Policy makers and development partners in Zimbabwe have become accustomed to making decisions on the basis of the annual Crop and Livestock Assessment report as well as the Zimbabwe Vulnerability Assessment Committee (ZimVAC) report which also conducts a rural livelihood assessment every year. Other sources of information for policy makers include demographic and healthy Surveys, the national census, the poverty assessment surveys and the national economic performance reviews. To what extent are the methods used in gathering information in all these surveys reliable?
The annual Crop and Livestock Assessment methodology comprises a sample of 30 households per farming sector per ward. Information is gathered through household interviews, observations and Key Informant Interviews around the country within at least three months. On the other hand, the ZimVAC conducts a rural livelihood assessment almost every year using a structured household questionnaire and focus group discussions within a few weeks. In 2014 ZimVAC data collection occurred from 9 to 21 May 2014 covering a total of 10 782 household interviews and 879 focus group discussions throughout the country. The ZimVAC assessment is based on ZIMSTAT sampling frame comprising a minimum of 15 Enumerator Areas (EAs) in each district. In each EA, 12 households are randomly selected and interviewed. Is it enough to rely on information gathered from 30 households and 12 households by the respective initiatives? Building on methods used by the Crop & Livestock Assessment and ZimVAC, the table below show a sample minimum and maximum percentage of respondents for selected districts and wards. Populations in each ward were gleaned from ZIMSTAT 2012 data:
Sample Minimum and Maximum %age respondents for selected districts and wards
Working with percentages, both sampling methods by ZimVAC and Crop & Livestock Assessment are less than 7% of the total households in each ward. This percentage might not be significant in reflecting the true picture on the ground. For instance, Chipinge Rural District’s smallest ward has 707 households and the largest has 4494 households. In this case, the sample percentage for the smallest ward, using Crop & Livestock Assessment, is 1.7% and for the largest ward it comes to 0.27%. Using ZimVAC, the smallest ward percentage representative sample is 4.24% and the largest is 0.67%. Significant information should come from percentages in the range of 10 to 30% and, being proportionate to number of households in the ward. Since wards have different numbers of households, there is need to consider proportionate representation. Trying to interview 30 households in a ward with only 25 households, for example Matobo, is more than 100 sample size and highly misleading.
Alternatives ways of enriching annual Crop & Livestock Assessments and ZimVAC
The way data is collected and processed into recommendations has an enormous bearing on how government and development partners meet their goals. An inadequate methodology produces flawed evidence which may result in missed opportunities and targets. These national information gathering initiatives should not be events but processes that tap into ordinary people’s ways of managing and sharing personal knowledge. Zimbabwe has structures that can easily be used to speed up and validate crop & livestock and ZiMVAC rural livelihood assessments. The ministry of agriculture has a presence from every ward up to national level. Same with all other government departments like the Ministry of Youth and Ministry of Women’s Affairs. Also, each rural community has farmer group leaders. Another structure is the local government which comprises village heads who, most of the time have information about their villages regarding population by gender and age, number of livestock, schools, land holdings, dams, irrigations schemes, boreholes and natural features like forests and rivers. All this information can easily enrich the crop & livestock assessments and ZimVAC so that information is not based on small samples.
Each district has development partners and NGOs that can provide assistance in kick-starting this process of data gathering and analysis to a point of empowering community knowledge centres at ward level where information is kept alive and fluid, not compressed into a portable document formats (pdf) documents which are currently accessible to a few people mostly based in urban areas. Such a process will go a long way in saving resources that would otherwise go towards baselines by several NGOs and private companies. Working through government departments such as ministries of youth and gender means information is collected already disaggregated by gender and age. Interventions that want to target youth or women can use this data without need to conduct separate surveys.
There is need for a simplified tool that can capture information from the farmer, rural entrepreneur and any actor. This information can be fluidly fed into a national database that speaks to the crop and livestock assessment.
Random sampling is rather inadequate in the modern digital age where information literacy is increasing. Random sampling can misrepresent what is on the ground. An alternative would be identifying and gathering information from categories of farmers in each ward. For instance, out of 100 master farmers, a sample of 10 is engaged. The same number can be sampled for medium farmers as well as for low producers. The total can then be expressed as a percentage of each category in the ward. It is very possible to have a total of master farmers; medium farmers and low producers in each ward rather than assuming farmers are the same. Such an approach can reveal the extent of required interventions in terms of food security and other support services. A community with 70% of master farmers able to produce enough for the ward requires a different intervention from the one where only 30% of the master farmers can meet local requirements.
Crop and livestock assessment should be a process from planting to harvesting so that we are able to analyse at what level the crops have been affected and why. Communities can tell whether crops have been affected at germination stage, tasseling stage or by too much rainfall at harvesting. Analysis has to come from communities. We can’t just generalize at national level because climate change is more of a local issue. That is why some wards receive rainfall while adjacent wards do not receive as much rainfall in a particular season. Such fine-grained information is critical for climate change interventions. If crops are affected at germination stage in three consecutive seasons, seed companies, hydrologists and farmers have to connect at that local level to address that particular challenge. It can’t be generalized. Climate change comes in cycles of less rain, interspersed with many good years in between. It has to be understood as more of a local problem than a national issue. Stepping up horticulture production to complement field crops has to be done based on accurate information. Besides being produced all year round, most horticulture commodities go together with main staples such as maize, sorghum and millet, leading to a holistic food basket.
Another option is getting local statistics through schools. School children or students can be requested to bring information from their parents regarding number and types of livestock, crops planted by area and amount of harvested commodities. Such information can be collected within a week and consolidated. The information can later be continuously updated. This way, crop & livestock assessments can become processes embedded in communities unlike the current situation where they are more like events happening over a few months conducted by experts from Harare.
At the moment, there doesn’t seem to be clear mechanisms for validating information gathered through ZimVAC and crop & livestock assessments at community level through feedback meetings. Having used random sampling, it would make sense for those responsible for ZiMVAC or crop & livestock assessments to go back to wards where information was collected and engage communities as a way of validating findings. During these events, communities can even provide additional information on their coping mechanisms before decisions are taken in the direction of food aid or importing maize.
What is also missing in most conventional reports is accurately information on the role of markets versus subsistence use of agricultural commodities. The role of remittances in wards and districts is also not captured yet most communities are now relying on remittances from South Africa, UK and other countries. Some households may reduce household consumption of a particular commodity like maize due to a new reliance on rice purchases. Reduction in area planted to maize can be by choice, for example, coping through remittances when a family member suddenly goes to work in South Africa.
If this process is enriched with updated analysis, droughts or floods will not come as a crisis or a surprise because regular data collection and feedback will provide signals before things go out of hand. We don’t have to wait until the ZimVAC report is published in order to take action or declare a drought. By that time, one community will have suffered beyond repair. Providing information ahead of time results in pro-active action. As information comes through, it should be possible to see types of commodities and volumes that are going to markets (local or urban). Some data can point to the role of local business people in supporting agriculture by either providing a market for local produce or supplying extending inputs on credit. The capacity of SMEs to organise and absorb local agriculture produce can also be revealed through elaborate ways of capturing data.
Riding on ICTs
Since random sampling tends to exclude some farmers, ICTs is now making it possible for local people to collect data. In this case they don’t just remain informants but co-creators of knowledge and shapers of their own destiny using data. A lot of information can be gathered and processed at growth points and rural business centres. Given the proliferation of ICTs each village or ward can gather, digitize and store its own data. Development partners can support building of local people’s capacity to gather their own data and conduct longitudinal studies while gathering and monitoring changes as they happen. With ICTs increasingly rendering information fluid, communities should be engaged so that they contribute to information as a flow not as a stock that can be condensed into a pdf accessible to a few technocrats. Engaging communities using their mobile technology enhances co-creation of real-time knowledge and information. Community dialogues can be used to consolidate and exchange information from various wards. Village development committees should be empowered to gather and semi-process valuable agricultural and rural development information using mobile technology. Such information can be consolidated once a fortnight at ward level.
The ministry of agriculture should be capacitated to embed a ICTs system through all these data gathering processes. Such an effort can be harnessed with the ministry of ICTs’ information centres programme. Agritex supervisors at ward level can be responsible for capturing and consolidating data at community information centres or in computers that are now available in most rural schools. Such content becomes very relevant for school children interested in agriculture and rural development instead of using text books with foreign content. A holistic data gathering and processing tool or framework can be designed by ICTs experts, local farmers as users, subject matter specialists (livestock scientists, agronomists, etc.), economists, entrepreneurs and policy makers. The tool can also be available in local languages.
Access to knowledge as power
Aggressive data collection and processing at local level will show viability points rather continue talking in terms of vulnerability. If we want to create wealth and not continue managing poverty we have to embrace a positive narrative. Evidence gathered throughout the year can show where support is required and when. For instance, areas with more grazing in much of the year may require more support to keep on supporting livestock before they become like the rest of the country. Data and evidence will also show at what point supplementary feeding is required. Making data fluid will avoid a situation where by the time the ZimVAC report is published 90% of the content is irrelevant.
Although farmers and agriculture markets are becoming hyper-linked – thanks to mobile phones, ICTs are yet to fully transform the way African countries like Zimbabwe gather agriculture and rural development information. Crop and Livestock assessments as well as the ZimVAC should tap into the way the agriculture sector now operates as networks where everyone can be a contributor within a transparent environment. A major source of knowledge is now diversity of ideas, and openness as well as an ability to make sense of what is going on. Farmers and traders are now able to set the context around them and build consensus around emergent practices. Policy makers and development partners can make better decisions by actively listening to farmers and traders as networked contributors who are closely in touch with their environment. With an informed perspective, they can propose changes and build consensus around suggested responses.
Farmers, traders and rural communities can learn more from the process of creating a report than from implementing recommendations from a report in whose codification they didn’t participate. They can remember how the process of codifying their experiences felt than reading someone’s explanation of their circumstances. Involving communities in crop & livestock and ZimVAC assessments will give communities the confidence of taking development matters into their own hands.
Charles@knowledgetransafrica.com / firstname.lastname@example.org
Clever@knowledgetransafrica.com / email@example.com
firstname.lastname@example.org / email@example.com
firstname.lastname@example.org / email@example.com
firstname.lastname@example.org / email@example.com
firstname.lastname@example.org / email@example.com
firstname.lastname@example.org / email@example.com
firstname.lastname@example.org / email@example.com
Laizah@knowledgetransafrica.com / firstname.lastname@example.org
eMKambo Call Centre:
0739 866 343-6