Baseline Evaluation of the Takunda 
Resilience Food Security Activity in 
Zimbabwe 
January 2023 | Volume I
IMPEL | Implementer-Led Evaluation & Learning Associate Award
ABOUT IMPEL
The Implementer-Led Evaluation & Learning Associate Award (IMPEL) works to improve the design and implementation of 
Bureau for Humanitarian Assistance (BHA)-funded resilience food security activities (RFSAs) through implementer-led 
evaluations and knowledge sharing. Funded by the United States Agency for International Development (USAID) BHA, IMPEL 
will gather information and knowledge in order to measure performance of RFSAs, strengthen accountability, and improve 
guidance and policy. This information will help the food security community of practice and USAID to design projects and 
modify existing projects in ways that bolster performance, efficiency, and effectiveness. IMPEL is a seven-year activity (2019–
2026) implemented by Save the Children (lead), TANGO International, Tulane University, Causal Design, Innovations for Poverty 
Action, and International Food Policy Research Institute. 
RECOMMENDED CITATION
IMPEL. (2023). Baseline Evaluation of the Takunda Resilience and Food Security Activity in Zimbabwe (Vol. I). Washington, DC: 
The Implementer-Led Evaluation & Learning Associate Award. 
PHOTO CREDITS
Takunda RFSA, Kudakwashe Murambadoro / CARE
DISCLAIMER
This report is made possible by the generous support of the American people through the United States Agency for 
International Development (USAID). The contents are the responsibility of the Implementer-Led Evaluation & Learning (IMPEL) 
award and do not necessarily reflect the views of USAID or the United States Government.
CONTACT INFORMATION
IMPEL Associate Award
c/o Save the Children
899 North Capitol Street NE, Suite #900
Washington, DC 20002
www.fsnnetwork.org/IMPEL
IMPEL@fsnnetwork.org
PREPARED BY:
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Acknowledgments i
ACKNOWLEDGMENTS
This report was written by Monserrat Lara, Daniele Barro, Stephanie Kabukwor Adjovu, and Antoine 
Guilhin from Innovations for Poverty Action; Lasse Brune (Northwestern University); Emily Beam 
(University of Vermont); and Craig McIntosh (University of California San Diego).
IMPEL | Implementer-Led Evaluation and Learning
ii Table of Contents
TABLE OF CONTENTS
Acknowledgments...............................................................................................i
List of Tables......................................................................................................iv
Acronyms...........................................................................................................vi
1. Introduction .................................................................................................1
2. Methodology................................................................................................3
2.1 Study Area ...................................................................................................................................3
2.2 Baseline Sample Size ...................................................................................................................3
2.3 Sampling Strategy........................................................................................................................3
2.4 Random Assignment and Balance ...............................................................................................4
2.5 Baseline Questionnaire Development.........................................................................................5
2.5.1 Household Survey ...................................................................................................................5
2.5.2 Anthropometric Survey...........................................................................................................6
3. Data Collection.............................................................................................7
3.1 Team Composition.......................................................................................................................7
3.2 Pilot Test Survey and Training .....................................................................................................7
3.3 Research Ethics and Data Quality Protocols................................................................................7
3.3.1 Ethics Review...........................................................................................................................8
3.3.2 Data Security and Encryption..................................................................................................8
3.4 Data Quality Protocols.................................................................................................................8
3.4.1 Audio Auditing.........................................................................................................................8
3.4.2 High-Frequency Checks...........................................................................................................8
3.4.3 Backchecks..............................................................................................................................8
3.5 Challenges in Data Collection ......................................................................................................9
3.5.1 General Background to the Survey .........................................................................................9
3.5.2 Absence of Children in the Anthropometry Sample ...............................................................9
3.5.3 Suspicion of Respondents Trying to Include Additional Children in their Household ............9
3.5.4 Poor Network Connectivity.....................................................................................................9
3.5.5 Rough Terrain..........................................................................................................................9
3.5.6 Access to Buhera District.......................................................................................................10
3.6 Challenges in the Sampling Strategy .........................................................................................10
4. Descriptive Statistics ..................................................................................11
4.1 Household Demographics .........................................................................................................11
4.2 Housing Quality .........................................................................................................................12
4.3 Water, Sanitation, and Hygiene.................................................................................................13
4.4 Sources of Income .....................................................................................................................14
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Table of Contents iii
4.5 Poverty Indicators......................................................................................................................16
4.6 Food Security.............................................................................................................................17
4.6.1 Food Consumption Score ......................................................................................................17
4.6.2 Food Insecurity Scale.............................................................................................................18
4.7 Children’s Nutritional Status .....................................................................................................18
4.8 Women’s Health and Family Planning.......................................................................................21
4.8.1 Women’s Diet Diversity.........................................................................................................21
4.8.2 Family Planning .....................................................................................................................21
4.9 Gender.......................................................................................................................................22
4.10 Women’s Empowerment...........................................................................................................23
4.11 Land and Agriculture .................................................................................................................25
4.11.1 Cultivated Crops................................................................................................................25
4.11.2 Fruit Trees.........................................................................................................................26
4.11.3 Farming During the Dry Season ........................................................................................27
4.11.4 Farming Inputs..................................................................................................................28
4.11.5 Farmer Groups..................................................................................................................29
4.11.6 Agricultural Sales ..............................................................................................................29
4.11.7 Off-Farm Business.............................................................................................................30
4.11.8 Agricultural Finance and Techniques................................................................................30
4.12 Asset Ownership........................................................................................................................31
4.13 Livestock ....................................................................................................................................32
4.14 Resilience...................................................................................................................................33
4.15 Access to Credit and Group Participation..................................................................................34
4.16 Financial Health .........................................................................................................................34
4.17 Savings and Loans......................................................................................................................35
4.18 Mental Health and Well-Being ..................................................................................................36
5. Next Steps..................................................................................................38
5.1 Process Evaluation.....................................................................................................................38
5.2 Outcome Monitoring Survey .....................................................................................................38
5.3 Final Evaluation .........................................................................................................................38
Annex A: Additional Descriptive Statistics.........................................................39
Annex B: Balance Test.......................................................................................52
VOLUME II
Annex C: Baseline Survey Tools
Annex D: Research Protocols
IMPEL | Implementer-Led Evaluation and Learning
iv List of Tables
LIST OF TABLES
Table 1. Surveys by province and district .....................................................................................................3
Table 2. Distribution of households by control and treatment group and disaggregated by SFECG-eligible 
and SFECG-ineligible stratum before baseline..............................................................................................4
Table 3. Distribution of households by control and treatment group and disaggregated by SFECG-eligible 
and SFECG-ineligible stratum after baseline.................................................................................................4
Table 4. Baseline instrument ........................................................................................................................5
Table 5. Demographics................................................................................................................................11
Table 6. Housing quality..............................................................................................................................12
Table 7. WASH indicators............................................................................................................................14
Table 8. Sources of income in the past 12 months.....................................................................................14
Table 9. Poverty indicators .........................................................................................................................16
Table 10. Food consumption in the past seven days..................................................................................17
Table 11. Food insecurity in the past 12 months........................................................................................18
Table 12. Young children: diet and health..................................................................................................18
Table 13. Anthropometric measurements..................................................................................................19
Table 14. Women's health ..........................................................................................................................21
Table 15. Family planning ...........................................................................................................................22
Table 16. Cash income and usage by gender..............................................................................................23
Table 17. Women empowerment...............................................................................................................24
Table 18. Women empowerment, women who earned cash ....................................................................24
Table 19. Crops cultivated...........................................................................................................................26
Table 20. Fruit-bearing trees.......................................................................................................................26
Table 21. Crops cultivated and type of irrigation used in the last dry season............................................27
Table 22. Farming inputs during the last rainy season ...............................................................................28
Table 23. Farmer groups.............................................................................................................................29
Table 24. Agricultural sales.........................................................................................................................30
Table 25. Off-farm business........................................................................................................................30
Table 26. Agriculture indicators in the past 12 months..............................................................................30
Table 27. Assets ..........................................................................................................................................31
Table 28. Livestock......................................................................................................................................32
Table 29. Livestock structures.....................................................................................................................32
Table 30. Resilience indicators....................................................................................................................33
Table 31. Gender access to credit and group participation........................................................................34
Table 32. Financial health ...........................................................................................................................35
Table 33. Savings.........................................................................................................................................35
Table 34. Loans ...........................................................................................................................................36
Table 35. Kessler Score, in the last 30 days ................................................................................................36
Table 36. Mental health in the last 12 months...........................................................................................36
Table 37. Food security...............................................................................................................................39
Table 38. Dietary diversity ..........................................................................................................................39
Table 39. Minimum acceptable diet, disaggregated...................................................................................40
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
List of Tables v
Table 40. Diet of minimum diversity, disaggregated..................................................................................41
Table 41. Diet of minimum diversity, by food group..................................................................................41
Table 42. WASH indicators full....................................................................................................................41
Table 43. Consumption, durables goods ....................................................................................................42
Table 44. Family planning detailed .............................................................................................................43
Table 45. Improved management practices or technologies on crops ......................................................44
Table 46. Improved management practices or technologies on livestock .................................................45
Table 47. Adaptive index.............................................................................................................................46
Table 48. Absorptive index .........................................................................................................................47
Table 49. Transformative index ..................................................................................................................47
Table 50. Education and training ................................................................................................................48
Table 51. Collective action..........................................................................................................................48
Table 52. Perceived economic ladder.........................................................................................................49
Table 53. Self-control..................................................................................................................................49
Table 54. Population lived under the poverty line, by province and district..............................................51
Table 55. Balance test.................................................................................................................................52
IMPEL | Implementer-Led Evaluation and Learning
vi Acronyms
ACRONYMS
BHA Bureau for Humanitarian Assistance
CI Confidence Interval
CPR Contraceptive Prevalence Rate
F&M Female and Male
FCS Food Consumption Score
FIES Food Insecurity Experience Scale
FNM Female no Male
HFC High-Frequency Checks 
HH Household
IE Impact Evaluation
IMPEL Implementer-Led Evaluation & Learning Associate Award
IPA Innovations for Poverty Action
IPM Integrated Pest Management
IRB Institutional Review Board
MAD Minimum Acceptable Diet
MDD-C Minimum Dietary Diversity – children
MDD-W Minimum Dietary Diversity – women
MNF Male no Female
MRZ Medical Research Council of Zimbabwe
NRM Natural Resource Management
ORT Oral Rehydration Therapy
PII Personal Identifying Information
PPP Purchasing Power Parity
QP Q-Partnership
RCT Randomized Controlled Trials
RCZ Research Council of Zimbabwe
RFSA Resilience Food Security Activity
SACCO Savings and Credit Cooperative Society
SFECG Supplementary Feeding and Expanded Care Groups
TVET Technical Vocational Education Training
USAID United State Agency for International Development
USD United States dollar
VC Village Cluster
VSLA Village Savings and Loan Association
WASH Water, Sanitation, and Hygiene
YSLA Youth Savings and Loan Association
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Introduction 1 
1. INTRODUCTION
Under the Implementer-Led Evaluation and Learning (IMPEL) Associate Award funded by the United 
States Agency for International Development (USAID) Bureau for Humanitarian Assistance (BHA), 
Innovations for Poverty Action (IPA) is conducting an impact evaluation of the Takunda resilience food 
security activity (RFSA) in Zimbabwe. CARE is implementing Takunda—“we have overcome” in Shona—
that aims to achieve sustainable, equitable, and resilient food, nutrition, and income security in 
Masvingo and Manicaland Provinces by improving income, nutritional status, and resilience to shocks 
and stressors of vulnerable households (HH), women, and youth. The 5-year Takunda RFSA has three 
main purposes:
• Purpose 1: Gender-equitable income for poor and vulnerable households. 
• Purpose 2: Health, nutrition, and sanitation for children, girls, and women. 
• Purpose 3: Resilience to shocks for poor and vulnerable households.
Each purpose encompasses various interventions, which may be layered to provide customized 
assistance to individuals depending on the number of targeting criteria they fulfill. Takunda will be 
implemented until the end of 2025 and will include the following core interventions:
• Life skills and business training, 
• Cash transfers, 
• Weirs irrigation and solar-powered gardens, 
• Farmer field business schools and producer groups strengthening, 
• Expanded care groups, 
• Village savings and loan associations (VSLAs) and youth savings and loan associations (YSLAs), 
• Training on health, nutrition, life, and leadership skills for adolescents, 
• Water, sanitation, and hygiene (WASH) interventions on infrastructure and community 
outreach, 
• Technical vocational education training (TVET), 
• Disaster risk management, and
• Natural resources management (NRM). 
The target population for the Takunda activity is extremely poor, chronically vulnerable, and high 
malnutrition-risk households living across Chivi and Zaka Districts (Masvingo Province) and Buhera and 
Mutare Districts (Manicaland Province). At baseline, 60% of the population lived under the $1.90 United 
States Dollar (USD) Purchasing Power Parity (PPP) poverty line, and 24% of the children were stunted in 
Masvingo and 35% in Manicaland. The eligibility of households for the various interventions depends on 
demographic characteristics and socioeconomic status.
Innovations for Poverty Action (IPA) is conducting an impact evaluation to assess the overall impacts of 
the Takunda activity. The study’s objective is to evaluate the cost-effectiveness of the Takunda activity
interventions in Chivi and Zaka districts (Masvingo Province) and Buhera and Mutare districts 
(Manicaland) on outcomes such as poverty reduction and child nutrition. 
IMPEL | Implementer-Led Evaluation and Learning
2 Introduction
IPA will measure the impact of the activity 36 months after baseline on the following primary outcomes: 
• Consumption,
• Dietary diversity,
• Child growth and development based on anthropometric measures,
• Food security,
• Assets value, and
• Subjective well-being.
The purpose of this report is to summarize the findings from the baseline survey. Section 2 of this report 
describes the research methodology, including the research design and the sampling strategy. Section 3
of the report discusses the field organization activities, including data collection and the supervision and 
monitoring of fieldwork. Section 4 describes the key findings of the baseline. The last section presents 
the next steps of the RFSA.
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Methodology 3 
2. METHODOLOGY 
2.1 Study Area 
The Takunda activity serves 301,636 individuals in 77,211 households, living in 92 wards spread across 
Chivi and Zaka Districts (Masvingo Province) and Buhera and Mutare Districts (Manicaland Province).
Within the 92 activity wards, villages are grouped into 564 village clusters (VCs). In this context, the 
randomized control trial (RCT) focuses on a subset of 161 VCs; this translates to 3,348 households living 
in 87 wards in 207 villages in the Masvingo and Manicaland Provinces. 
2.2 Baseline Sample Size
IPA field staff interviewed 3,348 households between April 30, 2022 and July 4, 2022. Enumerators 
conducted an average of 2.70 surveys per day. Table 1 below shows the number of completed 
household surveys by province and district. 
Table 1. Surveys by province and district
Province Districts Household Reached Completed
Manicaland
Mutare 1,265 100%
Buhera 1,007 100%
Masvingo
Chivi 543 100%
Zaka 533 100%
Total 3,348 100%
IPA’s field team also collected anthropometric measurements and captured the weight and height of
each child under 36 months. Of the 3,348 households surveyed, 1,646 (49%) had children eligible for 
anthropometric measurements. The anthropometric survey was not administered in 189 households
because the eligible children were absent. In total, 1,678 children in 1,457 households were weighed 
and measured. 
In addition, IPA randomly selected 42% of households to respond to the gender (cash), agriculture, 
sanitation and hygiene, and resilience modules; and 51% of households to answer the women’s health 
module and the module on contraceptives.
2.3 Sampling Strategy
Takunda is a multi-faceted activity comprising various interventions targeting households in the selected 
communities. The total number of participants and cost per participant of each intervention vary widely. 
Therefore, IPA’s survey strategy focused on the most “tightly” targeted1 component of the Takunda 
1 Here, “tightly targeted” refers to the fact that interventions differ in how precisely the target group can be described ex-ante
based on third-party easily verifiable characteristics. 
IMPEL | Implementer-Led Evaluation and Learning
4 Methodology
activity: Supplementary Feeding and Expanded Care Groups (SFECGs), which focuses on maternal and 
child nutrition for pregnant and lactating women and households with children under the age of 5. The 
SFECG intervention targeted at the SFECG-eligible stratum comprises an estimated 36% of all spending 
through Takunda activities. IPA targeted a random sample of 2,000 households from the SFECG-eligible
stratum and a random sample of 1,300 additional households from the “SFECG-ineligible” stratum. The 
SFECG-ineligible stratum refers to all other households eligible to participate in Takunda interventions 
but not eligible for the SFECG component, as described below.
To be SFECG-eligible, the household must have at least one pregnant or lactating woman or at least one 
child under 2 years old living in the household.
2.4 Random Assignment and Balance 
In consultation with CARE Zimbabwe and BHA, IPA designed an RCT to evaluate the impact of the 
Takunda activity and performed randomization at the VC level. RCTs are a rigorous evaluation method 
that can provide strong causal statements about whether an activity is achieving its goals. In an RCT, 
some members of a target population are randomly assigned to participate in an activity (treatment 
group) or not to participate (control group) during the study period. Because the selection between the 
two groups is random, they should, on average, be similar before the activity starts. Therefore, any 
difference between the two groups after the activity can be attributed to the activity itself. 
This design will allow measurement of the overall impact of the Takunda activity and determine which 
components are the most effective. For the impact evaluation and before the baseline, IPA assigned 
households to treatment and control groups by eligible and ineligible stratum, as indicated in Table 2
below.
Table 2. Distribution of households by control and treatment group and disaggregated by SFECG￾eligible and SFECG-ineligible stratum before baseline
Total Treatment Control
Number of VCs 161 88 73
SFECG-eligible households 2,000 1,212 788
SFECG-ineligible households 1,300 788 512
TOTAL target households 3,300 2,000 1,300
After collecting baseline data, IPA adjusted the number of households in each stratum and treatment 
group, as shown in Table 3. 
Table 3. Distribution of households by control and treatment group and disaggregated by SFECG￾eligible and SFECG-ineligible stratum after baseline
Total Treatment Control
Number of VCs 161 81 80
SFECG-eligible households 1,584 865 719
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Methodology 5 
Total Treatment Control
SFECG-ineligible households 1,764 966 798
TOTAL target households 3,348 1,831 1,517
The decrease in the proportion of SFECG-eligible households is due to an error in the sampling process 
that researchers addressed after they surveyed approximately 25% of the households. This led them to 
correctly label previously eligible households as ineligible. This process is explained more fully in 
section 3.6.
2.5 Baseline Questionnaire Development 
In consultation with CARE International, BHA, and the principal investigators, IPA developed the baseline 
questionnaire for the baseline survey of the Takunda Impact Evaluation. The primary objectives of the 
baseline survey are: (1) to assess the status of key indicators based on USAID's “Food for Peace 
Indicators Handbook” and other demographic variables to gain a better understanding of the prevailing 
conditions of the study population, (2) to improve the precision of the impact estimates, and (3) to 
collect necessary baseline information to measure heterogeneous treatment effects at endline. 
Enumerators administered the survey in Shona, the local language of Zimbabwe spoken in Masvingo and 
Manicaland. Responses were collected using the SurveyCTO data collection application. The baseline 
instrument had two parts: a household survey and an anthropometric survey. As per BHA guidelines, 
enumerators administered the household survey to the household head or a knowledgeable person in 
the household. In addition, specific survey modules were administered to selected household members, 
as shown in Table 4. The anthropometric survey was administered to children below 36 months.
2.5.1 Household Survey
The household survey was done through an in-person interview with an average duration of 1 hour and 
50 minutes. The instrument included the following modules: 
Table 4. Baseline instrument
Modules Module’s Respondent
Household Identification
Head of household or responsible adult household member if 
household head is absent
Household Screening
Consent for the main survey, GPS, 
audio recording
Verification and updating contacts
Household members
Cash for Asset
Children's Nutritional Status & 
Feeding Practices Primary caregivers of all children aged 0–59 months
IMPEL | Implementer-Led Evaluation and Learning
6 Methodology
Modules Module’s Respondent
Women's health, nutritional 
status, dietary diversity, and 
family planning
All women aged 18–49
Gender and cash All men and women who earned cash. No substitutes
Women empowerment Women in a relationship (married or living with a partner)
Household sources of income Head of household or responsible adult household member if 
household head is absent
Household food access
Person in charge of food preparation, who decides what food 
items the household should buy, what to eat on a specific day, 
and what quantities
Asset ownership
Head of household or responsible adult household member if 
household head is absent
Consumption
Livestock
Financial health
Basic farming
Agriculture All farmers with access to a plot of land and involved in the 
decision-making for that plot
Savings information
Head of household or responsible adult household member if 
household head is absent
Loan and savings
Mental health
Self-control
Water, sanitation, and hygiene
Resilience measurement
Access to credit and group 
membership 
Housing quality
Contraceptives All women aged 18–49
The modules were administered to household members as indicated in Table 4. Additionally, IPA 
administered certain modules (women’s health, gender, agriculture, sanitation and hygiene, resilience, 
and contraceptives) to a random subsample of 50% of the households to reduce overall survey length 
and respondent fatigue. See Annex 1 Baseline survey tools for a copy of the complete baseline survey.
2.5.2 Anthropometric Survey
The anthropometric survey was administered to all children aged between 0 and 36 months in sampled 
households. The anthropometric survey was administered by experienced nutritionists, one for each 
district team. The measurements were recorded on SurveyCTO at the end of the baseline questionnaire. 
The survey measurements included children’s heights, weights, mid-upper arm circumference (MUAC),
and edema; the measurements were recorded using SurveyCTO. 
IMPEL | Implementer-Led Evaluation and Learning
7 Data Collection
3. DATA COLLECTION 
3.1 Team Composition
The Takunda baseline data collection was led by IPA’s Senior Research Associate Daniele Barro. IPA’s
Senior Field Manager Patrick Simbewe led the training of the enumerators and the supervisors, and IPA 
Research Associate Monserrat Lara oversaw the data cleaning and programming of survey 
questionnaires. The principal investigators of this study are Emily Bream, Lasse Brune, Craig McIntosh, 
Dean Karlan, and Jessica Goldberg. 
Q-Partnership (QP) International was contracted by IPA for the primary data collection. QP organized a
team of 20 enumerators, four supervisors, eight anthropometric surveyors, one field manager, and two 
back checkers. The team was divided into four groups, with one group surveying each district. The 
baseline data was collected using the SurveyCTO application and uploaded to the SurveyCTO server 
every day of fieldwork.
3.2 Pilot Test Survey and Training 
To ensure data quality and well-functioning of the survey instruments, survey instruments were bench￾tested iteratively over more than two months before starting the data collection. IPA conducted a field 
pilot with 40 households in Mutare in April. The instrument was refined after debriefing with the field 
team after piloting. 
Enumerator training took place before the field pilot, running from 11th to 28th April 2022. The objective 
of the surveyor training was to equip the enumerators with the necessary skills to perform their role as 
part of a survey team and to become familiar with the baseline survey instrument and the guidelines 
that regulated the execution of the fieldwork. 
The training modules comprised theoretical, practical, and discussion-focused components. The theory 
component focused on fundamental principles and ethics of research, as well as fieldwork policies and 
protocols. The training facilitator went through the questionnaire, question by question, ensuring that 
each question was clearly understood and that enumerators could ask it the same way consistently in 
the language of the survey. The training agenda was designed to optimize discussions and to learn from 
previous personal experiences, as all the enumerators have previously been part of QP’s field surveys in 
the recent past. In addition, the training was as practical as possible regarding the use of devices, 
including anthropometry and tablet-based data collection and tracking of households. 
3.3 Research Ethics and Data Quality Protocols
To ensure research is conducted ethnically and that respondents’ privacy is protected, IPA implemented 
the following actions: 1) Institutional Review Board (IRB) review and approval of the study, and 2) data 
encryption. 
IMPEL | Implementer-Led Evaluation and Learning
8 Data Collection
3.3.1 Ethics Review
IPA submitted the research protocol, the survey instruments, and the consent form to the IPA IRB, the 
Medical Research Council of Zimbabwe (MRZ), and the Research Council of Zimbabwe (RCZ) for review 
and approval. These research ethics committees monitor the respect for the rights of the human 
subjects participating in the research. The data collection activities complied with the data security
protocol submitted to the IRB. 
3.3.2 Data Security and Encryption 
To ensure data security, IPA encrypted the data at all stages, and only researchers included in the IRB
approval were granted permission to access personally identifiable information (PII). Completed surveys 
remained encrypted as soon as enumerators finished an interview. Once transferred from the 
SurveyCTO server to IPA computers, Boxcryptor was used to keep files encrypted and secure on IPA’s 
institutional cloud account. 
3.4 Data Quality Protocols
To ensure the high quality of the data collected, IPA implemented the following actions: 1) audio 
auditing, 2) high-frequency checks (HFC), and 3) backcheck surveys. 
3.4.1 Audio Auditing
To ascertain data quality and with the authorization of the respondent, IPA randomly selected 10% of 
the surveys to be recorded and reviewed. The objective of audio auditing was to verify that interviews 
occurred and that none of the enumerators were fabricating data. Additionally, IPA used recordings to 
check the respondent's answers when there were discrepancies between the backcheck and the 
baseline survey.
3.4.2 High-Frequency Checks
IPA conducted daily high-frequency checks (HFC) to provide feedback to the field team and take 
corrective action when anomalies were found. IPA implemented HFCs to identify and resolve duplicate 
surveys, outliers in the data, logical inconsistencies, enumerators performing below or above average, 
and check compliance with protocols. When problems were detected, IPA research staff consulted with 
the survey team to clarify or correct the responses. HFCs were performed on incoming data using Stata, 
a statistical software program. 
3.4.3 Backchecks
The backcheck questionnaire consisted of a short version of the baseline instrument; a highly qualified 
surveyor revisited a random 10% subsample of households to administer this backcheck questionnaire. 
The goal was to check if the interview occurred, if there were discrepancies with the household survey 
administered by the enumerators, and to verify that enumerators followed protocols as per IPA 
guidance. Whenever discrepancies arose, supervisors, back-checkers, and enumerators met to 
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Data Collection 9
understand the origin of the discrepancy and find the correct answer. With this input, the IPA research 
team corrected the verified answer in the dataset and retrained enumerators if necessary.
3.5 Challenges in Data Collection
3.5.1 General Background to the Survey
During the period of data collection, the COVID-19 pandemic was still ongoing. However, in the regions 
sampled, most of the COVID-19 policies had been rolled back, there were no lockdowns and mask usage 
was rare. However, the economic impacts of the pandemic were quite acute: inflation was 66% in 
February, when the listing survey began, increased to 96.4% in April, and reached 256.9% in July when 
fieldwork finished. Similarly, fuel costs rose by over 30% during the same period.
There was also a false start to the rainy season in November 2021, after which the rainfall was erratic.
Anecdotally, many farmers encountered in the listing and baseline survey blamed this for causing a poor 
agricultural season.
3.5.2 Absence of Children in the Anthropometry Sample
Enumerators could not record the measurement of 262 children (14% of the 1,940 eligible children) 
because they were not present on the day of the interview. Some of the causes of absence reported by 
the households included travel to caregivers’ places of business or work, family events, or health 
centers.
3.5.3 Suspicion of Respondents Trying to Include Additional Children in 
their Household
When administering anthropometric surveys, the field team encountered cases where children initially 
recorded as household members seemed to belong to a different household. The absence of a birth 
certificate made the verification difficult. Therefore, the field team probed households and asked follow￾up questions about the children.
3.5.4 Poor Network Connectivity
Limited network connectivity slowed fieldwork, delaying form updates and household replacement 
updates. Teams had to travel tens of kilometers to obtain internet access needed to update forms,
curtailing productivity. If a team could not access the internet in the field, replacement surveys could 
not proceed. 
3.5.5 Rough Terrain
Teams were working in treacherous terrain, including in mountainous locations. In addition, the team in 
Zaka was warned of hyenas on the loose in the operating areas. This situation affected the movement 
plan and productivity. 
IMPEL | Implementer-Led Evaluation and Learning
10 Data Collection
3.5.6 Access to Buhera District 
District’s security agents denied access to the survey team in the Buhera district for 1 week. While 
negotiating access to Buhera with the District Development Coordinator and the head of the District 
Intelligence Officer, IPA deployed the Buhera research team to support the team working in Mutare.
District authorities finally granted the authorization to access Buhera on May 4, 2022, and fieldwork 
started on May 6, 2022. 
3.6 Challenges in the Sampling Strategy
The sampling objective of the impact evaluation initially targeted 2,000 households in the SFECG-eligible 
strata and 1,300 households in the SFECG-ineligible strata.
2 Deviations occurred during two separate 
phases of the research implementation: the listing exercise and the baseline survey. In the former case, 
the survey instruments were initially programmed to identify as eligible for the SFECG strata, 
households with children below 5 years old and not households with children below 2 years old. This 
divergence was addressed as soon as it was identified and affected roughly 25% of the listing sample. 
Therefore, the sampling strategy for the baseline was adjusted to account for this missing information.
3
 
The second deviation happened during the implementation of the baseline survey. Some households 
selected for the baseline did not match with the listing households. This issue led to an 11% difference 
between the number of households expected to be in the eligible stratum and actual households 
belonging to this stratum. However, the final difference was not only due to the mismatch but, in some 
instances, also to households misreporting the eligibility questions asked during the listing exercise, as 
discussed in section 3.4. 
The detour from the initial sampling plan may reduce the statistical power of the study to detect 
treatment impacts among the SFECG-eligible stratum,
4 but it does not affect the validity of estimates 
nor the overall power of the study.
2 Section 1.2 and Section 2, IMPEL. (2022). Sampling Plan and Pre-Analysis Plan: Takunda. Washington, DC: The Implementer￾Led Evaluation & Learning Associate Award.
3 The baseline sampling plan was implemented following an updated sampling randomization strategy to select the final sample 
for the baseline. 4 The sample size is lower for the CARE eligible stratum compared to the initial plan.
IMPEL | Implementer-Led Evaluation and Learning
11 Discriptive Statistics
4. DESCRIPTIVE STATISTICS 
This section offers a descriptive overview and findings for several key indicators of the 3,348 households 
interviewed in the study. The section covers demographics; housing; water, sanitation, and hygiene; 
sources of income; poverty levels; food security; children’s and women’s health; gender and women’s
empowerment; land and agriculture; assets; livestock; access to credit, savings and loans; mental health; 
and resilience.
The following tables present weighted means and their 95% confidence intervals (CI). Mean estimates 
are computed using sampling weights that reflect the probability of a given household being sampled 
and represent the population.5
4.1 Household Demographics
Table 5 shows descriptive statistics on household demographics. A minority of households are headed 
by women (37%), 69% of household heads are married and are aged 52 on average, and the average 
household size is five. Only half of the household heads have received some secondary school education
or more, and their main occupation is farming (56%). 
Table 5. Demographics
Description Mean CI Lower CI Upper N
Female HH head 37% 36% 39% 3,348
HH head age 52 51 52 3,348
HH head married or living together 69% 67% 70% 3,348
HH head education
No formal schooling 6% 5% 7% 3,348
Primary school completed 73% 71% 74% 3,348
Secondary school completed 26% 25% 28% 3,348
HH head occupation
Farming 56% 54% 58% 3,348
Self-employed non-farmer 25% 24% 27% 3,348
Unemployed 11% 10% 13% 3,348
Employed 3% 2% 3% 3,348
Other 5% 4% 5% 3,348
HH size 5.02 4.94 5.11 3,348
5 For more detailed information on the construction of particular indicators, please see BHA’s Indicator Handbook, Part I: 
Indicators for Baseline and Endline Surveys for Resilience Food Security Activities. 
IMPEL | Implementer-Led Evaluation and Learning
12 Discriptive Statistics
Description Mean CI Lower CI Upper N
Percentage of HH that have children …
Under 6 months 6% 5% 7% 3,348
between 6 months to 24 months 20% 19% 21% 3,348
between 2 and 5 years old 47% 45% 49% 3,348
between 6 and 18 years old 77% 76% 79% 3,348
Under 18 88% 87% 90% 3,348
Notes: The category “Farmer” includes commercial farming and subsistence farming. The category “Self-employed non-farmers” 
includes business people, boda drivers or taxi drivers, brewers, repairmen, market vendors, shopkeepers, fishermen, gold 
panners, carpenters, builders, mechanics, hairdressers, and miners. The category “Employed” include teachers, government 
employees, engineers, catering, non-governmental organization workers, and community health workers.
4.2 Housing Quality 
Most households own their house and live in mixed6 dwellings, predominantly with cement floors (82%), 
iron sheet roofs (48%), brick walls (61%), and three rooms on average. In addition, only 9% of 
households have access to electricity, and most of those with electricity access it through solar home 
systems or solar lantern lighting systems. 
Table 6. Housing quality
Description Mean CI Lower CI Upper N
Household’s tenure status7
Owner/purchaser without a title 58% 56% 60% 3,348
Owner/ purchaser with a title 36% 34% 38% 3,348
Other 6% 5% 7% 3,348
Type of dwelling
Mixed 53% 51% 55% 3,348
Detached 25% 23% 27% 3,348
Traditional 17% 15% 18% 3,348
Other 5% 4% 6% 3,348
6 A mixed dwelling is where a homestead has a combination of traditional (mud and thatch) and modern (brick wall with 
corrugated sheets, etc.) structures. The expectation is that most homesteads would classify as mixed because, for those with 
“modern” structures, there is often at least one dwelling/structure, usually the kitchen, which is mud and thatch/grass. 7 The majority of respondents to the survey were in a customary tenure setting, where the State owns the land but local 
traditional leaders (Chiefs, village headmen/headwomen, etc.) manage it. For example, when someone settles in a village, they 
need to get a letter from the village head and/or be written in the village register. Those who responded as "Owner/purchaser 
with a title" are actually formally registered with the traditional authorities and usually expected to pay an annual tax. Others
buy land or homesteads in the village, and these, too, have to be registered with the village head. Sometimes people register 
("Owner/purchaser with a title"), and sometimes they do not (Owner/purchaser without a title), but very few do not register 
with the Village head.
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 13 
Description Mean CI Lower CI Upper N
Main material for floor
Cement 82% 80% 83% 3,348
Earth/sand 10% 9% 11% 3,348
Dung 6% 5% 6% 3,348
Other 2% 2% 3% 3,348
Main material for roof
Metal/tin sheets 48% 46% 49% 3,348
Asbestos 36% 34% 37% 3,348
Thatch 15% 14% 16% 3,348
Other 2% 1% 2% 3,348
Main material for walls
Bricks 61% 59% 63% 3,348
Cement 35% 33% 37% 3,348
Other 4% 3% 5% 3,348
No. rooms in the HH 3.33 3.26 3.39 3,348
Access to electricity 9% 8% 10% 3,348
Sources of electricity:
Solar home system 39% 33% 45% 318
Solar lantern/Lighting system 37% 32% 43% 318
National grid 20% 14% 25% 318
Rechargeable battery 11% 7% 15% 318
Other 5% 3% 8% 318
Note: Households may have reported more than one source of electricity. Mixed includes detached and traditional.
4.3 Water, Sanitation, and Hygiene
To reduce overall survey length and respondent fatigue, the WASH section was administered to a 
random subsample of 1,420 households (42% of the sample). IPA collected data on four WASH 
indicators: (1) use of water treatment technologies, (2) practicing open defecation, (3) access to a basic 
sanitation service, and (3) availability of soap and water at a handwashing station on the premises. The 
WASH indicators are summarized below in Table 7. 
Only 6% of households used recommended household water treatment technologies, while 31% of 
households practiced open defecation. Interviewers observed the presence of water, soap, detergent, or 
another cleansing agent at the handwashing station in 9% of households. For more details about those 
indicators' subcomponents, see Table 42 in Annex A. 
IMPEL | Implementer-Led Evaluation and Learning
14 Discriptive Statistics
Table 7. WASH indicators
Description Mean CI Lower CI Upper N
[BL18] HH in target areas practicing correct use of 
recommended household water treatment 
technologies
6% 5% 8% 1,420
Chlorination 5% 4% 7% 1,420
Flocculant/disinfectant 1% 1% 2% 1,420
Filtration 0% 0% 0% 1,420
Solar disinfection 0% 0% 0% 1,420
[BL27] HH with access to a basic sanitation service 45% 42% 48% 1,420
Female and Male Adults (F&M) 45% 42% 48% 1,104
Adult Female no Adult Male (FNM) 43% 36% 49% 277
Adult Male no Adult Female (MNF) 51% 34% 69% 39
[BL19] HH in target areas practicing open defecation 31% 28% 34% 1,420
Female and Male Adults (F&M) 31% 28% 35% 1,104
Adult Female no Adult Male (FNM) 30% 24% 36% 277
Adult Male no Adult Female (MNF) 30% 15% 46% 39
[BL17] HH with soap and water at a handwashing 
station on the premises 9% 7% 11% 1,420
Female and Male Adults (F&M) 9% 7% 11% 1,104
Adult Female no Adult Male (FNM) 10% 6% 14% 277
Adult Male no Adult Female (MNF) 7% 0% 15% 39
Note: For the WASH indicators, we used a random subsample; 1,420 households were selected (42% of the sample).
4.4 Sources of Income 
Farming is the most important source of income or food for 40% of households. The most important 
sources of cash are non-agricultural self-employment (business) for 28% of the households and crop 
sales for 28%. 
Table 8. Sources of income in the past 12 months
Description Mean CI Lower CI Upper N
Sources of food/income
Crop production and sales 64% 62% 66% 2,814
Self-employment/own business (non-agricultural) 41% 39% 43% 2,814
Remittances 15% 14% 17% 2,814
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 15 
Description Mean CI Lower CI Upper N
Agricultural wage labor 13% 12% 15% 2,814
Livestock production/fattening and sales 13% 12% 15% 2,814
Non-agricultural wage labor 11% 9% 12% 2,814
Gifts/inheritance 10% 8% 11% 2,814
Salaried work 6% 5% 7% 2,814
Self-employment/own business (agricultural) 6% 5% 7% 2,814
Other 16% 14% 17% 2,814
Most important source of income or food
Crop production and sales 40% 38% 42% 2,814
Self-employment (non-ag.) 23% 22% 25% 2,814
Remittances 7% 6% 8% 2,814
Agricultural wage labor 6% 5% 6% 2,814
Non-agricultural wage labor 4% 3% 5% 2,814
Other 20% 18% 22% 2,814
Most important source of cash income
Self-employment (non-ag.) 28% 26% 30% 2,814
Crop production and sales 28% 26% 30% 2,814
Remittances 8% 7% 9% 2,814
Agricultural wage labor 6% 5% 7% 2,814
Non-agricultural wage labor 5% 4% 6% 2,814
Livestock production/fattening and sales 5% 4% 6% 2,814
Other 19% 17% 20% 2,814
Notes: Note that the sources of food or income sum more than 100% because some households had more than one source of 
income. The number of observations is 2,814 instead of 3,348 due to an error in skip patterns that was corrected after 1 week of 
surveying.
IMPEL | Implementer-Led Evaluation and Learning
16 Discriptive Statistics
4.5 Poverty Indicators
To assess the prevalence of poverty, IPA collected household-level consumption data to calculate three 
indicators: (1) daily per capita expenditures,
8 (2) the percentage of people living on less than $1.90 USD 
per day (at 2011 prices),
9 and (3) depth of poverty.
10 
The average daily per capita expenditure is $1.63. Households with at least one adult male but no adult 
female (MNF) reported more expenditures than other household types. Most households fall below the 
poverty line; 84% live below the $1.90/day 2011 poverty line. Refer to Table 52 in Annex A for details by 
province and district. 
Table 9. Poverty indicators
Description Mean CI Lower CI Upper N
[BL40] Daily per capita expenditures (as a proxy for 
income) in USG-assisted areas $1.63 $1.57 $1.70 3,348
Female and Male Adults (F&M) $1.50 $1.44 $1.56 2,571
Adult Female no Adult Male (FNM) $1.90 $1.73 $2.08 679
Adult Male no Adult Female (MNF) $2.72 $1.95 $3.48 98
[BL01] Prevalence of Poverty: people living on less than 
$1.90/day 2011 84% 82% 85% 3,348
Female and Male Adults (F&M) 86% 85% 88% 2,571
Adult Female no Adult Male (FNM) 77% 74% 81% 679
Adult Male no Adult Female (MNF) 63% 52% 73% 98
[BL02] Depth of Poverty of the Poor: mean percentage 
shortfall of the poor relative to the $1.90/day 2011 
poverty line
48% 46% 49% 2,837
Female and Male Adults (F&M) 52% 50% 53% 2,245
Adult Female no Adult Male (FNM) 36% 33% 38% 531
Adult Male no Adult Female (MNF) 22% 16% 29% 61
Note: The $1.90 threshold is inflated from 2011 to 2022 using the United States CPI to match the year of the data collection. No 
PPP adjustment factor was used, as the most recently available WB PPP deflator comes from 2018, before Zimbabwe instituted 
a new currency. The mean percentage shortfall of the poor indicates the percentage shortfall of the poor relative to the per 
capita PPP $1.90/day poverty line.
8 Per capita expenditure accounts for household consumption expenditures on food in the last 7 days, assets, and durable goods 
in the last 30 days, 3 months, 6 months, and 12 months.
9 PPP adjustment factor was not used. The most recent PPP deflator for Zimbabwe is from 2018, while a new currency was 
issued in 2019. This prevented an accurate PPP adjustment from being made.
10 Depth of poverty indicates that, for households that lie below the poverty line, how far they fall on average. Households with 
per capita consumption greater than $1.90 per day are not included in calculating mean percentage shortfall indicator. For more 
information on the construction of consumption poverty indicators, please see BHA’s Indicator Handbook, Part I: Indicators for 
Baseline and Endline Surveys for Resilience Food Security Activities. 
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 17 
4.6 Food Security 
This section describes the findings for food security, measured using the Food Consumption Score (FCS)
and Food Insecurity Scale (FIES). Table 37 in Annex A provides food security indicators details by 
household type. Additionally, IPA collected information on dietary diversity, see Table 38 in Annex A.
4.6.1 Food Consumption Score
The FCS is a weighted score measuring the diversity and frequency of the consumption of the different 
food groups ranging from 0 to 112, with higher scores indicating a higher degree of food security. The 
FCS is calculated by summing the household consumption of nine food groups (main staples, pulses, 
vegetables, fruit, meat and fish, milk and dairy, sugar, oil, and condiments) over the previous seven 
days.11 Households are categorized into three groups: poor consumption (≤ 21), borderline consumption 
(≥ 21.5 and ≤ 35), and (3) acceptable consumption (> 35), based on their FCS score. 
Households’ FCS is 43.14 on average, as shown in Table 11. Most households (59%) have an 
“acceptable” FCS. The FCS indicates that 4% of households showed poor food consumption, and 36% 
showed borderline food consumption. Main staples, vegetables, condiments, and oil are the most 
consumed food groups over a 7-day period. Refer to Table 37 in Annex A for details by gendered 
household type.
Table 10. Food consumption in the past seven days
Description Mean CI Lower CI Upper N
[BL10] Food Consumption Score (FCS) 43.14 42.46 43.82 3,348
[BL10] HH with poor FCS (0–21) 4% 4% 5% 3,348
[BL10] HH with borderline FCS (21.5–35) 36% 34% 38% 3,348
[BL10] HH with acceptable FCS (> 35) 59% 57% 61% 3,348
Over the past seven days, number of days HH
consumed ...
Main staples 6.83 6.80 6.86 3,348
Vegetables 6.07 6.00 6.13 3,348
Condiments 5.72 5.62 5.81 3,348
Oil 4.94 4.83 5.04 3,348
Sugar 3.72 3.61 3.84 3,348
Fruit 2.45 2.34 2.56 3,348
Pulses 1.79 1.71 1.86 3,348
Meat and fish 1.63 1.56 1.71 3,348
Milk and dairy 1.18 1.09 1.28 3,348
11 For more information on the FCS and FIES questionnaires and indicator construction, please see BHA’s Indicator Handbook, 
Part I: Indicators for Baseline and Endline Surveys for Resilience Food Security Activities.
IMPEL | Implementer-Led Evaluation and Learning
18 Discriptive Statistics
4.6.2 Food Insecurity Scale
The indicator for “moderate and severe food insecurity” measures the share of households that 
experienced food insecurity at moderate and severe levels in the past 12 months. The FIES comprises
eight questions that record difficulty accessing food due to lack of money or other resources. The results 
of the FIES module show that there are high levels of food insecurity. Overall, 82% of households 
experienced moderate or severe food insecurity, as shown in Table 37 in Annex A.
Table 11. Food insecurity in the past 12 months
Description Mean CI Lower CI Upper N
[BL06] Moderate and severe food insecurity, based on 
the Food Insecurity Experience Scale (FIES) 82% 67% 98% 3,348
Raw FIES Score (0–8) 5.74 5.66 5.81 3,348
During the last 12 months, because of a lack of money 
or resources, you, or others in your HH...
Ate only a few kinds of foods 90% 89% 91% 3,348
Were worried you would not have enough food to 
eat
89% 88% 90% 3,348
Were unable to eat healthy and nutritious food 89% 88% 90% 3,348
Ate less than you thought you should 88% 87% 89% 3,348
Had to skip a meal 70% 68% 71% 3,348
They were hungry but did not eat 57% 55% 59% 3,348
Did not have food 47% 46% 49% 3,348
Went without eating for a whole day 37% 35% 39% 3,348
Notes: The Raw FIES Score is a sum of the eight FIES binary questions (higher = more food insecure).
4.7 Children’s Nutritional Status 
To measure children's nutritional status, IPA collected child-level nutrition data to calculate four
indicators: (1) exclusive breastfeeding, (2) minimum acceptable diet, (3) diet of minimum diversity, and 
(4) children treated with Oral Rehydration Therapy (ORT). IPA administered the children’s nutrition
module to the caregivers of all children under 59 months who were present at the time of the survey. 
Table 13 shows that more than one-third (36%) of the children under 6 months are exclusively
breastfed. The baseline study data indicate that, among children between 6 and 23 months, 8% received 
a minimum acceptable diet (MAD), an indicator that tracks whether children had both sufficient 
frequency of meals and diversity of nutrients. Only one in four children between 6 and 23 months 
consumed a diet of minimum diversity (MDD-C). Furthermore, 22% of children under 5 suffered diarrhea 
2 weeks before the survey, and 69% of children with diarrhea were treated with ORT. See Table 39, 
Table 40, and Table 41 for further details on children’s diet and health. 
Table 12. Young children: diet and health
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 19 
Description Mean CI Lower CI Upper N
[BL13] Exclusive breastfeeding of children under six months 36% 30% 42% 276
Female 38% 30% 47% 146
Male 33% 24% 41% 130
[BL12] Children 6–23 months receiving a minimum 
acceptable diet (MAD) 8% 6% 10% 977
Female 7% 4% 9% 485
Male 9% 6% 12% 492
[BL39] Children 6–23 months consuming a diet of minimum 
diversity (MDD-C) 25% 22% 28% 977
Female 26% 22% 30% 485
Male 24% 20% 28% 492
[BL14] Children under five (0–59 months) who had diarrhea 
in the prior two weeks 22% 21% 24% 2,891
Female 22% 19% 24% 1,491
Male 23% 20% 25% 1,400
[Bl15] Children under five (0–59 months) with diarrhea 
treated with Oral Rehydration Therapy (ORT) 69% 65% 73% 662
Female 67% 61% 73% 323
Male 70% 65% 76% 339
Notes: The number of observations corresponds to the number of children.
IPA recorded anthropometric measurements of all children in the household under 36 months who were 
present at the time of the survey; 1,678 children were measured. The children’s heights and weights 
were recorded using standard anthropometric equipment (see Annex D for further details). Table 14
presents the findings disaggregated by gender and age.
Table 13. Anthropometric measurements
Description Mean CI Lower CI Upper N
[BL03] Wasted children (WHZ < -2) 2% 1% 3% 1,672
Female 2% 1% 3% 851
0–5 months 3% 0% 6% 141
6–11 months 3% 0% 6% 126
12–23 months 2% 0% 3% 325
24–35 months 2% 0% 3% 259
Male 3% 1% 4% 821
IMPEL | Implementer-Led Evaluation and Learning
20 Discriptive Statistics
Description Mean CI Lower CI Upper N
0–5 months 4% 0% 8% 125
6–11 months 4% 1% 6% 149
12–23 months 4% 0% 7% 314
24–35 months 1% 0% 3% 233
[BL04] Stunted children (HAZ <-2) 23% 20% 25% 1,674
Female 19% 16% 21% 853
0–5 months 6% 2% 9% 143
6–11 months 6% 2% 10% 126
12–23 months 19% 15% 23% 325
24–35 months 28% 22% 34% 259
Male 27% 23% 30% 821
0–5 months 7% 3% 11% 127
6–11 months 14% 8% 20% 148
12–23 months 28% 23% 34% 313
24–35 months 39% 32% 46% 233
[BL05] Healthy weight (WHZ ≤ 2 and ≥ -2) 92% 91% 94% 1,672
Female 94% 92% 95% 851
0–5 months 87% 81% 93% 141
6–11 months 93% 88% 98% 126
12–23 months 96% 94% 98% 325
24–35 months 95% 91% 98% 259
Male 91% 89% 93% 821
0–5 months 82% 75% 89% 125
6–11 months 90% 85% 95% 149
12–23 months 92% 87% 96% 314
24–35 months 95% 92% 98% 233
Notes: WHZ = weight-for-height z-score. HAZ = height-for-age z-score. Wasted is defined as having a WHZ less than -2. Stunted 
is defined as having a HAZ less than -2. Healthy weight is defined as having a WHZ greater than -2 and less than 2.
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 21 
4.8 Women’s Health and Family Planning
4.8.1 Women’s Diet Diversity
To reduce the overall survey length, the women’s health section was administered to a random 
subsample of households with at least one woman of reproductive age (between 18 and 49 years).
12
This section was administered to all women aged 18–49 within the selected households. In total, the 
data of 1,361 women from 1,061 households were recorded (32% of the sample).
As shown in Table 15, 31% of women attained a minimum dietary diversity (MDD-W), consuming at 
least 5 of 10 nutritionally diverse food groups during the previous day at the time of the survey. The 
primary food groups consumed were grains, white roots, tubers and plantains, dark green leafy 
vegetables, and other vegetables.
Table 14. Women's health
Description Mean CI Lower CI Upper N
[BL11] Women of reproductive age consuming a diet of 
minimum diversity (MDD-W) 31% 28% 34% 1,181
Group food:
Grains, white roots, tubers, and plantains 98% 97% 99% 1,181
Dark green leafy vegetables 73% 70% 76% 1,181
Vegetable 53% 50% 56% 1,181
Fruits 39% 36% 42% 1,181
Pulses 39% 36% 42% 1,181
Other vitamin A-rich fruits and vegetables 28% 25% 31% 1,181
Meat, poultry, and fish 24% 21% 27% 1,181
Dairy 18% 15% 20% 1,181
Nuts and seeds 6% 5% 8% 1,181
Eggs 5% 3% 6% 1,181
Notes: For the MDD indicator, we used a random subsample of households with at least one woman of reproductive age; 1,361 
household members of 1,061 households were selected (32% of the sample); the data is shown at the women’s level.
4.8.2 Family Planning
For family planning indicators, IPA gathered data from a random subsample of households with at least 
one woman of reproductive age in a union;
13 1,172 women were interviewed in 1,139 households (34% 
of the sample). 
12 Based on USAID's "Food for Peace Indicators Handbook,” this section should have been administered to all women aged 15–
49. However, following IRB rules, we are not allowed to interview individuals under 18. 13 In union means currently married or living together with their partner.
IMPEL | Implementer-Led Evaluation and Learning
22 Discriptive Statistics
Table 16 indicates high levels of knowledge about contraceptive methods, but only 50% of the 
respondents reported making decisions about contraceptive usage in the past 12 months. Family 
planning decision-making is lower for women between 18 and 29 years old (44%). The Contraceptive 
Prevalence Rate (CPR) is measured by the share of non-pregnant women of reproductive age (18–49 
years)
14 who are married or in a union who are currently using (or whose sexual partner is using) at least 
one contraceptive method is 71%. More details can be found in Table 44 in Annex A. Among women 
using contraceptive methods, almost all of them used modern methods.
Table 15. Family planning
Description Mean CI Lower CI Upper N
[BL36] Women in a union who have knowledge of 
modern family planning methods that can be used to 
delay or avoid pregnancy
95% 93% 97% 951
Women ages 18–29 94% 91% 97% 367
Women ages 30–49 96% 94% 98% 584
[BL37] Women in a union who made decisions about 
modern family planning methods in the past 12 
months
50% 46% 53% 951
Women ages 18–29 44% 38% 50% 367
Women ages 30–49 53% 48% 57% 584
[BL20] Contraceptive Prevalence Rate (CPR) 71% 67% 75% 797
Women using Modern Methods 99% 98% 100% 569
Women using Traditional methods 1% 0% 2% 569
Notes: Indicator BL20 was asked to non-pregnant women aged 18–49 
4.9 Gender 
The share of women and men in a union who earned cash in the past 12 months was measured for all 
women and men in a union; the other gender indicators were measured for a random subsample of 
households with at least one member in a union who earned cash in the last 12 months. In total, 391 
household members in 352 households (11% of the sample) were interviewed, 227 women and 164 
men.
Baseline data indicate a large gender gap in cash earned; more than half of men in a union reported 
earning cash in the past 12 months, compared to 28% of women. Among women in a union who earned 
cash, only 51% reported participating in decisions about using their partner’s self-earned cash.
14 Based on USAID's "Food for Peace Indicators Handbook,” this section should have been administered to all women aged 15–
49. However, following IRB rules, we are not allowed to interview individuals under the age of 18.
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 23 
Table 16. Cash income and usage by gender
Description Mean CI Lower CI Upper N
[BL32] Women and men in a union who earned cash in the 
past 12 months 40% 38% 42% 5,280
Female 29% 27% 31% 2,855
Male 53% 51% 55% 2,425
[BL33] Women in a union and earning cash who report part. 
In decisions about the use of self-earned cash 84% 79% 90% 227
< = 29 years old 83% 74% 92% 69
30–49 years old 85% 78% 92% 126
> 49 years old 84% 69% 99% 32
[BL34] Women in a union and earning cash who report part. 
In decisions about the use of spouse/partner’s self-earned 
cash
51% 44% 59% 227
< = 29 years old 51% 36% 65% 69
30–49 years old 53% 43% 64% 126
> 49 years old 44% 24% 64% 32
[BL35] Men in a union earning cash who report 
spouse/partner part. In decisions about the use of self￾earned cash
88% 83% 94% 164
< = 29 years old — — — 18
30–49 years old 89% 82% 95% 113
> 49 years old 92% 83% 100% 33
Notes: For the gender indicators, BL32 indicators were administered to each household member in a union. For the other gender 
indicators, we used a random subsample of households with at least one household member in a union who earned cash over 18 
years of age; 391 household members were interviewed in 352 households (11% of the sample), 227 women, and 164 men. Part. 
= participation.
4.10 Women’s Empowerment
The women’s empowerment module was administered to every woman in a union over 18: 2,346 
women answered the questions in this section, and Table 17 shows summary statistics for this module.
Almost half (44%) of women must ask permission to buy clothes for themselves, and only 30% of women 
are allowed to visit women from other villages without permission. Table 18 shows the same results but 
conditional on women who earned cash.
IMPEL | Implementer-Led Evaluation and Learning
24 Discriptive Statistics
Table 17. Women empowerment
Description Mean CI Lower CI Upper N
When HH makes a major purchase, the wife's opinion is heard in deciding what to buy
Yes, always 43% 41% 46% 2,346
Yes, usually 14% 12% 15% 2,346
Yes, sometimes 24% 22% 26% 2,346
Rarely 9% 8% 11% 2,346
Very rarely 1% 1% 2% 2,346
Never 8% 7% 9% 2,346
The wife has to ask other HH members for permission to buy clothes for herself
Yes 52% 50% 54% 2,346
No 44% 42% 46% 2,346
Have never bought 4% 3% 5% 2,346
The wife is allowed to buy things in the market without asking permission
Yes, always 51% 49% 54% 2,346
Yes, usually 10% 9% 12% 2,346
Yes, sometimes 14% 12% 15% 2,346
Rarely 6% 5% 7% 2,346
Very rarely 1% 1% 2% 2,346
Never 16% 15% 18% 2,346
The wife is allowed to visit women from other villages to talk to them without asking permission
Yes, alone, do not need permission 31% 29% 33% 2,346
Yes, alone, with permission 48% 46% 51% 2,346
Yes, but never alone 2% 1% 2% 2,346
Never 16% 14% 18% 2,346
Notes: This section was administered to each married woman over 18 years of age; the data is shown at the women's level. 
Confidence intervals for binary indicators are based on Normal approximations; for very small samples and indicator values near 
0 or 1, these confidence intervals can exceed 0 or 1, and in this table, confidence interval bounds are censored at 0 from below 
and 1 from above. 
Table 18. Women empowerment, women who earned cash
Description Mean CI Lower CI Upper N
When HH makes a major purchase, the wife's opinion is heard in deciding what to buy
Yes, always 38% 33% 42% 670
Yes, usually 17% 13% 20% 670
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 25 
Description Mean CI Lower CI Upper N
Yes, sometimes 26% 22% 30% 670
Rarely 9% 7% 12% 670
Very rarely 1% 0% 2% 670
Never 9% 6% 12% 670
Wife has to ask other HH members for permission to buy clothes for her
Yes 47% 43% 52% 670
No 51% 47% 56% 670
Have never bought 1% 0% 2% 670
Wife is allowed to buy things in the market w/o asking permission
Yes, always 48% 44% 53% 670
Yes, usually 12% 10% 15% 670
Yes, sometimes 14% 11% 17% 670
Rarely 6% 4% 8% 670
Very rarely 1% 0% 2% 670
Never 18% 15% 21% 670
Wife is allowed to visit women from other villages to talk to them w/o asking permission
Yes, alone, do not need permission 33% 29% 38% 670
Yes, alone, with permission 45% 41% 50% 670
Yes, but never alone 1% 0% 2% 670
Never 18% 14% 21% 670
Notes: Confidence intervals for binary indicators are based on Normal approximations; for very small samples and indicator 
values near 0 or 1, these confidence intervals can exceed 0 or 1, and in this table, confidence interval bounds are censored at 0 
from below, and 1 from above.
4.11 Land and Agriculture
4.11.1 Cultivated Crops 
IPA collected household-level information about basic farming practices, land ownership, and crop 
cultivation. A high percentage of households (82%) own land, and an even higher percentage (94%) 
report cultivating anything in the last 12 months. Among those households that engaged in cultivation, 
the most commonly cultivated crops are maize (76%), groundnuts (69%), roundnuts (54%), and 
sorghum/millet (53%). 
IMPEL | Implementer-Led Evaluation and Learning
26 Discriptive Statistics
Table 19. Crops cultivated
Description Mean CI Lower CI Upper N
HH owns the land, not including the plot where the 
house is 82% 81% 83% 3,348
Size of agricultural land (in acres) 4.80 4.65 4.96 2,723
HH cultivated anything in the last 12 months 94% 93% 95% 3,348
Crops cultivated during the last rainy season
Maize 76% 75% 78% 3,158
Groundnuts 69% 68% 71% 3,158
Roundnuts 54% 52% 56% 3,158
Sorghum or millet 53% 51% 55% 3,158
Cowpeas 47% 45% 49% 3,158
Vegetables 18% 17% 20% 3,158
Sweet potatoes 17% 15% 18% 3,158
Tomatoes 10% 9% 11% 3,158
Bean 7% 6% 8% 3,158
Rapoko 6% 5% 7% 3,158
Sunflower 6% 5% 7% 3,158
Other 17% 15% 18% 3,158
4.11.2 Fruit Trees
Furthermore, IPA collected information on fruit-bearing trees. Among all households that reported 
cultivating something in the past 12 months, three-quarters have fruit-bearing trees. Mango (70%), 
guava (36%), and musau (33%) are the main fruit trees harvested. 
Table 20. Fruit-bearing trees
Description Mean CI Lower CI Upper N
HH owns any fruit-bearing trees 78% 76% 79% 3,158
Kind of fruit trees:
Mango 70% 68% 72% 2,396
Guava 36% 34% 38% 2,396
Musau 33% 31% 35% 2,396
Lemon 24% 22% 26% 2,396
Paw 19% 17% 21% 2,396
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 27 
Description Mean CI Lower CI Upper N
Orange 19% 17% 21% 2,396
Mulberry tree 16% 15% 18% 2,396
Banana 14% 13% 16% 2,396
Mushuku 13% 12% 15% 2,396
Avocado 13% 11% 14% 2,396
Peach 7% 6% 8% 2,396
Nartjies 5% 4% 6% 2,396
Other 21% 19% 22% 2,396
Notes: This section was administered to HH, who cultivated anything in the last 12 months. Note that some households have 
more than one kind of tree. 
4.11.3 Farming During the Dry Season
About half (58%) of households cultivated any land in the last dry season, with the three most common
crops in the dry season being covo, tomatoes, and onions. Regarding the types of irrigation used, Table 
21 shows that 94% of the households used water cans to irrigate crops.
Table 21. Crops cultivated and type of irrigation used in the last dry season
Description Mean CI Lower CI Upper N
Cultivated any land in the last dry season 58% 57% 60% 3,158
Crops cultivated in the last dry season:
Covo 68% 66% 71% 1,844
Tomatoes 61% 59% 64% 1,844
Onions 46% 43% 48% 1,844
Tsunga 44% 41% 47% 1,844
Rape 38% 36% 41% 1,844
Beans 26% 24% 28% 1,844
Green vegetables 12% 10% 14% 1,844
Sweet potatoes 11% 10% 13% 1,844
Green maize 10% 8% 11% 1,844
Cabbage 9% 7% 10% 1,844
Carrots 8% 7% 10% 1,844
Other 17% 15% 19% 1,844
Type of irrigation used last dry season:
Water can 94% 93% 96% 1,844
IMPEL | Implementer-Led Evaluation and Learning
28 Discriptive Statistics
Description Mean CI Lower CI Upper N
Other 4% 3% 6% 1,844
None 2% 1% 2% 1,844
Notes: This section was administered to HH who cultivated anything in the last 12 months.
4.11.4 Farming Inputs 
Among households that cultivated something in the last 12 months, 13% hired any labor, 32% rented 
any farming equipment, and 44% rented any farming animals. Organic fertilizer is used by 74% of 
households, and inorganic fertilizer by 65%. 
Table 22. Farming inputs during the last rainy season
Description Mean CI Lower CI Upper N
Hired any labor to help with any farming tasks 13% 11% 14% 3,158
Rented any farming equipment 32% 30% 33% 3,158
Rented any farming animals 44% 43% 46% 3,158
Used any inorganic fertilizer 65% 63% 67% 3,158
...bought any inorganic fertilizer 21% 19% 23% 2,067
Source of this purchase:
Commercial provider 86% 83% 90% 408
Government outlet/extension services 8% 5% 11% 408
Cooperative 1% 0% 2% 408
Other 8% 5% 11% 408
Used any organic fertilizer 74% 72% 76% 3,158
...bought any organic fertilizer 4% 3% 4% 2,310
Source of this purchase:
Commercial provider 12% 3% 21% 81
Government outlet/extension services 8% 1% 15% 81
Cooperative 4% 0% 9% 81
Other 76% 65% 87% 81
Used packaged seeds 65% 63% 67% 3,158
...bought packed seeds 35% 33% 38% 2,039
Source of this purchase:
Commercial provider 91% 89% 93% 704
Government outlet/extension services 6% 4% 8% 704
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 29 
Description Mean CI Lower CI Upper N
Cooperative 1% 0% 2% 704
Other 3% 2% 5% 704
Used any pesticides or herbicides 26% 24% 28% 3,158
...bought any pesticides or herbicides 26% 23% 29% 827
Source of this purchase:
Commercial provider 84% 78% 90% 211
Government outlet/extension services 8% 4% 12% 211
Cooperative 1% 0% 3% 211
Other 7% 3% 11% 211
Notes: This section was administered to HH who cultivated anything in the last 12 months. Farming equipment includes ox carts, 
tractors, hand plows, ridgers, or other major farming equipment. Farming animals include cattle or donkeys. Confidence 
intervals for binary indicators are based on Normal approximations; for very small samples and indicator values near 0 or 1, 
these confidence intervals can exceed 0 or 1, and in this table, confidence interval bounds are censored at 0 from below 1 one 
from above.
4.11.5 Farmer Groups 
IPA asked about membership in farmer groups, cooperatives, and other group activities to households 
that cultivated crops in the 12 months before the survey. About one-quarter of households (22%) are 
members of farmer groups or cooperatives, and only 6% of households met in the previous rainy season 
to organize the sale of farm products as a group.
Table 23. Farmer groups
Description Mean CI Lower CI Upper N
HH is a member of a farmer group/cooperative 22% 20% 23% 3,158
Met with other farmers last rainy season to organize 
some sales as a group 6% 5% 7% 3,158
Activities performed to organize sales as a group:
Find markets or buyers with good prices 53% 45% 61% 187
Share transport to market 27% 20% 34% 187
Negotiate prices 15% 9% 20% 187
Call buyer to pick up the crops 12% 7% 17% 187
Other 14% 8% 20% 187
Notes: This section was administered to HH who cultivated anything in the last 12 months. 
4.11.6 Agricultural Sales
Table 24 shows that a small share of the households (12%) sold any crops. The most common buyers of 
crops are relatives or friends (40%) and local traders that don’t go to the market (33%). 
IMPEL | Implementer-Led Evaluation and Learning
30 Discriptive Statistics
Table 24. Agricultural sales
Description Mean CI Lower CI Upper N
HH sold any crops 12% 11% 14% 3,158
HH sold more than half of the total output 36% 31% 41% 389
Main buyer:
Relative/Friend 40% 34% 45% 389
Local trader not at the market 33% 27% 38% 389
Local trader at the market 10% 6% 13% 389
Out-of-town mobile trader 8% 5% 11% 389
Other 10% 7% 14% 389
Notes: This section was administered to HH who cultivated anything in the last 12 months.
4.11.7 Off-Farm Business
In the off-farm business module, IPA asked households about their experiences with owning a business. 
“Off-farm business” refers to non-agricultural income-generating activities, including those that produce 
or trade goods or services. The baseline data indicate that business ownership was low (14%), and 
among those owning a business, only 38% had inventories worth more than $50 USD. 
Table 25. Off-farm business
Description Mean CI Lower CI Upper N
Household operates a business 14% 13% 15% 3,348
Household operates more than one business 22% 17% 26% 468
Number. of years operating the main business 6.93 5.93 7.94 468
Value of business inventory is > $50 USD 38% 34% 43% 468
4.11.8 Agricultural Finance and Techniques
This section was administered to a random subsample of households with at least one farm worker in 
the household; 1,424 household members were interviewed in 1,227 households (37% of the sample). 
Access to financial services is limited; only 17% of farmers used any financial services in the past 12 
months, as shown in Table 26. All the farmers use at least three agricultural improved management
practices, and 76% use improved management practices or technologies for livestock. 
Table 26. Agriculture indicators in the past 12 months
Description Mean CI Lower CI Upper N
[BL29] Farmers who used financial services (savings, 
agricultural credit, or agricultural insurance) in the past 12 
months
17% 15% 19% 1,424
Female 18% 15% 21% 1,014
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 31 
Description Mean CI Lower CI Upper N
Male 15% 11% 19% 410
[BL21] Producers who have applied targeted improved 
management practices or tech. on crops 100% 100% 100% 1,355
Female 100% 100% 100% 962
Male 100% 99% 100% 393
[BL21] Producers who have applied targeted improved 
management practices or technologies. on livestock 76% 72% 79% 721
Female 74% 69% 78% 497
Male 81% 74% 87% 224
Notes: For the Agriculture indicators, we used a random subsample of households with at least one farm worker in the 
household; 1,424 household members were interviewed in 1,227 households (37% of the sample); the data is shown at the farm 
worker level. The number of observations for the BL21 indicator on crops corresponds to the number of farm workers growing 
crops, while the number of observations for the BL21 livestock indicator corresponds to the number of farm workers engaged in
livestock activities. Improved management practices includes: practices for cultivation (“Micro dosing,” “Manure,” “Compost,” 
“Planting basins,” “Mulching,” “Weed control,” “Dry planting,” “Ripping into residues,” “Clean ripping,” “Tied ridges,” “Pot￾holing,” “Crop rotations,” “Intercropping,” “Integrated Pest Management (IPM)”, “Early planting or planting with first rains”, 
“Use of improved crop varieties,” “Dead level contours,” “Ridging,” “Double dug beds/fertility trenches”), natural resource 
management practices (“Management or protection of watersheds or water catchments,” “Agro-forestry,” “Management of 
forest plantation/woodlands,” “Regeneration of natural landscapes,” “Sustainable harvesting of forest products,” “Development 
and implementation of NRM by laws”), and methods to store (“Locally made storage structures such as sheet metal silos,” 
“Sealed/air tight bags,” “Community storage facilities, including warehouse receipting,” “Use of solar or fuel-powered dryers to 
reduce post-harvest moisture,” “Seed or grain treatment techniques including botanical pest control agents or phytosanitary 
irradiation,” “Grain treatment with agro-chemicals,” “Other post-harvest practices that reduce pre-storage losses”).
4.12 Asset Ownership
IPA asked households about asset ownership. Nearly all households own a cellphone (91%), and half of 
the households own a solar panel.
Table 27. Assets
Description Mean CI Lower CI Upper N
HH owns:
Land 95% 94% 96% 3,348
Cellphone 91% 90% 92% 3,348
Hoe 80% 78% 81% 3,348
Ax 63% 61% 64% 3,348
Goat 54% 52% 56% 3,348
Solar panel 50% 48% 52% 3,348
Cattle 39% 37% 41% 3,348
Plough 38% 36% 40% 3,348
IMPEL | Implementer-Led Evaluation and Learning
32 Discriptive Statistics
Description Mean CI Lower CI Upper N
Wheelbarrow 30% 28% 31% 3,348
Radio 25% 23% 27% 3,348
Lounge suite 19% 17% 20% 3,348
Scotch cart/Water cart 17% 16% 19% 3,348
Knapsack sprayer 13% 12% 15% 3,348
Bicycle 13% 12% 14% 3,348
Cultivator 8% 7% 9% 3,348
Plantation/Orchard 8% 7% 9% 3,348
Television 6% 5% 7% 3,348
Donkey 5% 4% 6% 3,348
Other 15% 14% 16% 3,348
4.13 Livestock
Most households (87%) reported owning some livestock in the past 12 months. The main types of 
livestock owned are poultry (91%), goats (62%), cattle (45%), and turkey (20%). Regarding livestock 
structure, 77% of households own a bird pen, 54% own a goat house, and 44% own a cow house. 
Table 28. Livestock
Description Mean CI Lower CI Upper N
HH owned livestock in last 12 months 87% 86% 88% 3,348
Livestock owned by HH
Poultry 91% 90% 92% 2,907
Goats 62% 60% 64% 2,907
Cattle 45% 43% 47% 2,907
Turkey 20% 18% 22% 2,907
Donkey/mule 6% 5% 7% 2,907
Other 15% 14% 17% 2,907
Table 29. Livestock structures
Description Mean CI Lower CI Upper N
HH owns livestock structures:
Bird Pen/Coop 77% 76% 79% 3,348
Goat house/Goat pen 54% 52% 56% 3,348
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 33 
Description Mean CI Lower CI Upper N
Cow house/kraal 44% 42% 46% 3,348
Turkey 11% 10% 12% 3,348
Pigeons 9% 8% 10% 3,348
Other 9% 7% 10% 3,348
None 13% 12% 14% 3,348
4.14 Resilience
To reduce overall survey length and respondent fatigue, the resilience section—except for the indicator 
BL31—was administered to a random subsample of 1,420 households (42% of the sample). IPA collected 
data on seven resilience indicators: (1) households' ability to recover from shocks, (2) households` belief 
in local government responsiveness, (3) household participation in group-based savings, (4) adaptive, (5) 
absorptive, (6) transformative and (7) social capital indices. The resilience indicators are summarized by 
gendered household type below in Table 30. See Table 47, Table 48, and Table 49 in Annex A for more 
details on the sub-components of each indicator.
Table 30. Resilience indicators
Description Mean CI Lower CI Upper N
[BL23] Ability to recover from shocks and stresses index (2–6) 2.67 2.61 2.74 1,390
Female and Male Adults (F&M) 2.64 2.56 2.71 1,082
Adult Female no Adult Male (FNM) 2.73 2.57 2.89 270
Adult Male no Adult Female (MNF) 3.18 2.65 3.70 38
[BL24] Percentage of HH believing that the local government. 
will respond effectively to future shocks and stresses
64% 61% 67% 1,420
Female and Male Adults (F&M) 63% 59% 66% 1,104
Adult Female no Adult Male (FNM) 68% 61% 74% 277
Adult Male no Adult Female (MNF) 73% 57% 89% 39
[BL31] HH participates in group-based savings, micro-finance,
or lending programs 7% 6% 8% 3,348
Female and Male Adults (F&M) 8% 7% 9% 2,571
Adult Female no Adult Male (FNM) 6% 4% 7% 679
Adult Male no Adult Female (MNF) 2% 0% 5% 98
[BL08] Adaptive capacity index (0–100) 45.34 44.39 46.28 1,217
[BL09] Absorptive capacity index (0–100) 38.49 37.53 39.46 1,322
[BL25] Transformative capacity index (0–100) 38.54 37.45 39.64 1,267
IMPEL | Implementer-Led Evaluation and Learning
34 Discriptive Statistics
Description Mean CI Lower CI Upper N
[BL38] Index of social capital (0–6) 2.61 2.50 2.71 1,420
Bridging social capital Index (0–6) 2.38 2.27 2.49 1,420
Bonding Social Capital Index (0–6) 2.84 2.73 2.95 1,420
Notes: This section—except for the indicator BL31—was administered to a random subsample of respondents; 1,420 households 
were selected (42% of the sample). BL08 has fewer observations because one of the subcomponents only applies to some 
respondents; see Table 47 in Annex A. BL23, BL25, and BL38 indicators have different subsamples due to skipping pattern issues 
corrected after a few days.
4.15 Access to Credit and Group Participation
The gender access to credit and group participation module was administered to a random subsample of 
households with at least one household member in a union; 1,361 household members were 
interviewed in 1,061 households (32% of the sample). One-third of the respondents had access to credit; 
of those, 90% made credit decisions. 
Table 31. Gender access to credit and group participation
Description Mean CI Lower CI Upper N
[BL42] Women/men in a union with access to credit 33% 31% 36% 1,361
Female 34% 31% 38% 965
Male 32% 26% 37% 396
[BL43] Women/men in a union who make decisions 
about credit 90% 87% 93% 466
Female 89% 85% 93% 335
Male 93% 87% 98% 131
[BL41] Women/men in a union who are members of a 
community group 56% 53% 60% 1,014
Female 58% 54% 62% 715
Male 52% 46% 59% 299
Notes: For the Gender Access to Credit indicators, we used a random subsample of households with at least one household 
member in a union; 1,361 household members were interviewed in 1,061 households (32% of the sample).
4.16 Financial Health 
The financial health module was included to determine household access to financial resources to deal 
with emergencies. IPA asked the households how difficult it would be to come up with USD 20 within 30 
days and 7 days, as well as the source of this money. Three-quarters of households reported that it
would be “very difficult” to come up with that amount of money in the next 30 days, and almost all 
households (94%) indicated it would be “very difficult” to come up with that amount of money in the 
next seven days. The main source of funds is family, relatives, or friends (27% of households). 
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 35 
Table 32. Financial health
Description Mean CI Lower CI Upper N
Difficulty coming up w/ USD 20 in the next 30 days
Not difficult at all 3% 2% 3% 3,348
Somewhat difficult 25% 23% 26% 3,348
Very difficult 72% 71% 74% 3,348
Difficulty coming up w/ USD 20 in the next 7 days
Not difficult at all 1% 1% 1% 3,348
Somewhat difficult 5% 4% 5% 3,348
Very difficult 94% 93% 95% 3,348
Main source of funds
Family, relatives, or Friends 27% 25% 28% 3,348
Selling livestock 21% 19% 22% 3,348
Money from working 13% 12% 15% 3,348
Selling assets 7% 6% 8% 3,348
Bank, employer, or private lender (borrow) 5% 4% 5% 3,348
Other sources 11% 9% 12% 3,348
Could not come up with the money 21% 19% 22% 3,348
4.17 Savings and Loans 
More than half (56%) of households did not save money in the 6 months before the survey. Those who 
saved mostly kept their savings informally, either in their pockets or clothes or in a secret place at home. 
Table 33. Savings
Description Mean CI Lower CI Upper N
Has kept any savings in the past six months
No savings 56% 54% 58% 3,348
In your pocket, your clothes, or in a bag that you carry 25% 24% 27% 3,348
A secret place in your home 18% 16% 19% 3,348
VSLA 11% 10% 13% 3,348
Mobile money 7% 6% 8% 3,348
Other 9% 8% 10% 3,348
IMPEL | Implementer-Led Evaluation and Learning
36 Discriptive Statistics
Only 2% of households have obtained a loan from a bank, a microfinance institution, or a Savings and 
Credit Cooperative Society (SACCO) in the past 12 months, and just 27% of household members 
regularly save cash. 
Table 34. Loans
Description Mean CI Lower CI Upper N
In the last 12 months, obtained a loan from:
None 98% 97% 98% 3,348
SACCO 1% 1% 2% 3,348
Bank or microfinance institution 1% 1% 1% 3,348
Total loan (USD) $122 $47 $196 78
Regularly save cash 27% 25% 29% 3,348
4.18 Mental Health and Well-Being 
IPA collected information on respondents’ levels of distress in the past 30 days using the Kessler 
Psychological Distress Scale (K6). The K6 score is calculated by summing the responses from six 
questions, ranging from 0 to 24, with higher scores indicating higher levels of psychological distress, 
such as anxiety and depression. Table 35 and Table 36 show summary statistics from the mental health 
module. The average Kessler 6 score is 8.88, and 39% of respondents reported that “everything was 
difficult all the time.” In addition, half of the respondents felt “worried, tense, or anxious” most of the 
time for 30 days in the 12 months before the survey. Food shortage was a concern for more than half 
(55%) of the respondents.
Table 35. Kessler Score, in the last 30 days
Description Mean CI Lower CI Upper N
Kessler 6 (0–24) 8.88 8.70 9.06 3,348
The respondent felt ... all or most of the time
That everything was difficult 39% 37% 41% 3,348
Restless or fidgety 33% 31% 35% 3,348
So depressed that nothing could cheer you up 23% 21% 24% 3,348
Worthless 23% 21% 25% 3,348
Hopeless 22% 21% 24% 3,348
Nervous 12% 10% 13% 3,348
Table 36. Mental health in the last 12 months
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Descriptive Statistics 37 
Description Mean CI Lower CI Upper N
Had a period lasting 30 days or longer when the respondent
felt worried, tense, or anxious most of the time 51% 49% 53% 3,348
The period…
Still going on 70% 67% 72% 1,740
Still going on, but reduced 19% 17% 21% 1,740
Ended 11% 10% 13% 1,740
These worries interfered with their ability to carry out normal activities
A lot 62% 60% 65% 1,740
Some 18% 16% 20% 1,740
A Little 15% 13% 16% 1,740
Not at all 5% 4% 7% 1,740
Issues that sometimes are reasons for concern:
Food shortage 55% 53% 57% 3,348
Financial constraints 42% 40% 44% 3,348
Living situation 30% 28% 32% 3,348
Children's education 28% 26% 29% 3,348
Health 28% 26% 29% 3,348
Clothing 16% 15% 18% 3,348
Employment 10% 9% 11% 3,348
Domestic issues 11% 10% 13% 3,348
Other 16% 14% 17% 3,348
Nothing 6% 5% 7% 3,348
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Next Steps 38
5. NEXT STEPS
5.1 Process Evaluation
IPA will conduct a process evaluation to understand the extent to which the interventions are 
implemented as planned. Findings from this process evaluation will be critical for interpreting impact 
evaluation results. IPA will monitor Takunda’s implementation throughout the delivery period. IPA’s 
methodology for the process evaluation comprises a mix of site visits and observations, face-to-face 
interviews, discussion groups, desk-based research, and a review of existing reports and secondary data. 
The process evaluation will happen between April and September 2023. 
5.2 Outcome Monitoring Survey
IPA will conduct an outcome monitoring survey 1 year after interventions begin, around August 2023. 
The outcome monitoring survey will be administered to 3,348 households in all treatment and control 
areas. The objective of the outcome monitoring survey is to evaluate the short-term impact of the 
Takunda RFSA on the participants.
5.3 Final Evaluation
IPA will conduct a final evaluation survey approximately 3 years after the beginning of the activity to 
evaluate Takunda’s impact. It will happen around July 2025.
IMPEL | Implementer-Led Evaluation and Learning
39 Annex A: Additional Descriptive Statistics
ANNEX A: ADDITIONAL DESCRIPTIVE STATISTICS
Food Security
Table 37. Food security
Description Mean CI Lower CI Upper N
[BL10] Food Consumption Score (FCS) 43.14 42.46 43.82 3,348
Female and Male Adults (F&M) 43.56 42.75 44.36 2,571
Adult Female no Adult Male (FNM) 41.97 40.59 43.35 679
Adult Male no Adult Female (MNF) 41.77 37.91 45.64 98
[BL10] HH with poor FCS (0–21) 4% 4% 5% 3,348
Female and Male Adults (F&M) 4% 3% 4% 2,571
Adult Female no Adult Male (FNM) 6% 4% 8% 679
Adult Male no Adult Female (MNF) 8% 2% 14% 98
[BL10] HH with borderline FCS (21.5–35) 36% 34% 38% 3,348
Female and Male Adults (F&M) 36% 34% 38% 2,571
Adult Female no Adult Male (FNM) 38% 34% 42% 679
Adult Male no Adult Female (MNF) 36% 26% 47% 98
[BL10] HH with acceptable FCS (> 35) 59% 57% 61% 3,348
Female and Male Adults (F&M) 60% 58% 63% 2,571
Adult Female no Adult Male (FNM) 56% 52% 60% 679
Adult Male no Adult Female (MNF) 55% 45% 66% 98
[BL06] Prevalence of moderate and severe food insecurity in 
the household, based on the Food Insecurity Experience Scale 
(FIES)
82% 67% 98% 3,348
Female and Male Adults (F&M) 83% 65% 100% 2,571
Adult Female no Adult Male (FNM) 82% 47% 100% 679
Adult Male no Adult Female (MNF) 73% 0% 100% 98
[BL06] Raw FIES Score (0–8) 5.67 5.59 5.75 3,348
Female and Male Adults (F&M) 5.70 5.61 5.80 2,571
Adult Female no Adult Male (FNM) 5.63 5.45 5.81 679
Adult Male no Adult Female (MNF) 5.25 4.68 5.82 98
Notes: FIES = Food Insecurity Experience Scale. The Raw FIES Score is a sum of the 8 FIES binary questions (higher = more food 
insecure).
Table 38. Dietary diversity
IMPEL | Implementer-Led Evaluation and Learning
40 Annex A: Additional DescriptiveStatistics
Description Mean CI Lower CI Upper N
Food groups consumed yesterday by any member of 
the household...
Cereals 91% 89% 92% 3,348
Dark green leafy vegetables 78% 77% 80% 3,348
Oils and fats 78% 76% 80% 3,348
Spices, condiments, beverages 44% 42% 46% 3,348
Legumes, nuts, and seeds 37% 35% 39% 3,348
Fruits 33% 31% 35% 3,348
Vegetables 32% 31% 34% 3,348
White roots and tubers 23% 21% 25% 3,348
Sweets 21% 19% 23% 3,348
Rich vegetables and tubers 21% 20% 23% 3,348
Milk and milk products 21% 19% 22% 3,348
Vitamin-A-rich fruit 17% 16% 19% 3,348
Flesh meats 15% 14% 16% 3,348
Eat anything OUTSIDE the home 12% 11% 14% 3,348
Eggs 6% 5% 7% 3,348
Fish and seafood 5% 5% 6% 3,348
Organ meat 4% 3% 4% 3,348
Children Nutrition
Table 39. Minimum acceptable diet, disaggregated
Description Mean CI Lower CI Upper N
[BL12] Children 6–23 months receiving a minimum 
acceptable diet 8% 6% 10% 977
Minimum dietary diversity 
Breastfed 30% 26% 34% 592
Non-breastfed 2% 0% 3% 385
Minimum meal frequency
Breastfed 32% 28% 36% 592
Non-breastfed 1% 0% 2% 385
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Annex A: Additional DescriptiveStatistics 41 
Table 40. Diet of minimum diversity, disaggregated
Description Mean CI Lower CI Upper N
[BL39] Children 6–23 months consuming a diet of 
minimum diversity (MDD-C) 25% 22% 28% 977
Breastfed 30% 26% 34% 592
Non breastfed 18% 14% 22% 385
Table 41. Diet of minimum diversity, by food group
Description Mean CI Lower CI Upper N
[BL39] Children 6–23 months consuming a diet of minimum 
diversity (MDD-C) 25% 22% 28% 977
Food groups:
Breast milk 89% 87% 91% 977
Eggs 64% 61% 67% 977
Other fruits and vegetables 58% 55% 61% 977
Vitamin-A-rich fruits and vegetables 55% 52% 59% 977
Legumes and nuts 34% 30% 37% 977
Grains, roots, and tubers 32% 29% 35% 977
Dairy products 17% 14% 19% 977
Flesh foods 6% 4% 7% 977
WASH
Table 42. WASH indicators full
Description Mean CI Lower CI Upper N
[BL19] HH in target areas practicing open defecation 31% 28% 34% 1,420
[BL27] HH with access to a basic sanitation service 45% 42% 48% 1,420
Kind of toilet:
No facility/bush/field 31% 28% 34% 1,420
Ventilated improved pit latrine 30% 27% 32% 1,420
Pit latrine with slab 25% 22% 27% 1,420
Composting toilet 7% 5% 8% 1,420
Other 7% 6% 9% 1,420
Share toilet with other HH 29% 26% 32% 964
IMPEL | Implementer-Led Evaluation and Learning
42 Annex A: Additional DescriptiveStatistics
Description Mean CI Lower CI Upper N
[BL17] HH with soap and water at a handwashing station on 
the premises 9% 7% 11% 1,420
Handwashing station on the premises observed 41% 38% 44% 1,420
Presence of water at the place for handwashing 33% 29% 37% 601
Presence of soap:
None 71% 67% 76% 601
Soap, ash, or detergent (bar, liquid, powder, paste) 27% 23% 31% 601
Other 0% 0% 1% 1,420
[BL18] HH in target areas practicing correct use of 
recommended household water 6% 5% 8% 1,420
Chlorination 5% 4% 7% 1,420
Flocculant/disinfectant 1% 1% 2% 1,420
The main source of drinking water:
Unprotected well 36% 33% 39% 1,420
Protected well 28% 25% 31% 1,420
Protected spring 15% 12% 17% 1,420
Surface water 10% 8% 11% 1,420
Rainwater 5% 3% 6% 1,420
Other 7% 5% 8% 1,420
Notes: 'Share toilet with other HH' was administered only to households with any kind of toilet.
Consumption
Table 43. Consumption, durables goods
Description Mean CI Lower CI Upper N
In the past month, the household purchased any...
Communication (such as airtime) 57% 56% 59% 3,348
Transport 36% 34% 38% 3,348
Tariffs on mobile money transfers 7% 6% 7% 3,348
Fuel (paraffin, charcoal, and firewood) 5% 4% 6% 3,348
No HH purchases of this type 33% 31% 35% 3348
In the past three months, the household purchased any...
Personal care 50% 48% 52% 3,348
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Annex A: Additional DescriptiveStatistics 43 
Description Mean CI Lower CI Upper N
Other clothing expenses 10% 9% 11% 3,348
No HH purchases of this type 47% 45% 49% 3,348
In the past six months, the household purchased any...
Household operations (matches and soap) 79% 77% 80% 3,348
Education 54% 52% 56% 3,348
Clothing for children 25% 23% 26% 3,348
Footwear 22% 21% 24% 3,348
Medical care 19% 17% 20% 3,348
Clothing for ladies 13% 11% 14% 3,348
Clothing for men 11% 10% 13% 3,348
Household utensils 10% 9% 12% 3,348
No HH purchases of this type 11% 10% 13% 3,348
In the past 12 months, the household purchased any...
Funerals 37% 35% 39% 3,348
Festivals such as New Year, Christmas, Easter, Ramadan, 
Eid 32% 31% 34% 3,348
Home improvements or repairs 17% 15% 18% 3,348
Birth of a child, excluding hospital bills when the child was 
born 10% 9% 12% 3,348
No HH purchases of this type 39% 38% 41% 3,348
Family Planning
Table 44. Family planning detailed
Description Mean CI Lower CI Upper N
[BL36] Women in a union who have knowledge of modern 
family planning methods that can be used to delay or avoid 
pregnancy
95% 93% 97% 951
Knowledge of modern family planning methods score (1–13) 7.62 7.42 7.81 951
Percentage of women who have heard about …
Contraceptive pill 97% 96% 98% 951
Injectables 94% 92% 96% 951
Male condom 92% 90% 94% 951
Implants 91% 89% 93% 951
IMPEL | Implementer-Led Evaluation and Learning
44 Annex A: Additional DescriptiveStatistics
Description Mean CI Lower CI Upper N
Female condom 85% 82% 87% 951
IUD 77% 74% 80% 951
Female sterilization 66% 62% 69% 951
Lactational amenorrhea method (LAM) 50% 47% 54% 951
Male sterilization 38% 34% 41% 951
Standard days method 35% 32% 39% 951
Emergency contraception 18% 15% 20% 951
Diaphragm with spermicidal foam, cream, or gel 16% 13% 19% 951
Other modern methods 3% 2% 4% 951
[BL37] Women in a union who made decisions about modern 
family planning methods in the past 12 months 50% 46% 53% 951
Non-pregnant women 86% 84% 88% 951
[BL20] Contraceptive prevalence rate (CPR) 71% 67% 75% 797
Method used:
Contraceptive pill 64% 59% 68% 569
Injectables 21% 17% 24% 569
Implants 8% 5% 10% 569
Other 8% 5% 10% 596
Notes: This section was administered to a random subsample of women in union aged 18 to 49 years. The indicator [BL20] 
Contraceptive Prevalence Rate was administered to non-pregnant women.
Agriculture 
Table 45. Improved management practices or technologies on crops
Description Mean CI Lower CI Upper N
Practices/technologies for cultivation 
Weed control 92% 90% 93% 1,355
Manure 56% 53% 59% 1,355
Early planting or planting with first rains 55% 52% 58% 1,355
Ripping into residues 52% 49% 55% 1,355
Micro dosing 49% 46% 52% 1,355
Planting basins 46% 43% 49% 1,355
Use of improved crop varieties 43% 40% 46% 1,355
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Annex A: Additional DescriptiveStatistics 45 
Description Mean CI Lower CI Upper N
Intercropping 42% 39% 45% 1,355
Crop rotations 38% 35% 41% 1,355
Pot-holing 35% 32% 38% 1,355
Compost 33% 30% 36% 1,355
Dry planting 26% 23% 29% 1,355
Ridging 22% 20% 25% 1,355
Clean ripping 16% 14% 18% 1,355
Tied ridges 16% 13% 18% 1,355
Mulching 14% 12% 16% 1,355
Dead level contours 12% 10% 14% 1,355
Integrated pest management (IPM) 11% 9% 13% 1,355
Double-dug beds/fertility trenches 7% 5% 8% 1,355
Methods to store
Sealed/air-tight bags 46% 43% 49% 1,355
Grain treatment with agro-chemicals 8% 6% 9% 1,355
Other post-harvest practices that reduce pre-storage 
losses 7% 5% 8% 1,355
Seed or grain treatment techniques 6% 5% 8% 1,355
Locally made storage structures 2% 2% 3% 1,355
Use of solar or fuel-powered dryers 0% 0% 1% 1,355
Community storage facilities 0% 0% 0% 1,355
Natural resource management practices
Management or protection of watersheds 15% 13% 18% 1,355
Regeneration of natural landscapes 15% 13% 17% 1,355
Management of forest plantation, woodlands 7% 5% 8% 1,355
Development and implementation of NRM bylaws 7% 5% 8% 1,355
Agro-forestry 1% 1% 2% 1,355
Sustainable harvesting of forest products 0% 0% 1% 1,355
Table 46. Improved management practices or technologies on livestock
Description Mean CI Lower CI Upper N
Practices/technologies for livestock
IMPEL | Implementer-Led Evaluation and Learning
46 Annex A: Additional DescriptiveStatistics
Description Mean CI Lower CI Upper N
Vaccinations 42% 38% 46% 721
Homemade animal feeds 26% 22% 29% 721
Deworming 22% 19% 26% 721
Improved shelters 12% 9% 15% 721
Castration 9% 7% 11% 721
Fodder production, veld reinforcement with legumes 7% 5% 9% 721
Used the services of comm. animal health workers 7% 5% 9% 721
Animal feed supplied by stock feed manufacturer 7% 5% 8% 721
Dehorning 6% 5% 8% 721
Pen feeding 6% 4% 8% 721
The services of community animal health ext. worker 6% 4% 8% 721
Other 1% 0% 1% 721
Periodic replacement of male breeding stock 1% 0% 1% 721
Artificial insemination 0% 0% 0% 721
Natural resource management practices
Management or protection of watersheds 16% 13% 19% 721
Regeneration of natural landscapes 15% 12% 18% 721
Development and implementation of NRM bylaws 10% 7% 12% 721
Management of forest plantation, woodlands 6% 4% 8% 721
Agro-forestry 2% 1% 2% 721
Sustainable harvesting of forest products 0% 0% 1% 721
Resilience
Table 47. Adaptive index
Description Mean CI Lower CI Upper N
[BL08] Adaptive capacity index (0–100) 45.34 44.39 46.28 1,217
Index of aspirations/confidence to adapt 10.14 10.02 10.25 1,420
Index for bridging social capital 2.38 2.27 2.49 1,420
Index for Linking social capital 0.66 0.59 0.73 1,420
Social networking index 2.19 2.13 2.25 1,419
Education and training index 1.59 1.52 1.65 1,420
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Annex A: Additional DescriptiveStatistics 47 
Description Mean CI Lower CI Upper N
Livelihood diversification 1.86 1.80 1.93 1,420
Index for information exposure 7.08 6.78 7.37 1,411
Adoption of improved practices 0.95 0.94 0.97 1,227
Index of asset ownership 8.97 8.72 9.21 1,420
Index for access to financial institutions 0.80 0.75 0.85 1,420
Notes: This section was administered to a random subsample of respondents; 1,420 households were selected (42% of the 
sample). “Social networking index” and “Index for information exposure” has different sample because of a mistake in the flow 
of the survey. “Adoption of improved practices” have a different sample because not all households have farmers. Index of 
assets includes Land, Cultivator, Plough, Planter, Sheller, Harrow, Plantation/ Orchard, Incubator, Beehives, Scotch cart/Water 
cart, Wheelbarrow, Knapsack sprayer, Water pump, Donkey, Goat, Cattle, Generator, Solar Panel, Lounge suite, Bicycle, 
Television, Satellite Dish & components, Radio, Cell-phone, Refrigerator/Deep–freezer, Peanut Butter / Candle Making /Oil￾pressing machine, Cattle, Goats, Sheep, Donkey/mule, and Poultry.
Table 48. Absorptive index
Description Mean CI Lower CI Upper N
[BL09] Absorptive capacity index (0–100) 38.49 37.53 39.46 1,322
Number of informal safety nets available in the community 1.97 1.89 2.05 1,322
Bonding Social Capital Index 2.84 2.73 2.95 1,420
Household regularly saves cash 0.28 0.26 0.31 1,420
Access to remittances 0.13 0.11 0.16 1,420
Index of asset ownership 8.97 8.72 9.21 1,420
Index of shock preparedness and mitigation 0.39 0.36 0.43 1,420
Household-acquired crop insurance 0.00 0.00 0.00 1,420
Availability of humanitarian assistance from the 
government or non-governmental organization 0.25 0.22 0.28 1,420
Notes: This section was administered to a random subsample of respondents; 1,420 households were selected (42% of the 
sample). “Number of informal safety nets available in community” has a different sample because of an incorrect skip pattern 
that was corrected after a few days.
Table 49. Transformative index
Description Mean CI Lower CI Upper N
[BL25] Transformative capacity index (0–100) 38.54 37.45 39.64 1,267
Availability of formal safety nets index 0.72 0.65 0.79 1,420
Availability of markets within 5km of a village index 0.54 0.48 0.59 1,420
Access to communal natural resources index 0.29 0.24 0.33 1,322
Index for access to basic services 1.00 0.95 1.05 1,420
Access to infrastructure index 1.31 1.26 1.36 1,420
IMPEL | Implementer-Led Evaluation and Learning
48 Annex A: Additional DescriptiveStatistics
Description Mean CI Lower CI Upper N
Access to agricultural extension services 0.59 0.55 0.62 1,420
Access to livestock services 0.34 0.31 0.37 1,420
Index for bridging social capital 2.38 2.27 2.49 1,420
Index for Linking social capital 0.66 0.59 0.73 1,420
Index of types of collection action 0.47 0.43 0.51 1,420
Community-level gender equitable decision-making 
index 0.29 0.28 0.30 1,419
Index for local government responsiveness 0.66 0.62 0.70 1,420
Gender index 0.39 0.34 0.43 1,420
Participation in local decision making 0.45 0.42 0.48 1,267
Notes: This section was administered to a random subsample of respondents; 1,420 households were selected (42% of the 
sample). “Access to communal natural resources index” has a different sample because not all the households selected have a 
least one farmer member. “Community-level gender equitable decision-making index” has a different sample because it applies 
to women and men in union. “Participation in local decision making” has a different sample because of an incorrect skip pattern 
that was corrected after a few days.
Table 50. Education and training
Description Mean CI Lower CI Upper N
Household ever received...
any vocational (job) or skills training 19% 17% 21% 1,420
any business development training 9% 7% 10% 1,420
any early warning training 5% 4% 6% 1,420
any natural resource management training 18% 16% 20% 1,420
any adult education 8% 6% 9% 1,420
training in how to use your mobile phone to get 
market information 11% 9% 13% 1,420
Notes: This section was administered to a random subsample of respondents; 1,420 households were selected (42% of the 
sample).
Table 51. Collective action
Description Mean CI Lower CI Upper N
HH worked with others in the village to do something 
for the benefit of the community 37% 35% 40% 1,420
Activities:
Road maintenance/construction 66% 61% 71% 532
Repaired/built schools 13% 10% 17% 532
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Annex A: Additional DescriptiveStatistics 49 
Description Mean CI Lower CI Upper N
Repaired/built health posts or centers 12% 9% 15% 532
Improved community access to drinking water 11% 8% 14% 532
Other 22% 18% 26% 532
Notes: This section was administered to a random subsample of respondents; 1,420 households were selected (42% of the 
sample).
Perceived Economic Ladder
Table 52. Perceived economic ladder
Description Mean CI Lower CI Upper N
Rung where the HH…
...would place on the ladder in terms of financial status 2.55 2.49 2.62 3,337
...thinks it will be in terms of financial status in five years 4.03 3.93 4.13 3,005
...would place on the ladder in terms of self-satisfaction 3.75 3.64 3.85 3,306
...thinks it will be in terms of self-satisfaction in five years 5.09 4.98 5.21 2,979
Notes: The rungs for financial status go from 1 (poorest) to 10 (richest). The rungs for self-satisfaction go from 1 (very 
dissatisfied) to 10 (very satisfied). The sample is different among the variables because some participants refused to answer the 
question.
Self-Control
Table 53. Self-control
Description Mean CI Lower CI Upper N
Find hard time breaking bad habits
Not like me at all 74% 73% 76% 3,348
Not much like me 17% 16% 19% 3,348
Somewhat like me 3% 2% 3% 3,348
Mostly like me 4% 3% 4% 3,348
Very much like me 2% 1% 2% 3,348
Get distracted easily
Not like me at all 55% 53% 57% 3,348
Not much like me 23% 21% 25% 3,348
Somewhat like me 11% 10% 13% 3,348
Mostly like me 7% 6% 8% 3,348
Very much like me 4% 3% 4% 3,348
IMPEL | Implementer-Led Evaluation and Learning
50 Annex A: Additional DescriptiveStatistics
Description Mean CI Lower CI Upper N
Say inappropriate things
Not like me at all 60% 58% 61% 3,348
Not much like me 24% 22% 26% 3,348
Somewhat like me 14% 13% 16% 3,348
Mostly like me 2% 1% 2% 3,348
Very much like me 0% 0% 1% 3,348
Refuse things that are bad for me, even if they are fun
Not like me at all 14% 13% 15% 3,348
Not much like me 8% 7% 9% 3,348
Somewhat like me 9% 8% 10% 3,348
Mostly like me 26% 25% 28% 3,348
Very much like me 43% 41% 45% 3,348
Good at resisting temptation
Not like me at all 4% 3% 5% 3,348
Not much like me 6% 5% 7% 3,348
Somewhat like me 29% 28% 31% 3,348
Mostly like me 28% 26% 30% 3,348
Very much like me 33% 31% 34% 3,348
Have very strong self-discipline
Not like me at all 3% 2% 3% 3,348
Not much like me 4% 3% 4% 3,348
Somewhat like me 21% 19% 23% 3,348
Mostly like me 36% 34% 38% 3,348
Very much like me 37% 35% 39% 3,348
Pleasure and fun sometimes keep me from getting work done
Not like me at all 46% 44% 48% 3,348
Not much like me 24% 22% 25% 3,348
Somewhat like me 16% 14% 17% 3,348
Mostly like me 11% 10% 12% 3,348
Very much like me 3% 3% 4% 3,348
Do things that feel good in the moment but regret later
Baseline Evaluation of the Takunda RFSA in Zimbabwe (Vol. I)
Annex A: Additional DescriptiveStatistics 51 
Description Mean CI Lower CI Upper N
Not like me at all 50% 48% 52% 3,348
Not much like me 22% 21% 24% 3,348
Somewhat like me 18% 16% 19% 3,348
Mostly like me 8% 7% 9% 3,348
Very much like me 2% 1% 2% 3,348
Can't stop myself from doing something, even if I know it's 
wrong
Not like me at all 56% 54% 58% 3,348
Not much like me 23% 21% 25% 3,348
Somewhat like me 15% 14% 16% 3,348
Mostly like me 5% 4% 6% 3,348
Very much like me 2% 1% 2% 3,348
Often act without thinking through all the alternatives
Not like me at all 58% 56% 60% 3,348
Not much like me 25% 23% 26% 3,348
Somewhat like me 12% 11% 13% 3,348
Mostly like me 4% 3% 5% 3,348
Very much like me 1% 1% 1% 3,348
Table 54. Population lived under the poverty line, by province and district
Description Mean CI Lower CI Upper N
[BL01] Prevalence of poverty: Percentage of people 
living on less than $1.90/day 20 84% 82% 85% 3,348
Manicaland 87% 85% 88% 2,272
Buhera 90% 87% 92% 1,007
Mutare 84% 81% 86% 1,265
Masvingo 77% 74% 80% 1,076
Chivi 79% 75% 83% 543
Zaka 75% 71% 79% 533
IMPEL | Implementer-Led Evaluation and Learning
52 Annex B: Balance Test
ANNEX B: BALANCE TEST
Table 55. Balance test
(1) (2) (1) – (2)
Description Treatment Control Difference
Mean Mean (P-Value)
Female head of household 0.37 0.38 -0.02 (0.45)
Age of household head 51.70 51.76 -0.07 (0.94)
Married household head 0.69 0.68 0.00 (0.87)
Number of household members 5.13 4.91 0.22 (0.02)
Number of children under 18 in household 2.87 2.69 0.19 (0.05)
Number of rooms in main house 3.33 3.32 0.02 (0.86)
Daily per capita expenditures ($USD) 1.61 1.66 -0.05 (0.55)
Food Consumption Score (FCS) 42.65 43.69 -1.04 (0.33)
Own agricultural land 0.81 0.83 -0.03 (0.29)
Households cultivate anything in last 12 months 0.94 0.95 -0.00 (0.83)
Index of durable assets -0.03 0.00 -0.03 (0.18)
Own livestock 0.88 0.88 -0.01 (0.71)
N 1,831 1,517
Joint Test P-Value: 0.41