Mozambique 
1999-2000 Floods
Impact 
Evaluation: 
Resettlement 
Grant Activity
Emergency 
Recovery: 
Agriculture and 
Commercial 
Trade (ER: ACT)
U.S. Agency for 
International 
Development 
(USAID)
Abt Associates Inc.
(with Afrisurvey and 
Caresoft Lda.)
Agricultural Policy 
Development Project 
(APD)
USAID Contract No. 
PCE-I-00-99-00033-00
Task Order No. 803
July 2002
Mozambique Flood Resettlement Grant Activity i
PREFACE
Mozambique has a population of 17.2 million people, and a land area of 789 800 square 
kilometers. Although 45 percent of the land is considered suitable for agriculture, only four 
percent is presently cultivated. Over 81 percent of the labor force is engaged in some type of 
agriculture production, which reflects limited employment opportunities in the non-farm 
sectors despite the growth of the manufacturing sector in the past decade. Historically, 
Mozambique has been a major producer of cash crops including sugar, copra, cotton, cashew 
nuts, tea and tobacco. These activities declined during more than 15 years of civil strife that 
devastated the country's economy, following independence in 1975 (FAO, 2000).
After a prolonged civil war, the peace accord signed in October 1992 and the first multiparty 
elections in 1994 led to the current period of relatively stable government and buoyant 
economy accompanied by increased growth.
Under agreements with the IMF and the World Bank, market-based economic policies, 
including far-reaching structural reforms, have been implemented. In addition, the country 
has benefited from a foreign investment boom. Inflation has declined from an average of 50 
percent in 1991-95 to 2.9 percent in 1999. The value of exports has almost doubled from 
1995 to 1999 reaching a level of US$300 million. The exchange rate has stabilized over the 
same period. GDP growth has been particularly impressive, averaging 9.5 percent annually 
in 1996-1999 (EIU, 2001). 
Despite these gains, the country remains one of the poorest in the world, with a GDP per 
head of US$256 in 1999. In late 1999 and early 2000, economic progress was halted by
disastrous flooding created by heavy rains and cyclones, hitting hardest the five most 
southern provinces. The immediate focus by governments was on saving lives. The 
Government of Mozambique mobilized rescue resources from its neighboring countries and 
international donors (including the US Office of Foreign Disaster Assistance). 
Following this emergency response, the Government of Mozambique and many donors 
developed projects – some with objectives for short-term recovery and others for long-term 
sustainability – to assist the affected families living in the flood areas. The subject of this 
evaluation is USAID’s post-emergency recovery project – a unique Resettlement Grant 
Activity, not designed as a traditional development project, but rather to directly assist flood￾affected families in the recovery of their livelihoods.
This impact evaluation was carried out from November 2001 to May 2002 by a team 
contracted by USAID through the Agricultural Policy Development Project (APD) with Abt 
Associates Inc. The team included Afrisurvey (social science research firm), Caresoft Lda 
(database design and processing firm), Esther Kazilimani-Pale, and Liv Bjornestad. The 
entire team, including the 20 survey field supervisors and survey interviewers identified in 
Annex 3, is acknowledged and thanked, with particular gratitude directed to Ms. Ercilia 
Santos of Afrisurvey for her skill, dedication, and attention to detail. 
The team benefited from the experience and insights of those involved in the implementation 
of the resettlement grant program, most particularly Mr. Carvalho Neves who led Deloitte 
and Touche’s Project Management Unit.
Mozambique Flood Resettlement Grant Activity ii
We thank Ms. Christine de Voest, USAID/Mozambique Rural Enterprise Officer, for her 
guidance throughout, and Gale Rozell and Sarah Gavian for their constructive critiques of the 
draft report. 
To those Mozambicans who suffered in the floods, we are pleased that the resettlement 
grants contributed in some small way to an improvement in their lives.
John Miller
Team Leader
July 2002
Mozambique Flood Resettlement Grant Activity iii
TABLE OF CONTENTS
PREFACE...............................................................................................................i
TABLES, FIGURES, AND BOXES .....................................................................iv
ACRONYMS AND TERMS..................................................................................v
MAP………………………………………………………………………………..vi
ABSTRACT…………………….……………………………………………..…..vii
CHAPTER 1
BACKGROUND AND INTRODUCTION ............................................................1
CHAPTER 2
BENEFICIARY IDENTIFICATION PROCESS....................................................7
CHAPTER 3 
HOUSEHOLD IMPACTS ...................................................................................14
CHAPTER 4 
ECONOMIC EFFECTS .......................................................................................19
CHAPTER 5 
CONCLUSIONS AND OBSERVATIONS ..........................................................34
Annex 1. Reference Documents.........................................................................A-1
Annex 2. Persons Interviewed ...........................................................................A-3
Annex 3. Evaluation Team ................................................................................A-4
Annex 4. Survey Estimates and Sampling..........................................................A-5
Mozambique Flood Resettlement Grant Activity iv
TABLES, FIGURES, AND BOXES
List of Tables
Table 1-1 Direct Loss in the Agriculture and Livestock Sectors (US$ millions) 2
Table 1-2 Number of Surveys Conducted, by Province and Survey Type 6
Table 2-1 Village Chiefs Understanding of Selection Criteria 10
Table 2-2 Destruction of Small Farms 11
Table 2-3 Second Harvests 12
Table 2-4 Grant Recipients Spending Time in a Resettlement Camp 13
Table 3-1 Percent Use of Grants, by Province 17
Table 3-2 Percent Use of Grant, by Where Spent 18
Table 4-1 Inflation: Consumer Price Index, 1996-2000 (%) 20
Table 4-2 Retailers Reporting That More Money Was Spent During This Period 22
Table 4-3 Retailers Response to Types of Products Purchased by Grant 
Recipients 22
Table 4-4 Use of Grant, All Five Provinces 23
Table 4-5 Retailers Response as to Whether There Were Price Changes After the 
Floods 24
Table 4-6 Retailers Opinion of Price Changes After the Floods 24
Table 4-7 Retailers Opinion as to Reason for Price Change 25
Table 4-8 Price Increase Due to the Grant Money 26
Table 4-9 Real GDP Growth, 1997-2001 26
Table 4-10 Retailers Response as to What They Did With Money Earned 29
Table 4-11 GDP Per Capita, 1998 32
Table 5-1 Comparisons of Circumstances in Cash-Grant Programs in Selected
Country Disasters 38
List of Figures
Figure 2-1 Effects of the Floods 12
Figure 3-1 Percent Use of Grant, All Five Provinces 16
Figure 4-1 Monthly National Inflation Rate, 1999-2001 (in percentage points) 20
Figure 4-2 Use of Grant, by Where Spent 27
List of Boxes
Box 1 Grant Recipient Comments: The Grant Amount Was Not Enough 14
Box 2 Purchasing Patterns: Household Needs 15
Box 3 Purchasing Patterns: Personal Use 15
Box 4 Purchasing Patterns: Housing 16
Box 5 Purchasing Patterns: Agricultural Needs 29
Box 6 Grant Recipient Comments: Grants Were Used for Future Productive 
Earnings 30
Box 7 Grant Recipient Comments: Future Assistance is Expected 34
Box 8 Grant Recipient Comments: Cash is Better Utilized Than
Commodities 34
Mozambique Flood Resettlement Grant Activity v
ACRONYMS AND TERMS
BCI Banco Comercial e de Investimento (Commercial and Investment Bank)
CTA Confederação das Associações Económicas de Moçambique (Economic 
Associations of Mozambique)
DINAGECA Direcção Nacional de Geografia e Cadastro (National Directorate of 
Geography and Mapping)
ER: ACT Emergency Recovery: Agriculture and Commercial Trade Program
FDC Fundação de Desenvolvimento da Comunidade (Foundation for Community 
Development)
GRM Government of the Republic of Mozambique
MIC Ministry of Industry and Commerce
MZM Mozambican Meticais
NGOs Non-Government Organizations
PC Program Committee
PMU Project Management Unit (Deloitte and Touche, contracted by USAID)
STAE Secretariado Técnico de Administração Eleitoral (Technical Secretariat of 
Election Administration)
USAID United States Agency for International Development
lobolo dowry
Mozambique Flood Resettlement Grant Activity vi
Mozambique Flood Resettlement Grant Activity vii
ABSTRACT
In response to the1999-2000 floods in Mozambique, USAID’s Resettlement Grant 
Activity helped families return to their homes and farms and rebuild their lives, while 
at the same time, jump-starting economic activity, re-establishing effective demand and 
supply of goods and services.
Cash grants were distributed from December 2000 to April 2001 to more than 106,000 
rural families. Each head of household received 1,500,000 meticais, about US$92. 
The distribution of US$9.7 million was accomplished at 167 distribution sites that 
encompassed over 730 villages in 30 districts within the five affected provinces. 
Grants were delivered to the woman of the household. Families were able to chose for 
themselves which goods or services, or even savings were their highest priorities. 
Grant recipients for the most part met criteria based on, among others, location and 
damage to homes and crops. Beneficiary identification involved a village by village 
registration process, collecting names, inventorying losses, and verifying information 
with village elders. 
Grant recipients were poor, largely subsistence farmers, living in rural villages near 
their own fields. Many recipients live in areas often affected by extreme natural 
conditions. They suffered in the floods then, and are now suffering with drought. 
Grants were primarily spent on household goods (e.g., dishes, pots, pans, blankets), 
clothes, and livestock. The money was spent mainly near local distribution points, and 
thus remained in the region, stimulating sales and job creation by retail traders. 
Food prices increased substantially during and after the floods, but food inflation 
effects of the program were minor, given the tendency by households towards 
purchases other than food. The main cause of the marginal local inflation was the 
scarcity of products, highlighting the importance of accompanying such cash programs 
with complementary programs on the supply end. 
The program contributed to the revitalization of distribution networks in affected areas. 
The extra income that trickled up to local retailers in the form of increased business 
allowed them to restock their stores and repair damages caused by the floods.
The fundamental principle of a cash grant program – that without any conditions 
attached, households would make prudent use of the money – was confirmed. The 
view that women manage the money and choose the family priorities was also borne 
out.
The program helped stabilize affected households by providing them with income lost 
during the floods and allowed homes to be re-established, essential needs to be met, 
and productive, income-generating activities to be re-started. Those most severely 
affected by the floods were assisted, and in the absence of the grant program, would 
have been worse off. At the same time, however, the program reinforced a sense of 
expectation by villagers that others may be depended upon for their survival.
Mozambique Flood Resettlement Grant Activity 1
CHAPTER 1
BACKGROUND AND INTRODUCTION
The Floods. The floods that occurred in Southern and Central Mozambique in late 
1999 and early 2000 resulted in the displacement of 500,000 people, severely 
damaging housing, agricultural infrastructure, public buildings, schools, hospitals, 
water and energy supply systems, roads networks, railways and telecommunications. 
These losses represented an enormous setback for the Mozambican national economy 
and for the efforts achieved in the area of poverty reduction (World Bank, 2000).
According to initial post-flood assessments, the direct and indirect losses to 
Mozambique’s economy amounted to a significant US$600 million (more than 
double annual export earnings) including: losses of assets (direct costs), US$273 
million; reduced production (including the reduced stimulus to the economy), 
US$247 million; reduced exports, US$48 million; and increased imports for 
consumption, US$31 million (World Bank, 2001). Later, the real cost of replacing 
infrastructure greatly exceeded the estimated losses.
“The magnitude of the disaster affected economic activity in such a profound way –
with particular impact on agricultural and industrial production – and over such a 
large area, that the macroeconomic impacts in 2000 were enormous” (World Bank, 
2000). The floods resulted in a sharp fall in GDP from 7.5 percent in 1999 to 1.6 
percent in 2000, inflation reached a high of 12.7 percent in 2000 as compared with 
2.9 percent in 1999, and the exchange rate depreciated sharply at an annualized rate 
of 28.2 percent in 2000 from a rate of 7.7 percent in 1999 (Mozambican State Budget, 
2002).
The provinces most affected by the floods were Maputo, Gaza, Inhambane, Sofala 
and Manica. The total population in the five affected provinces is roughly five 
million people. According to the Instituto Nacional de Gestão de Calamidades 
(INGC), nearly two million people experienced economic losses. Breakdowns in the 
transportation system – resulting from destroyed roads, bridges and railways –
separated people in the affected areas from food, water and essential services (World 
Bank, 2000).
Flood and cyclone damage affected substantial areas of agricultural production in 
southern and central Mozambique, resulting in crop and livestock losses and damage 
to agricultural infrastructure and equipment. World Bank estimates suggest that 
losses reached almost US$58 million in the agricultural sector and US$8 million in 
the livestock sector. Of this, 47 percent resulted from smallholder losses, mostly in 
annual crop and livestock losses. Livestock losses were estimated at 20,000 cattle, 
4,000 goats, sheep and pigs, and 180,000 chickens. The cattle losses alone were 
feared to reverse the recovery of the pre-war herd. Smaller animals virtually 
disappeared in the flooded areas, removing a valued source of cash income, savings, 
and nutrition from the small farmer (World Bank, 2000).
The flooding devastated the affected areas (about 12 percent of the cultivated land 
and 90 percent of irrigated land in the five provinces) and caused considerable loss of 
life and property. The flooding covered planted areas, which mostly supported maize 
and rice. The largest impact was in Gaza (accounting for 43 percent of the flooded 
cultivated land), followed by Maputo (31 percent) and Sofala (18 percent). Assuming 
Mozambique Flood Resettlement Grant Activity 2
constant yields across hectares, up to 21 percent of expected agricultural production 
in the affected area was lost (World Bank, 2000). As shown on Table 1-1, flood and 
cyclone damage affected substantial areas of smallholder agricultural production in 
the flood-affected region of Mozambique, resulting in crop and livestock losses and 
damage to agricultural infrastructure and equipment.
Table 1-1. Direct Loss in the Agriculture and Livestock Sectors (US$ millions)
Smallholder
Annual crops in cultivation 20.95
Livestock 7.90
Capital assets 3.06
 Productive infrastructure 2.17
 Tools 0.89
Total 31.91
Source: MADR and Agricultural Provincial Directorates (DPADRs).
The Resettlement Grant Activity. Given the magnitude of the damage in a country 
as poor as Mozambique, USAID wanted to maximize the impact of resources it 
sought from the US Congress. USAID designed the Emergency Recovery: 
Agriculture and Commercial Trade Program (ER: ACT) as one part of a larger 
reconstruction program effort. ER: ACT1
 sought to: 
· help families devastated by the floods and cyclones return to their homes and 
farms and rebuild their lives, and
· jump-start economic activity to re-establish effective demand and supply of goods 
and services in affected areas.
Cash grants were provided directly to families whose homes and farms were 
destroyed in the floods. The grant was delivered to the woman of the household, 
primarily because it was likely that she would utilize the money in the best interests 
of the whole household. Families were able to choose for themselves which goods or 
services, or even savings, were their highest priorities in the aftermath of the floods. 
The grants were intended to generate local demand for goods and services and thus 
stimulate renewed economic activity.
USAID’s hypothesis at the time was that working on the supply side of economic 
activity alone via credits was insufficient to re-establish economic networks. A direct 
 
1
 ER: ACT had two other relevant, complementary components intended to increase the supply of 
goods and services in the flood damaged communities through the private commercial trading sector in 
the shortest possible time. First was a project designed to make short-term credit available for retailers 
and wholesalers to increase stocks in anticipation of increased demand from grant recipients (Inventory 
Credit Activity). Second was a project providing low-interest, term loans to rural commercial 
enterprises to repair facilities and equipment damaged or destroyed during the floods (Rural Enterprise 
Credit Activity). When it became clear in late 2000 that credit needs of wholesalers did not differ 
significantly from credit needs of other flood-affected enterprises, USAID combined the two credit 
programs into one, with the principle that both short term and long term credit could be provided. 
Mozambique Flood Resettlement Grant Activity 3
cash transfer program would have the fastest impact on recovery of the economy in 
the shortest amount of time and at a cost comparable or less than other traditional 
emergency food programs. This would provide immediate liquidity to families, 
thereby stimulating local markets and jump-starting the local economy.
The Resettlement Grant Activity distributed cash grants to slightly more than 106,000 
rural families. Each head of household received 1,500,000 meticais in a single 
payment, about US$91 to US$93, depending on the fluctuating exchange rate during 
the December 2000 to April 2001 distribution period. 
Originally expected to take place between August and October 2000, implementation 
was delayed by four months due to (a) US Congressional approval for supplemental 
funds; (b) the extra level of effort to establish the required beneficiary list; and (c) the 
complex logistic planning and security required for a distribution of this nature. 
Implementation took approximately four months and the last grant distributions were 
completed at the end of April 2001.
A Project Management Unit (PMU), led by Deloitte and Touche, was contracted for 
the organization and fieldwork associated with the Resettlement Grant Activity. The 
PMU’s work occurred in two distinct phases – beneficiary identification and cash 
distribution.
The principal challenge in identifying beneficiaries was to overcome the enormous 
information gaps that existed regarding affected households. This required a village 
by village registration process whereby names were collected, lists compiled, losses 
inventoried, and all then verified by village elders. About 220,000 potential 
beneficiaries were registered and surveyed in this manner, documenting the extent of 
damages to homes and crops, and the levels of assistance previously received. 
The second challenge during the beneficiary identification process was the 
establishment of eligibility criteria that were fair and transparent as well as 
operational at the field level. Based on information gathered by the field teams 
during the registration process, USAID and the Government of Mozambique (GRM) 
developed a set of criteria that were then tested during a pilot phase, refined, and 
became the basis for subsequent distributions; criteria are identified in Chapter 2. 
From close to 220,000 households registered, 114,432 recipient beneficiaries were 
ultimately determined to be eligible. 
The actual cash distribution represented an enormous planning and logistical 
challenge. The Program Management Unit developed procedures and signed 
agreements with public and private entities for transport, communications, security 
and banking services. Distribution teams were organized and traveled by road and 
helicopter to distribution sites. Advance teams were deployed to inform villages of 
the distribution dates and enlist village chiefs in organizing the recipients. USAID’s 
correspondent bank, Banco Comercial e de Investimentos (BCI), organized the 
delivery of pre-prepared cash packets and tellers to distribution sites. 
On distribution days, the village chiefs assisted the teams in organizing lines of 
eligible heads of household. A color-coded identification ticket was issued to each 
eligible, registered recipient and then a bank check. Each individual’s index finger 
Mozambique Flood Resettlement Grant Activity 4
was dipped in indelible ink. Recipients were then guided to representatives of the 
bank where they were able to cash their checks immediately. A detailed security 
program was organized with a local security firm contracted to provide those 
services. 
The distribution of US$9.7 million in local currency cash grants was accomplished 
within a four-month period at 167 distribution sites that encompassed over 730 
villages in 30 districts within the five affected provinces. On average over 1,100 cash 
grants were distributed daily in 94 distribution days. In total, 106,280 resettlement 
grants were distributed, representing 93 percent coverage of the eligible households; 
the remaining seven percent could not be located at the time of distribution. 
The Impact Evaluation. USAID believed that the economic stimulus of a direct 
cash grant program for flood victims would have a significant positive impact on 
restoring local economic activity. This Impact Evaluation is intended to test this 
hypothesis, assessing the impact of the resettlement grants on the lives of the 
recipients and on the economy of the affected areas.
The Evaluation is based on a review of documents (see Annex 1), a series of 
interviews with those involved in the project (see Annex 2), and on four sets of in￾depth sample surveys carried out in January and February 2002 with 630 households, 
44 village chiefs, 96 retailers, and 11 wholesalers.
Survey Design. Registration lists and baseline data from two surveys undertaken by 
the PMU included a detailed pre-grant survey of 3,500 families as well as a smaller 
survey of commercial traders in the impacted areas. Given that nearly 750 villages in 
five provinces were reached, the principal variable to the evaluation of the activity’s 
impact centered on the sample coverage and the sample size. The team defined both 
an appropriate number of villages to be covered and an appropriate number of 
households in each village. The estimates measuring the impact of the grant on 
households are based on a sample of 630 grant recipients. The precision of these 
estimates, stated in terms of the margin of error at 95 percent confidence level, is 
discussed in Annex 4.
Survey Writing. Four survey instruments were used to capture facts and opinions 
from a sample of those directly impacted by the grant program:
· Household Survey
· Village Chief Survey
· Retailer Survey
· Wholesaler Survey
Afrisurvey planned and implemented the field survey. After pre-testing in January 
2002, several types of changes were made to the survey instruments:
· As a result of a perception by respondents that the survey team represented a first 
step of a new grant-making program, the wording of the introduction was changed 
to make clear the purpose of the survey. Interviewer training emphasized 
anticipation and preparation for such an attitude.
Mozambique Flood Resettlement Grant Activity 5
· Originally designed in English, the surveys’ first Portuguese translation required 
adjustments to the exact and implicit meanings of several words and phrases.
· A few questions regarded as irrelevant were eliminated, and new, relevant ones 
were designed and added.
· The coding protocol was rationalized, and made consistent.
Survey Planning. Logistics planning was underway in December 2001 and January 
2002. Arrangements were made for vehicles, communications, camping and other 
field equipment, food and water, mapping, village identification and locations, and 
scheduling. Important communications and relationships were established with 
village chiefs and district administrators. From the beginning, seasonal rains and 
consequent severe travel difficulties were anticipated, so there was an urgency about 
planning and carrying out the surveys. Ultimately, the survey teams were hampered 
by heat, but not severe travel conditions.
During this phase, two important elements became evident, changing the survey from 
a strictly simple random sample to a systematic sample.2
· First, that limited time and logistical obstacles would not allow visits to targeted
villages that were so widely scattered. With the difficulties of travel and 
communications, it was not possible to adhere strictly to the intent of conducting 
proportionate numbers of interviews in the three categories of villages – very 
remote, remote, and accessible.
· Second, that it would be extremely difficult if not impossible to successfully 
locate and interview specific, individual grant recipients who could be identified 
in the PMU registry. As a consequence, surveyors would interview the most 
accessible and available grant recipients that could be identified upon arrival in a 
given village.
The original intention was to interview both the male and female heads of households 
in each of seven households in each village. After consultation with survey experts 
and USAID, the team decided to interview only the woman to avoid inconsistency, 
multiple interviews, or asymmetric data from polygamous men. As a result, 
household interviews were conducted with women grant recipients, consistent with 
the original rationale that delivered the grants to women.
In January 2002, some 20 interviewers participated in a four-day training session 
conducted in Maputo by the Abt team. They discussed program objectives, their role, 
interview techniques, language sensitivities, and field travel issues. Using the 
specially written training manual that ultimately became their field guide, they carried 
out mock interviews among themselves. They prepared for physical hardships in the 
field, and health and safety issues. Following a day in a rural village conducting 
practice interviews, the interviewers and staff re-grouped to make additional 
adjustments and plans.
 
2
 See Annex 4 for a discussion of Survey Estimates and Sampling.
Mozambique Flood Resettlement Grant Activity 6
Survey Implementation. The interviewers and supervisors were divided in two groups 
of teams, one group made up of two teams covering the provinces of Maputo and 
Gaza, the other group made up of three teams covering the provinces of Inhambane, 
Sofala, and Manica. Interviews were conducted over a two-week period in late 
January and early February 2002.
The importance the team placed on securing permission from village chiefs and 
district administrators was worth the effort, assuring survey practicalities and 
efficiencies. Advance visits and communications with village chiefs and district 
administrators succeeded in facilitating surveys in all but one case.3
As shown on Table 1-2, the initial target number of surveys was largely met. The 
target reflected the proportionate numbers of villages and grant recipients in each 
province – 14 villages, 14 village chiefs, and 630 households. The 96 retailer 
interviews conducted did not meet the target of 270 (six in each of 45 villages) simply 
because that number of retailers does not exist in and around the villages. The 11 
wholesalers interviewed represented half of the country’s 22 wholesalers, all based in 
Maputo or Beira.
Table 1-2. Number of Surveys Conducted, by Province and Survey Type
Provinces
Villages
Visited
Households
Surveyed
Village Chiefs 
Surveyed
Retailers 
Surveyed
Wholesalers
Surveyed
Gaza 12 174 11 14 0
Inhambane 5 72 5 18 0
Maputo 4 56 4 3 8
Manica 7 98 7 13 0
Sofala 17 238 17 48 3
Total 45 638 44 96 11
After survey reviews and cross-checking by field supervisors, the completed surveys 
were brought to Maputo for final reviews and collating. 
During this period, Caresoft Lda. designed and developed a comprehensive database 
for this project. It is a graphical, Windows-based user interface that employs the 
frames, panels, etc. used on most Microsoft-based software. It was developed using
Microsoft Visual Basic 6.0 programming language. This entailed developing active 
screens, database storage, linking active screens to database storage, developing and 
activating the “business logic”, and testing and debugging. Fields were created in the
database to relate to the active screen. A corresponding database entry was developed 
for each screen entry. Internal screen names were matched to the corresponding field 
name in the database. Database training and data entry took place over a two-week
period in late February 2002. 
 
3
 In one district in Manica, the deputy administrator would not let the team proceed to several villages; 
those villages were replaced with others elsewhere in Manica.
Mozambique Flood Resettlement Grant Activity 7
CHAPTER 2
BENEFICIARY IDENTIFICATION PROCESS
The Resettlement Grant Activity began with a target of 87,000 families. However, it 
quickly became apparent that the information base supporting the target number was 
poor and internally inconsistent. As a result, the program had to begin by surveying 
affected families, refining selection criteria and selection areas, and registering 
households. Finally, to verify selection, households were evaluated against the 
criteria, selected and targeted.
Geographic Location. Beneficiary identification was a difficult process. The 
original proposal outline simply stated that the resettlement grant would be directed 
toward assisting those “negatively affected” and required definition to determine a 
target group. The first phase therefore involved documenting the geographic area 
where the target group would be found. This entailed meeting with government 
officials, donor agencies, church groups, volunteer groups and non-governmental 
organizations (NGOs) that played an active role in the humanitarian relief effort of 
early 2000 or that were presently working within the affected areas. The Technical 
Secretariat of Election Administration (STAE) and the National Directorate of 
Geography and Mapping (DINAGECA) were also solicited for information 
pertaining to census statistics and geographic specifications of the affected areas.
Local disaster committees and NGOs based in the field were consulted. With this 
information and the PMU’s own research, 32 specific districts in five provinces were 
targeted:
Manica: Sussendenga, Mossurize, Machaze
Inhambane: Govuro, Maxixe, Vilankulos, Mabote, Inhassoro, Panda, Inharrime, 
Inhambane
Maputo: Manhica, Magude, Marracuene, Moamba, Boane, Matutuine, 
Namaacha
Sofala: Buzi, Machanga, Chibabava
Gaza: Chibuto, Mabalane, Chokwe, Chicualacuala, Massingir, Bilene, Guija, 
Xai-Xai, Massangena, Chigubo, Mandlakazi
Compilation and Verification of Lists. NGOs provided lists of affected people in 
the areas in which they were working and had access, but these lists did not provide a 
comprehensive accounting of affected people. NGOs were not working in areas, for 
instance, where helicopters dropped food aid during the flood emergency. 
As data were collected from a variety of sources, it needed to be verified, a process 
that involved entire communities. Monitors were selected, trained and sent to the 
communities to evaluate the actual situation and compare it with the information 
previously gathered. In certain instances where travel distances were extremely long 
or inaccessible, committees made up of village leaders were consulted.
Mozambique Flood Resettlement Grant Activity 8
The PMU informed those present at the community meetings of their purpose to 
collect information about those affected by the floods. Information gathered in the 
communities included the name and gender of the head of family and spouse (if 
applicable), whether the individual was single or widowed, the number of individuals 
within the immediate household and losses suffered due to the flooding and cyclones 
(e.g., loss or damage to infrastructure, cultivated fields, animals, goods, etc.). 
Information pertaining to those community members not present during the 
verification and registration process, yet affected by the flooding, was solicited in 
order to compile a comprehensive picture of the loss and damage in the area.
Besides verifying the community members affected by the natural disasters, the PMU 
visited individual households. In this personal setting, a baseline survey was
implemented aimed at gathering information pertaining to (a) the present status of the 
household; (b) whether the household was forced to flee its homestead during the 
flooding and if so, when it was able to return; (c) the type of assistance, if any, 
received by the household; and (d) the level of recuperation achieved. Household 
heads were also asked to prioritize the goods and services that would be acquired if 
they had the means to purchase or produce them.
Households Registered. Throughout the four-month verification phase, with the 
assistance of information solicited by the PMU from over 100 organizations, and 
applying the initial criteria used, 219,083 families were registered within the five 
designated provinces. The registered number far exceeded the estimated 87,000 
households originally targeted by USAID. The increase in households registered was 
attributed to several factors:4
· The PMU team entered areas not previously visited, principally due to lack of 
access (usually from standing water).
· Village residents in the most isolated areas were assisted with food aid dropped at 
designated distribution points via helicopters. These individuals – not formally
registered on any government or organization list and not included in published 
figures – were registered by the PMU.
· Additionally, particularly in the province of Sofala, those affected by Cyclone 
Eline were not formally registered. The cyclone destroyed virtually everything 
within its path in the districts of Machanga, Buzi and Chibababava. Households 
that received emergency assistance within these districts were those mostly
affected by the flooding of the Buzi and Save Rivers as opposed to Cyclone Eline. 
This under-reporting of victims of the cyclone also stretched into the province of
Manica, specifically in the districts of Machaze, Sussendenga and Mossurize. As 
a result, the inclusion of these households on the PMU verification lists resulted in 
elevated numbers in the two provinces.
 
4
 The PMU conducted this work following the spirit of the terms of reference, which had stated only 
that the target group were those who had been “negatively affected” by the floods. The 219,083 figure 
emerged from a loose application of criteria, rather than the “additive” approach that was intended and 
ultimately established by December 2000. The PMU embarked on Manica and Inhambane 
distributions with a simple application of any of the five criterion (the “either/or” rather than the 
“additive” approach). 
Mozambique Flood Resettlement Grant Activity 9
Determining Selection Criteria. The selection criteria, as defined in the original 
proposal outline, simply state that the resettlement grant would be directed towards 
assisting those “negatively affected” by the natural disasters of early 2000. The 
criteria evolved as the project went forward. With the compilation of lists it was 
possible to make better decisions and develop workable, reasonable, and transparent 
selection criteria. The eligibility criteria for the resettlement grants under ER: ACT 
were formally approved by USAID, Ministry of Industry and Commerce, and the 
Ministry of Women and Social Action.
Guiding Principles. Fairness and political acceptability were important guiding 
principles. Early in the design stage, it was decided that the grant money would be 
given to women heads of households. There were three main reasons for this:
· In most of the affected areas, the family is led by women due to men migrating to 
jobs.
· Polygamy is a common practice particularly among households in the central 
areas. Within a polygamous setting, the PMU considered wives and their children 
as the individual family unit. 
· Additionally, the project managers felt that women might be better than men in 
making spending decisions reflecting family priorities.
During the PMU pilot phase (in Manica), with the realization that the numbers could 
far exceed the 87,000 households that USAID had estimated, eligibility criteria were 
strictly enforced with two fundamental clarifications:
· Resettlement area – The people who received a cash grant could not also be 
located in a designated family resettlement area. This meant that they were better 
off than those who had not participated in resettlement programs.
· Second harvest – The second harvest indicator also allowed clarification on the 
help received. A second harvest was an indicator of a family receiving significant 
assistance such as seeds and tools donated by NGOs in the areas where they
worked.
The PMU and the Program Committee (PC) jointly refined the criteria presenting 
several draft versions prior to approval.
The final selection of criteria was influenced by data relating to results of second 
harvests throughout the country. This information was seen as a key indicator in the 
livelihood of individuals as over 80 percent of the households affected within the five 
provinces rely on agriculture as their only source of livelihood. Questions pertaining 
to the second harvest were incorporated into the baseline survey to assist in field 
assessments. In addition, recorded field observations depicting the overall situation 
within the areas and knowledge gained from interviews with GRM officials, NGOs, 
and other organizations working within targeted areas was taken into consideration. 
Overall results of the baseline survey were deemed as vital information in the 
selection process.
Mozambique Flood Resettlement Grant Activity 10
Technical Criteria. Since the program qualification criteria were additive, and not 
“either/or”, registered people had to meet all criteria, rather than any one of the 
criteria. In the end, this restricted the number of eligible participants. The final 
version of the selection criteria approved in January 2001 was the following:
1. The household was located in one of the districts affected by the floods and/or 
cyclones in the beginning of calendar year 2000 and which were on the list; AND
2. The household was outside a formal resettlement zone;5
 AND
3. The residence was destroyed or significantly damaged or the household was 
otherwise forced to resettle; AND
4. Crops were lost or damaged or other sources of income were disrupted or lost; 
AND, EITHER
5. The household did not have the means with which to produce a second harvest in 
calendar year 2000 or resided in an area where a second harvest in calendar year 
2000 was not possible; OR
6. The household did not meet all of the above criteria but it was located in a 
community of 300 households or less where at least 70 percent of the households 
in the community were affected as per criteria listed in 2 above.6
As planned, nearly all (93 percent) village chiefs participated in the selection of grant 
recipients. Their involvement was vital to the success of the program, and reinforces 
the importance of effective communication at the village level.
Nearly 81 percent of village chiefs reported that they understood the selection criteria, 
and found it easy to provide information to the grant resettlement team. 
Table 2-1. Village Chiefs Understanding of Selection Criteria
Understood selection criteria Number Percent
Yes 36 82
No 6 14
Don’t Know 2 4
Total 44 100
About 95 percent of village chiefs regarded the selection criteria as fair and believed 
that the selected families were in need. With only eight percent of chiefs reporting 
that some who received grants were not in need, they were effectively confirming that 
the selection criteria worked well. 
 
5
 Resettlement zones were areas selected by the government for permanent resettlement where families 
received significant and continuous assistance. Resettlement camps were areas where flood victims 
were provided refuge during and shortly after the floods.
6
 This sixth criterion was established to balance equity and efficiency. On the one hand the program 
didn’t want to foster community discontent by making grants to 70 percent or more of households in a 
small village where all are poor even though not all might be technically affected; and on the other 
hand, the PMU didn’t believe it was efficient to reach very small villages of fewer than 300 households 
where at least 70 percent of the households were affected.
Mozambique Flood Resettlement Grant Activity 11
The one reason given as to why some families may have been left out is that they 
were absent. However, some families who needed assistance were left out as they 
were in small villages (with fewer than 300 households where 70 percent or more did 
not meet criteria and therefore the whole village was left out). 
Negative aspects of the beneficiary identification process were few, but noteworthy. 
Many affected families were excluded, as they didn’t register. They were simply 
tired of registering and not receiving anything; they had previous experiences of 
signing up and waiting for assistance that never materialized.7
With these criteria, 114,000 households were eligible. Since funds for only 87,000 
had been anticipated for this activity, USAID re-aligned the budget of the economic 
recovery component in order to accommodate the additional 27,000 households. 
The disaster’s impact was severe, as shown on Figure 2-1 on the following page. 
Over 80 percent of beneficiaries lost all or a large part of their houses, food, goods 
and livestock. Equipment losses were more limited. Over 62 percent of beneficiaries 
reported that they did not own any equipment. Of the 38 percent of beneficiaries who 
reported owning equipment, 25 percent said they had lost their equipment during the 
floods.
Because virtually all grant recipients (96 percent) earned their income principally 
from agriculture, the destruction of their cultivated fields had severe effects, and was 
a major factor in meeting eligibility criteria. The survey results provide data allowing 
an evaluation of the success of reaching the target households. As shown on Table 2-
2 below, the beneficiaries from Gaza, Manica and Maputo provinces reported that all 
their fields were flooded and therefore crops destroyed. In Inhambane and Sofala 
provinces, well over 90 percent of respondents reported that their fields were flooded. 
Overall, nearly 98 percent of the respondents reported that their cultivated fields had 
been flooded.
Table 2-2. Destruction of Small Farms
Province Destroyed Small Farms Small Farms not Destroyed
Number Percent Number Percent
Total
Number
Gaza 173 100 0 - 173
Inhambane 68 97 2 3 70
Manica 98 100 0 - 98
Maputo 57 100 0 - 57
Sofala 225 94 13 5 238
Total 621 98 15 2 636
 
7
 For security reasons and to avoid a moral hazard problem, the PMU didn’t discuss the nature of the 
assistance during the beneficiary registration process. As a consequence, many potentially eligible 
people did not register; they didn’t realize that assistance would really be forthcoming.
Mozambique Flood Resettlement Grant Activity 12
Figure 2-1. Effects of the Floods
(percentages of grant beneficiaries suffering losses)
0
10
20
30
40
50
60
70
80
90
100
Seeds
destroyed
Food
destroyed
Animals
lost/died
Goods
destroyed
House
destroyed
Roof
destroyed
Equipment
destroyed
As shown on Table 2-3 below, nearly all grant recipients were unable to collect a 
second harvest. 
Table 2-3. Second Harvests
Did Not Collect 
Second Harvest
Collected Second 
Province Harvest
Number Percent Number Percent
Total
Number
Gaza 174 100 0 - 174
Inhambane 69 99 1 1 70
Manica 97 99 1 1 98
Maputo 57 100 0 - 57
Sofala 233 97 6 3 239
Total 630 99 8 1 638
Nearly all grant recipients were displaced from their homes for a period of time 
during and after the floods, but only about one-quarter of them spent time in 
resettlement camps. As shown on Table 2-4 below, most of those grant recipients 
who lived in a resettlement camp spent up to one year there.
Mozambique Flood Resettlement Grant Activity 13
Table 2-4. Grant Recipients Spending Time in a Resettlement Camp
Less than 
Two Weeks
Two Weeks to 
Three Months
Three Months 
to One Year
More Than 
One Year
Don’t Know
Province No. % No. % No. % No. % No. %
Gaza 7 18% 31 66% 29 64% 1 7% 7 70%
Inhambane 0 0% 5 11% 8 18% 1 7% 1 10%
Manica 10 26% 6 13% 4 9% 8 57% 0 0%
Maputo 0 0% 2 4% 2 4% 0 0.0% 1 10%
Sofala 22 56% 3 6% 2 4% 4 29% 1 10%
Total 39 100% 47 100% 45 100% 14 100% 10 100%
Conclusions. For the most part, the selection process was a success. It evolved 
during a time-sensitive situation, and with virtually no prior experience in identifying 
families in such remote areas affected by natural disasters. Identifying and 
registering victims was very difficult; little data about residents and flood victims was 
available. The international humanitarian relief organizations had no lists of affected 
families. Local efforts to help were sporadic with no real attempts to register affected 
people or families. As a result, beneficiary criteria had to be developed, tested, and 
agreed upon as part of the program. The surveys confirm that the criteria were 
stringently applied and that those who received grants were indeed the affected 
people. The beneficiary identification process can be characterized as appropriate 
and effectively administered. 
Mozambique Flood Resettlement Grant Activity 14
CHAPTER 3
HOUSEHOLD IMPACTS
Grant Recipients. Grant recipients were poor, largely subsistence farmers, 
inhabitants of rural areas, living in villages near their own fields. For them, the 1999-
2000 floods were indeed devastating and frequently fatal. In interviews, several 
referred to losing family members to the floods. 
Not surprisingly, nearly all (96 percent) recipients derived their principal source of 
income from agriculture. Salaried and commercial workers together represented 
nearly three percent.
Nearly all recipient families owned a bicycle. Some affected families in the southern 
provinces owned a plough animal and a sewing machine, but very few owned other 
significant personal items.
Destruction of family homes and livelihoods was high; homes and roofs were lost by
about 85 percent of households, food and goods were lost by about 90 percent, and 
seeds and livestock were lost by about 95 percent. Of those households that owned 
equipment, two-thirds of households lost it in the floods. Virtually all recipients lost 
their fields to the floods and were unable to collect a second harvest.
About one-quarter of recipients spent some time in resettlement camps. Nearly all 
recipients (99 percent) received assistance other than staying in resettlement camps 
and the grant itself. From other sources, they received food, construction materials, 
and health services. 
Many recipients live in areas often affected by extreme natural conditions. They 
suffered in the floods, and are now, in the same areas, suffering with drought. Many 
asked for immediate assistance.
Recipient Views. Grant recipients regarded the program as a success. Many took the 
opportunity of the presence of the survey team to express their gratitude. Nearly all 
recipients (98 percent) expressed 
satisfaction with the program. A small 
number of recipients in Gaza and 
Inhambane reflected dissatisfaction, 
considering that the process was not 
transparent and the grant amount was not 
sufficient for their needs. Villagers in 
Maxixe, for instance, thought the 
distribution process and the criteria for 
selecting beneficiaries was not clearly
explained to them. Villagers in 
Chicuecue thought there had been a 
degree of unfair manipulation based on 
personal relations. One recipient in Gaza was told that because she had no children, 
she could only receive half the grant; others thought that the prices for goods were 
much too high. 
Box 1. Grant Recipient Comments
The grant amount was not enough.
· I bought things I needed even 
though the money wasn’t enough.
· I bought the little I could because 
the money was little.
Mozambique Flood Resettlement Grant Activity 15
Recipients were clearly pleased with the grant program, particularly as it compared 
with other assistance programs. More than 80 percent thought it better than other 
programs in terms of getting needed items (and 16 percent thought it at least as good 
as other programs). About a third of recipients, however, credited other programs’ 
selection and distribution process as good as this program’s process. One recipient 
reported that distribution was difficult because the village secretary was drunk on that 
day.
Nearly all recipients were women, as intended. Not surprisingly, however, several 
recipients gave the grant to their husbands, and were not sure how the money was 
spent. A Sofala recipient gave her grant to her husband out of fear of being beaten.
Comments taken from one of the “open-ended” survey questions described suffering 
and loss, but revealed villagers mostly grateful, often spiritual, recounting their 
dancing and singing upon receipt of the grant. Recipients verified several program 
assumptions and hypotheses, as illustrated by their comments in the boxes below.
Use of Grants. In general, 
grants were spent on what 
program designers 
envisioned. Most was spent 
on household goods, 
clothes, and livestock. 
Expenditures for food were 
high as well, and many 
recipients expressed 
particular appreciation for 
being able to buy food, but 
food was not the most 
purchased good. 
Expenditures on seed and 
construction materials were 
substantial. 
As shown on Figure 3-1 on the 
following page, recipients used the 
grants primarily for household 
goods (e.g., dishes, pots, pans, 
blankets), secondarily for clothes 
and livestock, and then for food, 
seed, and construction materials.
Anecdotal evidence gathered by 
the survey team suggests that some 
recipients used the grant to 
purchase their divorce (buy back 
their lobolo) from their husbands; 
one reported to the survey team 
that she did exactly this. Based on 
Box 2. Purchasing Patterns
Household Needs
· When choosing food items, recipients purchased 
mostly sugar, salt, and cooking oil, and not water or
alcohol.
· Among consumer products, soap and matches 
dominated, while paraffin was less in demand.
· Most clothes were purchased for the recipient or her 
children, very few for the husband.
· Among household goods, dishes, pans, and blankets 
dominated, while few water containers were 
purchased.
Box 3. Purchasing Patterns
Personal Use
· Virtually all grant resources were spent within the 
family, with some minor amounts given to teachers 
or chiefs.
· Very little grant money was used to repay debts and 
loans.
· Health expenditures were more for services than 
medicines; many recipients received free medicines 
from other programs.
· School supplies rather than schoolbooks were the 
main education expenditure; many received donated 
schoolbooks from other programs.
Mozambique Flood Resettlement Grant Activity 16
the prevalence of polygamy, principally in Sofala and Manica, project managers 
assumed male grant recipients would purchase new wives, but no evidence of this 
was found. Generally, recipients (86.5 percent) say they made the purchases that they 
desired.
Figure 3-1. Percent Use of Grant, All Five Provinces
0
5
10
15
20
25
Household Goods
Clothes
Livestock
Food
Construction Materials
Seeds
Consumer Products
Payment for Services
School Expenses
Equipment
High Cost Goods
Health Services
Medicines
Paid Debts
Raw Materials
Other
Gave Money
Box 4. Purchasing Patterns
Housing
· Grants used for labor were mostly spent on 
house repairs, not agricultural labor.
· Many recipients rebuilt their homes with 
traditional, local materials. Construction 
purchases that were made were mostly for 
doors, nails, roofing, reeds, and poles; little 
was spent on paint, cement, and blocks.
· Some materials were provided by other 
assistance programs, but some materials –
zinc roofing material for one – were often 
re-sold by flood victims because it was not 
what they needed.
Mozambique Flood Resettlement Grant Activity 17
Provincial Variations. Expenditure patterns varied by province, in some cases quite 
substantially, as shown on Table 3-1 below.
Table 3-1. Percent Use of Grants, by Province
Gaza Maputo Manica Inhambane Sofala Total
Food 25% 63% 23% 6% 39% 11%
Consumer Products 21% 3% 21% 1% 54% 7%
Clothes 26% 10% 18% 8% 37% 17%
Seeds 31% 5% 16% 3% 45% 10%
Livestock 25% 3% 17% 10% 45% 14%
Household Goods 14% 11% 16% 7% 52% 22%
School Expenses 20% 0% 7% 3% 69% 3%
Medicines 50% 0% 7% 0% 43% 1%
Health Services 62% 0% 5% 5% 29% 1%
Equipment 12% 34% 18% 5% 30% 3%
Payment for Services 26% 3% 10% 16% 46% 3%
Construction Materials 40% 17% 3% 22% 18% 7%
Gave Money 0% 0% 0% 0% 0% 0%
Paid Debts 40% 0% 20% 40% 0% 0%
High Cost Goods 4% 11% 22% 9% 53% 2%
Raw Materials 13% 0% 13% 25% 50% 0%
Other 0% 0% 0% 0% 0% 0%
Location of Expenditures. The use of the grant also varied by where the recipients 
actually spent the money. As Table 3-2 below shows, most village purchases were 
made for food, consumer products, clothes, seeds, and service payments. High cost 
goods and school expenses were made in the district capitals. A significant 
percentage of livestock, household goods, and equipment were purchased in Maputo, 
but most reported such purchases were made in Zimbabwe. 
These geographic patterns of expenditures almost certainly reflect normal spending 
patterns of households. Those that spent the grant in Zimbabwe normally purchase 
those goods there. Whether the pattern is because someone in the household is 
employed in that location or because they usually purchase those goods at a certain 
source, there is no special significance to household expenditures outside of 
Mozambique, or far from home. Border crossing is common, more so of course by 
those who live nearest the borders of Zimbabwe and South Africa. The more 
significant pattern is that in which the grant was spent in the village itself, or in the 
vicinity of a grant distribution point. 
Mozambique Flood Resettlement Grant Activity 18
Table 3-2. Percent Use of Grant, by Where Spent
Village Distribution 
Point
Nearby 
City
District 
Capital
Beira Maputo South 
Africa
Zimbabwe Total
Food 45 2 20 17 4 4 1 8 100%
Consumer 
Products
57 2 17 13 3 2 0.3 6 100%
Clothes 58 2 16 13 4 2 0 5 100%
Seeds 61 1 17 11 3 3 0 4 100%
Livestock 8 0 0 26 0 0 0 66 100%
Household 
Goods
10 0 0 20 0 0 0 70 100%
School 
Expenses
0 0 0 100 0 0 0 0 100%
Medicines 0 0 0 0 0 0 0 0 0
Health 
Services
46 0 4 30 20 0 0 1 100%
Equipment 6 0 0 0 0 26 0 68 100%
Payment for 
Services
62 0 2 17 18 0 0 1 100%
Construction 
Materials
0 0 0 0 0 0 0 0 0
Gave Money 0 0 0 0 0 0 0 0 0
Paid Debts 0 0 0 0 0 0 0 0 0
High Cost 
Goods
20 0 0 80 0 0 0 0 100%
Raw Materials 0 0 0 0 0 0 0 0 0
Other 0 0 0 0 0 0 0 0 0
Mozambique Flood Resettlement Grant Activity 19
CHAPTER 4
ECONOMIC EFFECTS
USAID’s Resettlement Grant Activity was intended to provide an efficient and 
effective alternative to traditional relief programs. It was hypothesized that as a result 
of the multiplier effect, a cash transfer program (of US$9.7 million) would provide 
significantly greater impact on local economic activity relative to other programs; 
local demand for goods and services would in turn stimulate economic activity as 
merchants and laborers expanded their activities. 
The concept of cash relief following disaster situations has been an infrequent donor 
practice as a tool of emergency aid by both bilateral and multilateral donor 
institutions. In general, this type of aid has been controversial and treated with 
caution.
Nevertheless, it is generally recognized that in terms of theory and practice there 
appears to be a strong case for cash-based responses, at least to food emergencies 
where the supply and market conditions are appropriate. Practical experience to date, 
albeit limited, suggests that direct cash distribution, in the right circumstances and 
with careful planning and monitoring, can be more timely, less costly and more 
empowering to local communities than traditional food distribution.8
“The distribution of cash or a package based upon economic value could be more 
efficient in terms of logistics, and provide [beneficiaries] with much greater ‘value’ 
and a flexibility to use the aid to meet whatever their specific needs” (Wilson, 
1991:14).
The evidence concerning the impact of cash disbursement on economic activity is less 
clear. Over the past 20 years, cash has been used as a form of assistance for 
rehabilitation in a small number of emergency and post emergency situations. The 
underlying motivations for providing this type of aid are apparent: 
· empowerment of the recipient to prioritize goods and services according to
household needs;
· transfer of income in cases where individual/household income is halted and lost 
over a significant period of time; and
· stimulation of demand through a transfer of income, which in turn stimulates 
supply and hence economic activity, is re-established in affected areas. 
On the other hand there are arguments against this type of post-disaster aid:
· perceived lack of capacity on the part of the recipient to spend the money 
effectively;
· lack of access to goods and services;9
· inflation exacerbated by injection of cash; and
 
8
 Buying Power: The use of cash transfers in emergencies, British Red Cross, November 2000. This 
Red Cross document, cited elsewhere in this chapter, is a comprehensive literature review of most cash 
relief programs that have been implemented since the 1970s. 
9
 In very underdeveloped economies or in post-emergency situations, thin or missing markets mean 
many goods and services may not be available at any price. In the case of floods, this is especially the 
case for food.
Mozambique Flood Resettlement Grant Activity 20
· concerns about corruption, diversion and leakage of funds.
This chapter focuses first on economic effects including inflation, price gouging and 
economic stimulation (including recovery of distribution networks, recovery of
production capacity of local households, and multiplier effects). Then, an analysis is 
presented of the costs and benefits of the program.
Inflation. Inflation is a national macroeconomic concept while the Resettlement 
Grant Activity was localized in select districts of five provinces. It is widely 
recognized that the floods of 2000 had a determinant role in the abnormally high 
inflation rates experienced throughout that year, disrupting the progress made in this 
area by the Government of Mozambique.10 Structural adjustment and stabilization 
programs implemented over the past decade resulted in a significant reduction in 
inflation (as measured by the Consumer Price Index, CPI) throughout the late 1990s, 
as shown on Table 4-1.
 Table 4-1. Inflation: Consumer Price Index, 1996–2000 (%)
1996 1997 1998 1999 2000
CPI (%) 46.9 6.4 0.6 2.0 10.3
As Figure 4-1 illustrates, monthly inflation exhibited abnormally large fluctuations 
during the first quarter of the year 2000, coinciding with the impact of the floods. 
Global inflation figures were clearly affected by the price increases of foodstuffs, 
which account for 65 percent of the baskets of goods used in the inflation 
calculations. However, given the level of aggregation of the data available and the 
fact that it only reflects prices in the three major cities (Maputo, Beira, and Nampula), 
one cannot derive a direct correlation with the impact of the floods of 2000. This is in 
contrast to prices for other goods used in the CPI calculation, which remained stable 
(health, education and transport services) or fell (clothing and footwear).
Figure 4-1. Monthly National Inflation Rate, 1999-2001 (in percentage points)
Monthly National Inflation rate, 1999-2001 (Percentage Points)
-7.0
-5.0
-3.0
-1.0
1.0
3.0
5.0
7.0
Jan-99
Mar-99
May-99
Jul-99
Sep-99
Nov-99
Jan-00
Mar-00
May-00
Jul-00
Sep-00
Nov-00
Jan-01
Mar-01
May-01
Jul-01
Sep-01
Nov-01
Month
percentage Points
Total (100%) Foodstuff (65%) Other (35%)
Mozambique Flood Resettlement Grant Activity 21
The supply shortages that caused a jump in food prices were due to poor or lost 
harvests and loss of household stocks of main agricultural foodstuffs in flood stricken 
areas. This, in turn, resulted in market shortages in large parts of the south and 
central regions of the country. Estimates from the Food and Agricultural 
Organization (FAO) place cereal and bean production throughout the year 2000 at 
levels eight percent below those achieved during 1999. At the same time, the 
disruption of transport infrastructures throughout the southern and central regions 
reduced capacity to transport cereals, grains and other foodstuffs to flood stricken 
areas, either through imports of these goods from neighboring countries (such as 
Zimbabwe and South Africa) or from surplus producing regions in the north of the 
country.
Impact on Inflation of the Grant Resettlement Program. Previous experiences with 
cash distribution programs have highlighted the inflationary pressures that arise from 
cash deliveries in local economies where food prices are rising rapidly due to overall 
food shortages. A cash-for-food project in Ethiopia found that local food prices 
increased more as a result of food shortages than the cash transfer itself (Red Cross, 
2001).
The amount of local temporary inflation would be a function of the supply and variety 
of goods in the market at the time of the grant distribution, distance to neighboring 
markets, and the price setting behavior of national, regional and local traders. In the 
Mozambican case, the impact on inflation of USAID’s Resettlement Grant Activity 
should ideally be assessed taking into account both the demand and supply
components of the program. On the one hand, increased demand was created by the 
Resettlement Grant activities and, on the other hand, supply was facilitated by another 
component of USAID’s ER: ACT program – the Rural Enterprise Loans activity. 
It should also be noted that USAID’s emergency response also included US$35 
million for road rehabilitation (Inchope-Gorongoza; Chokwe-Macarretane; Guijá￾Chibuto; EN1 in Nova Mambone – taking place in areas where grants were 
distributed) and US$35 million for Limpopo rail rehabilitation (Maputo￾Macarretane). 
Inflationary effects of the Resettlement Grant Activity should be analyzed and 
compared in the context of similar monetary or goods injections by relief efforts 
undertaken by other international agencies and NGOs during the floods. 
Inflationary dynamics can be mitigated, in the short and medium to long term, if 
parallel actions are targeted to reestablish supply/distribution chains and production 
capabilities in flood stricken areas. This allows the supply of foodstuffs in the 
immediate term and the rehabilitation of agricultural production capacity in the longer 
term. 
Evidence from the Sample. This section uses the results from the survey undertaken 
in districts that benefited from the Resettlement Grant Activity to explore the possible 
effects of inflation caused by the injection of cash into the affected areas.
 
10 See 2000 Annual Report by Banco de Mozambique; also FAO and World Bank reports.
Mozambique Flood Resettlement Grant Activity 22
The results from the survey suggest there was an increase in consumer spending 
during the period that cash disbursements were made. As presented in Table 4-2, up 
to 65 percent of local retailers that participated in the survey thought that local people 
spent more during that period, and only six percent believed the opposite. 
A significant proportion of Resettlement Grant beneficiaries’ spending went to goods 
other than food (see Table 4-3). Thus, when retailers were asked what type of 
products had been purchased by beneficiary households, the various types of food 
products received 48 percent of total responses, other consumable goods 20 percent, 
household items 10.5 percent, equipment six percent, seeds five percent and clothes 
and construction materials 2.7 percent each.
Table 4-2. Retailers Reporting That More Money Was Spent During This Period
Number Percent
Yes 63 65
No 06 6
Don’t Know 02 2
No Answer 26 27
Total 97 100
Table 4-3. Retailers Response to Types of Products Purchased By Grant Recipients
Number of 
Responses
(multiple)
Percent
Food 691 48.2 
Construction materials 38 2.7
Seeds 68 4.7
Other consumables 284 19.8
Clothes 38 2.7
Household items 151 10.5
Medicine 1 0.1
Livestock 39 2.7
Equipment 86 6.0
Large items 0 0.0
School material 38 2.7 
Total Products 1434 100.0 
The results obtained from retailers’ responses, above, differ somewhat from the 
consumption patterns expressed by beneficiary households that participated in the 
survey, below. Differences are accounted for by the facts that:
· households bought goods from many sources other than the retailers interviewed;
· most of the retailers interviewed were primarily merchants of food; and 
· figures in both tables should not be read as the proportion of grant money that was 
spent on each type of good, but rather as the number of responses that each type 
of good received in the survey.
Mozambique Flood Resettlement Grant Activity 23
As shown on Table 4-4 below, 18 percent of purchases consisted of food, 14 percent 
of purchases went to acquiring seeds, 14 percent to household goods, 11 percent were 
purchases of clothes, six percent were purchases of livestock, 5.5 percent were 
acquisitions of construction materials and four percent of purchases consisted of 
different types of equipment.
Table 4-4. Use of Grant, All Five Provinces
Use of Grant
Number Percent Value (MZM) Percent 
Value
Food 1763 17.9 49,361,817 6.8%
Consumer Products 1360 13.8 30,428,940 4.2%
Clothes 1046 10.6 88,068,100 12.1%
Seeds 1391 14.1 56,642,500 7.8%
Livestock 626 6.4 101,643,150 13.9%
Household Goods 1336 13.6 101,780,750 13.9%
School Expenses 359 3.6 11,660,046 1.6%
Medicines 137 1.4 2,857,500 0.4%
Health Services 214 2.2 3,748,000 0.5%
Equipment 360 3.7 29,398,000 4.0%
Payment for Services 252 2.6 45,124,000 6.2%
Construction Materials 541 5.5 63,119,700 8.6%
Gave Money 163 1.7 24,992,000 3.4%
Paid Debts 57 0.6 7,704,000 1.1%
High Cost Goods 175 1.8 14,929,000 2.0%
Raw Materials 53 0.5 5,740,000 0.8%
Other 20 0.2 93,925,335 12.9%
Total 9853 100.0 731,122,838 100.0%
A further attempt was made to quantify the value of the purchases made by 
households in the sample with the money received from the program. The results 
corroborate the view that only part of the money was used for food purchases (seven 
percent), the rest being spent on other consumer goods (clothes 12 percent of total 
expenditure; other 13 percent; seeds eight percent; livestock 14 percent; and 
household goods 14 percent). 
In this sense, the food inflation effect of the program may not have been very
significant, given the purchasing patterns described above, which show a clear 
tendency towards purchases of items other than food. The cash disbursement 
approach allowed beneficiaries to make the expenditure decisions that reflected their 
needs as a result of damage brought about by the floods. As the above responses 
suggest, these do not necessarily consist only of food and other consumption goods, 
but also of producer goods (equipment, seeds, and livestock) suggesting a possible 
longer-term development consequence of the program. 
Mozambique Flood Resettlement Grant Activity 24
Nevertheless, there still appears to have been post-flood inflationary dynamics in the 
districts that participated in the program according to interviewed retailers. It is not 
clear whether these were solely related to the impact of the Resettlement Grant 
Activity. More than 63 percent of survey participants acknowledged the existence of 
price changes in the aftermath of the floods, and a large majority of these (88 percent) 
thought that prices had increased. In addition, we would expect that in isolated 
regions the impact of scarcity of goods would result in higher inflation rates than in 
communities with better communications to major distribution centers, though no 
data is available on other flood-affected areas to confirm this point. 
In this context, there appear to be regional differences in the post-flood price 
dynamics. Only 29 percent and 0 percent of retailers interviewed in the southern 
provinces of Gaza and Maputo, respectively, believed that there had been any price 
changes, in comparison to percentages as high as 100 percent, 75 percent and 63 
percent in Inhambane, Manica and Sofala, respectively. This may well reflect better 
links of the two southern provinces with the main national distribution hub (Maputo) 
and the South African market.
Table 4-5. Retailers Response As To Whether There Were Price Changes After The 
Floods
Gaza Inhambane Maputo Manica Sofala Total
No. % No
.
% No. % No. % No. % No
.
%
Yes 4 28.6 17 100 0 0 6 75 22 62.8 49 63.7
No 9 64.3 0 0 3 100 2 25 12 34.3 26 33.7
Don’t Know 1 7.1 0 0 0 0 0 0 0 0 1 1.3
No Answer 0 0 0 0 0 0 0 0 1 2.9 1 1.3
Total 14 100.0 17 100 3 100 8 100 35 100.0 77 100.0
Table 4-6. Retailers Opinion of Price Changes After the Floods
Number Percent
Increased 43 87.6
Decreased 3 6.1
Don’t Know 2 4.1
No Answer 1 0.2
Total 49 100.0
When asked their opinion to reasons that could explain price variation, 21 percent of 
the retailers who thought that prices had indeed increased attributed the increase to 
the amount of money in circulation (which could have been the direct result of the 
Resettlement Grant Activity itself). However, other factors appeared to have played 
an equally important role, such as the scarcity of products (35 percent of responses), 
higher prices at source (19 percent) and increased transport costs (seven percent). 
Mozambique Flood Resettlement Grant Activity 25
Finally, a significant portion of these retailers (19 percent) was unable to provide an 
explanation for the price increases. Nonetheless, the large percentage of responses 
pointing to the scarcity of products as a major source of inflation highlights the 
importance of accompanying cash disbursement programs with other complementary 
programs on the supply end through the restocking of goods, increased access to 
markets and improvement and rebuilding of distribution channels connecting these 
districts to food surplus regions or to international markets.
Evidence of price gouging. The results of the surveys of retailers and wholesalers 
do not provide any conclusive evidence of price gouging. Although there might have 
been a certain degree of opportunistic behavior by local traders, the available 
evidence (see Table 4-7) suggests that the main cause of inflation was the scarcity of 
products.
A significant number of interviewed retailers (21 percent) did believe that the price 
increases were mainly caused by the increased amount of money in circulation in the 
flood-affected areas, suggesting the existence of price gouging. In this respect, at the 
microeconomic level there would be no economic rationale for retailers to increase 
prices solely as a result of increased money supply in circulation, unless there was a 
supply constraint. However, scarcity of goods appears as the main cause of inflation 
in the retailer survey.
Table 4-7. Retailers Opinion as to Reason for Price Change
Number Percent
A lot of money circulating 9 20.9
Scarcity of products 15 34.9
Prices high at source 8 18.6
Increase in storage price 0 0
Transport costs 3 7.0
Other 8 18.6
Total 43 100.0
As Table 4-8 shows, it is interesting to note that of the 11 wholesalers that 
participated in the survey, nine of them (representing 82 percent of the sample) 
responded that they had not increased their prices as a result of the Resettlement 
Grant Activity. This fact does not necessarily imply that wholesalers did not increase 
prices for other reasons (e.g., supply shortages, increased transport costs, or even 
general price gouging not exclusive to the program), but it does reinforce the sense 
that in the event of price gouging, this was probably limited to local, informal traders 
and formal retailers. This finding is not surprising given the proximity and familiarity 
of local operators to the program; with their knowledge of the cash injection, they had 
a good perception of opportunistic gains to be made from it. Competition among 
wholesalers, and the fact that they know each other’s costs and prices, also served to 
keep the inflationary impact low.
Mozambique Flood Resettlement Grant Activity 26
Table 4-8. Price Increase Due to the Grant Money
Number Percent
Yes 1 9.1
No 9 81.8
Don’t Know 1 9.1
No Answer 0 0
Total 11 100.0
Evidence of stimulating the economies of villages and communities affected by 
the floods. As presented in Chapter 1, the floods of 2000 had a very significant 
negative economic impact on the country, severely affecting economic growth and 
interrupting the past buoyant growth trend. During 2000, the Mozambican economy 
as a whole grew by barely 1.6 percent, against double-digit rates achieved in previous 
years. Given an overall annual population growth estimated at 2.3 percent, this 
translates into a significant fall in GDP per capita during that year.
Table 4-9. Real GDP Growth, 1997-2001
1997 1998 1999 2000 2001
Growth Rate 11.06% 12.05% 7.33% 1.60% 14.80%
Both the Government of Mozambique and the international donor community have 
widely recognized that post-flood reconstruction could not be limited to short-term 
emergency and famine relief, but should also encompass reconstruction efforts aimed 
at stimulating economic activity, rebuilding market infrastructures and, thus, attaining 
previous levels of economic performance. Evidence for 2001 suggests that this was 
successfully achieved, with GDP growth rate at its highest in recent years.
Evidence from the Survey. The results of the survey conducted in districts that 
benefited from the Resettlement Grant Activity provide some insightful evidence of 
the impact of this program in terms of stimulus to local economies. In particular, 
three aspects provide evidence – multiplier effects, recovery of distribution networks, 
and recovery of production capacity of local households.
Indirect Demand Response-The Multiplier Effect. In order to fully assess the 
economic impact of Resettlement Grant Activity, it is necessary to explore the 
indirect effects it might have had through the multiplier effect11, and investigate to 
what extent this effect remained within the local or regional economy. 
A first step in this direction is to examine where purchases of goods by grant 
recipients took place. To this effect, Figure 4-2 shows the distribution of grant 
purchases by place where the money was spent. A significant amount of goods were 
 
11 In economic theory, the multiplier effect is the amount by which a change in autonomous 
expenditure (such as the one being analyzed here) is magnified or multiplied resulting in a more than 
proportional increase in spending.
Mozambique Flood Resettlement Grant Activity 27
purchased within the household’s geographic region: 55 percent of purchases were 
done in their own village, two percent at the grant distribution point, 16 percent in a 
nearby locality and 14 percent in the district’s capital. Overall, 87 percent of 
purchases took place within district perimeters. The remaining 13 percent of 
purchases either took place in the country’s two main urban areas, Maputo and Beira, 
or in neighboring Zimbabwe.
Figure 4-2. Use of Grant, by Where Spent
These figures do not necessarily mean that goods purchased were locally or nationally 
produced, especially in the Mozambican case where a significant amount of 
consumption and investment goods are imported. Nonetheless, the figures show that, 
in a first stage, the program’s money was spent mainly in local distribution points, 
and therefore remained in the region, stimulating sales, income gains, and job 
creation by store owners and their employees. This effect should not be undervalued 
given the evidence presented in previous sections regarding expenditure patterns of 
sample retailers.
There is some evidence that a significant part of the goods purchased by grant 
recipients were of Mozambican origin. According to retailers’ responses, on average 
about 50 percent of purchases corresponded to domestically produced goods. 
However, this figure should be viewed with caution given that (a) it is the result of 
interviews held several months after grants were distributed, (b) only 55 of the 97 
Nearby City
16%
South Africa
0%
Village
55%
Swaziland
0%
Zimbabwe
6%
Maputo
3%
Beira
4%
Grant Distribution Point
2%
District Capital
14%
Mozambique Flood Resettlement Grant Activity 28
participating retailers responded to the question, and (c) the survey sample covers 
only a small proportion of potential purchasing channels used by households. 
Although both these figures should be viewed with caution, they do point to the fact 
that a significant proportion of household expenditure was used in the purchase of 
domestically produced goods. A significant part of the grants was spent on purchases 
of goods – such as sugar, salt, cooking oil, soaps, livestock, pots and pans – for which 
there is a relatively competitive production capacity and supply in Mozambique. 
Recovery of Distribution Networks. Similar to other developing countries, 
distribution networks play a key economic role in rural areas of Mozambique, 
connecting rural local economies to regional, national and international markets. In 
addition to providing consumption and investment goods to local households and 
production units, they are fundamental in the process of surplus commercialization 
both within and outside the region. Efficient distribution channels allow a better flow 
of goods in and out of the districts easing pressures on prices. The need therefore to 
recover these networks was widely acknowledged by the government and donor 
community as fundamental to the process of economic recovery. It should be noted 
that Mozambique’s transport and marketing infrastructure was poor prior to the 
floods. Furthermore, it cannot be assumed that grant funds would have a direct 
impact on major infrastructure (e.g., roads, ports, etc.) reconstruction. Nonetheless, 
grant funds can help to stimulate the market chain and provide capital to traders by 
stimulating economic activity. It is this facet that will be investigated.
The evidence provided by interviews with local retailers suggests that a majority of 
them benefited from the Resettlement Grant Activity. Fully 70 percent of 
interviewees affirmed that their businesses had benefited from the program; 11 
percent thought that the program had not affected their business. Most retailers 
thought that the money brought in by the program revitalized the local economy, 
stimulating business and improving the business environment from which they 
benefited through increases in their business sales.
Furthermore, it appears that the proceeds from the extra-business the program 
generated were mainly reinvested in enhancing business and distribution capacities, 
rather than on purchases of consumption goods or savings. As presented on Table 4-
10, of the 64 retailers in the sample that agreed to give details of what they had done 
with the extra money earned from the Resettlement Grant Activity, 39 (equivalent to 
60 percent) confirmed that they had used it to increase stocks, 28 percent used the 
extra money to buy new items for their stores, and in five cases the extra income had 
been used to open new shops. Although, 43 percent of interviewees responded that 
they had used this money to purchase personal items, in many cases such items 
consisted of investment goods for use in other productive activities, such as livestock 
(sheep, goats and pigs), farming material (farming equipment, seeds, etc.), repair of 
damaged equipment, and in one case the purchase of a pick-up van. In other cases, 
the extra income was used to rehabilitate the homes of local retailers, also affected by 
the floods, and to buy food to alleviate hunger. These aspects are discussed in further 
detail below in the context of possible multiplier effects.
Mozambique Flood Resettlement Grant Activity 29
Additionally, some local retailers were able to improve their creditworthiness as a 
result of the improved business environment, in the form of additional credit (as 
reported in six cases) and improved credit conditions (ten respondents).
The evidence provided by the survey suggests that the grant resettlement program 
contributed to the revitalization of distribution networks in affected areas and, to 
some degree, in enhancing them, via small investments in retailer establishments and 
complementary assets and increases in the number of operators. Thus, the extra 
income that trickled up to local retailers in the form of increased business allowed 
them to restock their stores, repair damages caused by the floods, and invest in other 
small improvements.
Table 4-10. Retailers Response as to What They Did With Money Earned
Number of positive 
responses
Percent of positive 
responses
Bought items for the store 18 27.7
Increased stock 39 60.0
Saved money 5 7.7
Opened a new shop 5 7.7
Bought equipment 1 1.5
Bought personal items 28 43.1
Other 0 0.0
Number of respondents 65 100.0
Don’t know 32
Total Interviewees 97
Overall, retailers have a positive perception of the Resettlement Grant Activity and 
might have helped to reduce price increases in these districts. Some of the 
interviewees that had witnessed or participated in 
other disaster relief programs based on direct food 
distribution (17 out of 97, albeit a small sample) 
were able to comment on both types of 
approaches. Nearly half of those (eight) thought 
that cash distribution had benefited more their 
business than direct food aid. Three others found 
no substantial differences in terms of how the two 
sets of approaches affected their businesses and 
only four held that the program had left them 
worse off than food aid programs.
Rebuilding Productive Capacity-Investing for the 
Future. Another crucial element in the process of 
economic recovery and business stimulation is the 
rehabilitation of local production capacities. The 
floods that affected central and southern 
Mozambique during the year 2000 not only 
affected crops and local harvests, reducing the 
Box 5. Purchasing Patterns
Agricultural Needs
· When choosing seeds, recipients 
purchased mostly sorghum, some 
maize and groundnut, but little 
vegetable, millet, or cowpea.
· Several recipients reported 
receiving seeds from other 
assistance programs. 
· Those that spent grant funds on 
livestock bought mostly chickens 
and some goats, but few cows and 
pigs.
· Most expenditure for equipment was 
for agricultural implements; very 
little was spent on lanterns, 
bicycles, fishing lines, and batteries.
Mozambique Flood Resettlement Grant Activity 30
availability of food crops in local communities, but also caused a great deal of 
damage to the productive infrastructure, reducing the ability of affected communities 
to provide themselves with a sustainable livelihood in the future.
In this respect, interviews held with family households and local retailers in areas that 
benefited from the Resettlement Grant Activity show that part of the program money 
that circulated in those districts was used, directly or indirectly, for the purchase of 
investment goods and/or improvement and repair of damaged assets (land, equipment, 
livestock, etc.). This suggests that beneficiaries allocated a significant part of the 
grant money to rebuilding productive capacity and investing for the future. For 
example, according to retailers’ accounts 
(Table 4-3), an important portion of 
beneficiary households’ purchases went to 
purchases of equipment (six percent), the 
acquisition of seeds (five percent), and 
buying livestock (three percent). 
Furthermore, as shown on Table 4-4, 
households claim to have spent an average 
of 14 percent of the money on the purchase 
of livestock, nine percent on construction 
materials, eight percent on seeds, and four 
percent on acquiring equipment. 
Nonetheless, most of the grant went to 
purchases of food, acquisition of household 
goods and other consumption goods and 
clothes, presumably lost during the floods. 
The increase in economic activity caused by 
the impact of the program also allowed 
local retailers to undertake productive investments other than retail improvements 
with the extra revenue that was generated. These consisted of repairing damaged 
buildings and equipment such as tractors and other farm utensils. Some interviewees 
actually engaged in new farming and husbandry activities (four retailers claimed to 
have started cultivating a new field and two started raising goats and other animals). 
Finally, as has been pointed out above, retailers used much of the extra revenue 
indirectly generated by the program to restock their stores, repair flood related 
damages and upgrade and open new distribution facilities. 
Constraints, Benefits, and Costs of the Program. 
Program Constraints. Despite the fact that cash aid is infrequently used, there is 
growing donor interest in tackling post-emergency situations from the perspective of 
livelihood or household rehabilitation. Key concerns must still be addressed about 
the relative merits of this type of aid in the areas of:
· the effectiveness of the intervention (will sufficient goods be attracted to the 
market to satisfy beneficiary needs);
· the effects of the market (will the injection of cash exacerbate inflation); and
· the way beneficiaries elect to spend the money in the way intended. The option of 
cash-based programs has often been dismissed because of the underlying 
Box 6. Grant Recipient Comments 
Grants were used for future productive 
earnings.
· The money was very good; I was 
able to open a business that I’m still 
running today.
· The money allowed me to open a 
fish business; I’m very happy with 
this.
· It helped me a lot because my 
husband was able to start a small 
carpentry business and it is this 
income that is helping us at home.
Mozambique Flood Resettlement Grant Activity 31
assumption by donors that recipients of cash settlements are not capable of 
determining their best needs and interests, and will squander the money on goods 
that do not benefit the family. From the evidence of other programs (and indeed 
this program as well), however, this contention has been demonstrated to be 
completely unfounded. As mentioned previously, numerous studies on the coping 
strategies of famine victims, refugees and other vulnerable groups have confirmed 
that affected communities are more than capable of determining their best needs 
and interests. 
The anticipated risk that money would be squandered by male household members on 
goods that would not directly benefit the household was made moot in the initial 
program design of the Resettlement Grant Activity. Female heads of household 
received the grant, and indeed, households reported that 20 percent of consumption 
was spent on food goods and of that amount only 1.5 percent was spent on alcohol.
The question of whether sufficient goods can be attracted to the market to satisfy 
beneficiary needs is crucial to the effectiveness of this type of program. In this case 
USAID addressed this issue by tackling supply side constraints in conjunction with 
providing cash to stimulate demand. 
The question of whether this particular injection of cash exacerbated inflation cannot 
be directly answered, but from the analysis above on inflationary impacts, we can 
assume that this was not strongly the case.
Costs of Resettlement Grant Program. It has been demonstrated (Red Cross, 2000, 
p.28) that the transaction costs of cash distributions are substantially lower than those 
of commodity distribution. Overhead costs associated with food aid (handling, 
transport, storage and administration) on average range from 40 to 50 percent of the 
total aid provided. Cash, on the other hand, has a minimal transaction cost. As a 
result, the unit cost per beneficiary head is significantly less than commodity 
distribution. Importantly, less expenditure on overheads could mean beneficiaries 
receive a greater overall proportion of the money donated. Of course the reduction in 
transaction costs provides a significant benefit to the donor as their costs are also 
significantly reduced.
Given the complex logistics and security arrangements planned and implemented, the 
costs of identifying and registering flood victims and then distributing cash grants 
represented a reasonable portion of USAID’s funds. The USAID/PMU contract for 
US$3 million was to manage USAID’s US$32 million program (of which 
approximately US$10 million was in grants to affected families and US$22 million 
was in loans to affected enterprises). Implementation cost about ten percent of the 
value of the grants themselves, quite a reasonable portion, particularly when 
contrasted with the costs associated with food distribution. 
It has been demonstrated that the transaction costs of cash distributions are 
substantially lower than those of goods distributions. Cash on the other hand has a 
minimal transaction cost. As a result, the unit cost per beneficiary head is 
significantly less than goods distribution. Importantly, less overhead costs generally 
mean that beneficiaries receive a greater overall proportion of the money donated; an 
analysis of costs in Geramider in northern Shoa, Ethiopia, estimated that in an 
Mozambique Flood Resettlement Grant Activity 32
equivalent food distribution, the value of food received by beneficiaries would have 
been around 90 percent of the costs of the cash distribution (UNICEF, 1998). 
Benefits of the Resettlement Grant Program vis-à-vis Other Programs. A positive 
aspect to this type of assistance relates to its consistency with the Mozambican 
Government’s desire that flood reconstruction efforts would avoid significant 
deviations from the existing policy and reform agenda. In particular, emergency aid 
should have been in line with the shift from a supply-driven approach dominant in 
Mozambique during previous emergencies to a demand-driven approach involving 
effective community participation in making decisions (World Bank, 2001). 
A further positive aspect of implementing this type of program concerns the benefits 
to the recipient economy of not receiving food aid from donor countries. Cash should 
also have less of a distortionary effect on the domestic food market. As demonstrated 
previously, cash can have certain stimulation effects on the market.
Lastly, we infer that the income transfer covered only a small portion of the nominal 
loss of income from the floods. The US$92 cash grant per family represents roughly 
half of the yearly per capita income of the average Mozambican in the five affected 
provinces (see Table 4-11), but, depending on household size, represents a much 
smaller share of a family’s income. Thus, the income lost during the floods (for up to 
six months) was only partially recovered by the cash transfer. 
Table 4-11. GDP Per Capita, 1998
Province GDP per capita (US$)
Maputo 174
Inhambane 170
Manica 184
Sofala 306
Gaza 147
Average 196
Conclusions. The Resettlement Grant Activity was successful in achieving its 
objectives of implementing an efficient and effective alternative to traditional relief 
programs in terms of providing short-term social protection from income losses as a 
result of the floods. In addition, the evidence suggests that the program was to some 
degree successful in stimulating economic activity in the affected areas targeted by 
the program in the short term. 
Food prices did increase substantially during and after the floods. However, the 
evidence suggests food inflation effects of the program were minor, given the clear 
tendency by households towards purchases other than food. Notably, food 
inflationary effects were greater in the northern markets, possibly reflecting their lack 
of proximity to the large and developed markets of Maputo and South Africa enjoyed 
by the southern markets. Most suppliers pointed to the scarcity of food as the prime 
reason for the increase in prices, and hence inflation, underscoring the importance of 
accompanying cash disbursement programs with other complementary programs to 
impact the supply side. Such programs would focus on direct food relief or through 
Mozambique Flood Resettlement Grant Activity 33
improvement and rebuilding of distribution channels connecting the districts to food 
surplus regions or to international markets.
The survey results do not provide any conclusive evidence that there was detectable 
price gouging by local traders as a result of increased money in circulation or scarcity 
of goods. 
In terms of the program’s impact on economic stimulation, the results are quite 
favorable. The recovery of distribution networks appears positive. Most retailers that 
participated in the interviews believed that the money brought in through the program 
revitalized the local economy, stimulating business and improving the business 
environment from which they benefited from an increase in sales. The evidence 
provided by the survey suggests that the grants contributed to the revitalization of 
distribution networks in affected areas. To some degree, the grants also enhanced 
networks through small investments in retail establishments and complementary 
assets and increases in the number of operators.
The evidence of rebuilding productive capacity also demonstrates positive results 
given that in the areas where money circulated, it was used directly or indirectly for 
the purchase of investment goods and/or improvements and repairs to damaged assets 
(land, equipment, livestock, etc.). Moreover, since retailers used much of the extra 
revenue indirectly generated by the program to restock their stores, repair flood￾related damages and upgrade and open new distribution facilities, the program had 
positive linkages for non-recipients of rebuilding productive capacity.
While it was impossible to quantify the multiplier effect, there were several signs of 
productive downstream spending of grant funds. The evidence, albeit not conclusive, 
points to the fact that a significant proportion of household expenditure was used for 
the purchase of domestically produced goods and, moreover, that the consumption 
goods purchased were for items for which there is a relatively competitive production 
capacity and supply in Mozambique. 
A cash transfer program is an efficient and cost-effective way to impact affected 
people in a short amount of time. 
It is difficult to measure the benefits of this type of program vis-à-vis other programs 
in the absence of available data. However, we have suggested that the program was 
in line with that of the Mozambican Government’s development strategy and in some 
ways is an improvement over the traditional use of food distribution.
The program helped stabilize the affected households by providing them with income
lost during the floods and positively empowered households to rebuild their own 
futures.
Mozambique Flood Resettlement Grant Activity 34
CHAPTER 5
CONCLUSIONS AND OBSERVATIONS
Overall Conclusions. The Resettlement Grant Activity was an important and 
effective program of humanitarian assistance with immediate economic impact in 
flood-affected villages. Although the Resettlement Grant Activity can more 
accurately be characterized as a “recovery” activity, rather than a “resettlement 
activity”, it was successful, clearly in assisting flood-affected families, and somewhat 
in reinvigorating the economy. The grants provided a life-saving safety net for some 
families, and contributed to the rebuilding of lives and incomes for most others. It 
allowed homes to be re-established, essential needs to be met, and productive, 
income-generating activities to be re-started.
Unavoidably, the program reinforced a sense of expectation by villagers that others 
may be depended upon for their 
survival.
A small number of eligible, affected 
families probably did not receive 
grants, and a small number of 
ineligible families probably did receive 
grants, a risk that USAID recognized it 
was taking. Those most severely
affected by the floods, however, were 
assisted, and in the absence of the grant 
program, they certainly would have 
been worse off.
The fundamental principle of a cash 
grant program – that without any 
conditions attached, households would 
make prudent use of the money – was 
confirmed. In general, the view that 
women manage the money, and choose the 
family priorities was also borne out. 
However, women in the more remote 
villages, often in polygamous marriages 
and in general subservient to the men, had 
less control over grant expenditures than 
other women.
Management. The program was effective 
at establishing eligibility criteria, 
identifying affected families, registering 
them, and ultimately distributing cash to 
them. Without any model upon which to 
base its planning, the PMU managed 
complex logistics and communicated 
directly with communities. The PMU and 
Box 8. Grant Recipient Comments
Cash is better utilized than commodities.
· I appreciate the cash assistance 
because I bought what I needed at 
home.
· I appreciate it because it was cash 
that we didn’t expect.
· I appreciate it and was very happy to 
get the cash because it was very 
important.
· I appreciate the cash because I had 
nothing.
· I’m happy to have received cash; I 
was able to buy what I needed; it was 
very important.
Box 7. Grant Recipient Comments 
Future assistance is expected.
· It helped for purchases at home, but 
it would be good if they returned.
· We ask the entity that gave us cash 
to do it again because right now we 
are dying from the drought.
· I appreciate the help and hope that 
they come back to help us again.
· I really appreciate it and hope that 
they don’t get tired of helping us.
Mozambique Flood Resettlement Grant Activity 35
USAID teams were very professional, with defined roles for many functions and a 
good information flow. Logistics and security were well managed. The distribution 
process was remarkably efficient. USAID was fortunate that the PMU had a very 
effective and efficient management team. The PMU leadership was instrumental in 
program success.
There were large information gaps among NGO/PVO lists of affected families. 
NGOs generally operate in more densely populated areas, while the floods affected 
the more remote areas where NGOs aren’t present. Thus, compiling lists and 
registering victims was time-consuming and costly, but ultimately effective. The 
involvement of village chiefs and local administrators was vital to both beneficiary 
identification and ultimate distribution. 
Timing. The uncertainty of when and how much money was going to be available by 
the US Government delayed planning and implementation. USAID’s determination 
to expedite contracting and implementation was admirable. 
The time lag had an unforeseen impact. While many programs were targeting the 
same flood-affected families, many victims grew disaffected with the meetings and 
registrations involved. Thus, by the time the PMU deployed its field teams, many 
potential beneficiaries were not interested in registering given their skepticism 
regarding future assistance. Also, due to the time taken to define the target group, 
grant distribution began (in Manica, in December 2000) before the target group was 
conclusively identified (in January/February 2001). 
The time lag from the floods in February to the grant distribution in December had 
different impacts in different provinces. Inhambane, Sofala, and Manica provinces 
have one annual season of harvest, while Gaza and Maputo provinces have two. 
Thus, the grants were distributed in time for the second harvests in the south, and 
during the lean season in the north.
Costs. One of the single largest direct costs of project implementation was for 
vehicle rental. Vehicle purchase would have been substantially cheaper. Also, labor 
costs were higher than planned simply because of the time it took to identify and 
register flood victims according to proper criteria. Nevertheless, costs for program 
implementation – about ten percent of grant amount – appear quite reasonable.
Concurrent Activities. Two program objectives were not met, without apparent 
negative consequences. First, USAID sought to provide inventory credit to retailers 
so that they would have stock to sell; by the time any retailers received any credit, 
most grants had long been spent. In addition, most retailers did not need credit; they 
had excess and unsold inventory in stock. Second, USAID and the government 
sought to disseminate information to retailers so that they would be prepared for trade 
in districts where grants were to be distributed; grant recipients in many villages, 
however, were unable to make the purchases they wanted in the places they could get 
to. 
Post-Emergency Strategies and Activities. As the Government of Mozambique and 
donors were providing life-saving assistance to flood victims, they also began to think 
about what kind of programs would be appropriate after the emergency assistance. 
Mozambique Flood Resettlement Grant Activity 36
The cash distribution program ultimately implemented was one of several options 
considered. Among the issues considered were:
· family/household impact (to identify needs; replace lost goods and income)
· family composition (male/female; polygamous situations)
· economics (to restore jobs and production)
· availability of goods and services (location of commercial sector and traders; 
location of assistance)
· programmatic budget limitations (to maximize impact; numbers of people to be 
assisted)
· logistics and management (to develop a viable program)
Vouchers. One of USAID’s first considerations was to use a voucher system in 
which cash vouchers or commodity vouchers would be distributed. Both were 
rejected. A cash voucher system would have been too easily abused. A commodity 
voucher system would have required an analysis to identify items most needed. 
There were concerns that a secondary market for vouchers would develop, negating 
their intent. In addition, relationships with shops and vendors would have to be 
established.
Commodities. A program of commodity distribution was a natural (and common) 
choice. USAID had such experience and distribution agents (particularly NGOs) 
seemed well located and prepared to implement. 
Commodities would have to be identified, specified, ordered, delivered, and 
distributed. In the absence of sound analysis, the selection of which commodities to 
purchase and distribute would have been difficult. Delivery of commodities to 
remote locations would have been challenging, and some network of local 
organizations and small enterprises would have to be enlisted.
Most of the likely commodities would have to be imported; those on the local market 
would be very expensive. In any case, if commodities were provided, the program 
would be undercutting the private sector.
Credit. Providing credit directly to flood victims was not seriously considered; the 
flood victims were extremely poor, and unlikely ever to be able to repay a loan. 
Credit for viable businesses, however, to use to replenish and increase their inventory
was regarded as an appropriate solution, but only if consumers had money to spend. 
Thus, ultimately the grant program converged with the inventory credit program and 
loan program. 
Capital Projects. Traditional, but unsustainable, post-emergency assistance could 
have been programmed for replacing lost infrastructure and buildings (e.g., bridges, 
roads, schools). In fact, most donor assistance was targeted for such capital projects. 
The effective ones will have incorporated mechanisms into the projects to raise funds 
for operations and maintenance.
Cash. The advantages of a cash distribution program were several. It is 
fundamentally non-directive. That is, it relies on each household to make prudent 
Mozambique Flood Resettlement Grant Activity 37
decisions. It avoids or at least minimizes opportunities for corruption. And, a cash 
program, with all its problems of security and logistics, is efficient. 
Replicability. Cash grants have been effectively used following disasters in some 
countries, as presented in Table 5. The circumstances under which cash distribution 
programs may be appropriate following emergency disaster assistance vary, and 
issues that should be considered include: 
· Alternatives – can NGOs or other institutions provide more appropriate 
assistance?
· Intent – are grants intended for recovery of homes and livelihood, that is, not 
intended to encourage behavioral changes (such as giving up arms)?
· Cost effectiveness – are implementation costs too high a proportion of total 
costs?
· Use of cash – are goods, services, land, etc. available for purchase?
· Consultation – are implementing institutions able to communicate and 
understand the needs of the affected population?
· Beneficiary identification – can affected victims be identified (or is virtually 
everyone a victim)?
Other Lessons. Important conclusions from the survey about the real needs of poor, 
rural villagers can be applied to the design of traditional development projects.
The grant activity had an unanticipated consequence. In the context of forming 
policy about Mozambican private sector participation and capacity, the program 
contributed to ideas about increasing local content in projects, about using local 
subcontractors, and about empowerment.
USAID is currently providing support for four emergency mitigation activities – the 
Famine Early Warning System Network, Improved Varieties of Sweet Potato and 
Cassava, Development of Mozambique Seed Protocol, and Regional Linkages. 
USAID should consider “disaster preparedness, prevention, and mitigation” as a new 
or crosscutting sector of focus. 
The US Federal Emergency Management Agency’s grant-making program following 
natural disasters in the US provided an important model for USAID’s planning in 
Mozambique. US agencies should collaborate to identify other programs that provide 
lessons for development assistance, both in the US and in other countries. USAID 
and US domestic agencies would benefit from sharing development approaches. 
Mozambique Flood Resettlement Grant Activity 38
Table 5-1. Comparisons of Circumstances in Cash-Grant Programs in Selected Country Disasters
Circumstances Mozambique Mozambique Ethiopia Ethiopia Côte 
d’Ivoire
Cameroon Chad
Disaster and 
rationale
floods, 
recovery 
assistance
post-war, 
demobilized 
soldiers, cash 
for arms
Food 
shortages
post-war, displaced 
families, deportees, 
families of 
deceased
Highway 
construction
pipeline 
construction in 
agricultural 
areas
oilfield and pipeline 
construction in 
agricultural areas
Location remote country-wide Wollo Restricted
Population 
characteristics
poor, rural rural, peasant Shanty 
towns
Definition of 
target group; 
ability to 
identify victims
definable, 
but difficult 
to identify
definable and 
identifiable
Definable 
along 
highway
Cash provided US$97 US$160-$320
Costs US$10 m
Families 
assisted
106,000 470,000
Use of cash purchase 
goods and 
services
returning to 
farms
Purchase of 
goods and 
food
relocation; 
home 
construction
compensation 
for land
compensation for 
loss of crops
Implementation managed 
effectively 
by USAID 
and PMU
direct 
government
Save the 
Children 
(UK)
direct government 
and World Bank
direct 
government 
and World 
Bank
direct government 
and World Bank
Ability to 
provide timely 
assistance 
relatively 
timely
difficult Effective
Mozambique Flood Resettlement Grant Activity A-1
Annex 1. Reference Documents
Audit Report, USAID Resources Managed by the Project Management Unit / Deloitte 
& Touche. Sithole Chartered Accountants, SA, draft 12/01.
Banco de Moçambique, Relatório Annual. Bank of Mozambique, 2000.
Buying Power: The use of cash transfers in emergencies. British Red Cross. 
November 2000.
Cash For Food, Project Evaluation of Arsi I, II & III Sites, UNICEF, Addis Ababa, 
1988.
Concurrent Audit of USAID/Mozambique’s Resettlement Grant Distribution Activity, 
Office of USAID Regional Inspector General, July 31, 2001.
Hunger and Public Action, Dreze, J. and Sen A., Clarendon Press, Oxford, 1989.
Economist Intelligent Report, Quarterly Report 2001:1. The Economist, EIU.
Electronic Newsletter. State of the Environment Reporting Network for Southern 
Africa, Number 8, and 2001.
Emergency 2001 International Appeal by the Government of Mozambique in 
Collaboration with the United Nations. United Nations, February 2001.
Final Report, Agriculture Commerce and Trade Program, Component I: Resettlement 
Grant, Deloitte and Touche, May 2001.
Food/cash for Work Interventions in Famine Mitigation, Famine Mitigation Strategy 
Paper, Bryson, J. and Hansch, S., U.S. Foreign Disaster Assistance, Washington, 
1993.
Food Economy Baseline Profile of the Limpopo River Basin Complex: Gaza Province 
Mozambique. FEWS NET. January 2001. 
Impact Assessment Machanga District, Sofala Province. World Vision Mozambique 
Emergency Flood Relief, World Vision, 2000.
International Reconstruction Conference Rome 3-4 May 2000: Six Months Later. 
Government of Mozambique, May 2000.
The Mozambican State Budget. Government of Mozambique, 2002.
A Preliminary Assessment of Damage from the Flood and Cyclone Emergency of 
February-March 2000. World Bank, March 2000.
Mozambique Flood Resettlement Grant Activity A-2
Social Safety Nets for Poverty Alleviation in Southern Africa, ESCOR Report R7017. 
Institute for Development Studies, Sussex. Devereux, S. 2000a.
Special Report: FAO/WFP Crop and Food Supply Assessment Mission to 
Mozambique. FAO, May 2000.
Survey of Damages to the Commercial Trading Sector Caused by the Flooding and 
Cyclone Eline in Central and Southern Mozambique. CTA, TechnoServe, USAID. 
April 2000.
USAID Implementation Letters:
No. 20 (07 Dec 00) eligibility criteria
No. 21 (07 Dec 00) implementation plan, first phase, Manica and Sofala 
provinces
No. 34 (21 Feb 01) revised eligibility criteria
No. 35 (21 Feb 01) implementation plan, second phase, Gaza province, 
some of Sofala
No. 36 (22 Mar 01) implementation plan, final phase, Maputo province
Mozambique Flood Resettlement Grant Activity A-3
Annex 2. Persons Interviewed
Cynthia Rozell Mission Director, USAID
Christine de Voest Rural Enterprise Officer, USAID
Phillip Tonks ER: ACT Activity Manager, USAID
Gale Rozell USAID
Abdul Magid Osman Presidente do Conselho de Administração
Banco Comercial e de Investimentos (BCI)
Carlos Morgado Ministro da Indústria e Comércio
Luis Eduardo Sitõe Direcção de Relações Internacionais
Ministério da Indústria e Comércio
Abdul Azim Commercial Director, Universal Commercial 
(Wholesaler)
Carvalho Neves Coordinator, Project Management Unit, Deloitte & 
Touche
Michele Gross Resettlement Grants Coordinator, Project Management 
Unit, Deloitte & Touche
Sol de Carvalho Project Manager, Promarte
Machado de Graça Department Manager, Promedia
Jamú Hassan Administrator, MOPAC, Sociedade Comercial e de 
Investimentos, Lda. (Wholesaler)
Riaz Merchant Director, Operations, Ayul Trading Company 
(Wholesaler)
Mozambique Flood Resettlement Grant Activity A-4
Annex 3. Evaluation Team
John Miller Team Leader
Marilyn Wilkinson Survey Design
Jovito Nunes Afrisurvey, Managing Partner, Anthropologist
Ercilia Santos Afrisurvey, Project Director, Regional Supervisor
Virgilio Amorin Afrisurvey, Field Logistics Coordinator, Regional Supervisor
Esther Kazilimani-Pale Analyst
Oladapo Carew Caresoft Lda., Manager, Database Designer
Liv Bjornestad Economist
Field Supervisors:
Rosa Tembe
Eric Herbet
Jonas Cossa
Leonel de Sousa
José Carlos Mboa
Survey Interviewers:
Esmeralda Banze
Zedequias Manganhela
António Francisco
Jorge Mugarre
Marciano Manuel
Enelda Zunguze
Luís Osvaldo
Deolinda Chirindza
Vicente Chau
Deodato Nhabinde
Tânia Mondlane
Ana Jona
Idalina Chapanga
Eleutério Macanjo
Ana Mabunda
Drivers:
João Simbine
Agostinho Mucumbe
Carlos Agostinho
Joaquim Manhiça
Sebastião Langa
Mozambique Flood Resettlement Grant Activity A-5
Annex 4. Survey Estimates and Sampling
Precision of Survey Estimates. The estimates measuring the impact of the grant on 
households given in Chapter 1 are based on a sample of 630 grant recipients. The 
precision of these estimates will be stated in terms of the margin of error at 95 percent 
confidence level. The margin of error depends on the variance of the estimates. 
As indicated in Chapter 1, the sample of recipients was formed by first selecting 
villages and then recipients within selected villages. Therefore, the sample of 
recipients is not a simple random sample but a clustered sample. The variance of the 
estimates based on a clustered sample is generally higher than under simple random 
sampling. The ratio of the variance of the estimates under the design to the variance 
under simple random sampling is called design effect. For this survey, we have 
assumed a design effect of 1.4. To assess the reliability of the estimates under the 
sampling design adopted for the survey, we first divide the sample size by the design 
effect to get an effective sample size. Using this sample size we determine the 
margin of error assuming simple random sampling. The effective sample size for this 
study is 455. The margin of error of the estimates based on grant recipients is 
calculated assuming a sample of 455. 
Two types of estimates were produced based on the responses from the recipients. 
The first type is estimates of percentages of respondents saying “yes” or “no” to 
various types of questions. For example, what percentage of respondents lost homes 
and roofs? The second type is the percentage of total grant spent on various items, 
e.g., food, clothing, etc. 
The margin of error of the estimates of characteristics which are measured by “yes” 
or ‘‘no” type responses depends on the estimated percentage and the sample size. For 
example, 85 percent of the recipients in the sample lost homes and roofs. A 95 
percent confidence interval for the percentage of recipients who lost homes and roofs 
in the population of recipients is 85 plus or minus 3.3. That is, we are 95 percent 
confident that the percentage of recipients who lost homes and roofs is contained in 
the interval 81.7 percent to 88.3 percent. Similarly, we are 95 percent confident that 
the percentage of recipients who lost food and goods is 95 percent plus or minus 2.7. 
The interval is 92.3 percent to 97.7 percent. The margin of error for such percentages 
is around plus or minus three to five percentage points.
The precision of the estimates relating to percentage of the total grant spent on 
various items depends on the variability of this characteristic between recipients in 
addition to the sample size. For example, based on the data from all five provinces, 
we see that 10.2 percent of the grant was spent on food. If we assume that the range 
of the percentage of grant spent on food varies from 0 percent to 100 percent 
(maximum), then the margin of error for this percentage at 95 percent confidence 
level is plus or minus 1.5 percentage points. The other estimates of percentage of 
grant spent on various estimates have a similar margin of error or even better. 
Therefore, the numbers given in the tables are fairly precise.
Survey Sampling. The survey utilized a “systematic sampling” methodology. Like 
random sampling, systematic sampling is also a method of scientific sampling. In 
Mozambique Flood Resettlement Grant Activity A-6
simple random sampling (generally called random sampling), every person or 
household in the population has the same chance (probability) of being included in 
the sample. Also, since there are many possible samples that could have been 
selected, every possible sample has the same chance of being selected as our sample.
In systematic sampling also, every household or person has the same chance of being 
included in the sample but it is much easier and simpler to implement. An example 
of systematic sampling follows:
Suppose we have a population of 100 households. We want to select ten households. 
First, we compute a sampling interval which is equal to the population size divided by 
the sample size. In our example, this is 100/10 = 10. We select a random number 
between 1 and 10, say 4. We select the 4th household on the list and every 10th 
household thereafter. Therefore, the households selected in the sample are the
households which are 4th, 14th, 24th, 34th, 44th, 54th, 64th, 74th, 84th and 94th 
household on the list.
If the random start was 8, then we would have selected the 8th household, 18th 
household, 28th household, etc., on the list, the last one being the 98th household.
In summary, systematic sampling is a method of sampling in which the first unit on 
the list is selected at random and then every kth unit thereafter where "k" is the 
sampling interval which is the ratio of the population size to sample size.