The Starter Pack Program in Malawi

Material Information

The Starter Pack Program in Malawi implications for household food security
Gough, Amy Elizabeth
Publication Date:
Physical Description:
xii, 240 leaves : ill. ; 29 cm.


Subjects / Keywords:
Agricultural Education and Communication thesis, M.S ( lcsh )
Dissertations, Academic -- Agricultural Education and Communication -- UF ( lcsh )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Thesis (M.S.)--University of Florida, 2002.
Includes bibliographical references (leaves 234-239).
General Note:
General Note:
Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.
Statement of Responsibility:
by Amy Elizabeth Gough.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services ( with any additional information they can provide.
Resource Identifier:
002822771 ( ALEPH )
50388958 ( OCLC )
ANV1298 ( NOTIS )

Full Text

I would first like to thank my mom, Mary Bigler, and dad, Paul Gough, for their
understanding, laughter, and energy. Their adaptability has allowed me to strive for what I hope to accomplish. Additionally, Omer Bigler has been my biggest encouragement and inspired me to follow the roads in front of me. Thanks also go to Joanne Gough, who
has reminded me of the value of patience, timing, and companionship, which truly helped me work so hard; and to Kelli F. Chavez for her own ambition and acceptance.
I would also like to thank my committee chair, Dr. Ricky Teig, for his continuous
support and faith, and Dr. Peter Hildebrand for his expertise, optimism, and for the tremendous amount of time he devoted to this project. I also would like to thank Dr.
Christina Gladwin, who allowed this research to be possible, and Dr. Glenn Israel for his assistance.
The most gracious thanks are also extended to all those in the Ministry of
Agriculture in Malawi for their warm welcome and attention to this research. Also, a special thanks goes to all who worked towards this project in Malawi, particularly Edith Kanyenda. Edith's passion for this project has truly inspired this work. The villages of Sesse II and Bwetu provided never-ending inspiration and revealed the realities surrounding this project. Tendai Mataya, David Nkaonja, Calvin Makoko and Titus Mvalo created a foundation in Malawi allowing all things possible. Thanks go also to C. Clark, T.Benson, and those in Malawi who shared their work towards this project.
I would like to thank those who have created a foundation for my learning and

knowing how to help. Finally, a special appreciation goes out to F. Conr for reminding
me that I know where I come from, what I represent, and what I can accomplish.

The Current Problem of Chronic Food Security in Malawi ......... ..... ..... 10
The Need for Research and Considered Methodologies ........................ 18
2 LIT1E1R.ATU RJE R~EVIE3w ....... **. ............. ............. ............... .. ... .. ... .. .... ........ .25
Theoretical Considerations in Data Collection and Analysis................ 25
Theoretical Understanding of the Starter Pack Program in Malawi ..... 37
STJIA T~~EER PACK .............................................45
The Recent Situation of Smallholders in Mvalawi........................... ........45
Th Starter Pack ad VToucher P'rogramxs ..................... ... 63
Anticipating the Impact of the Starter Pack:
Research Goals uand Objectives ............................................................ 95
Pesearchl M ethodology. 7.. ....................................9

Analyzing the Impact on the Starter Pack Using the Linear Program. 138 Measuring the Potential of the Starter Pack Program........................... 147
Results of LDecision Tree Analysis ........................................................ 167
Implications of Starter Pack Distribution Methods ................................ 172
Recommendations for the Starter Pack Program
in the M a1lawvianr 3Cntext.......................................................................... 183
D FOOD INSECURE EPA'S IN MALAWI 1998/1999 (MOAI)................ 199

1-2: Average Monthly Rainfall in Southern Region: Liragnwe.......................6
1-5: I~el1?tJaryr 1I1~LIily I~ainf~ll iii D~I~onela. .. .......... . . . a...... .. . a.... a as a a a a a a a a a a. nsa...... 6
1-6: I'vI~rrch 1II~aily I~ain17all in I'vlponela ~ ~ .a* ...... 6 1-7: Percentage Growth in Agricultural Gross Domestic Product in
I\/lLIlavvi.... a. *.. ...aa..*aaaaa.a**s. a a a as a. ...a...... a.... ........................................ ..*..** . .... 10
3-1: World Prices; Fe1urijar~r 2000............................ ................. ........ 49
3-3: Tobacco Seasonal Calendar ...................... ..................................... ........... 50
3-Li.: Cassava Seasonal Calendar ~ 50
3-7: Initially Designated Inputs and Suppliers for 1999/2000 Starter 3-8: Actual Inputs and Suppliers for 1999/2000 Starter Packs ..................~...67
3-9: Rate of Malawi Kwacha During Starter Pack Distribution

3-12: M~ponela Starter Pack Voucher M'atrx ........... ............ .. ..... .. .... .... .. .... 75
3- 13: M~vzimrb a Starter Pack V1oucher Mvatrix ..... ...... ....*... .. . ... . ...... .. .. .. ...... '76
3-14: Rate of Malawi Kwacha to US Dollar in
3-15: Rate of Malawi Kwacha to US Dollar Prior to
4-1: Subject Selection by Region, Actual Input Received,
&~ H-ousehold HFeald 104
4-2: Characteristics and Location of Interviewed Starter Pack Voucher
4-3: Characteristics and Location of Interviewed Flexi Voucher
Hlouseholds.............*................ ....... ...............................e...a.........e* ** ** * ** s * 1(06
4-4: Characteristics and Location of Interviewed Starter Pack Recipients... 106 4-5: Characteristics and Village Location of Interviewed Households
4-6: Famine Early Warning System (FEWS) Reported Maize Yields In 4-7: Hybrid Maize Yield in Fertilizer Verification Trials ............................. 113
4-8: Hybrid Maize Yield Conversion from Hectares to Acres...................... 114
4-9: Final Crop Yield Information for All Regions in Linear Program........ 114 4-10: Reported and FED Suggested Maize Consumption............................. 116
4-11: Individual Consumption Requirenients................................................ 117
4-12: H~ousehold Maize Requiremrents ........................................................... 118
4-13: Decision to Sell or Redeem Starter Pack Voucher............................... 131
4-14: Decision tn Sell Flexi Vnucher Nnw or Redeem Voucher---------.....132

5-2: Linear program Feasibility Results and Field Determined Wealth
Index Ra k n .............................................14
5-3: Characteristics of Households in Linear Programming Analysis:
5-4: Characteristics of Households in Linear Program Analysis:
B y RIegion ............................................................ ...... ............................ 146
5-5: Characteristics of Households in Linear Program Analysis: 5-6: Household Average Annual cash with varied Starter Pack Inputs:
Central Region (Households with average annual cash < USD 5-7: Average Year-End Cash with Varied Starter Pack Inputs:
Central Region (Households with average annual cash > USD
5-2:0verag Ire n= )Cash........Among...... Houshol Groups... (n-47)..................... 153
5-9: Pere Increases in Cash Among Household Groups (n=47) ............. 153
5-10: Actual Cash Increase with Five years of flexi Voucher (n=47) .......... 155
5-11: Combined household Maize Distribution Before, During, and After 5-12: Combined Household Maize Distribution Before, During, and After
Starter Paclk Voucher Distribution (n=417) ............................................ 158
5-13: Combined Household Maize Distribution Before, During, and After 5-14: Total Six-Year Maize Production of Forty-seven Households With
1Four-Y!er Starter Pack Distribution ........................... 162

5-16: Total Six-Year Maize Production of Forty-Seven Households With
IFour-Yers FElexi V~oucher DIIistribution .................... .... .................... .... 163
5-17: Predicted Six-Year Maize Production with Various
5-18: Average Increase in Total Seven-Year Maize Production with Starter
Pack and Flexi Voucher Inputs: By Household Group (n=47) ........... 165
5-19: Percent Increase in Total Seven-Year Maize Production with Starter
Pack and Flexi Voucher Inputs: By Household Group (n=47) ........... 165
5-20: Household Average Increase in Annual Maize sales with Five Years
o~f Starter Pack IIIistributicon ..........................................................aeea**** 167

ADD Agricultural Development Division
ADMARC Agricultural Development and Marketing Corporation
DFID Department for International Development
EPA Extension Planning Area
FAO Food and Agriculture Organization
FEWS Famine Early Warning System
FSRE Farming Systems Research and Extension
MoAI Ministry of Agriculture and Irrigation
MCP Malawi Congress Party
MK Malawi Kwacha
PBS Net Productivity Enhancing Safety Net
RDP Rural Development Project
SL Sustainable Livelihoods
UDF United Democratic-Front

Abstract of Thesis Presented to the Graduate School of the University of Florida in
Partial Fulfillment of the Requirements for the Degree of Master of Science
Amy Elizabeth Gough
May 2002
Chair: Dr. Ricky Telg
Major Department: Agricultural Education and Communication
Due to increasing population, decreasing soil fertility, and limited fertilizer use,
Malawian smallholders suffer from chronic food insecurity. During both the 1998/1999 and 1999/2000 planting seasons, the Government of Malawi, in cooperation with donor agencies, responded to smallholders desire for assistance with the Starter Pack Program. Aimed at increasing food security during and after the program, the program provided hybrid maize seed, urea, 23:21:0+4s, and groundnuts to 2.86 million farming households.
In 1999, a pilot program distributed vouchers, instead of staffer packs, primarily
redeemable for the same inputs. A limited number, however, were redeemable for either a starter pack or goods from participating retail stores.
It is likely that in future years the starter pack program will have to reduce expenditures, consequently reducing the quantity of inputs, the size of the target

for the greatest increase in food security upon receiving starter pack inputs, and c) what inputs prove most beneficial in increasing both short and long-term food security..
This research explores the potential of the starter pack to increase food security of Malawian smallholders. Data were collected from forty-seven households, disaggregated by head of household, geographic location, and type of inputs received. Predictions of potential increases in discretionary cash and maize production were made based on simulations of households created using an ethnographic linear program.
Potential increases in cash or maize production differ with the type of inputs
received. Discretionary cash increases are greatest when households receive vouchers with the option for either agricultural inputs or goods. The value of cash increases is similar to the value of the inputs, indicating little potential for future production increases. Increases in maize production are greatest when households receive starter packs, an effective tool for increasing national maize production. Although all household types demonstrate similar patterns of change in discretionary cash and maize production, the relative impact of these benefits is greater for impoverished households.
Suggestions for the starter pack program foctds on a reducing the target population to the most impoverished households. This will allow for distribution of a larger quantity of inputs to each household and a more effective distribution process. Future starter pack programs should target households with less than two acres of available land, as they are the most chronically food insecure, If the intention is to increase immediate food security of these land-constrained households, vouchers redeemable for either agricultural or
h-~hl nd eno.tnetero -d t tn here.

Malawi is a small African country located in the southern section of the continent, bordered by Tanzania to the north, Mozambique to the east and southwest, and Zambia to the west. Agriculture provides employment for nearly 90% of all households, accounts for 40% of the GDP, and generates 77% of the revenue from Malawi's exports (Sahn and Arulpragasam, 1991). Typical crops grown for consumption are maize, cassava, groundnuts, and pulses. Other crops include tobacco, tea, sugarcane, cotton, bananas, vegetables, and rice. As of July 2000, the Ministry of Agriculture and Irrigation (MoAI) estimated that 2,786,576 farming families existed in the country. Other livelihood activities include timber, charcoal, fisheries, ganyu (informal) labor, and making and selling bricks. With 87% of Malawi's population residing in rural areas, food security in Malawi is far less than adequate. The Malawi government, as well as donor agencies, have attempted to introduce programs to both improve smallholder production and to increase food security.
In both 1998/1999 and 1999/2000 planting seasons, the Ministry of Agriculture
and Irrigation, in collaboration with numerous international agencies, implemented a plan to distribute "starter packs" to all farming households. These packs contained two kilograms hybrid maize seed, 10 kilograms 23:21:0+4s, five kilograms urea,. and two

of the targeted farmers. The suggestion for the starter pack program was presented in 1998 by Charles Mann, who stated that national food security could be best achieved by distributing hybrid seed and fertilizer to all Malawian farmers. The program began under the assumption that by distributing appropriate modem technologies to smallholders (in the form of inorganic fertilizer, maize seed, and legumes) Malawi' s food insecure households would experience an increase in annual yields, improved soil fertility, and ultimately improved food security. Distribution of starter packs resulted in a total of 2.53 million farming households receiving inputs during 1998/ 1999 and 2.86 million farming households during 1999/2000. The objectives of the starter pack distribution in 1999/2000 were: "a) to assist fill the food gap; b) to promote crop diversification; and c) to promote the concept of soil fertility improvement" (Clark et a., February 2000 p. ii).
The 1999 starter pack program included a pilot project designed to distribute up to 50,000 starter packs through existing private-sector retail outlets. Selected registered households did not receive traditional starter packs, rather vouchers to be redeemed at local retailers. On 49,000 of these vouchers, the words "starter pack voucher" were printed. These were redeemable at local trading center retailers for only starter packs. On the remaining 1,000 vouchers, the words "flexi voucher" were printed, indicating that vouchers could be redeemed at local trading centers for either a starter pack or goods valuing up to Malawi Kwacha (MK) 450.00.'
The primary purpose of this research is to evaluate the impact of both the starter pack and voucher program as a tool of increasing food security and promoting
sustainable lieihos-mn -ua Maawa -mlhle frs In anlvi h

program, this research also compares differences in the potential impact of the three distribution methods utilized for the 1999/2000 starter pack program (starter pack, starter pack voucher, and flexi voucher). Although implementation of the starter pack resulted in 2.86 million households receiving inputs, the verdict is still questionable as to whether
the starter pack helped the poor households reduce food insecurity in both the short and long term.
In order to assess the impact of the starter pack program according to the perspective of smallholder farmers throughout Malawi, household level data was
collected and disaggregated both by head of household and the type of inputs received in association with the starter pack program. The primary goal of household data collection included eliciting complete household level data in order to understand existing livelihood systems and the methods in which starter pack inputs were utilized ultimately to determine the program's impact on food security. Additionally, secondary
data served as baseline data with which to understand the productivity trends of these households, the national context of the starter pack program, cultural, seasonal, and farming trends, and fundamentally the overall livelihood systems by which these households function.
The Country
Within the relatively small nation of Malawi, one fifth of the actual country is
comprised of Lake Malawi, which stretches almost 600 kilometers along the east side of the country. To the west of the lake, plateaus reach between 915 and 1220 meters. The

Malawi experiences a considerably high population and population density.
Despite discrepancies as to the actual population of Malawi, the National Statistics Office (NSO) projected the 1998 mid-year population at 9.8 million, with an annual population growth rate of 1.9%. In 1999, 87% of the population lived in rural areas (Central Intelligence Agency ECIA], 1999). The population density in persons per square kilometer is: 46 persons/km in the northern region, 114 persons/km in the central region, and 144/km in the southern region. Estimates of the number of households living below the poverty level are debatable, with the most recent results compiled by the National Economic Council's (NEC) Integrated Household Survey (IHlS). The IHS estimates the number of people living below the poverty line in Malawi (1997-1998) to be 64.0 percent nationally, and 43.7 percent in urban areas.
Throughout Malawi, agriculture provides employment for nearly 90% of all households, accounts for 40% of the GDP, and generates 77% of the revenue from Malawi's exports (Sahn and Arulpragasam, 1991). The estimated 2,786,576 farming families In Malawi was a figure based on registration for the starter pack distribution, and may in fact be much higher, as registration records appear to have missed a number of farming households.
Agricultural Administrative Boundaries
Agriculturally, the country is divided into eight regions, known as Agricultural Development Division's (ADD), consisting of Karonga, Mzuzu, Kasungu, Lilongwe, Salima, Machinga, Blantyre, and Shire Valley. ADDs are managed by program

divided into smaller sections called Extension Planning Areas (EPA), which are managed by development officers. There are a total of 175 EPAs in the country. Finally, these planning areas are subdivided into sections, which are coordinated by field assistants who are in contact with farmers. There are a total of 2,029 sections in the country (figure 1-1).
Extension Planning Area (EPA) 175
Rural Development Project (RDP) 31
Agriculture Develop ment Division (ADD) 8
I Ministry of Agriculture (MoAI)...
Figure 1-1: Agricultural Divisions in Malawi Climate
Temperatures in Malawi vary depending on altitude. Mountain areas exhibit
cooler temperatures ranging from 14.4 and 17.8 degrees Celsius, while temperatures in some valleys reach up to 37.8 degrees Celsius.
Annual rainfall ranges from 635 to 3050 mm per year. In average years most
parts of the country receive adequate rainfall for farming. The southern region receives a bit more rainfall during the dry season than do the central and northern regions. A sample rainfall curve from the 1999/2000 Lirangwe EPA (in the southern ADD of

months of the rainy season are illustrated utilizing the 1999/2000 rainfalls from the central region of Mponela (figures 1-3 through 1-6).
1999/2000 Rainfall in Southern Region:
E 60 ... ....... -:: :---Rainfall
400 .0 0 a.
Figure 1-2: Average Monthly Rainfall in Lirangwe, 1999/2000
December January
40.00 U, _December 400 :, : {mJanuaryI
Figure 1-3: December Rainfall Rates Figure 1-4: January Rainfall Rates
in Mponela in Mponela
February March
60.00 60.00
40.00 40.00
20.00 200
Figure 1-5: February Rainfall Rates Figure 1-6: March Rainfall Rates
in Mponela in Mponela
Te pas fieveshaervae a vait of enirnena inlene

smallholder farmers. The 1999/2000 season was characterized by a positive marking in the Southern Oscillation Index (SOD. Generally, a positive marking is associated with wet weather conditions over parts of Southern Africa. A positive SOL tends to support
cooling of the waters in the Central and Eastern Pacific Ocean (i.e. La Nina). The result of this situation generally presents above average rainfall of an erratic nature. Rains (and therefore planting) in the 1999/2000 season began in late November; approximately November 26, 1999.
Reforestation is currently a large concern of the Ministry of Agriculture and
Irrigation. Protected lands throughout the country currently include national parks, forest reserves and game reserves. National parks include Kasungu National Park, Lake Malawi National Park, Liwonde National Park, Nyika National Park, and Luwawa National Park. Forest reserves include Dzalanyama Forest Reserve and Chikangawa Forest Reserve. Game reserves include Lifupa Game Reserve (within Kasungu National Park) and Majete Game Reserve. The Government of Malawi has recently implemented a MK 5000 penalty for individuals caught deforesting. The problem of deforestation is the result of households searching for firewood and charcoal.2 "Deforestation has been a problem in Malawi for a long time because of the sheer need for wood fuel which still gives Malawi ninety percent of its energy, a fact that is likely to survive the millennium. The introduction of a multiparty system in 1994 corresponded with dramatic incidences of license, as when trees were cut in Lilongwe city's fuel wood reserve plantations and

the unfettered advent of daytime roadside piles of wood and charcoal for sale" (Lwanda, 1999 p. 33). Though most forests tend to be in the northern region of the country, those areas suffering the consequences of deforestation tend to be areas exhibiting high population density.
Politics and Agriculture in Malawi
In 1891, the Queen of Britain proclaimed what is now Malawi as a protectorate. Malawi received independence in 1964, which initiated the beginning of Dr. Kamuzu Banda's regime. Dr. Banda's thirty-year leadership bordered on tyranny, yet served successful in developing strong relations with heavily industrialized areas. Dr. Banda, a representative of the Malawi Congress Party (MCP), remained in power until 1994; when Dr. Bakili Muluzi from the opposition party of the United Democratic Front (UDF) obtained power.
During Malawi's colonial era, from 1891 until independence in 1964, Malawi
experienced minimal growth in the industrial sector, and grew into a society dominated by subsistence agriculture. Initial measures towards formal development of social welfare programs emerged during the colonial era, with the Colonial Development and Welfare Act of 1940, which aimed at improving rural livelihood sustainability through agricultural development. It was assumed that higher incomes from a market-oriented agricultural sector would be used to buy goods and services, which would in turn raise people's living standards (Lwanda, 1999).
Dr. Kamuzu Banda's regime from 1964 to 1994 displayed slow but steady growth

predominately emphasizing the expansion of agricultural production. This plan was predominately an economistic approach to social development. During the 1970s, agricultural development remained the top priority of the Government of Malawi (still under the leadership of Dr. Kamuzu Banda and the Malawi Congress Party). With ultimate goals primarily residing in financial independence from fonner colonial powers, education in agriculture, health, and other social programs were emphasized. During the 1980s, the most notable policy driven agricultural development occurred with the expansion of the estate sector of tobacco, due to the government's encouragement of burley and flue-cured tobacco production. With the large demand on the foreign market for these types of tobacco, the government targeted considerable policy towards such expansion.
Recent changes, most notably the 1994 transition from Dr. Banda's regime to the democratic regime of president Dr. Bakili Muluzi, have altered the agricultural sector of Malawi. Political influence on agriculture, such as liberalization of the market in 1994, the removal of fertilizer subsidies, reduction in credit opportunities, increased prices of maize enhanced by a devalued Malawi Kwacha and a series of changes in the tobacco sector have resulted in decreased smallholder food security. During the 1 990s, the arrival of the multiparty system in Malawi brought additional shifts in governmental agricultural priorities. Smallholders were recently permitted participation in production of burley tobacco with the hopes of strengthening existing rural livelihood systems and providing alternative income sources for smallholder farmers. It was also during these

nation immensely dependent updn male out-migration for employment, leaving the responsibility of subsistence farming heavily on Malawian women. Male out-migratioin has since reduced, meaning new definition of family structures.
The Current Problem of Chronic Food Insecurity In Malawi
The Food and Agricultural Organization of the United Nations (FAG) defines food security as all people at all times having physical and economic access to needed food (FAG, 1998). The objective of food security has been defined as "assuring to all human beings the physical and economc access to basic foods they need" (Thomson & Metz, 1997). During the 1964-1994 Banda regime in Malawi, few questioned the extent of food insecure households. Malawi, then, was assumed to be relatively food secure. In the early 1980s, Malawi experienced economic growth due to successful tobacco estate production and export, along with the introduction of hybrid maize. Dr. Banda' s emphasis on export crops strengthened international investment. This growth, however, was predominately the result of the estates rather than smallholders (figure 1-7).
Percentage Growth in Agricultural GDP in Malawi
6....-U[ Estates
o0 U Smaliholder
Figure 1-7: Percentage Growth in Agricultural Gross Domestic Product in Malawi Source: World Bank, 1995.

safety net program aimed at improving the productivity, increasing the food security, and encouraging potentially sustainable livelihood systems for Malawi' s resource-poor farmers. Government has prioritized increasing agricultural production, considering that over 80% of the population reside in mural households, and most face chronic food insecurity for two to five months every year (Gladwin et al., 2001). Smallholder farmers, facing limited opportunities for improving yields, sought government, donor, and private assistance to overcome these shocks and stresses. Obstacles to Malawian Food Security: Shocks and Stresses to Existing Systems
Shocks and stresses contributing to the deterioration of the sustainability of the Malawian smallholder livelihood systems include recurrent devaluations of the Malawi Kwacha (MK), depleted soils resulting in poor yields, collapse of the credit system, increased population (resulting in land constraints and higher population densities), instability within the tobacco market structure and regulations, imperfect markets, and insecure health conditions, particularly as a result of chronic malnutrition and the frequency of HLV/AIDS within Malawi. These factors, considered shocks and stresses to the system of smallholder farmers, have pushed farmers away from the sole activity of farming and towards secondary or tertiary activities to help them improve food security.. Devaluations of local currency
Global economic changes have resulted in the devaluation of the Malawi Kwacha and have created extreme shocks and stresses to livelihood activities in both rural and urban Malawi (Gladwin et al., 1999). During the 1999/2000 starter pack distribution, the
Malaw .aca ranged fro aprxmtl 38-5-M pe US dolr Arcual inu

Panner maize seed running MK 595; a fifty-kilogram bag of urea fertilizer running MK 825; and a fifty-kilogram bag of 23:21:0 +4s running MK 780. Considering the devaluation of the Malawi Kwacha, these prices simply create unrealistic situations for smallholders to purchase inorganic fertilizer. Smallholders who engage in any type of off-farm employment are equally challenged, as recent devaluations in the Malawi Kwacha have included no correspondent increase in paid wages.? Decreasing yields
One commonly identified production problem is that of decreasing yields over the past ten years. Farmers in all regions of the country complain of diminished soil fertility due to the increased cost of inorganic fertilizers (Benson 1999). Additionally, land constraints reduce the amount of food smallholders can produce on their own farms. Smallholders, particularly those in the southern region of Malawi and those residing relatively close to urban centers, often farming on 0.3 hectares of land or less, suffer the consequences of land constraints. Fanners also suffer from labor constraints, brought about by a lack of draft power, inadequate health conditions, and 11W/AIDS. Other production complaints stem from poor rainfall patterns, droughts in 199 1/1992 and 1993/1994 seasons and unreliable market prices and structure, particularly within the dominating cash crop of tobacco.
Prior to the 1980s, fertilizer subsidies were provided to smallholder farmers in
order to encourage fertilizer use and increase agricultural output, primarily maize (Sahn and Arulpragasam, 1991). These subsidies were phased out in the 1980s as part of

Aruipragasam, 1991). The subsidy removal (along with rising world fertilizer prices) has decreased the profitability of using fertilizer on food crops (Sahn and Arulpragasam, 1991). Fertilizer use has dropped since 1994 in Malawi (Gladwin et al., 1997). Collapse of the credit system
During the 1980s Malawi exhibited a variety of organizations providing fertilizer to smallholder fanners. Although not all areas were exposed to credit organizations or clubs providing access to fertilizer, households that found access to credit had an exceptionally high recovery rate on loans at well over 90 percent.4 Malawi became a model for providing access to credit at low interest rates to African smallholders (Gladwin, 1999). In 1994, as Malawi moved into a multiparty system, default rates continued to grow. Subsidized interest rates, previously close to ten percent, grew to anywhere from thirty to fifty percent during this time (Gladwin, 1999). The collapse of the credit system has left farmers with extremely limited access to credit, subsequently decreasing yields, and further affected the integration of cash crops in smallholder production.
Increased population and land scarcity
Within many Malawian households, the available land for production is often so
minimal that these land constraints create barriers to the farmer's livelihood system. The amount and type of land allocated to a particular household can be considered the result of a number of influences including geographic region. The northern region, for
4Acrigto interviews here, a number of different situations repeatedly occurred surrounding household
dynamic in obaiin crdt atcual fo fetilize us. Man lon wer rvie to woen whil

example, is less densely populated than thecentral or southern region. This could result in greater (adequate) land allocation in the northern region, which could in turn result in increased options in allocation of cropland. However, although the northern region exhibits a lower population density, it is comprised of large amounts of forestland, creating the potential for equally challenging land constraints.
Households in the southern region of Malawi demonstrate great frustration because of the high population density and subsequent land constraints. Similarly, farming households located close to urban centers, such as Lilongwe (central), Blantyre (south), and Mzuzu (north) often exhibit similar land constraint problems. Changes in market structure
Market liberalization occurred in 1994 throughout Malawi. Prices of all crops, with the exception of maize, are presently liberalized. Due to the recent occurrence of this liberalization, the long-term results, benefits, or hardships, cannot yet be observed. During this interim, farmers must work to learn how to best manipulate these once nonexistent markets. At present, many smallholder farmers consider market prices unreliable, particularly when smallholders consider the increasing fertilizer prices and decreasing yields. Similarly, instabilities in local markets damage the reliability of household systems. FAO considers availability, stability, and access when determining degrees of food security (Thomson & Metz, 1997). Due to the undetermined stability of the Malawian agricultural marketing systems, mural households are often lacking availability, stability, and access to producing and purchasing these basic foods due to
thi-w intai-t in-nrtn enitn e. innm- Th.vrl re .t1

markets as a source of income. The changes brought about by mere introduction of these markets have created a new set of strategies implemented by Malawian rural fanners. These changes, then, can be considered a stress or shock to the households currently operating within the newly emergent market structure.
A maize price-band system is still operated by the government, with the intention of eliminating extreme price variation resulting from differences in agricultural productivity, seasonal availability, and/or regional availability. Markets for cash crops, however, are extremely volatile and therefore greatly limit farmers' ability to consistently profit from selling cash crops.
Social developments
The high rates of H1V/AIDS throughout Malawi have also affected household livelihood systems through changes in household available labor, male out-migration, household remittances, intra-household dynamics, and additional household financial stresses. The once frequent male out-migration has become limited with the emergence of HIV/AIDS. Malawi' s previous out-migration to South African mines, for example, was stunted when many Malawian males were denied access to work in South Africa due to the fear of rapidly increasing rates of HIV/AIDS.5 The immediate impacts of HIV/AIDS are inescapable to subsistence farmers throughout Malawi. Nationwide, farmers of all economic or social classes have felt the impact of HIV/AIDS as labor pools are decreased and financial costs (such as funeral costs and mainstream or traditional doctors visits) are increased. The decreasing labor pool holds great potential to greatly

Similarly, malnutrition among both adults and children is considered an
influential factor adding to the problems of disease, hard labor, and early and frequent pregnancies among women, which all contributed to the poor health of many rural adults in Malawi (UNICEF, 2001). Although malnutrition is by no means a new social development, it further threatens the consistency of the Malawian labor force.
Additional forces have further compounded the instability of food security among Malawian households. Domestically, barriers include availability of off-farm income, changes in social structure, and of course political and ethnic barriers. Internationally, Malawian smallholders are subject to the impacts of decreased foreign investment, fads in donor interventions, and alterations in international relations with countries such as Mozambique, South Africa, Zimbabwe, and additional countries with which Malawi shares their resource pool for both agricultural outputs and labor inputs.
Beneath all factors potentially limiting food security rests the problem of
consistently decreasing yields and the lack of sustainability among the livelihood systems of Malawian smallholder farmers. Smallholder farmers in Malawi experience low productivity and low earnings leading to chronic food insecurity, and face little opportunity for employing different strategies. Without food security, the Malawian people are forced to seek governmental assistance in the form of food for work programs, welfare programs, and free input programs (e.g., safety net programs). Government intervention, such as the starter pack program, ams to increase food security by empowering smallholders and increasing production. Ideally, this would increase food

The Starter Pack: A Tool For Household Food Security and Enhanced Productivity
Considering the farming and livelihood systems of smallholders in Malawi, and the recent shocks and stresses to those systems, policy planners have been experimenting with ways to decrease the chronic food insecurity and vulnerability of the poor in Malawi. Given the considerable magnitude of chronic food insecurity, suggestions for improving food security included consideration of a "safety net" program. Safety net programs function under the preliminary assumption that overall sustainability depends upon increasing the resilience of the most marginal population within a particular area. One such effort to eliminate chronic food insecurity was the "starter pack" program first implemented in the 1998/1999 season.
Designed to jump start yields, the starter pack program distributed five kilograms urea and ten kilograms 23:2 1:0+4s, two kilograms of either groundnuts or soybeans, and two kilograms hybrid maize seed to 2.86 million farming families during both the 199 8/1999 and 1999/2000 farming seasons. The inputs selected aimed to provide each recipient with adequate maize supply for planting 0.1 hectares of land. The objectives were: "a) to assist fill the food gap; b) to promote crop diversification; and c) to promote the concept of soil fertility improvement" (Clark, February 2000, p ii). After two-years of starter pack distribution, the starter pack program was credited as being a major contributor to the national maize surplus in the year 2000.
The starter pack program was repeated in 1999/2000 along with a pilot starter
pack voucher project. The purpose of the starter pack voucher pilot project was to "test

2000). At three selected test sites, the pilot project tested the number of distributing outlets, timing of voucher distribution, and method of transporting starter packs to retail outlets. The pilot voucher distribution also held the potential to measure the priorities of smallholder Malawian farmers by analyzing the inputs selected.
Future pans for starter pack distribution include plans of reducing the target
population to forty-percent of the population, and reducing the amount of inputs included, with suggestions from the Ministry of Agriculture and Irrigation that communities themselves should select the reduced target population while prioritizing widows, femaleheaded households, and theelderly. Other considerations include possibilities of future inclusion of agroforesty inputs incorporated into the staffer packs.
The Need for Research and Considered Methodologies
Considering the farming and livelihood systems of smalholder farmers in
Malawi, and the recent shocks and stresses to those systems, policy planners have been experimenting with ways, like introduction of the starter pack program, to decrease the chronic food insecurity and vulnerability of the poor in Malawi. Given the considerable magnitude of food insecurity, suggestions for improving food security have included considerations of a "safety net" program. Safety net programs function under the preliminary assumption that overall sustainability depends upon increasing the resilience of the most marginal population within a particular area. Safety net programs, as a type of social welfare program, target marginalized populations expecting that empowerment of these groups will result in subsequent sustainability.
-l' c10oPto fr9etr~ ~l"nnra xa -1t e -v -hrl Man n QQ'

seed and fertilizer to all Malawian farmers. This essential foundation for broad-based income growth in Malawi was expected to result in a subsequent bumper maize harvest, thereby reducing inflation of maize prices, and ultimately reducing food insecurity. The starter pack and voucher programs were indeed implemented as such a nationwide safety net, and hold great potential, if implemented correctly, to become a targeted, productivity enhancing safety net (Devereux, 1999; Gladwin et a., 2001).
In this context, recommendations leading to the development of the starter pack. program came with the understanding that smallholder farmers needed some sort of "safety net" program to empower households. Even stronger recommendations suggested designing programs with the potential to increase the productivity of households suffering the most Severe economic strains. Unfortunately, virtually all smallholder farmers in Malawi suffer the most severe economic strains. Due to the diminishing maize production within Malawi, the ideal smallholder strategy was a comprehensive one as "it is not just the poor who need a safety net... Malawi itself needs a safety net" (Mann, 1998 p. 7).
Contextualizing Research: Overview of Sustainable Livelihoods (SL) Approach
With the hopes of understanding the multiple dimensions of food insecurity the
sustainable livelihoods (SL) framework built on Chambers (1997) attempts to construct a multidimensional understanding of poverty and food insecurity by understanding the complex dimensions of household life. Understanding the vulnerability of rural, food insecure households is intended to create entry points from which to design assistance

and structural forces such as political influences. Upon considering these influences, the livelihood strategies and outcomes adopted by farmers can be both described and rationalized. In evaluating a system utilizing the sustainable livelihoods methodology, systems are first evaluated according to their livelihood activities. Following this, SL research investigates methods of coping and adapting to shocks and stresses and eventually finds appropriate entry points for potential reinforcement.
Whether shocks experienced by households are economic (devaluation, removal
of fertilizer subsidies), bio-physical, or agro-climatic (droughts or floods), the foundation of the sustainable livelihoods approach rests in understanding the types of strategies a household can employ in order to empower itself and become a sustainable livelihood system. "A livelihood is sustainable when it can cope with and recover from stresses and shocks, and maintain or enhance its capabilities and assets both now and in the future, while not undermining the natural resource base" (DFID, 1999 p.5; adapted from Scoones, 1998). The preliminary assumption of the SL approach with regard to safety nets is that the overall system's sustainability depends on increasing the resilience of the most marginal or threatened population in a particular area (Gladwin, 2000). The sustainable livelihoods approach attempts to provide an analytical structure for understanding the multidimensional aspects of livelihood systems. The difficulty in sustainable livelihoods research lies in the difficulty in operationalizing the findings. Beginning with an analysis of the vulnerability of the poor, who usually suffer from many shocks and stresses to their livelihood systems, the sustainable livelihoods

The Need For Research: Understanding the Potential of the Starter Pack
The primary purpose of this research is to evaluate the impact of both the staffer pack and voucher program as a tool of increasing food security and promoting sustainable livelihoods among rural Malawian smallholder farmers. Unsatisfactory levels of food security in Malawi brought about the need for the starter pack. As stated, the future of the starter pack program within Malawi is undecided, and therefore it is of great use to understand what types of households were able to improve the sustainability of their livelihood systems, and what types of households can be predicted to do so with future distribution. Since inputs were received through means of traditional starter pack receipt, starter pack voucher receipt, or flexi voucher receipt, there exists great potential for varying degrees of assistance experienced at thehousehold level. Each distribution method required different responsibilities for household members in receiving inputs (picking up packs, redeeming vouchers, etc.) and provided households with various opportunities (selling vouchers, trading, receiving either goods or staffer packs, etc.). By understanding the impact that each distribution method had on various types of households, this research can assist in identifying types of households who may perhaps best benefit from particular types of distribution. The potential impact that various distribution methods of starter packs have on particular types of households can be predicted according to household domains. Such domains may include households represented by female or male-headed households, those engage in particular activities (for example off-farm employment), those in particular regions, those within a particular
a*icltra zoe or evnths wihi a -atca noliica zoe

Second, due to inclusion of analysis of the flexi voucher program (and the
opportunities to buy/sell starter packs) this analysis can extract the self-perceived input needs of rural Malawian households. By observing decisions made by recipients of flexi vouchers and questioning these recipients as to the criteria utilized in making these decisions, desired needs of smallholder farmers can be extracted: Further, by understanding the household level decision-making process regarding soil fertility amendments, or inputs received through voucher distribution, this research hopes to reveal the self identified priorities of smallholder fanners particularly with regard to desired inputs.
Third and final, this research can assist in understanding implications of the
starter pack and voucher pilot project as a prototype of a safety net for Malawi. Research here, then, can assist in future policy development regarding safety net voucher projects. The starter pack did not function as a typical safety net program, in that the target population included all smallholder farmers in Malawi rather than farmers considered the poorest of the poor. Traditional safety net programs do not target an entire population, rather only predetermined food insecure households. Although necessary, future funding will presumably not be adequate to continue such assistance on a permanent basis. Once the needs of smallholder farmers are understood, safety net programs face the immense challenge of targeting assistance. This research aims to understand the essential components, primarily the type of inputs necessary to assist households in both creating sustainable livelihoods and increasing food security. Future safety net programs, then,

Research Approach: Strategies Employed
The primary goal of data collection included eliciting complete household level data in order to understand both existing livelihood systems and the methods in which inputs provided by the starter pack program were utilized. Between July 8, 2000, and October 16, 2000, the researcher elicited data through ethnographic participant field interviews, field notes, and collection of secondary data.
Data collection was conducted through on farm interviews of a total of forty-seven households nationwide. Households were selected to represent a variety of economic categories, a variety of compositions, and households receiving each of the possible inputs distributed through the 1999/2000 starter pack program. A questionnaire was utilized to gather household specific information regarding farming system activities, livelihood activities, and activities related to the starter pack program. Information regarding farming systems, livelihood systems, baseline production information, additional relevant household or community information obtained during informal interviews, utilization of field notes and collaboration with secondarydata worked together to provide household level data depicting the existing livelihood systems of selected Malawian households.
Household level data was utilized to construct a linear program, or simulation of the entire household livelihood system including activities, constraints, and opportunities. Linear program data analysis allowed for simulation of households in order to understand the likelihood of livelihood sustainability both with and without starter pack inputs.

factors, preferences, and other unpredictable circumstances not foreseen by the linear program analysis. Together, these tools allowed for an understanding of the potential of the starter pack inputs as a tool to promote livelihood sustainability on both an ideal and practical context. By doing so, this research will create the analytical tool with which to
*understand the multiple dimensions in of the starter pack program within the systems of Malawian smallholder households.

Theoretical Considerations in Data Collection and Analysis
In implementing the starter pack program, it was assumed that distribution of
inputs (particularly fertilizer) to smallholders for a five year period would help fill the food gap and improve soil fertility enough to result in increased subsequent yield (Mann 1998). Distribution In 1999/2000 marked the second year of the starter pack program, and the first year of the pilot voucher project. Although future plans are far from definite, it is likely that the program will include plans to reduce the size of the target population and the quantity of inputs. To effectively target this program, it is important to understand the influence that distributed inputs have at the household level. Further, understanding the perceived impact of this program can assist in understanding the priorities of Malawian smallholders, and assist in future program development.
In planning, conducting, and evaluating research surrounding the starter pack distribution, methodology and framework of both the Farming Systems Approach (FSRIE) and the Sustainable Livelihoods (SL) approach are utilized here. Farming Systems Approach
Farming systems methodology has not been a "'top-down"~ approach where
information comes strictly from researchers, through extension agents, to farmers: rather

(FSRIE) is an approach to technology generation, evaluation, and delivery. It is applied, farmer-oriented, agro-biological research, supported by the SOCiOeconormc sciences in a team effort that includes extension responsibilities" (Hildebrand and Poey, 1985 p. ix). Fanning systems research and extension often utilizes participatory methods in order to best understand the systems existent to the client, or fanner. Participatory methods consist of open ended interviews, farmer trials, farmer analysis of trial results, mapping and modeling, matrix scoring and ranking, seasonal calendars, and other approaches that help researchers understand farmers and their systems (Chambers, 1997). Farming systems research thus utilizes a bottom up approach in order to understand the household objectives and constraints as perceived by the farmer.
Farming Systems Research and Extension has generally focused on limited
resource smallholder farmers who would be categorized in the poorest stratification of the population (Hildebrand and Russell, 1996: Bembridge, 1986). Diversity further exists within farmers categorized into this poorest stratification (Chambers, 1997). Chambers (1997) further noted that there is diversity within communities and villages, based on age, gender, ethnic group, and income level.
Farming Systems Within Livelihood Systems: The Sustainable Livelihoods Approach
One outgrowth of the farming systems approach was the sustainable livelihoods
(SL) approach that originated with the thinking of Robert Chambers during the mid1980s as a bottom-up, multidisciplinary, multidimensional approach to the treatment of poverty (Chambers 1997). Poverty was not seen as unidimensional, nor was it attributed

(decreases in maize yields, increases in input prices). Increased attention was placed on vulnerable households' assets and resources which make it easier for farmers to invent and use indigenous livelihood strategies intended to cope with these sudden shocks and stresses and even sustain their livelihood systems (Chambers, 1997). The sustainable livelihoods approach is described as:
perhaps the clearest formulation of this bottom up, decentralized, democratic and
at heart relativistic approach to development. In a sustainable livelihoods
approach, development and aid help the poor ensure that their own ways of
making a living (livelihood systems) remain viable and remunerative in response
to context-specific changes in the bio-physical, economic, political and cultural
environments in which local farming, gathering, hunting, fishing and
handicraft/artisinal systems are embedded. (University of Florida Soils CRSP,
2000 p. 3)
The sustainable livelihoods framework exhibits a prioritized focus on how the
poor manage their assets to sustain their livelihoods and cope with the shocks and stresses facing them. As a bottom up approach, the SL framework emphasizes assisting resource poor households according to their own priorities, opportunities, and available resources prior to developing resource and technology innovations. The aim of the framework is to help the poor design productive programs so that they become stakeholders in the development process and thus become empowered.
Conceptually, livelihood systems are comprised of the livelihood activities in which the poor engage in order to make their living. A livelihood is "considered sustainable when it can cope with and recover from stress and shocks, maintain or enhance its capabilities and assets and provide sustainable livelihoods at the local and global levels in the long and short tern" (DFLD, Sustainable Livelihoods Guidance

As an important part of the design process, the sustainable livelihoods
framework attempts to understand the structure of an existing system in order to identify appropriate entry points for livelihood supporting activities while considering that these systems of the poor do not necessarily function within a simple linear manner. The sustainable livelihoods approach involves a people first approach that can be utilized in project planning, research methodology, and program implementation. It seeks to reorient the development process away from top-down, technocratically planned interventions, in favor of finding ways for donors to enhance already-existing modes of making a living and systems for managing human and natural resources (Gladwin, 2000). Understanding and differentiating between livelihood systems and livelihood strategies is therefore of great importance to the sustainable livelihood approach. The particular strategies and activities employed by households represent choices made from all options and activities available to a household. These selected activities are considered livelihood strategies and the entirety of options available to a household constitute the livelihood system. Within a livelihood system, a variety of strategies are available, and the sustainable livelihoods approach works to understand some of the decision-making criteria households utilize in order to select such livelihood strategies from an entire livelihood system.
Following is a description of the livelihood system developed by a southern Malawian household. Within the Makina household, the primary and secondary livelihood activities can be observed, along with the non-linear structure of strategies
- nl-v-d-tn imr. -on C -riv o th- -n h-.eod uthr bot th- rmr

those dependent upon the specific opportunities, obstacles, and networks available to
each member of the Makina household.
The Livelihood System of the Makina Household
The livelihood system of the Makina household is based on the primary livelihood activity of subsistence farming. The Makinas
have quite a large plot of land for a household in the southern
region; approximately two acres, and a small dimba garden directly
outside their brick home. Cecilia, age 28, her husband Rafael,
and Precious, age 10, are responsible for farming all household
land. During the 1999/2000 season, the Makinas did not hire any
additional labor, or 'ganyu', however children within the Makina
household did partake in ganyu labor.
Cecilia Makinas dimba includes seasonal rotations of mustard
leaves, cabbage, cowpeas, tomatoes, onions, pumpkins and of
course maize. Most of the maize produced in the dimba is eaten
fresh rather than processed. Cecilia interoropped bananas and
sugarcane, both utilized predominately for small scale sale,
around the entire border of the dimba. Inside, she has a
combination of cowpeas and maize, with small plots of land
devoted to numerous pulses. These are all used predominately
for home consumption, sale within the village, and sale at
Luchenza trading center, located approximately eleven kilometers
away. Crops produced in the dimba are usually grown with
saved seed and occasional purchases. Cecilia purchased mustard
January 2000, for example, costing MK2O.00 (USD $0.40). This
allowed her to plant approximately five square feet of mustard,
which was sold for 0.50 tambala per three leaves. The Makinas
two acres of land consists almost predominately of maize and cassava, with scattered cowpeas. Within these two acres, the
Makinas rotate a plot of fallow land, because of household labor constraints. The remainder of the Makinas' land is scattered with
mango and some banana trees. The nearest mill for grinding
maize is approximately a three-kilometer walk away. Cecilia usually
her daughter, Precious, with a small sum of money to the
mill as necessary for grinding. Vegetables and other crops
Cecilia decides to sell are sold in the trading center of Luchenza.
This trip consists either of an initial three- kilometer walk to the main
road, where transport can be obtained for around MK2O or takes
aDrximtel the hour to walk Ceii collects-woo

The livelihood systems of rural farmers in Malawi, like that of the Makina family are complex; considering the multiple sources utilized to access food, inputs, and income. Sources of capital and social capital extend through boundaries of formal sectors, and include numerous livelihood strategies functioning within the informal sector.
The sustainable livelihoods approach encourages consideration of a wide range of factors as they shape and influence livelihoods with the attempt to provide long term impact on subsections of a livelihood system. The non-linear approach serves to highlight the complexity of the actual system without constructing a static portrait of such. DFID's sustainable livelihood approach specifies nine related concepts that will be described here. These concepts build upon one another to assist planners in understanding a variety of aspects of the household livelihood system. Capital and assets
Capital and assets are identified within sustainable livelihoods in order to assist in creation of a proper understanding of how the poor can manage to cope with sudden shocks and long-term stressors. Capital and assets are considered the building blocks of livelihood systems; and include natural resources, social ties, human skills, physical infrastructure, and financial opportunities. Coping and adaptive strategies
Coping and adaptive strategies are indigenous strategies developed by resource poor farmers. The importance of differentiating the two kinds of strategies lies in determining the need for understanding long term versus short-term situations and

strategies are considered long-term strategic adaptations to long-term and gradual stresses, such as increases in fertilizer prices or interest rates, or changes in geographic landscape. When alterations in farming systems or shifts in primary or secondary livelihood activities are employed, it is assumed that the household's ultimate goals are to enhance capacities to make a living based on the objectives of conserving household resources (land, labor, cash) in order to reduce the risk of exposure to these very same shocks and stresses. Livelihood activities are often divided into primary, secondary, and tertiary activities, all employed to reduce chronic food insecurity. Assets necessary to modify coping or adaptive strategies through changes in primary, secondary, or tertiary livelihood activities are identified here.
Urban coping strategies in Malawi have been identified as including: renting out houses on plots for plot owners, renting on a crowded plot to save money, moving to squatter areas where rent is cheaper, building houses for rent in squatter areas, using unprotected sources of water in order to save on water charges, walking to work, not sending children to school, sending some children to live in rural areas, walking long distances to collect firewood, collecting sawdust from sawmills, using maize husks, beer cartons or any possible combustible items instead of wood, and lighting as few fires as possible per day. (Roe & Chilowa, 1990). Similarly, rural coping strategies include performing ganyu labor in exchange for food, eating wild tubers, eating additional greens such as cassava or pumpkin leaves, eating unripe fruit such as mangoes, vending of products, petty trading, beer brewing, selling firewood, grass, woven baskets, mats, clay

begging, women tie a cloth around the stomach to reduce the feeling of hunger, and engagement in gambling" (Meuller-Glodde, 1998 p. 57). Empowerment
Empowerment of the poor is the ultimate aim of the sustainable livelihoods
approach. Empowerment activities, in this context, are considered activities that can be performed utilizing local resources and have the potential to enable problem solving by the poor themselves to improve their own quality of life. Therefore, a development project does not come into a region to distribute a new technology in a top down approach; instead it should work to understand the livelihood systems of the poor in order to empower them to change their livelihood strategies and outcomes. Entitlements
Entitlements are considered the assets and income streams available to an
individual especially those that vary dependent upon an individuals position or status. Resource availability is rarely equal, and entitlements include opportunities that, unlike empowerment activities, are institutionally based. Equity and inequity
An understanding of both equity and inequity issues is necessary to comprehend
options for individuals working towards improving livelihood systems. Although equity and inequity are somewhat underemphasized within the sustainable livelihood theory, potential categories include gender, class, generation, occupation, tribe, family, political affiliation, or marital status, marital status within polygamous unions, or changing marital

Governance at the national and local level are crucial in understanding the
livelihood of a household, as well as the livelihood strategies approach. Naturally, the success of a livelihood system is heavily influenced by political regulations at multiple levels; such as reciprocity, shared norms and values, traditional leadership, trust, familial ties, and institutionalized rules or regulations. Inclusion of such creates an understanding of regulations and expectations placed upon individuals within a livelihood system. Livelihood systems
Conceptually, livelihood systems are comprised of the assets and activities in
which the poor engage in order to make a living, and in Malawi' s case reduce chronic food insecurity. A livelihood "is considered sustainable when it can cope with and recover from stress and shocks, maintain or enhance its capabilities and assets, and provide sustainable livelihoods at the local and global levels in the long and short term" (DFTD, Sustainable Livelihoods Guidance Sheets; 2000). Identifying and understanding livelihood systems includes a cooperative understanding of the existing social and political structures, as well as the households opportunities within those existing structures.
A clear definition of resource poor rural households' level of poverty must be
defined in order to assess the sustainability within a livelihood system. Numerous criteria are utilized world-wide in defining poverty, such as: percentage of underweight children

differentiation according to direct and indirect measures of poverty determination. Area specific recommendations result from understanding of such differences and characterization of households or even particular regional areas begins with utilization of such indicators in identifying relevant characteristics of economically impoverished households.
Resilience is defined as "a measure of a socio-ecological system's ability to
recover from shocks and stresses" (DFID, 1999 p. 3.6). Resilience is considered within the ecological subsystem within which households operated, and is determined by the ability of the poor to bounce back within the social subsystem. It is further determined by the capacity of people and institutions to self-organize and design self-sustained adaptations resulting from interactions between social, economic, and ecological systems. Resilience further characterizes the ability of households to maintain their current status of economic well-being.
The sustainable livelihoods approach and logical framework can be used to design, manage, and evaluate projects and programs (DFLD, 2000 p. 3.6). The SL approach is used to construct a framework that is used throughout the project from the design to the evaluation stage of a program. Construction of the framework though livelihoods analysis (figure 2-1) results in the formation of a log frame to determine subsets of issues. Additionally, participatory methods are widely used in all stages of the SL approach; e.g., problem trees assist to express the emerging cause and effect
-elati-n-hia aach 'issue idnife bv the .mtciat thmslvs a fcs ru

meeting is recorded, ranked in self-perceived hierarchy and finally, the causal relationship is explored (figure 2-1).
iiiao .
Figure 2-1: Example of a Problem Tree Source: DFID, 2000 p 3.6
Although cause and effect relationships are determined and a holistic picture of
the sustainable livelihood system is created, the sustainable livelihood approach does not attempt to provide holistic solutions. Entry points are established in order to focus attention on a sub-set of issues. Particularly, these entry points are useful in designing safety net programs targeting a subsection of resource poor farmers. The strategies implemented by households experiencing food insecurity may be short or long-term strategies, and serve to create informal safety nets for many households. Illustrating the

The Makina Household Livleihood System and Strategqies
Cecilia Makina lives in the southern region of Malawi, in the village of Chimwanga. Her husband Rafael moved from his nearby village to Cecilia's home village of Chimwanga when the two were wed, and together they have two living children. Daughter Precious, aged ten, is a vital part of the Makina livelihood system, while four month Chifundu is the newest addition to the family. Two years ago, Cecilia lost a six-year old son to an unknown sickness. The Makinas primary livelihood activity is that of subsistence farming. As the primary subsistence farmer working to complete Production Activities within the household, Cecilia struggles for enough time to handle her reproduction and production activities. Ten-year old Precious attends school occasionally, however her primary responsibilities lie in production as well. She is also responsible for her share of cooking, cleaning, transporting maize by foot to the mill and collecting water. Precious also works as ganyu. labor in the village, earning the household pails of maize. Four month old Chifundu, when he reaches the age of around five, will begin assuming his share of production activities as well.
The Makina household currently includes two goats and two chickens who are not layers. Due to consistently decreasing yields experienced by Cecilia and Rafael, coupled with the addition of newborn Chifundu, two goats were sold in November 1999 to supplement household income. Similarly, two chickens were eaten for special occasions in late 1999, and six chickens were sold between August 1999 and August 2000. Prior to sale of these chickens, Cecilia sold eggs for MK 2.00 per egg (approximately USD $0.04). Cecilia does not plan to purchase any additional chickens until at least 2001, due to the high incidence of Newcastle Disease during the late months of the year.
Cecilia's secondary livelihood system involves her informal employment in 'wooling.' Wool is brought to Cecilia, and she in turn makes baby clothes and blankets. This income, according to Cecilia, is most prevalent during 'baby season', but provides income year-round

Theoretical Understanding of the Starter Pack Program in Malawi
Implementation of the starter pack program in Malawi resulted 2.86 million
farming households receiving some type of starter pack inputs.' But did it help the poor develop enhance productivity and therefore hold potential as a tool to improve livelihood sustainability? That is the important question addressed by this research. At a glance, the starter pack distribution program was promulgated as a technology transfer program to get farmers to start using fertilizer (Mann, 1998). However, farmers in Malawi have been using fertilizer since the 1960s (Gladwin 1991). It can also be considered a political tool; as the starter pack program was first implemented during the election year of 1998/1999 after both parties had, in the previous election, promised free fertilizer to win votes. Additionally, the political leverage of the donor community can be observed, as they evidently are wary of providing long term free inputs programs, however in essence cast a political vote during implementation of their programs. The difficulty in defining the application of the starter pack is further compounded by evident regional disparities and distribution inconsistencies noted nationwide. Were these inconsistencies the result of political stronghold regions, a result of disparities propagated by traditional authorities, or simply the result of inadequate funds for equal distribution?
Defining the starter pack as a particular type of program (safety net, free input, etc.) has thus become extremely difficult, particularly due to the unique nature of the program. Significant ambiguity surrounds the implementation of this program, leaving room for questions regarding the actual purpose of the program; was it a political tool, a
froinl roaaaeft o rcrmn ohilo nfr nrn-h-t .

of some smallholders increased due to the starter pack program, the initial political, social, and agricultural motivations created great confusion to farmers and administrators regarding precisely what the starter pack program truly was.
We turn now to the potential of the starter pack program. This type of
agricultural input distribution appears to have great potential to become a successful safety net program, with an appropriately targeted audience. Distributing staffer packs to a subsection of the population in order to reduce chronic food insecurity may be a potential application for the starter pack program and the voucher program that have been piloted thus far. The success of each of these programs, as perceived by the farmers, will be evaluated here in order to understand the potential impact of the programs continuation.
What is a Safety Net?
Safety net programs function under the preliminary assumption that overall
sustainability depends upon increasing the resilience of the most marginal population within a particular area. Safety net programs, as a type of social welfare program, target marginalized populations expecting that granting entitlements to the poor will empower these groups and result in subsequent sustainability. Both informal and formal safety nets theoretically function to reduce chronic food insecurity within both rural and urban households in Malawi.
Both formal and informal safety nets are tools to reduce chronic food insecurity among smallholder Malawian households. Formal safety nets, functioning as welfare

themselves function to reduce chronic food insecurity, often through diversification of capital, social capital, and informal networks of friends and neighbors sharing food. These informal safety nets are usually developed by impoverished households themselves over time as a resulting action to stresses, such as decreased yields, or developed out of urgency as a result of immediate shocks to the livelihood system, such as droughts or floods. Introduction of a successful formal safety net strategy happens only after the informal safety nets whether ultimately constructive or destructive are deemed lacking by the targeted population, or have disappeared entirely. Understanding informal safety nets is crucial to implementing successful formal safety nets, if they are to be used as complements and not substitutes for the informal networks already in place. This understanding of existing formal and informal safety nets as well as their sustainability, can be taken aboard in the development of a (formal) safety net program. (Tkoka & Mvula; 1999).
Recent research in both rural and urban Malawi has explored the efficiency of existing informal safety nets (Devereux, 1999; Gladwin et al. 1999), in large part to determine appropriate formal safety net proposals. There is considerable evidence that both rural and urban households find infonnal transfers of food, cash, and credit to be decreasing in both frequency and effectiveness. Households adjusted to the shock of devaluation of 1998, and stresses of increased fertilizer, oil, and credit prices by increasing off-farm employment, informal employment (ganyu), gifts from relatives and friends, food rationing, and withdrawing children from school (Devereux, 1999). The

between rich and poor or among the poor themselves, appear to be declining over time, partly as a general consequence of commercialization and partly because deepening poverty means that the economic basis for redistribution is contracting" (Devereux, 1999 p. 1).
Both formal and informal safety nets within Malawian society influence the
sustainability of the resource poor farmers. Informal safety nets are seen to be useful in both rural and urban households, and often include transactions between these two. However, the security and consistency of these informal safety nets is questioned in times of extreme economic strain. "In a concern of deepening food insecurity and livelihood vulnerability, this suggests a role for formal transfers to supplement the inadequate incomes of the poor" (Devereux, 1999 p. 52). Only if a formal safety net program helps the poor achieve sustainable livelihoods, can it function as an effective safety net. It will not improve the sustanability of livelihood systems, however, if it only increases the dependency of the poor on government handouts, and thus decreases their feelings of empowerment. Designing a safety net program utilizing an understanding of the vulnerability of resource poor farmers is therefore necessary. This should include knowledge of the livelihood system, resources, and constraints faced by the household, and the livelihood activities developed by the poor themselves to cope with shocks and stresses they normally face. "Bearing in mind that the poor have been there for a long time although the numbers are increasing, it is possible that the poor employ a number of strategies to survive; both positive and negative. A safety net program can then possibly

Productivity enhancing safety nets (PES Nets)
Introduction of a formal safety net is often coupled with or enhanced by
introduction of a productivity enhancing safety net (PES net) (Devereux 1999; Gladwin et al. 2001). In Malawi, where over 80% of the households are rural farming households, increased agricultural production can be considered necessary in order to generate an operative economy at the national level. Increased agricultural production is necessary in order to improve the financial situation of both Malawians and Malawi. By introducing a safety net that intends to increase agricultural productivity, rural households should increase their household productivity and ability to be resilient, while at the same time becoming more empowered.
Is the starter pack a safety net?
Resource poor farmers in rural Malawi have proven vulnerable to socio-political influences resulting in chronic food insecurity within rural Malawi, particularly the aforementioned recurrent devaluations of the Malawi Kwacha (MK), depleted soils resulting in poor yields, collapse of the credit system, increased population, instability within the tobacco market structure and regulations, imperfect markets, and insecure health conditions, particularly as a result of chronic malnutrition and the frequency of HIV/AIDS within Malawi. Given these influences, smallholder farmers are forced to seek government, donor, and private assistance to overcome these shocks and stresses
In efforts to offset the high cost and risk of obtaining agricultural inputs,
particularly for resource-poor households, the Malawi Government responded with the

Malawi in 1998/1999 and 1999/2000 provided free agricultural seed and fertilizer with the hopes of increasing productivity, both at the household and aggregate levels. By enhancing productivity, this type of safety net holds the potential to empower households to create and even maintain sustainable livelihoods. Due to the distribution of productivity-increasing inputs within the program, the starter pack program could be
considered a prototype of a PES net program, with still ambiguous intentions. The starter pack therefore cannot solely be entitled a safety net program, due to the ambiguity of the ultimate goals of the program; for example were they political or socio-economic, and were they aimed increasing national or household food security? Further, this program was not designed as a safety net, as all/smaliholder farmers were targeted. The
decision to include all smaliholder households may have been the result of cultural considerations, such as the assumed jealousy potentially arising from a targeted distribution, or political considerations, such as politicians' claims to distribute free inputs to households supporting particular political parties. Keeping in mind, however, that
"Malawi itself needs a safety net" (Mann, 1998), the starter pack was in fact presented and implemented as some variation of a safety net program. The Role of the Starter Pack Program in Food Security
Due to the instability of the Malawian agricultural system, mural households are often lacking availability, stability, and access to basic foods, resulting in chronic food insecurity. During the 1964-1994 Kamuzu Banda regime in Malawi, few were allowed tn nuiestinn the extent nf fnnd insecuire hnousehnlds or regins of the cniintry. Ma1awi,

depletion, the removal of fertilizer subsidies, the collapse of the credit system, a recent and constant devaluation of the Malawi Kwacha, land scarcity, alterations in market structure, and changes in agricultural policy, and the social developments. Levels of food security
Household food security implies different levels of security for various
households. In a study conducted in rural Zomba, households defined food security "by giving characteristics of what they would consider a food secure household" (Tkoka & Mvula, 1999). Some characteristics were prioritized only by men, some only by women, and some by both. Often in Malawi, households with adequate maize supply are considered food secure. Characteristics identified by households themselves, however, included: the ability to buy and apply fertilizer, maintain healthy household members, purchase household goods (from the local parastatal Agricultural Development and Marketing Corporation, ADMARC), possess livestock, own or participate in ownership of business, and finally the ability to cultivate additional fields.
The Committee on World Food Secunity defines food security as "physical and economic access to adequate food for all household members, without undue risk of losing such assess" (Thomson & Metz 1997). Intra-household dynamics at the household, or even community level, further complicate creation of complete food security at the individual level. National food security does not infer household food security, as poor distribution of an adequate food supply at the national level may constrain food security at the household level. Similarly, stresses such as increases in

Numerous countries or regions often boast national food security, or a food surplus, while consecutively demonstrating a chronically food insecure population. National food security is described as "a satisfactory balance between food demand and food supply at reasonable prices" (Thomson & Metz 1999). The impact of the staffer pack program, however, may be evident at either the household or national level or both. As in most input programs, design and implementation of the project ultimately determine the subsequent impact.
The starter pack program, regardless of its lack of clarity, its objectives, or its
definition, aimed to narrow the food gap within smallholder Malawian households. Input programs, whether safety nets, PES nets, free input, or technology programs, ultimately aim to provide short-term strategies that empower resource poor households to create for themselves long-term sustainable livelihood systems. But do they? This analysis considers the implications of the starter pack program in relationship to the programs' stated goals, not necessarily those of a safety net. Success, in this evaluation, will include the success of the starter pack program as a tool in reducing the food gap and increasing household food security. The importance of considering safety nets, however, lies in the potential of this program to become a targeted inputs program, and ultimately a md~del productivity enhancing safety (PES) net. The remainder of this research will address the impact of the starter pack program on the sustainability of Malawian farming households, along with its potential as such.

The Recent Situation of Smaliholder Farmers in Malawi
To best understand actual and potential impact of the starter pack program, it is
necessary to explore the realities of Malawian smallholder farmers. The current situation of these farming households results from aforementioned political, cultural, and historical events, and is presently reflected in the existing systems both farming systems and entire livelihood systems. Additionally, the opportunities and barriers including the role of market forces, soil fertility, credit opportunities, and political realities, presented to each household play a large role. It is these systems, opportunities, and barriers that have led to the current situation; one of low production and subsequent food insecurity. Farming Systems
Despite obvious variation among households, villages, or regions, there are some universal attributes shared by many Malawian smallholder farmers. The main livelihood activity nationwide is subsistence farming. Beyond this, there are agricultural differences that are regionally specific due to climate, agricultural suitability, and preference. There do exist some general trends, or farming characteristics, among the majority of smallholder producers.
General characteristics of Malawian smallholder farmers.

tons during the 1998/1999 farming season to 72,300 metric tons during the1999/2000 farming season. "Remaining the staple crop, maize now occupies eighty-five percent of smallholder cropland" (Benson et al., 1998; p 10). The Ministry of Agriculture and Irrigation estimated smallholder maize production at 2.4 million metric tons in the 1998/1999 season and 2.5 million metric tons in the 1999/2000 season. Despite farmers' complaints of diminishing yields, these numbers resulted in a national surplus of maize during the 1999/2000 season. The predominant cropping activity consists of intercropping maize with groundnuts or other legumes. Generally, common beans (Phaseolus Vulgaris) are planted in the wetter parts of the country, and pigeon peas (Cajanus Cajan) in the drier parts; particularly in the south. In some areas, cowpeas are also utilized. Many households, particularly those with land scarcity, utilize dimba, or moist gardens located near the house, to grow a variety of vegetables. Additional consumption crops grown in regular or dimba land include groundnuts, cassava, pulses, tomatoes, sweet potatoes, pumpkins, chili peppers, pumpkins, bananas, sorghum, and miller.
Cash crops are limited, but include sugarcane, cotton, tea, and the dominant cash crop of tobacco. Types of tobacco grown in Malawi include Flue Cured, Burley, Northern Division Dark Fired, Southern Dark Fired, Sun Air, and Oriental, with smallholder farmers having increased production of burley tobacco in recent years.
Households commonly own chickens for production and sale, egg consumption and sale, or as liquid assets used for funerals or other cultural events. Draft power is
limited, nat5ul in the sot whr lan cosrit are motsvr.uoti

Those possessing draft animals may rent out oxcarts for transport to those who can afford it. However, a majority of smallholder farmers do not have access to draft power. Pigeons or guinea fowl are also a significant part of the farming system, especially considering the comparatively low purchasing and maintenance cost of these animals. Ownership of swine is minimal, although numerous programs are working towards implementing animal husbandry programs geared particularly towards women.
Household distribution of land is varied, and both patrilineal and matrilineal
systems exist. Many adolescent children receive a small portion of land with which to produce minimal household crops. In less densely populated areas, a married couple may have land belonging to both the husband and wife, further differentiating labor responsibilities. In areas of extremely high population density, one member of the household may possess land in a home village and return to farm this land during planting/harvest season.
Responsibilities of obtaining wood and water are predominately those of the
women and children within the household, with the exception of the excessive amounts of water necessary for tobacco production. This task is usually shared among all household members. Gender disaggregation within Malawian society is reflected by agricultural roles, however it varies by ethnicity, land availability, availability of off-farm employment, and even simply from household to household. Both men and women are deeply involved in both production and processing, with some roles gender disaggregated and some shared. Processing of maize, for example, is predominantly a female

unimodal rainfall pattern and additional seasonal characteristics. The dry season in Malawi begins in May and ends in October; while the rainy season endures from midNovember until April (CIA, 1999). Most agricultural work occurs only during the rainy season with the harvest arriving in April/May. The Ministry of Agriculture and Irrigation stresses the importance of promoting crops to be grown in the dry season. However, with the exception of dimbas, Malawi is dominated by a single growing season due to its single rainfall pattern. In the dry season, usually in July or August, land is prepared by burning crop residue and turning soil under. In September/October, ridges are made in preparation for planting, which occurs after the first rains; usually in mid November. Fertilizer use
Area-specific inorganic fertilizer recommendations for fanners were established by Benson, 1997 and further revised in 2000 (Benson 2000). These recommendations specified, by region; the amounts of fertilizer considered optimal for households producing maize for both market sale and home consumption. Most smallholder farmers in Malawi are currently utilizing lower rates of inorganic fertilizer than those recommended. Benson states that although present fertilizer use on maize faces difficulties, this use must increase in order to improve food insecurity (Benson; 2000). These short-term difficulties include the lack of cash and credit available for fertilizer. Selected input prices from the local retail outlet of Farmers World, from February 2000 are listed below (figure 3-1). Prices shown represent the item cost when purchased with cash. Prices of fertilizer acquired with credit (not shown) are slightly higher than those of

Type Input Price in Malawi Kwacha (MK)
Urea (50 kg bag) MK 840.00
CAN (50 kg bag) MK 650.00
DAP (50 kg bag) MK 820.00
23:210 + 4s MK 780
Maize producer price (50 kg bag)_____ MK 210.00
Maize consumer price (50 kg bag) MK 425.00 (drops to appx. MK 250.00
____ ____ ___ ____ ____ ___ ____ ___ between August and February) Figure 3-1: Farmers World Prices; February 2000 USD $1.00 = approximately MK 47.00
Farming calendar
The farming calendar throughout Malawi varies with region and crop (figures 3-2 through 3-6). As maize, groundnuts, and soybeans were included in the starter pack, the farming calendar for these crops must be considered here. Planting maize, and pulses intercropped with maize, occurs with the first rains, usually mid to late November. If applicable, first fertilizer application occurs in December, and a second application in January. Utilizing local maize, households usually begin eating fresh maize in March, followed by harvest in May. Planners of the starter pack distribution intended that starter packs be distributed in full by the end of November, allowing for timely planting. Some households claimed, however, that inputs were received much later than this; with some reports as late as January. For this reason, the time of input receipt in relationship to the farming calendar must be considered as a decision-making criterion. During analysis here the seasonal calendar will begin in July, when preparation for maize planting occurs.
The following seasonal calendars were elicited from thirty individuals in the
central region, reporting on tasks required on a monthly basis. Tasks performed in some

Month Activity Month Activity
July end of harvesting; selling; July' harvesting
transportation of market sold August harvesting
_______maize finishes September clearing land
August clear land (minimal households; October ridging
usually households devoted to November cutting planting
dimba at this time) December planting
January weeding
September completion of clearing land;
land preparation; some begin Februao' ~ banking...
March weeding._.
______row making April weeding_.
October row making a hcigfrmtrt
November planting (first rains); replanting June hckn ormtui)
if necessary ue aretn
December weeding; fertilizing (if Figure 3-4: Cassava Calendar
applicable) Mot ctvt
January weeding; fertilizing (if Julyh "Atingorslln
______applicable); some bindingJu __February fresh maize consumption Augut.egn.elin
March fresh maize consumption; September selig
_______harvest early, maturing varieties October land clearing
April harvest dry, maize November planting (with rains)
May, stalking December weeding/begin binding
June selling (may continue year January binding
round dependant on yield) Februa_.....ry monito growth
Figure 3-2: Maize Calendar March final stg;monitor
__ __ _ __ __ __ __ __ prlripe and ready mo io
Month Activity May, begins to dry; monitor
July_.reparation June begin harvesting
Au ust nreyFigure 3-5: Soybean Calendar
Setme repare beds
October preare land; transplant Month Activity_.
November monitornguy igig.rudnt
December monitoring..August sellin....
JanuT First leaves redmonitorSeFerayFirst leaves red;mntrOctober row making..
March Second leaves ready; hretNovember planting...
April Binding Deebr;edn
May.. Binding; trnpotn January... weeding;, binig rp
June Bidn;transprtngFebru-.' binding-.
Figure 3-3: Tobacco Calendar: March leaves begin to grow
April wait for ripening; monitor

Livelihood systems
Livelihood systems of smallholder farmers naturally revolve around the primary
livelihood activity of subsistence farming. Regional variation does occur, as evident for
example in the strong reliance on subsistence fishing within households residing on the
eastern coast of Malawi, along Lake Malawi. The predominant secondary livelihood
activity is that of 'ganyu' or informal labor. Ganyu work generally consists of working / for other Malawian households on an unpredictable basis. Additional secondary
livelihood activities range from farming timber, charcoal, fisheries, to making and selling bricks, baby clothes, furniture, or selling items resale or as shop owners. Similarly, there
is regional variation amongst secondary or even tertiary livelihood activities. It is the
combination of these primary, secondary, tertiary livelihood activities coupled with the
numerous adaptive and coping strategies adopted by smallholders that creates their
complete livelihood system (Scoones, 1998).
Livelihood strategies
Farmers' livelihood systems throughout the country are reflective of both cultural
tendencies and smallholders' efficient use of natural resources. Throughout the country
there are a number of shared adaptive livelihood strategies. Such strategies have evolved
from farmers' reactions to long-term shock or stress and are currently an integral part of traditional livelihood systems nationwide. These strategies are not short-term solutions,
rather strategies that have been woven into modern Malawian culture. Such strategies
include ownership of cattle, chickens, and guinea fowl as liquid assets to be sold in times

rural/urban codependency and participation in organized projects (i.e. field assistant monitored seed multiplication projects, animal husbandry project; donor intervention projects).' Short term coping strategies common throughout Malawi include, but are not limited to planting of quickly maturing maize in order to harvest maize in February, utilization of handouts from NGO's, churches, donors, and government, sharing of maize during the hungry season (November March), and theft of crops.
The southern region. In the southern region, of Malawi, "the past decade has
been one of considerable change but many of the strategies employed continue to prove effective. The main changes seen throughout the southern region of Zomba, for example, include the spread of burley tobacco growing, an influx of traders at harvest times, an increase in the scale and intensity of crop trading among local traders and farmers, and a perceptible growth in trading centers and local markets. Persistent patterns include the diversification of agricultural production and of income strategies, the signal importance of maize production in households at all levels of wealth, and the skewed distribution of income" (Peters, et al. 1995)
The central region. Both smallholder and estate farmers throughout the central region of Malawi grow a considerable amount of tobacco. Within the central region, livelihood strategies among the population of smallholders may exhibit some homogeneity. Unfortunately, the tobacco market has provided less than consistent opportunities to smallholders, perhaps creating a variety of coping strategies.
The northern region. In the northern region coping strategies tend to include
utition of tibe forsts seli r of chrol n cr-dvesfcain-1i ea

been exposed to shifts in deforestation, reforestation, and charcoal policy. This region is also less densely populated, eliminating many pressures associated with land scarcity. Market Opportunities for the Smaliholder Farmer
Market forces have changed enormously since the late 1980s, the period of socalled 'market liberalization' in Malawi. The impact of this force includes both obstacles and potential solutions to increase the productivity of smallholders. The potential of the starter pack program on increasing food security is subject to these types of forces. An understanding of the changes in the markets, particularly for maize, tobacco, and additional consumption crops, is in order. Long-term solutions are dependent upon Malawians producing successful and sound adaptive strategies, rather than coping strategies, able to remain in operation within the formal existing structure.
Market liberalization, occurring predominately in 1994, has subjected Malawian smallholders to a variety of new systems and forced changes in agricultural production goals. At tetime of starter pack distribution, prices of all crops, with the exception of maize, had been completely liberalized. Markets apart from maize, particularly for the dominant cash crop of tobacco, are extremely inconsistent and therefore greatly limit farmers' ability to utilize cash crops. The arrival of the multiparty political system introduced a free market as well as a system reliant on private traders. The maize priceband system is operated by the government, with the intention of eliminating extreme price variation resulting from differences in agricultural productivity, seasonal availability, or regional availability. The market opportunities available toe smallholders
in-ec h rt i u dh mllode- A -lcto -fcomo-mrkt avilbl

The Malawi government presently operates the maize price band system. The price band system has four objectives:
1. To encourage use of purchased farm inputs by assuring farmers a market
for their maize at harvest time at an assured producer floor price;
2. To protect consumers of maize by releasing maize from the Strategic
Grain Reserve (SGR) into the market at a target ceiling price to assure that
maize is available at reasonable prices;
3. To provide enough scope between the floor and ceiling prices so that
private traders can profitably buy maize at harvest time and store it
properly for sale in the hungry season;
4. To operate the band system in a way that facilitates maximum opportunity
for the private sector to participate in government purchase and sale
Maize produced by subsistence farmers is utilized predominately at the village
level, but some also finds its way into urban markets. Household production of maize is generally consumed within producing households, or shared among family members in rural areas (excluding times of extreme hunger). Market prices of maize vary with location and month (Appendices A B). The 2000 year was marked with a national maize surplus, and little variation in the prices of purchasing maize. Tobacco
Serving as the largest, and in some areas the sole cash crop produced, the
importance of tobacco cannot be overlooked. Subsistence farmers currently involved in tobacco production have recently been subjected to numerous changes within the tobacco market structure, stunting the economic growth for the smallholder tobacco farmer. This is of great importance, particularly among farmers within the central region, which, due

smallholder tobacco farmers. Predominately, smallholder maize production is that of burley tobacco (Appendix C).
Prior to 1997, the tobacco market within Malawi functioned largely in part due to the success of the 'Phantom Train' which operated between Salima and Lilongwe. This railway system allowed operators to bring tobacco directly to the auction floor in Lilongwe. Through an agreement with the tobacco auctioneer of Auction Holdings, Malawi Railways, and the farmers, tobacco produced by smallholders was transported directly to the auction floors. Auction Holding took a proportion of the earnings, as did Malawi Railways, leaving the farmer with a somewhat predictable income. Use of the railway system allowed tobacco to be transported to Lilongwe in a timely manner; an important consideration when dealing with a delicate product like tobacco leaves.
In 1998, Mozambique purchased all shares of the 'Phantom Train', eliminating
efficient transport of tobacco to the auction floors. The Tobacco Association of Malawi (TAMA), a parastatal organization, proposed an alternative strategy, employing local companies to transport tobacco by truck. In efforts to reduce congestion on the auction floors, this arrangement did not permit trucks with less than five tons to approach Auction Holdings. TAMA established depots nationwide, in order for smallholder farmers to drop tobacco. The parastatal organization TAMA then hired local transporters to deliver tobacco to Auction Holdings. The transporter charged TAMA a per bale rate, and in turn farmers were charged established rates per bale. Due to the transition from railway to vehicle, long delays occurred in transporting tobacco from depots to the
autolor.Tesl--------------ay was rce eanig fr smllhldr, et

spoilage grew, and in May 1999 a new alternative, the intermediate buying system, was introduced.
The intermediate buying system for tobacco has been subjected to great criticism
from local entrepreneurs and questioned by the Ministry of Agriculture and Irrigation due to the inadequate returns experienced during the first year of existence. Under the intermediate buying system, the 1999/2000 tobacco industry profits were substantially diminished. The minimized income generated from Malawi' s largest export cash crop sector co-occurred with the most recent devaluation of the Malawi Kwacha and further decreases in tobacco production quality and profits are expected in upcoming years. Under the intermediate buying system operating during the 1999/2000 farming season, buyers hoping to bring tobacco to the auction floors paid non-refundable registration fees of MK1500 as a payment committing the 'farmer' to producing a specified amount of at least 2000 kilograms (twenty bales) of tobacco. 2With this, the Malawi Government would presumably be able to accurately predict estimated national production and therefore function competitively within the world market. Farmers committing to purchase are generally those who were previously large-scale producers. Many of these producers found it more profitable to decrease their own planting of tobacco and simply purchase and transport tobacco from smallholders to Auction Holdings, where tobacco companies purchased tobacco in US dollars. Auction Holdings then deducted withholding taxes, handling charges, and loan repayment where applicable and paid the 'farmer.' Finally, these intermediate buyers would return the farmers' share to

Smallholders selling tobacco to intermediate buyers were usually given a timeperiod when these buyers would return with the smallholders' share of profit. Many smallholder tobacco farmers interviewed during the course of this study surrendered their tobacco in May and were anticipating payment somewhere between August and October. (Generally, smallholders should receive payments shortly after auction floors close, which in 2000 occurred on August 18). Due to possible congestion, transportation breakdown, spoilage, unknown withholding taxes subtracted by Auction Holdings, and unforeseen prices on the auction floors resulting from varying grades of Burley (the most commonly grown smallholder tobacco), Flue Cured, Northern Division Dark Fired, Southern Dark Fired, Sun Air, and Oriental Tobacco, the economic return per kilogram was generally unknown during pre-season. Prices imposed by the transporter were not fixed, and rates were usually established based upon distance to market, size of bale, number of total bales, and individual. Smallholders were at best provided with a rough yet unconfirmed estimate of the payment to be returned.
The Ministry of Agriculture and Irrigation has considering providing an
alternative to the 1999/2000 intermediate buying system due to the failed economic returns. The stability of tobacco as a smallholder cash crop is questionable due to the unknown nature of the system to be in effect in upcoming seasons.
The smallholder role in tobacco production is extremely important on a national
scale, however does not necessarily prove equally vital as a source of household support. Burle'y tobacco is by far the most common tobacco produced by subsistence farmers

Holdings) in 1999/2000. The income actually earned by the farmer following this transaction varies greatly and exhibited little consistency. Considering the nearly annual changes in the formal tobacco market structure, the tobacco market presently holds little reliability as a predominant income-generating activity. Additional crops
Additional significant crops produced within Malawi include groundnuts, pigeon peas, beans, cassava, tomatoes and other vegetables, sorghum, and sugarcane, just to name a few. Many more food sources, such as bananas and mangoes are cultivated. These crops are used for home consumption, trade within or between villages, and with the opening of the market have become available for sale at local markets. There is little price regulation or government incentive for producing such crops, however, they do serve as a crucial component of most households' diets. Fertilizer Incentive Programs
Several programs in recent years have attempted to "introduce" farmers to inorganic fertilizer with the hope that farmers will consequently continue utilizing inorganic fertilizer in order to increase yields. It is relevant, however, that smallholder farmers have been utilizing fertilizer since the 1 960s (Gladwin; 1991). Current programs, then, are best considered according to their ability to actually provide smallholder farmers with inorganic fertilizer, or increased access to such. A few existing programs are explored below.
1. A promotional ADMARC program in 1999/2000 allowed farmers purchasing 1

more than 1 kilogram did not necessarily result in receiving greater amounts of free burley tobacco seed. There was no repayment necessary upon utilizing the program.
2. On a larger scale, the starter pack program distributed 5 kilograms urea and 10 kilograms of 23:2 1:0+4s (along with 2 kilograms of either groundnuts or soybeans and 2 kilograms hybrid maize) to 2.86 million f families during the 1998/1999 and 1999/2000 farming season. The inputs provided were aimed to provide each recipient with adequate maize supply for 0.1 hectares of land. The objectives were: a) to help fill the food gap; b) to promote crop diversification; and c) to promote the concept of soil fertility improvement (Clark, 1999). The starter packs are credited as being a major contributor to the national maize surplus in the year 2000. The Ministry of Agriculture & Irrigation had plans to downscale the distribution by 40% for the 2000/200 1 season. Targeted households were to be selected by the communities themselves, with widows, female headed households, the elderly, and self identified resource poor households being considered priorities. Packs will also include a reduced amount of maize and fertilizer.
3. In early 2000, ADMARC markets began selling fertilizer in 1 kg, 2 kg, 5 kg, and 10 kg increments. This was done in order to accommodate farmers who are unable to purchase large amounts of costly fertilizer. Prices are directly correspondent with dividing a 50 kg bag of fertilizer into the desired allocation.
Organic fertilizer is of minimal use by smallholder farmers in Malawi. Land
constraints in Malawi, particularly in the southern region, appear to be a major constraint to utilizing organic nitrogen sources to improve crop yields (UF Soils CRSP, 2000).

higher population density in Malawi (UF Soils CRSP, 2000). With limited land, the decision to place some land into fallow or an alley crop would take food away from the household, and smallholders would not be able to afford this. Further, if farmers did plant legumes such as pigeon peas, they consumed the peas instead of turning the whole plant under as a green manure (Uttaro, 1998). Additional reasons for minimal use of organic fertilizer include: a) lack of ownership of animals; b) lack of money to pay for transport of animal manure; c) lack of labor for compost or similar exercises. Fanners who do have access to animal manure tend to use the majority on land close to the home (usually dimba land) in order to avoid cost, or additional labor, of transport. Current credit opportunities
The collapse of the credit system in 1994 has left smallholder farmers asking for increased access to fertilizer. Some existing (and previously existing) credit orgaizations lack the ability to monitor recipients, and therefore have created a trend of non-repayment. Limited credit opportunities are, however, available. Below are some examples, not exhaustive, of credit options available to farmers.
FINCA. Providing credit to groups of women (20-30 individuals) who, ideally, are already operating a business, FIINCA has been operating in Malawi since 1995. Currently there are approximately 547 village banking groups utilizing FICA. FINCA provides a group loan, operating on the idea of group guarantee, paid back over a period of sixteen weeks. Groups are selected by village members themselves and are not monitored in any form by FINCA. Although restricted to women, the lack of a

first week after initial payment receipt. There is a non-fixed group interest fee as well that varies with groups (example: receipt of a MK 5000 loan in the southern region results in a one time group interest fee of MK 400.) FINCA prides itself as holding a 98% repayment rate among current members.
ADMARC. Credit programs offered by ADMARC are somewhat restricted. Additionally, there is a lack of trust of ADMARC markets among many smallholder farmers; decreasing the number of farmers who apply for ADMARC credit. Credit opportunities through ADMARC include: a) the Productivity Investment Program (APP) funded by the European Union; b) the Maize Contract Program (implemented in 1998 for large-scale maize production and both small and large-scale tobacco production) funded through ADMARC; and c) the Tobacco Growers Program, funded through ADMARC.
With this, ADMARC hoped to provide inputs to farmers with large amounts of land (greater than five hectares) and subsequently purchase the product from these farmers. The majority of ADMARC credit programs operate with 20% interest requiring payment after harvest, and for programs like the Maize contract program, ADMARC "shall at its sole discretion select suitable commercial farmers to participate in the program" (section 4.1 of contract for Maize Contract Program; ADMARC). Due to the subsequent drop in producer prices of maize, farmers have not been satisfied with the results of many of these programs.
Malawian Political Influence
The influence of political structure affects the decisions and opportunities of

understand the historical trends of the relationship between national policy and the role of agriculture in order to further explore impacts at the household level. Banda s agricultural policy (1964-1994)
Dr. Kamuzu Banda's regime consisted of a single party rule. Only a brief
discussion of his leadership as relevant to agricultural development is included here. Most observers agree that democratic freedoms were suppressed; yet Banda promoted agriculture and the intensification of subsistence crops, not cash crops, on smallholder farms. Smallholder farmers were not allowed to produce burley tobacco; the export crop reserved for large farms constituting the "estate" sector in Malawi. This restriction encouraged a bipolar distribution of land size with many smallholdings owned by the poor and a small number of large flourishing farms in the "estate" sector (Lele 1990).
Banda prioritized development of both infrastructure and agriculture. Banda' s
contribution was marked by construction of numerous roads, draining of marshlands and development of fishing (Virmani, 1992). The Situational Analysis of Poverty, a document produced by Banda's Malawi Congress Party in 1993, identified lack of credit facilities, lack of land, inadequate infrastructure, and weak institutional structures as being crucial in contributing to the perpetuation of rural poverty (Lwanda, 1999 p. 2022).
During this time, the formal market structure continued to cater predominately to large estates and foreign investment. Banda "discarded the principles of Pan-Africanism in order that the atmosphere for investment from the West may be kept clear" (Lwanda,

agriculture in order to maximize exports, particularly of tobacco and sugar, is a subject that has been dealt with comprehensively elsewhere (Pryor 1990; Mhone 1992)). In the last four years (1988-1992) there has been some switch to maize and rice as these two crops became more profitable, particularly in neighboring countries, in the famine ridden middle nineties' (Lwanda, 1999 p. 34).
Muluzi' s agricultural policy (1994-present)
The arrival of the multiparty system in Malawi brought additional shifts in governmental agricultural priorities. Smallholders were permitted to participate in production of burley tobacco with the hopes of rural development and alternative income sources for small farmers were introduced. It was also during these years that additional pressures such as population, soil fertility depletion, and land scarcity worsened.
The long-term results of many of these alterations, such as market liberalization remain to be seen. However, the short-term results have resulted in an alteration of farming or livelihood systems for the rural poor. In research conducted in two villages in Malawi in 1999, 5 1.9% of the 104 households interviewed said they had changed their cropping pattern that particular year (Tsoka and Mvula, 1999). It is unclear whether such changes reflect coping strategies or adaptive strategies; it is clear, however, that new shocks and stresses are resulting in new livelihood outcomes.
The Starter Pack and Voucher Programs It is within this context of enormous changes in product and input markets
during the decade of the 1 990s that we can understand the rationale for the starter pack

legumes), Malawi's food insecure households would experience an increase in annual yields, improved soil fertility, sustainable livelihoods, and ultimately improved food security. However, as stated in chapter two, the program's aims were questionable and its objectives were ambiguous. Was it a safety net program or simply a technology transfer scheme or just another free-inputs program? Implementation of the Starter Pack Program
The starter pack targeted all farming households in Malawi during both the 1998/1999 and 1999/2000 planting seasons excluding the 36,000 estate holders in Malawi (defined as estate holders having cash income from tobacco). 3 The objectives of the starter pack distribution in 1999/2000 were: "a) to assist fill the food gap: b) to promote crop diversification; and c) to promote the concept of soil fertility improvement" (Clark et al., February 2000). Packs were to be distributed to every smallholder-farming household throughout the country. Program coordinators included relatively better off smallholders, or those with more than two hectares of available land. It was assumed that doing so would add "only 12% to the program (while) the cost of excluding them (would) omit from the program many of the most promising farmers, community leaders, and innovative elements" (Mann, 1998). Further, it was assumed that inclusion of these farmers would reduce leakage from resource poor to resource wealthy farmers. The Ministry of Agriculture and Irrigation reported distribution to farming households resulted in a total audience of 2.53 million in 1998/1999 and 2.86 million in 1999/2000.

Starter pack contents
Contents of the starter pack were intended to allow households to plant 0.1
hectare of land. Limiting contents to a production unit of 0.1 hectare was expected to nmimze the perception of the starter pack as a "free inputs" program. Initially intended contents for 1999/2000 starter packs included 10 kilograms of 23:21:0+4s, 5 kilograms urea, 2 kilograms of hybrid maize seed, and 2 kilograms groundnutsi
Maize and fertilizer were included to increase household food production.
Groundnuts were included to improve soil fertility, due to the declining soil organic matter levels in Malawi. Benson (1999) reviewed research measuring soil organic matter comparing data between Blantyre, Kasungu, and Lilongwe agricultural development divisions. Tests sampled the top 15 cm of soil, and found decreasing soil organic matter (Benson, 1999), confirming the need for a soil fertility-enhancing program. Similarly, fertilizer recommendations established by Bensons research provided the foundation for the included amounts of urea and 23:2 1:0+4s. Acquiring inputs
The Ministry of Agriculture and Irrigation provided open opportunity for
suppliers to place bids on costs of providing inputs. Following this, initial suppliers were selected from the proposed bidders. It is unclear as to the particular criteria utilized in selecting these suppliers. In some cases, the decision was not based solely on financial competition, as some suppliers with comparatively low costs were not selected. Ultimately, the Ministry of Agriculture and Irrigation determined the final selection of
su4ler asolos

Inp9ut Su pplier 1: Supplier 2: Supplier 3:
23:21:0+4s Norsk Hydro RAB (3000 Farmers; World
(6000 metric metric tons) (18000 met ric
tons) tons)p
Urea Farmers, World n/a n/a
(6000 metric
Hybrid Maize Pannar Farmers World n/a
(4132.81 metric (1552 metric
tons) tons)
Groundnuts based on local n/a n/a
_____ _____ ____ availability _ _ _ _ _ _ _ _
Figure 3-7: Initially Designated Inputs and Suppliers for 1999/2000 Starter Packs
At the time of pack assembly two important changes occurred. First, it became evident that, due to loss and spillage, additional maize was necessary for distributing inputs to the entire target population. Hybrid maize was not available at this time, so OPY maize was donated by PROSCAP. Second, inadequate crop predictions resulted in inadequate groundnut supply. Additional groundnuts were purchased, and some packs were reassembled with soybeans (figure 3-8).
Efforts to provide worthwhile inputs in a timely manner resulted in the existence of a variety of starter packs. The collaboration of organizations involved in pack assembly and distribution resulted in packs passing through the hands of many individuals (assemblers, re-assemblers, distributors, retailers, non-governmental organizations) before reaching targeted household. This left much room for error and theft. Registered households were not able to anticipate receipt of a starter pack,

In ut Supplier 1: Supplier 2: Supplier 3: Comments
23:21:O+4s Norsk H ydro RAB (3000 Farmers
(6000 metric metric tons)p World (18000
__ __ __ __ _ __ __ __ __tons)_ __ __ m etnc tons) _ _ _ _ _ Urea Fanners World n/a n/a
(6000 metric
_ _ _ _ _ _ _ to n s) _ _ _ _ _ _ _ _ __ _ _ _ _
Hybrid Mz Pannar Farmers World PROS CARP
(4132.81 metric (1552 metric (17.7 metric
tons) tons) tons OPV)
____ ___ ____ ___ donated __ _ _ _ _ Groundnuts Unknown ADM ARC n/a Ina deq uate supply
(1000 metric prompted inclusion of
tons) purchased so ybeans
___ ___ ___ ___ ___ ___ ___ ___ by govern m ent _ _ _ _ _ _ _ _
So ybeans Farmers World Transglobe Zimb~abwe *decision to use
(500 metric (400 metric (2000 metric so ybeans resulted in
tons) purchased tons) purchased tons) WFAD the costly exercise of by government by government purchased 're packing' a number
rof packs
Figure 3-8: Adtual Inputs and Suppliers for 1999/2000 Starter Packs
Limitations in transportation equipment, for example, resulted in distribution
vehicles beginning in the southern region of the country, then moving to the central
region, and finally to the northern region. Logistical difficulties also occurred due to the
hiring of additional laborers to drive vehicles, often resulting in theft of part or all of the
starter packs. The importance in these logistical difficulties lies with the consequent
confusion regarding content of starter packs. Finally, logistical difficulties based simply
on the need to reduce the size of the target population after starter pack registration
occurred (due to MoAI perceived over-registration), resulted in various methods of
eliminating households and consequent insecurity among registered households. There
wa1fe*ra ofuinsronig reiin.tts as.vien frmoer.see

supposed to receive one this year."'5 This household, in fact, was not issued a starter pack during the 1999/2000 season; however did receive one as a gift from a family member of another village, who, though not registered, received one starter pack. Political. context of starter pack program
In order for starter packs to successfully function as a safety net allowing a
technology trial experiment to be carried out by the farmers, and avoid being considered a free input program, political affiliation was to be omitted from the starter pack program. Unfortunately, registration for receipt of a starter pack was carried out in late May 1999, just a few weeks before a national election. Households may have assumed the starter pack registration exercise was a political exercise, or an attempt to sway the voting population. Perhaps because of this, representation of the number of farming households may have been somewhat inaccurate.
Description of Starter Pack Voucher Trial
The 1999/2000 starter pack program included a pilot project designed to distribute inputs to up to 50,000 households through vouchers redeemable at private-sector retail outlets rather than receipt of traditional input packs. Selected registered households did not receive traditional starter packs, rather vouchers to be redeemed at local retailers. On 49,000 of these vouchers, the words "Starter Pack Voucher" were printed. Such vouchers were redeemable at local trading center retailers for only starter packs. On the remaining 1,000 vouchers, the words "Flexi voucher" were printed, along with a star symbol in order to accommodate illiterate recipients. These limited flexi vouchers were redeemable
-tlaltai rrn~ fr nl alrdn nRAf lI(lr -' -g .1 tr -tna tr

with region, but included various combinations of stores including ADMARC, Farmers World, PTC, McConnell, and Chipiku. The logistical operation of the pilot voucher project was handled by the Department for International Development (DFLD), and the project itself designed by DFLD consultants Paul Hamnett and Elizabeth Cromwell (Overseas Development Institute). A previously designed document entitled "Design for a Pilot Voucher Scheme for Starter Pack IF' was followed in designing the utilized methodology.
Malawi Kwacha per USD$1 .00 During Starter Pack
1999/2000 Distribution and Redemption
47.50- ...>
46.50 $.w
Cu 44.50
45.00: 0: 0
445 it' 0) 020--Mlw
S -- --- r" Kwacha/ $1.00
Figure 3-9: Rate of Malawi Kwacha During Starter Pack Distribution and Redemption
The purpose of the starter pack voucher pilot project was to "test the capability of the national retail chains to transport, store and distribute packs to recipients, and to examine the various modalities of distribution" (Killick et al., February 2000). At each of the three selected test sites, number of distributing outlets, timing of voucher distribution, and method of transporting of starter packs to retail outlets was tested.
Implementation of the starter pack voucher and the flexi voucher provided

of cost effectiveness when utilizing a voucher system for distribution, the impact a voucher system can have on fraud related activities, and the impact a voucher system can have on efficiency of distribution. Inclusion of flexi vouchers allowed for potential assessment of the priorities of smallholder farmers identified by smallholder farmers themselves, the impact flexi voucher redemption may have on local retailers, and the financial and social effectiveness of utilizing flexi vouchers to encourage fertilizer use.
The hierarchy within Malawi' s department of agriculture influenced selection of villages to receive both starter pack vouchers and flexi vouchers. The designation process included.many aspects of the ministry of agriculture, from large to small divisions of the agricultural force.
Understanding selection of voucher pilot project areas
Villages in the southern, central, and northern regions of the country were
selected to participate in the pilot voucher project. Selection utilized both geographic and population criteria. Selection of general regions to receive vouchers was reportedly based on proximity (< 10 kilometers) to a trading/voucher redemption center and village size. Selection of households within 10 kilometers of trading centers was intended to reduce household redemption related costs, while selecting villages of similar size to the number of vouchers to be distributed was intended to simplify delivery of vouchers to retailers.
Attempts were made to universally distribute solely one of the two types of
vouchers within a single village.7 Selection resulted in villages surrounding the southern

trading center of Luchenza, the central trading center of Mponela, and the northern
trading center of Mzimba.
The starter pack logistics unit aimed to distribute a nationwide total of 49,000
starter pack vouchers and 1,000 flexi vouchers. Target figures were set at distribution of
20,000 vouchers in the southern region, 20,000 in the central region, and 10,000 in the
less densely populated northern region. Final selection resulted in a target population
including 19,026 households in the southern region, 16,717 households in the central
region, 6,203 households in the less densely populated northern region (total 41,496).
Actual recipients of vouchers are as follows:
__________South Central North Total
Intended SPV 18,6265 9" 16,317"' 6003 40,946
R ecip ients: _ _ _ _ _ _ _ _ _ _ _ _
Actual SPV 18,565 16,069"1 573612 40,370
R ecipients: _ _ _ _ _ _ _ _ _ _ _ _ _
Intended FV 400 400 9200's 1000
R ecipients: _ _ _ _ _ _ _ _ _ __ _
Actual FV 356 412 198 966
R ecipients: _ _ _ _ _ _ _ _ _ _ _ _ _
Figure 3-10: Reported Receipt of Starter Pack Voucher (SPV) & Flexi voucher (FV) Source: Final Report: Implementation of Starter pack Scheme 1999/2000 SPLU; February 2000
The southern region of Luchenza. Blantyre ADD sought to distribute vouchers
near the trading center of Luchenza. Luchenza town itself lies on the junction of three
rural development projects; Thyolo, Mulanje, and Shire Highlands. The Ministry of
8 Number based on total registered families in selected EPA's (19,026) minus 400 flexi voucher recipients. 9 ac of five retailers received 4000 packs + excess of 5% (4200 total) Making surplus 1000. 2000 packs ?icked up in surplus by Funeral & Relief Fund
16,317 familes registered in the selected EPA, Mponela; however 400 should have received flexi
vouchers; leaving a remaining 15,917 to be recipients of SP vouchers. World Vision actually distributed 1,8 voces(up 1 64 12 of the 164--------p------------------------------------------------------------rSP

Agriculture and Irrigation opted not to include Shire Highlands in the voucher trial. Within the two remaining rural development projects, selected subsections -known as extension planning areas- were chosen based on proximity to the town of Luchenza, where recipients were required to travel for voucher redemption. Within Thyolo RDP, two EPAs of Khonjeni and Matapwata were selected. Within Mulanje RDP, two EPAs of Thuchila and Msikawanjala were selected. In total, this included 19,026 registered households. The southern region involved complex organization in designating villages to receive vouchers. This complication was primarily because the town of Luchenza lies on the border of two RDPs, resulting in increased difficulty in geographic division. All households to receive vouchers were intended to be within 10 kilometers of central Luchenza town, although some households were up to at least 12 kilometers away (figure 3-11).
The central region of Mponela. Kasungu ADD initially designated areas surrounding the small town of Dowa to receive vouchers, however due to the low population of farm households in this region the staff was forced to reselect. Choosing to work with the Dowa West RDP, the representatives from the Ministry of Agricultural of Irrigation selected the entire Mponela EPA. Field assistants and staff at Dowa West RDP were extremely helpful in allowing for increased understanding of the selected area due to their strong organization and involvement in the area. Selection of the entire Mponela EPA was expected to simplify the distribution exercise. This included 16,317 registered farm families in 229 villages. Mponela EPA lies within 10 kilometers of Mponela

The northern region of Mzimba. Near Mzuzu, staff at the Mzuzu ADD cooperated with those at the Central Mzimba RDP in selecting the entire area of Kazomba EPA. This selection was made based on the fact that Mzimba town, the largest town and trading center within the area, lies within this particular EPA. Unfortunately, it was overlooked that Kazomba EPA is extremely long and narrow, and therefore not all voucher recipients were within 10 kilometers of Mzimba town and trading center. Some voucher recipients were required to travel up to 40 kilometers in order to redeem vouchers (figure 3-13)."14
Starter pack and voucher project operation
To conduct distribution and redemption of voucher inputs, project coordinators utilized various mechanisms to create fluid distribution. Selected criteria are included here; as these criteria are prioritized by both the Malawi starter pack logistics unit (responsible for distribution) and this research.
Number of distributing outlets. At the southern region pilot location of
Luchenza, five retail outlets were involved in voucher redemption. This equated to approximately 1000 packs being collected from each outlet on each of four days. At the central region pilot location of Mponela only one retail outlet was selected, equating to about 4000 packs per day being collected from the outlet on each of four days. Finally, in the northern pilot location of Mzimba one retail outlet was again selected. Due to the decreased number of packs distributed in the northern region, this equated to distribution of approximately 1500 packs on each of four days.

Program Manager
~RDP's: Thyolo & Mulanje
Thyolo Time: November 7-12 Mulanje
RDP NGO to distribute vouchers: World Vision RDP
subcontracted Khonjeni EPA and Thuchila EPA
--1 Voucher distribution: 18,965 of 19,026 registered
World Vision Funeral and Touch of Faith World Vision
Matapwata EPA Relief Fund Thuchila EPA Msikawanj ala
Khhonjeni EPA EPA
SMagunda Makane S LooaChikuli S Chonde
ScincinSection Section Section
9 Villages 7 Villages 12 Villages 11 Villages 13 Villages
3,562 2,759 5,303 3,186 4,216
families families families families families
Vochr recipient travels
_____ to appropriate retail outlet --t redeem pack
ADMARC Agora Chi piku McConnnel PTC
4,185 4200 4200 4184 4172
Packs delivered to b yMalawi Fertilizer Co. from Limbe ( appx 40 kcm)
Pack Distribution:
Stores not req uired to arrange transport, but req uired to unload trucks.
Each store distributed 4000 (1000 over 4 days (received 5% extra for loss)t.
Packs collected & distributed by, Funeral & Relief Fund.
Fi --,ure4 3-1UdrtadC Luhez Voche Ditrbuio

Program Manager: Mr Khonje
Wezi Girls Educational Foundation- selected to
distribute vouchers.
Voucher distrbtn: 5936/6003 rgstd families in RDP
RDP's: Kazomba
Time: November 28-Dec 1
Wezi Girls EF:
Kazomba EPA
WGEF not normally active in Mzimba; as it is
Mzuzu based; however utilized PA's successfully M zimba Section
Total Villages: 1 Peter Ndawandawa I
Chipikuae treve~ 6300 pak
Packs transported from Kanengo to Mzimba by Chipiku Pack Distribution:
Chipiku stores solely responsible for trasport. Transported 6300 on time Surplus pick up not effective; packs fell to Mzuzu ADD (WGEF)

Program Manager: Mr Khonje
Wezi Girls Educational Foundation- selected to
distribute vouchers.
Voucher distrbtn: 5936/6003 rgstd families in RDP
RDP's: Kazomba
Time: November 28-Dec 1
Kazomba RDP Wezi Girls EF: Kazomba EPA WGEF not normally active in Mzimba; as it is
Mzuzu based; however utilized FA' s successfully Mzimba Section STotal Villages: 1 Peter Ndawandawa
Chipiku received 6300 packs
Packs transported from Kanengo to Mzimba by Chipiku Pack Distribution: Chipiku stores solely responsible for trasport. Transported 6300 on time Surplus pick up not effective; packs fell to Mzuzu ADD (WGEF)

Timing of voucher distribution. At Luchenza, vouchers were to be completely distributed to households immediately prior to the dates designated for redeeming inputs. This would leave vouchers in the hands of the recipients for a maximum of five days. At Mponela voucher distribution was planned to overlap pack collection by two days, leaving vouchers with recipients for a maximum of three days. At Mzimba, recipients were to collect their packs the day after receiving their vouchers, meaning households were to have vouchers in hand for only one day. Coordinators planned this variation in distribution in order to allow for future evaluation of alternative voucher use, such as buying and selling, as related to the amount of time the vouchers were actually in the hands of household members.
Registration processes. Registration for starter packs occurred in late May 1999. Because this was only a few weeks before a national election, planners feared that households affiliated this registration process within a political context. Starter pack registration was carried out with the cooperation of staff from local rural development projects, area field assistants, and village headmen. As the registration occurred nationwide, all areas utilized available personnel and resources to complete registration in a timely manner. This resulted in the existence of a variety registration processes. Differences among villages, local political structure, and local agricultural staff further exacerbated variations in registration processes. Many households were dissatisfied with registration procedures. Various registration methods reported by households interviewed in this research include great variety and include approaches in which:

3. FA registered households by traveling door to door without headman.
4. FA registered households according to headman's recommendation.
5. FA conducted registration in company of political representative (by traveling door to door).
6. FA not involved in registration process; headman or headman representative registered households and reported to FA.
7. Headman registered households through meeting at central village location; with no FA present.
8. Headman registered households by traveling door to door.
9. Headman registered households by sending selected representative door to door.
10. Headman conducted registration in company of political representative.
11. No registration process.
12. Registration records from 1998 utilized (either in place of 1999 registration, or after conducting 1999 registration). 13. Registration qualification dependent upon work program within village (per headman).
14. No FA assigned to area at time of registration; villages either did not register or headman utilized any previously described method. 15. Headman not involved in registration process; FA utilized any previously described method.
Responsibility of distributing starter pack inputs to villagers. Traditional
starter pack distribution utilized the assistance of Ministry of Agriculture and Irrigation
staff at all levels of government agricultural organizations, the Malawi Army, selected
retail outlets, and numerous non-governmental organizations. In short, the large task of
distribution was extremely cooperative. In some areas packs were distributed to villagers
at a central location, usually a school or community center. In other areas, household
members were responsible for obtaining starter packs from local retail outlets.
Transporting the packs was a difficult task, and required retail outlets, such as Farmers
World, to employ additional labor (e.g. truck drivers) and government assistance (e.g.
army trucks). Noteworthy but unsubstantiated complaints related to distributing packs to

1. Additional, part-time, labor hired by retail outlets led to increased
theft and sale of starter packs, or starter pack contents.
2. Utilization of additionally allocated vehicles was unevenly
distributed: with vehicles moving from southern to central to northern regions upon completion of each. The stated result was an untimely distribution in many regions.
3. The role of non-governmental organizations (NGO) in distribution was often over emphasized, leading to transportation difficulties within NGO's; again resulting in untimely distribution in some areas.
4. High number of networks necessary for distribution of bulky packs led to increased opportunity for breakdown; perhaps resulting in increased theft and sale of starter packs, or starter pack contents. 5. Due to organizational difficulties, a number of starter packs were assembled and transported without any legume. The Ministry of Agriculture allocated additional legumes (often requiring additional packaging or additional transportation) to necessary ADDs. Area dependent, some ADDs utilized their own labor for this second distribution, while some requested assistance from the NGO assigned to starter pack distribution in the designated area. 15
Usually, members of a locally designated NGO traveled to the village, often
working in collaboration with local field assistants or village headmen in order to
distribute vouchers to registered households.
In correspondence with the seasonal calendar, the end of November 1999 was
identified as the target completion date for starter pack distribution. Actual completion
dates for distribution of starter packs is debatable. According to various-nongovernmental organizations and farmers, smallholder farms had obtained all packs by
early January. Results from the Ministry of Agriculture and Irrigation (MoAI) show all
packs distributed to recipients by the target date in late November.
Timeliness of starter pack vouchers and flexi vouchers was somewhat easier to
monitor, as recipients were granted only a roughly three-day duration in which to redeem
vocs VoceSeiinsana tohv rcie vocs v th taee daes

however, there were cases of retailers being inadequately stocked with packs or goods for redemption thereby limiting voucher recipients' choices. Responsibility of distributing starter packs to retailers. Voucher recipients, and in some areas starter pack recipients, were required to travel to retail outlets in order to receive inputs, making the task of adequately transporting goods to retailers extremely important. Responsibilities differed somewhat by region, but were shared between staff at various levels within the Ministry of Agriculture and Irrigation (agriculture development divisions, rural development projects, etc.), contracted NGOs, and retailers (figures 3-11 through 3-13). Generally, selected stores received starter packs free of charge, and were then responsible for appropriate storage and distribution.
At Luchenza the starter pack logistic coordinators assumed responsibility for
transporting packs to the retailers. This model simply tested the ability of retailers with varying storage capacities to receive, unload, and store packs prior to distribution. At Mponela, retailers were responsible for transporting the high volume of packs to outlets. At Mzimba, retailers were responsible for transporting this smaller volume of packs to outlet (Killick et al., February 2000).
Measures taken to ensure efficient voucher redemption. All vouchers included appropriate redemption information printed directly on the front. This information included valid redemption dates, redemption center(s), indication of the type of voucher, and flexi vouchers included the amount of value (MK450.OO). Security measures were considered in organizing the redemption process for both starter pack and

In each area, vouchers were color coded as a means to indicate the designated
redemption date. Particular color vouchers were redeemable only on particular dates, and limited amounts of each color voucher were distributed regionally. Additionally, recipients were to be reminded by headman, field assistant, or NGO that particular colors were only redeemable on particular days.
Some distribution areas experienced confusion in designating an individual
responsible for instructing recipients of this color-coding system, and many villagers had to make multiple trips to redeem their vouchers. Fortunately, retail outlets were accommodating, and often stayed open additional days to accommodate those who were unable to redeem vouchers because they arrived on the wrong date.
The second measure utilized to assist in functional voucher redemption was the use of a serial number system. Each voucher was marked with a unique serial number. This assisted in tracking vouchers and served as a tool reducing distribution center crowds. In regions where more than one retail outlet redeemed vouchers, each outlet was responsible for a different block of numbers, eliminating the potential problem of all voucher holders attempting to redeem vouchers fr6m the same retail center. In some areas, households receiving vouchers believed this numeric verification system allowed retailers to verify the registered voucher 'owner' with the individual redeeming the voucher. Considering that selling vouchers was illegal, this may have decreased fraud surrounding vouchers.
The third tool used to reduce error in distributing vouchers was a carbon copy
voch Vochr cosse ofl .he ncoies: twoofwh ch thhuehl surnee to

and returned the second copy to the Ministry of Agriculture and Irrigation (MoAI) in
order for the ministry to total all redeemed vouchers. Each copy included a unique serial
number to decrease fraud. In some areas, duplicate copies were not properly attached
bound, and amidst large crowds at redemption centers villagers often obtained additional
packs by paired remaining receipts of different serial numbers.
Despite vast differences in coordination and implementation, the Makina
household provides a perspective of one starter pack experience.
The Makina Houseohid Starter Pack Experience
As residents of Chimwanga village in the southern region of Malawi, the Makina household was selected to receive a flexi voucher. Registration process was carried out by a local village chairman for the United Democratic Front (UDF), and a representative from the Ministry of Health. The two individuals registered households by traveling door to door. At the time of registration, Cecilia's husband was not home, so Cecilia herself registered. The flexi voucher was obtained from the village headman, in the presence of a local NGO, at a central location only a couple of kilometers from Cecilia's home. Cecilia felt lucky to have received a flexi voucher over a starter pack as it guaranteed receipt of goods.
Members from the NGO present explained the "flexi" concept of receiving either goods or a starter pack to recipients at this time. The color-coding identification of the vouchers, which was intended to specify valid redemption dates, was unfortunately not explained to recipients at this time. The September 1999 passing of the Chimwanga field assistant left the villagers without a field assistant during starter pack distribution. At the designated redemption time, Cecilia was suffering from pregnancy complications, so her husband traveled to redeem the voucher. Rafael traveled to ADMARC, which he assumed would be the least congested retail outlet. Unfortunately, Rafael was turned away because the voucher he had was color -coded for a different day. Rafael's second trip to ADMARC was equally unsuccessful, however on the third day he was able to redeem the flexi voucher for a starter pack, as was the most common choice within the southern region during the 1999/2000 voucher trial. Cecilia says she opted to redeem her voucher for a pack, rather than goods, because she was in need of food. Similarly, she opted not to sell her voucher or pack because her primary concern was food rather

Anticipating the Impact of the Starter Pack: Household Considerations
Agricultural production and subsequent household food security in Malawi have proven vulnerable to recurrent devaluations of the Malawi Kwacha (MK), depleted soils resulting in poor yields, collapse of the credit system, increased population (resulting in land constraints and higher population densities), instability within the tobacco market structure and regulations, imperfect markets, and insecure health conditions, particularly as a result of chronic malnutrition and the frequency of HIV/AIDS within Malawi. These obstacles led to a variety of options at the household level, upon receiving inputs. Based upon the described systems existing among Malawian smallholders and the opportunities and barriers presented to them, a variety of factors are considered in exploring the criteria for utilizing starter pack inputs. For purposes here, factors potentially influencing utilization of starter pack inputs are considered to be based upon financial considerations
- such as national economic considerations, access to credit, land availability, labor availability: social considerations- such as regional distribution of ethnicity, regional distribution of wealth, gender division, traditional cultures; and of course agricultural considerations- such as suitability of starter pack inputs, soil fertility issues, and correspondence of inputs with the farming calendar. Financial Considerations Involved in Decision Making
National economic forces influencing availability of credit, prices of purchased household items, and established prices of institutions purchasing from smallholders (such as parastatals like ADMARC) impact the sustainability of a household. Household
disins suha the deiso to utlz *tre oak0ot invrosmnes

Devaluation of currency and inflation of goods
National economic forces, such as the devaluation of the Malawi Kwacha (MK) have created extreme shocks and stresses to livelihood activities. During starter pack distribution and voucher redemption, the Malawi Kwacha was at approximately MK 40.00 (figure 3-14 and 3-15). As the MK continuously devalued, little or no correspondent increase in paid wages was in effect. Taxation on paid employees throughout Malawi was reduced in 2000 from 38% to 35%; displaying little impact on subsistence farmers. Similarly, no decrease in input cost or household items occurred.
Malawi Kwachal $1.00 USD
90- l
a0 ......................
3 0
-.-4--- Malawi Kwacha/ $1.00 USO Date Figure 3-14: Rate of Malawi Kwacha (MK) to US Dollar (USD) in January 1995 January 2001
August 1999
5 0 _.. ._ ....
30 ......5.8.....
30 4
1 0 .......... ........ ..
8/20 8/21 8/22 8/28/4 8/25

Income distribution
Targeting economically stressed households in Malawi has historically been an
extremely difficult task. Criteria determining households in need of assistance are often determined by a variety of criteria rather than utilizing consistent criteria td determine trends. Additionally, the lack of research conducted during the Banda regime has given Malawi only a few years to define the accurate economic status of smallholders. Finally, the number of impoverished households is simply so high in Malawi that even that upon clear identification of a population in both need and want of intervention, resources are not necessarily available in adequate abundance.
Systematic identification of economically impoverished households from
previously conducted research considered here according to three different units, household consumption patterns, extension planning area comparisons (regional), and self perceived indicators of wealth.
The Integrated Household Survey, conducted by the National Economic Council (NEC) measured caloric and protein intake of households nationwide, providing results signaling those households under severe economic stress. Comparisons among households were intended to identify relative economically impoverished areas based on household consumption patterns. Preliminary findings from NEC (August 16, 2000), for example, show individual poverty headcount (%) to be comparatively high in the districts of Mwanza, Ntcheu, Thyolo, and Phalombe.
Second, The Ministry of Agriculture and Irrigations comparative evaluations of

Malawi during the 1998/1999 season include six EPA's within Kasungu agricultural development division, five within Machinga agricultural development division, two within Lilongwe agricultural development division, and one within theSalima development division (Appendix D)
A third methodology of analyzing regional differences in income distribution is through direct involvement of households themselves. Household members were asked to compare signs of material wealth as potential wealth indicators. Communities generally establish such indicators over time, and sub-cultural understanding of these indicators often results in establishment of community hierarchy. Utilizing these indicators can assist in classifying households into domains, allowing for further comparison of decisions surrounding utilization of staffer pack and voucher inputs. Potential wealth indicators for households in the northern, central, and southern regions of Malawi can include ownership of cattle, chickens, pigeons, guinea fowl, etc., outside employment, household size, soil quality, access to fertilizer, access to credit, available labor, ownership of a bicycle, ownership of a business, and health of household members. Indicators denoted here include credit, availability of land, and available household labor.
Credit. The aforementioned credit situation of the Malawian smallholder farmer is of relevance here. The desire for inorganic and organic fertilizer, increased access to credit, and a stable market economy created the need or desire for some intervention assistance. All recipients have traditionally, of course, been subjected to numerous shocks and stresses that create unique situations, and in effect determine the effectiveness

Available land. The amount of available land can seriously limit the amount of food produced and the food security of a household. Similarly, the amount of land available to a household can play a significant role in a household's decision to plant, sell, trade, consume, or sell starter pack inputs. The amount and type of land allocated to a particular household can be considered the result of a number of influences including population density within a geographic region. The less densely populated northern region, for example, exhibits greater (adequate) land allocation in the northern region, which can in increase options surrounding utilization of the starter pack or voucher inputs.
Significantly densely populated areas exhibiting frustration created by land
constraints include the southern region, as well as villages or households located close to urban centers, such as Blantyre (south), Lilongwe (central), and Mzuzu (north). Remember that the starter pack voucher and flexi voucher areas selected by the Ministry of Agriculture and Irrigation were geographically close to trading centers; the significance of land constraints may be severe in utilization of inputs as intended by the starter pack program. Other potential differences accounting for land distribution in a particular area include: culture, relationship to village headman, accessibility of local industry and off-farm employment, and political influence.
Results of land constraints can be devastating to production and soil fertility.
Land shortages appear to be the major constraint to using organic nitrogen sources to improve crop yields (UF Soils CRSP, 1997). With limited land, putting some land into

afford. Further, if farmers did plant legumes such as pigeon peas, they consumed the peas instead of turning the whole plant under as a green manure (Uttaro, 1998).
Labor constraints and household size. The Ministry of Agriculture and Irrigation assumes an average of five persons per household in rural households. According to the Statistical Booklet on Poverty (NEC, January 2000) the dependency ratio, or "the ratio of total number of children under 15 years of age and people over 65 years to the number of adults in the economically productive ages of 15-64" is 1.3, indicating a large dependent population.
Household size is important in understanding potential labor constraints and
variations in labor divisions. Female labor constraints can be considered a contributing factor to the low yields experienced by women. Women may not have necessary time to spend in their own gardens in order to produce higher yielding crops (Bembridge, 1987). Financial Considerations of the Makina Household
Within the Makina household, the majority of household goods and school related materials are purchased by Rafael with money earned from selling dimba crops. Although Rafael is 'not very interested' in agriculture, he sells crops at Luchenza trading center. The majority of crops are produced in small quantities in Cecilia's dimba. Cecilia and Rafael have determined thal bananas, mustard, and eggs earn Cecilia income. Though contents of the starter pack are none that provide Cecilia with direct income, she utilized the contents for production. Without access to fertilizer, the starter pack was desirable; however the Makina household did not receive any yield from starter pack maize. ADMARO's decision to begin selling fertilizer in 2 kg, 5kg, and 10 kg increments, with directly proportioned prices could become an option for the Makina's.
Cecilia requires 75 kg of maize per month to feed her household, obtained In large part from her own production and earned through her and Precious' ganyu work. Additional maize is purchased by Rafael. Cecilia estimates that Rafael spent a total of MK1 20.00 during 1999 on school related costs. WAith the exception nf transporting maize tn the mill, npaying for grindling, and