kcal person/day 2487 2186 2037 1885Per cap. protein cons. (g) 58.4 46.8 59.7 46.3Of which: animal protein (g) 10.4 6.0 6.4 8.8

Note: peak food aid year: Ethiopia through 2001 (India food aid 1966 est. as total cereal imports).Source: calculated from FAO food balance sheet data files.

Table 2Food aid in study countries, average 1995–2004

India Bangladesh Ethiopia Zambia

Total food aid (2000 ton) 280 624 880 59CommodityCereal (%) 84.9 99.0 95.8 84.7

Mode of supply

Direct transactions (%) 82.6 92.3 73.7 48.9Local purchases (%) 17.4 0.7 18.8 32.5Triangular transactions (%) 0.0 7.0 7.4 18.6

Category of use

Emergency (%) 1.8 13.0 77.9 57.9Programme (%) 2.0 23.9 2.7 1.6Project (%) 96.2 63.1 19.5 40.5

Source: calculated from WFP/INTERFAIS, June 2006.


can help stabilize supplies, weakening the case for food aid and food-assisted programsas opposed to cash transfers, but viability of such trade varies across these countriesaccording to infrastructure, private market development and government trade and pric-ing policies.

Food aid flows

Food aid has been a major source of cereal availability in peak food aid years, from12% in Bangladesh in 1987/88, a year of major floods, to 28% in Zambia in the 1992drought year. In recent years, though, food aid has not been a major source of cereal avail-ability in the South Asian countries, essentially zero in India and only 1.9% in Bangladeshin 2001, but remains significant in Ethiopia and Zambia, 5.7% and 3.8% of cereal availabil-ity, respectively. Almost all of this food aid is in the form of cereals. Local purchasesaccounted for nearly 20% of food aid in Ethiopia and India from 1995 to 2004 and morethan 30% in Zambia. In Bangladesh, almost all food aid is ‘‘direct transactions’’ (food aidimported from donor countries) (Table 2; See Fig. 1 for trends of country details onmodality of supply).

Emergency food aid predominates in Ethiopia and in years of major floods in Bangla-desh and droughts in Zambia. The remainder of food aid has been project food aid inrecent years; program food aid has been discontinued in these countries (Table 2). The dis-tribution of food aid by category of use mirrors the patterns in the respective regions andglobally.5 In sub-Saharan Africa as a whole, emergency food aid nearly doubled from themid-1990s to 2003–2004 (Shaw, 2002). In both continents, program food aid has dimin-ished substantially, falling to only about 5% and 10% of total food aid in sub-SaharanAfrica and Asia, respectively.

5 This decline in program and project food aid is in part due to a long-term shift away from food aid fordevelopment to food aid for relief, as well as changes in domestic agricultural policies in the OECD countries thatreduced surpluses and support for food aid (Shaw and Clay, 1994; Clay, 2003). Note that the distinction betweenprogram and project food aid, however, varies somewhat by donor (Barrett and Maxwell, 2005).

India: Food aid deliveries by mode. 1995-2004

0

50

100

150

200

250

300

350

400

450

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

(Th

ou

san

d t

on

s)

Direct Local Triangular

Source: WFP/INTERFAIS. June 2006

Bangladesh: Food aid deliveries by mode. 1995-2004

0

200

400

600

800

1. 000

1. 200

1. 400

1. 600

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

(Th

ou

san

d t

on

s)



Ethiopia: Food aid deliveries by mode. 1995-200

0

500

1. 000

1. 500

2. 000

2. 500

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

(Th

ou

san

d t

on

s)



Zambia: Food aid deliveries by mode. 1995-2004

0

20

40

60

80

100

120

140

160

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

(Th

ou

san

d t

on

s)



Fig. 1. Food aid deliveries by mode (direct transfers, local purchases and triangular transactions) in India,Bangladesh, Ethiopia and Zambia.


Staple food production, trade and availability

Production and availability

Domestic food production in the study countries is the major source of food supply,and fluctuations in production are a major cause of instability in food availability. Chronicproduction deficits, as reflected in large-scale food imports and low overall availability ofmajor staples, are typically a major rationale for large-scale programmatic food aidinflows. Acute production declines, often related to poor weather (floods and droughts)have led to massive emergency food aid efforts.

In India and Bangladesh, government investments and technical change have resultedin large gains in overall production and ultimately a reduction in large-scale food aidinflows. In contrast, production growth in Ethiopia and Zambia has been at best sporadicand availability of food continues to be a major constraint on food security, particularly indrought years (Fig. 2).

For India, food aid was a major source of total food availability from the mid-1950s tothe late 1960s, but a desire to avoid the uncertainties and political conditions tied to foodaid flows led the country to adopt a food self-sufficiency policy in the mid-1960s. Public

India: Cereal production, 1950-2003

0

50

100

150

200

250

1951

1955

1959

1963

1967

1971

1975

1979

1983

1987

1991

1995

1999

2003

mn

to

ns

Rice Wheat Coarse cereals

Source: FAOSTAT

Bangladesh: Foodgrain Production and Food Gap

Boro

Aus

Aman

Wheat

Food Gap

0

5

10

15

20

25

30

1980

/81

1981

/82

1982

/83

1983

/84

1984

/85

1985

/86

1986

/87

1987

/88

1988

/89

1989

/90

1990

/91

1991

/92

1992

/93

1993

/94

1994

/95

1995

/96

1996

/97

1997

/98

1998

/99

1999

/00

2000

/01

2001

/02

2002

/03

Year

(mill

ion

to

ns)

Source: Dorosh, del Ninno and Shahabuddin (2004)

Ethiopia: Cereal production, 1980-2003

0

2

4

6

8

10

12

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

(mn

to

ns)

Wheat Maize Sorghum Other Grains

Source: FAOSTAT

Zambia: Cereal and Cassava Production, 1980-2004

0

500

1000

1500

2000

2500

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

Th

ou

san

d M

etri

c T

on

s

Maize Other Cereals Cassava

Note: Cassava production expressed in grain equivalents, calculated as weight of root times 0.25.Source: FAOSTAT

Fig. 2. Trend in cereal production in India, Bangladesh, Ethiopia and Zambia.


investments in irrigation, agricultural research and extension, and subsidies to fertilizerand agricultural credit led to a rapid spread in green revolution technology. The govern-ment intervened heavily in rice and wheat markets by restricting private sector purchases,storage and grain movements, while providing support to producers (particularly in keyproducing states) through procurement at official minimum support prices. Together theseagricultural technology and price policies succeeded in rapidly raising production of riceand wheat, enabling the country to raise availability of grain and reduce food aid andcommercial imports to insignificant shares of total supply (Table 3). By the mid-1990s,excess availability of food (as reflected in rising public stocks) had replaced food supplyshortfalls as a major food policy issue. This new situation called into question the wisdomof continuing large subsidies on agricultural inputs, particularly for wheat and riceproduction.6

6 ‘‘The subsidies have reached a point where their beneficial effects seem to be outweighed by the adverse effectsin terms of macroeconomic imbalances, slowing down of investment in agriculture, inefficiency use of resources atthe farm level and degradation of the environment, and reduction in employment.’’ (Rao, 1994, p. 235). See alsoGulati and Rao (1994) and World Bank (2004).

Table 3Cereal production, imports and food aid – India, 1951–2004

Average quantity (million ton) Trend growth

1951–1967 1968–1990 1991–2004 1952–67 1968–90 1991–2004

Production 60.9 113.2 178.2 2.85 3.01 1.36Rice 29.6 50.6 81.4 3.16 2.88 1.18Wheat 9.5 33.9 65.6 3.59 5.53 2.41Coarse cereals 21.7 28.7 31.3 2.12 0.71 �0.16

Exports 0.3 0.6 2.4 – 1.60 15.17Imports 3.7 2.1 0.7 10.16 �12.59 �5.26

Food aida n.a. 0.7 0.2 n.a. �7.90 �4.36Net supply 56.7 100.5 153.8 3.27 2.69 1.21Net supply/personb 136.6 149.9 159.4 0.79 0.49 �0.43c

Imports/net supply 2.1% 1.6% �1.4% – – –

Source: Indian economic survey (various issues), WFP INTERFAIS.a Food aid: 1970–90 and 1991–2002 averages; most imports of cereals in the 1951–67 were food aid.b Includes changes in public stock and 12.5% deduction from production for feed, seed and wastage.c Growth rate for 1991–2003.


Bangladesh has achieved similar success in raising domestic production and reducingthe role of food aid in domestic food supplies, but with a more market-oriented approachthat has involved lower fiscal costs of government procurement of grains. As in India,expansion in irrigation (particularly private tubewells after relaxation of restrictions onimported pumps in the late 1980s), increased use of fertilizer and improved seeds enabledthe country to reach its rice and wheat production targets by 2000. Increases in winter(boro) season rice production not only added to total supplies, but reduced the seasonalityof production and helped stabilize total supplies and prices, particularly following periodicpoor monsoon season (aman) rice harvests due to excessive floods (Ahmed et al., 2000). Incontrast with India, however, stability in food supplies was provided mainly through foodaid from the mid-1970s to the early 1990s, and since the early 1990s, via private sectorimports (Table 4).

Table 4Cereal production, imports and food aid – Bangladesh, 1981–2005

Average quantity (1000 ton) Trend growth

1981–1990 1991–2000 2001–2005 1981–1990 1991–2000 1991–2005

Production 16,103 20,172 26,559 2.48 2.10 3.02Rice 15,011 18,772 25,156 2.63 1.70 3.04Wheat 1092 1400 1403 0.15 7.93 2.50

Net imports 1838 2079 2550 1.53 7.30 5.51Food aid 1222 929 367 2.19 �3.91 �8.80

Net supplya 16,331 20,234 26,453 2.28 2.78 3.27Net supply/person 164 168 196 0.10 1.06 1.61Imports/net supply 11.1% 9.9% 9.6% �0.74 4.40 2.17

Source: Authors’ calculations. Food planning and monitoring unit, Bangladesh Ministry of Food.a Includes 10% deduction from production for feed, seed and wastage, but does not include changes in gov-

ernment stocks.


Ethiopia’s gains in food production, unfortunately, have been much more modest thanthose of India or Bangladesh. A much drier climate, with relatively far fewer opportunitiesfor irrigation and multiple-cropping, greatly reduces the potential for the massive gains infood production achieved in South Asia. Economic reforms and increased adoption ofimproved seeds and fertilizer have enabled grain production to keep pace with a rapid2.7% population growth rate since 1994, but only after a major decline in per capita grainproduction over the previous two decades. Because high transport and marketing costs,and restrictive trade policies have prevented private sector imports from making a signif-icant contribution to supplies, food aid continues to be a major source of grain supply inEthiopia, even during normal years. And in years with major droughts, the country isheavily dependent on food aid to provide emergency relief (Table 5).

In Zambia, food production has actually declined substantially since 1990, due largelyto the withdrawal of government subsidies on inputs and an end to pricing policies thathad favored maize production at the expense of other agricultural commodities (Table6; Jansen, 1988; Jayne and Jones, 1997). As in Ethiopia, high transport and marketingcosts in Zambia raise prices of imported grain, though the Republic of South Africa isa relatively close source of potential import supply in most years. Private imports havealso been inhibited by substantial uncertainty regarding government policy related to vol-umes, quantities and sales prices of government commercial imports, and domestic pricesrose substantially above import parity levels during the 2001–02 drought. An effectiveearly warning system (FEWSNET) has helped planning government commercial importsand food aid. Nonetheless, food availability has declined substantially during droughtyears (Table 6).

Disincentive effects of food aid on domestic production

Food aid is often criticized for its potential disincentive effects on domestic produc-tion because it may lead to reduced market prices for producers and because it may also

Table 5Cereal production, imports and food aid – Ethiopia, 1980–2003

Average quantity (1000 ton) Coefficient of variation

1980–1989 1990–1994 1995–1999 2000–2003 1980–1995 1996–2003

Production 5628 5505 8234 8915 0.11 0.12Wheat 725 816 1126 1427 0.13 0.15Maize 1435 1569 2663 2987 0.22 0.15Sorghum 1096 713 1481 1474 0.35 0.22Other 2373 2407 2963 3027 0.12 0.09Imports 504 823 506 1005 0.51 0.47Food aida 517 816 531 910 0.45 0.63Net Supplyb 5569 5777 7917 9029 0.10 0.12Net supply/capita (kg) 128.3 106.9 130.6 134.0 0.16 0.10Net imports/supply 9.1% 14.2% 6.6% 11.2% 0.49 0.44

Source: FAOSTAT, WFP and authors’ calculations.a Food aid, imports and net supply figures are through 2002.b Includes 10% deduction from production for feed, seed and wastage, but does not include changes in gov-

ernment stocks.

Table 6Maize production, net imports and food aid – Zambia, 1980–2003

Average quantity (1000 ton) Trend growth

1980–1993 1994–2003 1980–2003 1980–1993 1994–2003 1980–2003

Production 1231 1044 1153 2.70 �2.27 �0.49Net imports 243 175 214 �2.75 6.28 �2.60Food aid 124 30 85 �8.87 25.50 �8.78Net supplya 1351 1115 1252 2.61 �0.94 �0.68Net supply/person 183 111 153 �0.51 �2.80 �3.28Including cassava 198 132 171 �0.10 �2.08 �2.60Net imps/net supply 18.7% 15.6% 17.4% �5.22 7.30 �1.94

Source: FAOSTAT, WFP and authors’ calculations.a Includes 10% deduction from production for feed, seed and wastage, but does not include changes in gov-

ernment stocks.


reduce government incentives for production-enhancing investments and policy reforms.The experience of the countries discussed here, however, show that these adverse effectsare not inevitable. The strongest evidence for the complementarity of food aid-supportedprograms without producer disincentives is the case of Bangladesh, where sustained foodaid flows of wheat over 25 years coincided with doubling of rice production and an evenmore rapid gain in wheat production. From 1979/80 to 1999/2000, Bangladesh morethan doubled its wheat production (from 829,000 ton to 1.84 million ton in 1999/2000), in spite of large food aid wheat flows that averaged 1.03 million ton per year.India also avoided adverse effects of food aid after the late 1960s through a determinedeffort to achieve self-sufficiency and eliminate major food aid inflows. In contrast, inEthiopia, availability of food aid may have reduced government incentives to invest,and late arrivals of food aid in 2001 contributed to steep reductions in producer prices.7

Cross-country econometric analysis that includes Ethiopia, however, finds no evidence ofdisincentive effects on output (Abdulai et al., 2005). Zambia’s food aid imports havebeen large relative to domestic production only during drought years. Apart from somelocalized disincentive effects of distribution of late arrivals of food aid in the early 1990s,the low levels of food aid inflows in most years strongly suggest that food aid was not amajor cause of agricultural stagnation or the government’s insufficient investment inagriculture.

Three factors were key in enabling India and Bangladesh to achieve these productiongains while receiving and ultimately reducing food aid inflows. First, Bangladesh andIndia maintained a political will and had donor support for long-term investments inproduction, including agricultural research, extension, irrigation and rural roads.Achieving similar production gains in sub-Saharan Africa will require similar publicinvestments in research and extension, though opportunities for irrigation and multi-ple-cropping are significantly less than in South Asia. Second, food aid flows were smallrelative to the size of total consumption, even in Bangladesh where food aid flows aver-aged about 600,000 ton per year in the late 1990s. In countries, where the size of food

7 Nonetheless, there is no evidence that food aid caused a greater disincentive effect on production for farmersin villages receiving food aid than in regions not receiving food aid (Abdulai et al., 2005).


aid shipments is large relative to the size of the markets, and especially where the foodaid commodity is a close substitute for major domestically produced staples, the risk ofadverse price effects on production incentives are especially great. Third, food aidinflows were channeled through a public distribution system, with adequate public stor-age and careful management of the timing of arrivals of food aid and the distribution offood. Food aid distributed mainly through emergency relief programs in Ethiopia hasbeen less effectively managed, though apart from a market price crash in 2001, thereis no evidence that food aid flows have had major price disincentive effects at a nationallevel since the early 1990s.

Markets and price stability

In both India and Bangladesh, domestic procurement, public distribution and govern-ment stocks have played a central role, not only in price stabilization, but in provision offood for chronic poverty alleviation and maintenance of stocks for emergency relief. In theabsence of budget constraints, the government can both procure unlimited quantities offood in local or international markets and, if necessary, subsidize the sales or distributionof this food to consumers. However, budget constraints (and in the short-run, storage con-straints), often limit government interventions.

In India in the mid-1990s, the government’s willingness to procure unlimited quanti-ties of rice and wheat at high support prices (though mainly in only a few states) led tosharp increases in quantities procured, a massive build-up of stocks and huge fiscal costswhen the grain was ultimately distributed at subsidized prices (World Bank, 2004). InBangladesh, tighter budget constraints and a willingness to give a greater role to the pri-vate sector kept the size of government’s interventions (and their impacts on marketprices) relatively small. Thus, procurement decisions were largely driven on quantity tar-gets for public distribution and stocks, and the procurement price is not meant to func-tion as a floor price.

Government procurement of key food commodities (e.g. wheat and maize) to boostproducer prices (and perhaps guarantee a price floor) could help maintain producer incen-tives in Ethiopia, as well. Government procured grain could be used for relief purposesand consumption smoothing, replacing imported food aid, particularly in the food deficitregions in northern Ethiopia. In other regions, where market supplies of food are moreplentiful, the government is now beginning to experiment with safety nets that rely moreon income transfers rather than food transfers (see the discussion of the productive safetynet programme (PSNP), below.) Wholesale open market operations to augment supply inhigh-priced markets are another option. To the extent that cash-based safety nets are inplace, and marketing reforms and investments can bring transactions costs and importparity prices down, private sector imports could help stabilize markets in years of majorproduction shortfalls and prevent large price spikes (particularly in years of low or med-ium world prices).

Zambia’s negative experience with large-scale intervention in domestic food marketsin the 1980s illustrates potential pitfalls with this strategy, however. Government inter-vention to a large extent crowded out private trade and led to high marketing costs.Political pressure for low consumer prices led to huge budget subsidies. Agriculturaltrade liberalization in the early 1990s has helped reduce marketing costs and eliminated

Table 7Monthly prices and variation of major staples in selected countries, 1996–2002 (Dollars/ton)

India Bangladesh Zambia Ethiopia

Delhi National ave. Lusaka Addis

Wholesale coarse rice Wholesale coarse rice Retail white maize Retail maize

Average price 218.7 240.9 191.8 127.7Standard deviation 23.0 32.0 59.4 39.6Coefficient variation 0.105 0.133 0.310 0.310Maximum 266.0 307.1 352.1 225.7Minimum 184.0 193.5 100.9 55.7Max/min 1.45 1.59 3.49 4.05Max/mean 1.22 1.27 1.84 1.77

Source: Authors’ calculations.


consumer subsidies, though after input and output subsidies for maize production wereremoved, output declined sharply (Jayne and Jones, 1997). Less costly and less trade-dis-torting means of government interventions to support domestic producers, includinginvestments in market infrastructure and increased provision of agricultural extensionservices, may have been more efficient alternatives than government direct marketinterventions.

As indicated in Table 7, both price levels and price stability vary markedly in thestudy countries, reflecting the combined effects of differing market structure, externalshocks and policy responses. Large scale food market interventions and long termincreases in production have resulted in long-term declines in the price of food in Indiato below import parity levels in most recent years. Moreover, prices of rice (expressed inUS dollars) in India are extremely stable. Bangladesh rice prices average about 10%higher than in India (in dollar terms at market exchange rates), and exports from Indiahave helped to stabilize rice prices at the import parity level in Bangladesh in years ofmajor production shortfalls.

Staple food prices in Zambia (white maize) have averaged about 10% lower than riceprices in India, though these prices are nearly 50% higher than average maize prices inEthiopia over this period. These relatively high prices of maize in Zambia in part reflectsupply shortages following major droughts.8 Moreover, food prices are much more vari-able in Zambia and Ethiopia than in India and Bangladesh, with coefficients of variationsroughly 3 times greater in the two sub-Saharan African countries than in Bangladesh andIndia (Table 7). Ethiopia has had substantial negative price swings in this period, reflectingthe production surges and market gluts in 2001–02. In Zambia, price spikes have been themajor source of instability: the maximum price of maize was over four times the minimumprice and the ratio of the maximum price to the mean price was 1.84, slightly higher thanin Ethiopia (1.77) and considerably above Bangladesh (1.27) and India (1.22). Thus, priceinstability has continued to plague Zambia and Ethiopia, in spite of the sizeable food aidflows to these countries.

8 Price levels measured in foreign currency (as opposed to Zambian kwacha) are, of course, also influenced bynominal exchange rates.

India: Calories from Cereals and Other Sources. 1980-2001

0

250

500

750

1.000

1.250

1.500

1.750

2.000

2.250

2.500

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

Year

Cal

ori

es/P

c/D

ay

50

55

60

65

70

75

80

85

90

95

100

Per

cent

Rice Wheat O. Cereal Rest Percent from Cereal


Bangladesh: Calories from Cereals and Other Sources. 1980-2001

0

250

500

750

1.000

1.250

1.500

1.750

2.000

2.250

2.500

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

Year

Cal

ori

es/P

c/D

ay

50

55

60

65

70

75

80

85

90

95

100

Per

cent

Rice Wheat O. Cereal Rest Perc Cereal


Ethiopia: Calories from Cereals and Other Sources. 1980-2001

0

250

500

750

1.000

1.250

1.500

1.750

2.000

2.250

2.500

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

Year

Cal

orie

s/P

c/D

ay

50

55

60

65

70

75

80

85

90

95

100

Maize Sorghum Wheat O. Cereal Rest Percent from Cereal


Zambia: Calories from Cereals and Other Sources. 1980-2001

0

250

500

750

1.000

1.250

1.500

1.750

2.000

2.250

2.500

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

Year

Cal

orie

s/P

c/D

ay

50

55

60

65

70

75

80

85

90

95

100

Per

cent

Maize Wheat O. Cereal Cassava Rest Percent from Cereal


Fig. 3. Calories from cereals and other sources in India, Bangladesh, Ethiopia and Zambia, 1980–2001.


Food consumption and safety nets

Overall trends in calorie consumption in the study countries have been largely deter-mined by trends in cereal and other food production (Fig. 3). India and Bangladesh expe-rienced a steady increase in caloric consumption since 1980, with a declining share ofcereals. In Ethiopia, consumption of cereals increased in the 1990’s but consumption ofnon-cereals stagnated, leading to an increase in total calories and an increased share of cal-ories. In Zambia, by contrast, total caloric consumption and share of cereals have shown asteady downward trend. However, these measures of calorie consumption essentiallyreflect calorie availability and mask significant differences in household consumption pat-terns between rural and urban areas, across geographical regions, and by income (See delNinno et al. (2005) for a discussion of variations of calorie consumption across householdgroups in the study countries.).

To enhance consumption of vulnerable groups, each of the study countries has imple-mented large-scale safety nets involving food distribution at various times.9 In India, thecornerstone of the safety nets system is the public distribution system (PDS). Even after

9 For an analysis of the role of safety nets in very low income countries for poverty reduction and food security,see Subbarao and Smith (2003b).


reforms in 1995 designed to better target clearly identified households, the PDS still cov-ers about 600 million people at a cost of 30 billion rupees ($700 million). Even with thissystem, though, an estimated 200–300 million people, 20–30% of the population, con-sume inadequate amounts of calories. In addition to the PDS, India has a long historyof implementing public works (PW) and other safety net programs, some financedentirely by the GOI, some partially with cost-sharing by states. Since 2001, governmentfoodstocks have been used for the Sampoor Grameen Rozgar Yojana (SGRY) employ-ment program10 that is implemented by state governments, which received an allocationof 5 million ton each. Other programs supported by food grain stocks include mid-daymeals, nutrition program, food for work, and Antyodaya (destitute feeding program).11

Nonetheless, the persistence of food insecurity for millions of households despite theavailability of food grain stocks exceeding 50 million ton has renewed much debate onthe role of ‘‘Right to Food’’ as a human right (Dev, 2003; and Gaiha, 2003). As a result,a new law, the Employment Guarantee Act, has been passed to provide guaranteedemployment for 100 days at minimum wage for persons below the poverty line; imple-mentation modalities are being worked out.

Like India, Bangladesh has a large-scale public foodgrain distribution system (PFDS).In contrast to India, however, Bangladesh greatly reduced the role of subsidized rationsales by eliminating its major rationing channels in the early 1990s, and replacing theseprograms with targeted programs (food for work, food-for-education and programs tar-geted to poor rural women) (Ahmed et al., 2000). Since 2000, food aid (which accountedfor about one-third of grain distributed through the PFDS in the mid-1990s) has declined,and Bangladesh has further reduced the scale of the PFDS, particularly for food for workprograms. This shift away from cash sales and food for work has improved the overalleffectiveness of targeting, since small transfers targeted towards rural women are moreeffective in raising food consumption (as in the Bangladesh Vulnerable Group Develop-ment program) than are food for work programs (which involve relatively large paymentsin-kind) (del Ninno and Dorosh, 2003).

In Ethiopia, although according to official policy, about 80% of food aid is to be dis-bursed through food for work programs12 (i.e. employment generation programs, EGS,and food for work, FFW). In reality, a large proportion of this food aid has been dis-tributed in the form of free transfers, because of administrative constraints, lack of non-wage (cash) components for materials and, in emergency relief situations, by the poorhealth of recipients (Subbarao and Smith, 2003a).13 The remaining 20% of food aidhas gone through free distribution programs14 (gratuitous relief, GR, and school feedingprograms, SFP). Some evidence suggests that food aid programs have been effective in

10 The SGRY replaced two other employment programs, namely the Employment Assurance Scheme and theJawahar Gram Samridhi Yojana (Jawahar Village Development Program).11 Few rigorous evaluations of these various programs are available (with the exception of PDS), though several

ad hoc evaluations did point out to several deficiencies of implementation of work fare programs. More researchneeds to be done to assess the cost and targeting efficiency of food-based and cash-based programs currently inoperation in various states (Radhakrishna et al., 2004).12 Ethiopia’s official food aid policy states that no able-bodied person should receive food aid without working

on a community project in return.13 Because much of the food for work grain is distributed without any work actually being done, the public

investment benefits of this program is greatly reduced.14 Free distribution programs distribute cereals like wheat, maize and sorghum.


reducing household vulnerability in Ethiopia (Dercon and Krishnan, 2004 and Quisum-bing, 2003), that it has had a positive impact on child nutritional status (Quisumbing,2003 and Yamano et al., 2005), and that it has been allocated mostly to the pooresthouseholds.15

There has been an increasing concern in Ethiopia, however, that ‘‘national policy-mak-ers and the international community have adopted a ‘crisis management’ strategy that hasresulted in dependence on food aid, rather than addressing the underlying problems ofpoverty and livelihood insecurity’’ (Devereux et al., 2004). To address this concern, amajor change in the government’s approach to safety nets introduced in 2005, the newproductive safety net programme (PSNP), involves greater use of cash transfers and alonger term program focus. The program is designed to enhance longer term food securityof 5 million chronically food-insecure people through cash and food transfers. Emergencyfood assistance programs will also continue, but at a lower level – only 2.2 million peoplein 2005 (compared with six to seven million people in previous recent years).16

Zambia had large and untargeted urban maize flour subsidies in the 1980s that contrib-uted to large fiscal deficits (Jansen, 1988). Maize flour subsidies for urban households wereeliminated in the early 1990s, but these programs were not replaced with alternative large-scale safety net programs. Following several drought-related production shortfalls, thecountry has relied on food aid inflows distributed through public works and transfer pro-grams (Jayne and Jones, 1997; Jayne et al., 2005).

In sum, well-designed safety nets can effectively target transfers to poor households,especially after a crisis, help enhance food security, and also build human and physicalcapital. For example, food-for work (FFW) programs in both India and Bangladesh,including the employment guarantee in the state of Maharashtra, India, have not onlyincreased incomes and access to food of poor households, but have contributed to buildinginfrastructure. However, specific design details such as the level of wage rates, degree oflabor intensity of projects, provisions for non-labor costs and gender issues are crucialto the effectiveness of these programs. In contrast, ration programs are difficult to operateefficiently especially since when there is a substantial difference between ration sales pricesand open market prices, leakages can be substantial, as evidenced by Bangladesh in thelate 1980s and many other countries (including Mozambique in the early 1990s17).

However, even in India where a massive public food distribution system is in place, eco-nomic growth (particularly growth in real wages and non-agricultural output), rather thanincreases in public transfers, were the major factors in raising real incomes of the poorabove a minimum food consumption-based poverty line (Datt and Ravallion, 2002). ThisIndia experience, as well as a lack of counter-examples where safety net programs alonewere sufficient to provide food security for the poor on a sustained basis, suggest that longterm improvements in food security require sustained increases in real incomes of thepoor.

15 Gilligan and Hoddinott (2004) provide evidence that GR resources were better targeted to the poor than EGSprogram, which were allocated largely to people able to work. Levinsohn and McMillan (2005) also found thathouseholds at all levels of income benefit from food aid and that – somewhat surprisingly – the benefits godisproportionately to the poorest households. However, questions have been raised about the effectiveness ofgeographical targeting of the food aid (Jayne et al., 2001).16 FAO/WFP (2005) and various government project papers.17 Dorosh et al. (1995).


Government response to crises

Government response to major production shortfalls has varied both in instrumentschosen and in the overall effectiveness of the interventions. In addition to or in place offood aid, the case study countries have utilized three other main approaches to addressshortages in availability: distribution from large-scale national food stocks, public sectorcommercial imports, and promotion of private sector imports. To address the problem ofaccess to food by the most vulnerable, governments of the case study countries have reliedlargely on public employment schemes and food transfers.

India’s negative experience in the mid-1960s with uncertain food aid flows thatdepended on donor political considerations was instrumental in creating the political willto support policies to increase domestic production and achieve national food self-suffi-ciency through investments in irrigation, agricultural research and extension. Having suc-cessfully increased domestic production of rice and wheat, by the mid-1970s India was ableto rely mainly on drawdown of accumulated stocks for addressing production shortfalls,coupled with self-targeting emergency relief programs (mainly food-for work), even fol-lowing the 1987 ‘‘drought of the century’’ that contributed to grain production declinesrelative to 1985/86 of 5.4 and 7.7 million ton in 1986/87 and 1987/88 (3.9% and 5.6%,respectively).

Bangladesh suffered a major famine in 1975 following flood-induced production short-falls in late 1974, as the country lacked foreign exchange for imports and governmentstocks for targeted distribution. Like India, it subsequently undertook major investmentsin irrigation (including major private sector investments in tubewells), agricultural tech-nology and roads, and succeeded in substantially raising rice and wheat production. Par-ticularly important was the increase in winter season crops (boro rice and wheat) thatprovided a second harvest following monsoon season production shortfalls, as experienceddue to major floods in 1988. Large public stocks, government commercial imports andfood aid inflows helped stabilize market prices and permitted large-scale public distribu-tion of grains, as well. Ten years later, following the 1998 floods, however, private sectorimports of rice and wheat, made possible by the trade liberalization of the early 1990s,played the major role in stabilizing grain prices, even without large-scale public distribu-tion of grain or food aid flows (Dorosh, 2001; del Ninno et al., 2001; Dorosh et al., 2004).

In addition, the government was very effective in quickly scaling up existing programsto flood-affected households through cash transfers via the gratuitous relief (GR) and inkind transfers to poor women via the vulnerable group feeding (VGF) (del Ninno et al.,2001). In the initial flood period, immediate relief through the gratuitous relief programwent mainly to seriously flood-exposed households. Vulnerable group feeding transfersstarted later and were better targeted to the poor than to the flood exposed households.38.8% of the households in the bottom quintile received grain transfers compared to17.2% and 11.2% in the top two quintiles. In all, total direct transfers (mainly throughthe VGF program) added an estimated 20–25 calories/day to consumption of recipients(del Ninno et al., 2003). Yet, government transfers were small relative to the needs ofhouseholds, as indicated by the share of the transfers compared to the monthly expendi-ture (del Ninno and Dorosh, 2001).

Ethiopia has been far less successful in maintaining food security following majordrought-induced production shortfalls. Major droughts in Ethiopia occurred in 1971–75, 1984–85, 1999–00 and 2002–03, but nearly every year since 1980 at least some parts


of the country have been affected by drought.18 The 1984–85 drought had particularly tra-gic results, as a 30% decline in cereal production (relative to the average of the three pre-vious years), combined with the reluctance of the government of that time to accept foodaid and slowness of donor response. Both nominal and real prices of cereals increasedsharply, and at least one million people are estimated to have died (Von Braun et al.,1998).

In the 1990s, Ethiopia’s new government and the donor community implemented a ser-ies of reforms to minimize the impact of possible droughts the occurrence of another wide-spread famine. These included the adoption of the National Policy on Disaster Preventionand Management (NPDPM) in 1993 which was complemented by the early warning sys-tem (EWS), the Emergency food security reserve (EFSR) and the national disaster preven-tion and preparedness fund (NDPPF).19

Although the system improved coordination of relief efforts, a major food crisis againoccurred in 1999–2000 as a result of a series of repeated weather shocks, lack of marketaccess for poor households in many regions, and the negative impact of restrictions ontransport of grain due to the conflict with Eritrea.20 Estimates of the number of peoplein need of food rose from about 2 million in early 1999 to 10 million in July 2000, beforedeclining to 3 million in December 2000. The situation was exacerbated by the fact that thegovernment was unable to intervene quickly because stock reserves had not been replen-ished. Eventually, by late 2000 interventions by the government and the international com-munity brought the situation under control, but the late arrival and distribution of foodaid well into the following season reportedly contributed to a sharp drop in producerprices (Hammond and Maxwell, 2003).

A third major crisis occurred in 2002–03, following yet another drought that led to asteep decline in maize and teff production, substantial loss of livestock, and rapid foodprice increases. Though this time, the food aid response appears to have been adequate,the non-food assistance was not, reflecting a weakness in the early warning systems thatremains largely focused on food production (Lautze et al., 2003).

Like Ethiopia, Zambia suffered two major droughts in recent years: in 1991–92 and2001–02 (as well as a more localized drought in 1994–95).21 In an effort to protect foodsecurity during these droughts, Zambian governments attempted to increase food suppliesthrough a combination of government commercial imports, food aid and private sectorimports (and bans on exports). Food and cash transfer programs were implemented aswell, in an effort to increase access to food by food-insecure households. These policiesappear to have been relatively more successful in maintaining availability of food in1991–92 than in 2001–02, in spite of improved early warning capacity.

18 Most droughts have been concentrated geographically in the central and northeastern highlands and in thesouthern lowlands.19 While the system has been very successful in avoiding a major crises, it is considered to be still weak in

pastoral areas and in areas either not traditionally prone to famine or those historically marginalized (Lautzeet al., 2003). Some report that the previous years estimates are a major determinant of current year’s figures (Clayet al., 1999; Lautze et al., 2003).20 Between 1998 and 2000 the war with Eritrea killed tens of thousands of people, forced about 350,000 people to

leave their homes, damaged the country’s infrastructures and encouraged many foreign donors to freeze foreignaid to the country.21 In 1994–95, Zambia suffered a moderate drought that mainly affected the Southern and Western provinces.

National cereal production fell by about 300,000 ton (about 30% of 1994 cereal production of 1,190,000 ton).


Because of the 1991–92 drought, maize production in Zambia was only 600,000 ton,half that of the previous year, and less than one-third of the record harvests in 1998and 1999. A massive and timely international response averted famine, prevented off-farmmigration, and facilitated recovery (Callihan et al., 1994). Food aid grain shipments toZambia of 500,000 ton were supplemented by 700,000 ton of commercial imports, so that,ultimately, cereals availability per capita in 1991/92 (211 kg/person) was greater than inthe previous two years, and only slightly less than the average for 1980–93 (216 kg/person).

The 2001–02 crisis, caused by poor rains over two consecutive crop years (2000–01 and2001–02) was more widespread, causing severe production and supply shortfalls in Zambiaand elsewhere in southern Africa (World Bank, 2002). A combination of food aid(176,000 ton of food aid in 2000–01 to 2002–03 period) and commercial imports in2001–02 (367,000 ton of commercial imports) offset part of the production shortfall. Deliv-ery of this food aid and planning for imports to the southern African region in advance ofdepletion of local stocks were facilitated by the early warning system (FEWSNET;USAID, 2003).22 Nonetheless, these interventions were not of sufficient size or timelinessto prevent a sharp increase in maize prices (by up to five times the five-year average) andmarket supply shortfalls, particularly in some southern provinces (WFP, 2002). As dis-cussed above, lack of transparency of government policy that inhibited private sectorimports was a major factor in discouraging private sector imports and thus reducingavailability.

Food aid made an important contribution to household food security in Zambia in2002 when it provided about 5% of the annual caloric requirements in most regions andup to 10–15% of total annual caloric requirements in the Southern region (ZambiaVAC, 2003). Following a smaller, regional drought in 1994–95, substantial sales of foodaid (mainly in small towns along rail lines) also took place, but the quantities were oftentoo large, arrived too late, and were poorly targeted Devereux (2000). The longer termsimpact on public works programs and transfers remains open to question, however.

Conclusions and policy implications

Critics of food aid often point out the numerous implementation problems experienced,including disincentive effects of production, disruption of markets and poor targeting. Ouranalysis, while not ignoring these potential negative aspects of food aid, finds that experi-ences with food aid in India and Bangladesh have been generally positive, though devel-opment and food security outcomes have been less encouraging in Ethiopia and Zambia.We find that the recipient country government policies, particularly those related to agri-cultural production, food markets, safety nets and emergency response, are generally moreimportant determinants of food security outcomes than are food aid policies, in both theshort and long run.

Policies in the four countries considered in this study differ most markedly in termsof production programs and policies (Table 8). Both India and Bangladesh investedheavily in production of major staples and in rural infrastructure and have reaped

22 Strategic reserves were very small and did not play a major role in the response to the crisis (Tschirley et al.,2004).

Table 8Summary of food security and food aid policies in study countries

Production/supplypolicies/programs

Production/supply role offood aid

Markets and tradepolicies/programs

Consumption policies/programs

Consumption role of foodaid

India Highly successful Minor role Highly restrictive policies Partially successful Minor role

Massive publicinvestments in R&D andirrigation achieve grainself-sufficiency in 12 years

Food aid a minor sourceof supply after 1970. Foodfor work used to buildroad and irrigationinfrastructure

Until 1990 GOI hadsevere trade restrictions inplace, includingmovement restrictions,private stock limits and aban on private imports

Large-scale PublicDistribution System withration sales, FFW;Targeting introduced1995 due to increasingcosts of PDS

Well-targeted FFWprograms, but smallcoverage

Bangladesh Highly successful Significant role Liberalized policies Partially successful Significant role

Large public investmentsin R&D; privateinvestments in tubewellsand input trade reformlead to rice self-sufficiencyin 2000

Large food aid inflowsuntil 2001 helped supplyand provided financialsupport. Food for workused to build roadinfrastructure

Relative smallgovernment interventions;major domestic andinternational tradeliberalization in early1990s

Major ration salesprograms replaced bywell-targeted food-basedsafety nets in early 1990s

FFW and various well-targeted, food transferprograms, including Foodfor Education (from 1994until 2002)

Ethiopia Not successful Major role Highly restrictive policies Partially successful Major role

Little investment in R&D;few opportunities forirrigation; modestincrease in productionsince 1994 (mainly maize)

Food aid flows are amajor source of supplyespecially in droughtyears; little or noproductive use of food aid

Several restrictionstowards private sectortrade

Food aid-supportedprograms saved lives inemergencies

Large, untargetedprograms, mainly throughfree food transfers

Zambia Not successful Significant role Highly restrictive policies Partially successful Significant role

Steep decline in maizeproduction after 1990 dueto reduced market priceincentives, problems withinput supply and droughts

Food aid flows significantonly in drought years;productive use of food aidnot a major source ofinvestment

Liberalized trade regimesince early 1990s;government marketinterventions inhibitprivate international trade

Large urban maize floursubsidies eliminated inearly 1990s; very limitedprograms remain.

Food aid transfers animportant, thoughlimited, safety net indrought years

C.

del

Nin

no

eta

l./

Fo

od

Po

licy3

2(

20

07

)4

13

–4

35

431


major gains in production; with relatively little investment and technical change, foodproduction in Ethiopia and Zambia stagnated over much of the period. Differencesin agro-climatic conditions and levels of public investment (especially regarding irriga-tion and agricultural research and extension) are the major proximate causes of thisagricultural stagnation. To what extent food aid adversely affects these government pol-icies remains an open question, though at least in the case of Zambia, the relatively lowvolumes of food aid in most years would suggest that its influence on policy has beenminimal.

Only Bangladesh has fully liberalized external trade in major grains, a policy which hasenabled private imports to stabilize domestic markets after major production shortfalls.State interventions in markets through large scale procurement and ration shop sales haveproved excessively costly and were eliminated or greatly curtailed in the early 1990s inZambia and Bangladesh; India introduced a targeting system in the mid-1990s to reduceits fiscal costs, but has retained the PDS ration sales, nonetheless. Other safety net pro-grams, including both food for work and cash-based programs, have been implementedin all these countries, though in no case has chronic poverty and food insecurity been ade-quately addressed.

Thus, the experience from India and Bangladesh in South Asia suggests that food aid,properly utilized, can contribute to increased food security in both the short- and long-term, provided it is used effectively as part of an integrated program of development.Increasing production and raising rural incomes require sustained government publicinvestments in agricultural technology and extension, irrigation (where feasible), and mar-ket infrastructure. The evidence also suggests that while government interventions in mar-kets can be helpful for maintaining adequate price incentives for producers and protectingconsumers from price spikes, facilitating the development of private markets and liberal-izing trade can increase the effectiveness and lower costs of price stabilization. Becauseindividual countries in SSA vary significantly from those in South Asia, these broad policyand program options to guarantee food availability need to be carefully adapted to specificlocal conditions.

It is worth emphasizing that food aid is not the only, or in many cases, the most efficientmeans of addressing food insecurity. The benefits of well-managed and timely food aidinflows in short-term emergency situations where markets are not functioning and house-hold access to food is extremely limited are clear. In the medium to long run, however,market mechanisms are generally more efficient in addressing food availability constraints.Food security at the household level is then mainly constrained by access to food, which isclosely linked to household incomes. Thus, sustained pro-poor growth is needed alongwith appropriate safety nets. None of the study countries have yet fully succeeded in sat-isfactorily solving this problem of access to food by the poor.

The Asian and SSA experiences presented suggest that food aid that supports buildingof production and market enhancing infrastructure, is timed to avoid adverse price effectson producers, and is targeted to food insecure households can play a positive role inenhancing food security. Most important, however, is a balanced, mutually-reinforcingmix of policies and programs that address both the production and marketing constraintsto food availability and that raise the real incomes of the poor and thereby increase theiraccess to food. Food aid can play a supporting role in achieving food security, but shouldbe seen neither as the scapegoat where policies have failed nor the centerpiece for successin the future.


Acknowledgements

The authors thank Harold Alderman, Steven Haggblade and two anonymous reviewersfor their detailed comments and suggestions. Helpful comments were also received fromMargaret Grosh and participants at a July 2005 workshop at the World Bank. Researchsupport by Mahfuzur Rahman and Alessandra Marini is also gratefully acknowledged.We also thank George Simon from WFP for providing data tables from ITERFAIS.Financial support for this work was provided by the Africa Human Development (SocialProtection) department of the World Bank. Nonetheless, the opinions expressed here arethose of the authors and do not necessarily reflect those of the World Bank, its executivedirectors or the countries they represent. The usual disclaimers apply.

References

Abdulai, A., Barrett, C., Hoddinott, J., 2005. Does food aid really have disincentive effects? New evidence fromsub-Saharan Africa. World Development 33 (10), 1689–1704.

Ahmed, Raisuddin, Haggblade, S., Chowdhury, Tawfiq Elahi, 2000. Out of the Shadow of Famine: EvolvingFood Markets and Food Policy in Bangladesh. Johns Hopkins University Press, Baltimore, Maryland.

Barrett, C. 2006. Food Aid’s Intended and Unintended Consequences. Agricultural and Development EconomicsDivision ESA Working Paper No. 06–05. FAO, Rome.

Barrett, C., Maxwell, D., 2005. Food Aid After Fifty Years: Recasting Its Role. Routledge, New York.Callihan, D., Eriksen, J., Herrick, A., 1994. Famine Averted: The United States Government Response to the

1991/92 Southern Africa Drought. Management Systems International, Evaluation Synthesis Report Preparedfor USAID/Bureau for Humanitarian Response. Management Systems International, Washington, DC.

Clay, E., 2003. Responding to change: WFP and the global food aid system. Development Policy Review (5–6),697–709.

Clay, E., Stokke, O. (Eds.), 2000. Food aid and Human Security. EADI. Book series 24. Frank Cass, London.Clay, E., Molla, D., Habtewold, D., 1999. Food aid targeting in Ethiopia: a study of who needs it and who gets it.

Food Policy 24 (4), 391–409.Clay, E., Riley, B., Urey, I., 2005. The Developmental Effectiveness of Food Aid – Does Tying Matter? OECD

DAC publication.Datt, G., Ravallion, M., 2002. Is India’s economic growth leaving the poor behind? Journal of Economic

Perspectives 16 (3), 89–108.del Ninno, C., Dorosh, P., 2001. Averting a food crisis: private imports and public targeted distribution in

Bangladesh after the 1998 flood. Agricultural Economics 25, 337–346.del Ninno, C., Dorosh, P., 2003. Impacts of in-kind transfers on household food consumption: evidence from

targeted food programs in Bangladesh. Journal of Development Studies 40 (1), 48–78.del Ninno, C., Dorosh, P., Smith, L., Roy, D.K., 2001. The 1998 Floods in Bangladesh: Disaster Impacts,

Household Coping Strategies, and Response. International Food Policy Research Report 122. Washington,DC.

del Ninno, C., Dorosh, P., Smith, L., 2003. Public policy, markets and household coping strategies in Bangladesh:avoiding a food security crisis following the 1998 flood. World Development 31 (7), 1221–1238.

del Ninno, C., Dorosh, P., Subbarao, K., 2005. Food Aid and Food Security in the Short and Long Run:Country Experience from Asia and sub-Saharan Africa. Social Protection Discussion Paper No. 0538. WorldBank.

Dercon, S., Krishnan, P., 2004. Food aid and informal insurance. In: Dercon, S. (Ed.), Insurance AgainstPoverty. Oxford University Press, Oxford.

Dev, Mahendra., 2003. Right to Food in India, Center for Economic and Social Studies, Hyderabad.Devereux, S., 2000. Social Highlights summary of Research Report for the Department for International

Development (DFID), ESCOR Report R7017.Devereux S., Sharp, K., Amare, Y., 2004. Destitution in Wollo, Ethiopia. IDS Research Reports – 55.Dorosh, Paul A., 2001. Trade liberalization and national food security: rice trade between Bangladesh and India.

World Development 29 (4), 673–689.


Dorosh, Paul, del Ninno, C., Sahn, David E, 1995. Poverty alleviation in Mozambique: a multi-market analysisof the role of food aid. Agricultural Economics 13 (1995), 89–99.

Dorosh, Paul A., Quazi, Shahabuddin. 2004. Price Stabilization and Food Stock Policy. In: Dorosh et al. (Ed.),The 1998 Floods and Beyond: Towards Comprehensive Food Security in Bangladesh. IFPRI and UPL,Dhaka, Bangladesh.

Dorosh, Paul A., del Ninno, C., Quazi Shahabuddin (Eds.). 2004. The 1998 Floods and Beyond: TowardsComprehensive Food Security in Bangladesh. International Food Policy Research Institute (Washington,DC) and University Press Ltd., Dhaka, Bangladesh.

Dorosh, Paul A., Ahmed, Raisuddin., del Ninno, C., Murshid, K.A.S., 2004. Overview of the BangladeshFoodgrain Economy. In: Dorosh et al. (Ed), The 1998 Floods and Beyond: Towards Comprehensive FoodSecurity in Bangladesh. IFPRI and UPL, Dhaka, Bangladesh.

FAO, 2005. FAOSTAT, several years.Gaiha, R., 2003. Does the Right to Food Matter. Economic and Political Weekly, October, XXXVIII (40).Gilligan, D., Hoddinott, J., 2004. Using Community Targeting to Provide Drought Relief: Evidence From

Ethiopia. International Food Policy Research Institute, Washington, DC.Gulati, A., Rao, C.H.H., 1994. Indian agriculture: emerging perspectives and policy issues. Economic and

Political Weekly 31 (December), A158–A169.Haddad, L., Frankenberger, T., 2003. Integrating Relief and Development to Accelerate Reductions in Food

Insecurity in Shock-Prone Areas, USAID Office of Food for Peace, Occasional Paper No. 2, Washington DC,USA.

Hammond, L., Maxwell, D., 2003. The Ethiopian crisis of 1999–2000: lessons learned, questions unanswered.Disasters 26 (3), 262.

Hoddinot, J., Cohen, M., Bos, S., 2004. Re-Defining the Role of Food Aid. IFPRI, Mimeo.Jansen, D.J., 1988. Trade, Exchange Rate, and Agricultural Pricing Policies in Zambia. World Bank.

Comparative Studies, World Bank, Washington, DC.Jayne, T., Jones, S., 1997. Food Marketing and pricing policy in eastern and southern Africa: A survey. World

Development 25 (9), 1505–1527.Jayne, T., Strauss, J., Yamano, T., Molla, D., 2001. Giving to the poor? Targeting of food aid in rural Ethiopia.

World Development 29 (5), 887–910.Jayne, T.S., Tembo, G., Nijhoff, J., 2005. Experiences of food market reform and price stabilization in eastern

and southern Africa. Food Policy 31 (4), 328–342.Lautze, S., Aklilu, Y., Raven-Roberts, A., Young, H., Kebede, Girma., Leaning, J., 2003. Risk and Vulnerability

in Ethiopia: Learning From The Past, Responding to the Present Preparing for the Future. A report for theUSAID.

Levinsohn, J., McMillan, M., 2005. Does Food Aid Harm The Poor? Household Evidence from Ethiopia. NBERWorking Paper 11048. Cambridge, MA.

Pillai, N., 2000, Food aid for development: a review of the evidence. In: Clay and Stokke (Eds.).Quisumbing, A., 2003. Food aid and child nutrition in rural Ethiopia. World Development 31 (7), 1309–

1324.Radhakrishna, R., Rao, K.H., Subbarao, K., 2004. Reducing Vulnerability and Increasing Opportunity:

Towards more effective use of India’s Safety Net Programs. The World Bank, Mimeo.Rao, C.H.H., 1994. Agricultural Growth, Rural Poverty and Environmental Degradation in India. Oxford

University Press, Delhi.Shaw, D., 2002. Food aid in sub-Saharan Africa: policy lessons for the future. Canadian Journal of Development

Studies XXIII (3), 571–599.Shaw, D., Clay, E., 1994. World Food Aid. James Currey, London.Shaw, D., Clay, E., 1998. Global food hunger and food security after the World Food Summit. Canadian Journal

of Development Studies (39), 55–76.Subbarao, K., Smith, W.J., 2003a. Safety Nets versus Relief Nets: Towards a Medium Term Safety Net Strategy

for Ethiopia. World Bank Mimeo.Subbarao, K., Smith, W.J., 2003b. What Role for Safety Net Transfers in Very Low Income Countries? (Paper).

Social Safety Net Primer Series. World Bank, Washington, DC.Tschirley, D., Nijhoff, J.J., Arlindo, P., Mwinga, B., Weber, M.T., Jayne, T.S., 2004, Anticipating and

Responding to Drought Emergencies in Southern Africa: Lessons from the 2002–2003 Experience, MichiganState University. Prepared for the NEPAD Regional Conference on Successes in African Agriculture, 22–25November 2004, Nairobi, Kenya.


USAID, 2003. Analysis of Food Security, Health and Nutrition in Zambia. Prepared by: USAID/WDCHA/FFPand AFR Bureau Team for USAID Zambia. Mimeo circumstances. In: Webb, P., Von Braun, J., Yohannes,Y. (Eds.), Famine in Ethiopia: Policy Implications of Coping Failure at National and Household Levels,IFPRI Research Report 92.

Von Braun, J., Teklu, T., Webb, P., 1998. Famine in Africa: Causes Responses and Prevention. InternationalFood Policy Research Institute and John Hopkins University Press, Baltimore, Maryland.

Webb, P., Von Braun, J., 1994. Famine and Food Security in Ethiopia, Lessons for Africa. Wiley, London.WFP, 1994. Food Aid for Development. Asia and Pacific Regional Seminar, Beijing, China, March 1994. World

Food Programme, Rome.WFP, 1995. Food Aid for Humanitarian Assistance. In: Proceedings of the UN World Food Programme Africa

Regional Seminar, Addis Ababa, Ethiopia, February 1995. World Food Programme, Rome.WFP, 2002. Facts and Figures Following the Emergency Food Need Assessment Mission in Aug–Sept 2002.

WFP, Rome.WFP INTERFAIS, 2006.World Bank, 2002. Technical Annex on a Proposed Credit in the Amount of SDR 22.8 Million and Grant in the

Amount of SDR 15.2 Million to the Republic of Zambia for the Emergency Drought Recovery Project,Report No. T7559-ZA.

World Bank, 2004. India: Re-energizing the Agricultural Sector to Sustain Growth and Reduce Poverty. OxfordUniversity Press.

Yamano, T., Alderman, H., Christiaensen, L., 2005. Child growth, shocks, and food aid in rural Ethiopia.American Journal of Agricultural Economics 87 (2), 273–288.

Zambia Vulnerability Assessment Committee (VAC), 2003. Livelihood and Vulnerability Assessment. FinalReport, Lusaka.

food aid, domestic policy and food security: contrasting ... · tries. this study reviews the...

Documents