National Gemini Micro and Small Enterprise Baseline Survey 2000
Malawi, 2000
Identification
MWI_2000_MSE_v01_M
National Gemini Micro and Small Enterprise Baseline Survey 2000
| Name | Country code |
|---|---|
| Malawi | MWI |
Enterprise Survey [en/oth]
This survey complements the survey carried out in 1992. As the methodology has evolved and improved, we have been able to get a better feel for the entire sector as well as how it has changed over the last 8 years. In addition, we have a new picture on several other issues: primary production activities as MSE and the question of which MSE are being most affected by HIV/AIDS in Malawi.
The second Malawi nationwide GEMINI Micro and Small Enterprise (MSE) baseline survey was carried out between October and December 2000. The survey was conducted by a multi-disciplinary team from Ebony Consulting International, the National Statistical Office (NSO), Kadale Consultants, and Wadonda Consult and was funded by the British Department for International Development (DFID).
The general objective of this survey was to provide an overview of the MSE sector in Malawi, a sector that has often been missed by more conventional survey methods. This survey has therefore captured business start-up, employment growth, and reasons for business closure, impact of HIV/AIDS and the contribution of the sector to the national income. The results from this survey will be of great use by the government as well as non-governmental organisations in developing programmes and policies that support the development of the private sector as well as projects initiated to alleviate poverty through income generating activities.
The survey team visited over 26,000 households and enterprises throughout the country. They collected data from both rural and urban areas through a stratified random sampling methodology. It involved a complete enumeration of any household or enterprise income-generating activity within the selected Enumeration Areas.
Sample survey data [ssd]
Micro and Small Enterprises
Households
Scope
This scope of this survey includes
- Business start-up
- Employment growth
- Rasons for business closure
- Impact of HIV/AIDS
- The contribution of the sector to the national income.
Coverage
National
The survey team visited a stratified sample of over 22,000 households and small businesses to identify active business activities of all kinds employing fewer than 50 employees. This study also enumerated on-farm agricultural activities, as long as 50 percent of the production was sold and the household earned more than MK 6,000 from the sale of the produce.
Producers and sponsors
| Name | Affiliation |
|---|---|
| National Statistical Office of Malawi | Government of Malawi |
| Name | Role |
|---|---|
| Ebony Consulting International | Training NSO Staff in GEMINI Methodology |
| Kadale Consultants | Survey Implementation |
| Wadonda Consult | Survey Implementation |
| Name | Role |
|---|---|
| Department for International Development | Funding |
Sampling
Stratified Random Sampling
The 2000 Baseline Survey in Malawi is part of an established method of carrying out surveys of micro and small enterprises known as the GEMINI approach. GEMINI-style surveys generally employ a stratified cluster sampling approach. This means that enumeration areas (EAs) are randomly selected within each stratum. To maximise the precision of the survey estimates, two issues present themselves. First, the strata should be as different from each other as possible. Second, the differences between enumeration areas within each stratum should be relatively small. In Malawi, seven strata were selected in consultation with the NSO, Wandonda, and other local experts. Urban areas are defined to include Malawi's four main cities: Blantyre, Lilongwe, Mzuzu, and Zomba. Small towns are considered to be district headquarters and trading centers. Because rural areas bordering on lakes are thought to be very different from all other rural EAs, these were put into a separate stratum. Within each stratum, a sample of EAs was selected using simple random sampling.
Urban commercial and industrial areas were sampled in a slightly different manner. Because the EAs that comprise these areas are large in size and small in number, these EAs were subdivided. Once this subdivision was accomplished, areas were randomly selected. Because the sampling procedure was conducted in accordance with established statistical methods, the results can be extrapolated to accurately represent the country as a whole.
During the fieldwork, each household and shop in a selected EA was visited. Some sites had one or more income-generating activity, others had none, and other sites were closed to our enumerators (either because no one was there who has knowledge of the business, or because the respondent refused to cooperate). This information, taken in conjunction with the information that is known about the total number of households in each stratum, allows the sample results to be extrapolated to the national level.
Because of the method employed, it is possible to calculate weights for each stratum and apply these to the data from the survey. Briefly the procedure involves weighting each stratum, taking into account both the probability of a household’s being selected, and the fact that in each enumeration area certain households and shops were closed to enumerators. Once these weights are calculated, survey results can be extrapolated to the national level, and an estimate of the total number of MSEs in each stratum can be made. Estimates of the total number of enterprises involved in each activity (e.g., tailoring, selling curios, repairing electronics, etc.), the total number of MSEs run by female proprietors, the number of MSEs by location, etc. will be accurate.
Survey instrument
This survey used three different instruments to collect the data. An existing business questionnaire (EBQ) was the main instrument and a closed business questionnaire (CBQ) covered businesses that had closed in the last 5 years. In addition, a tally sheet recorded the number of businesses that were identified (IBs) but not fully enumerated; those were not unidentifiable (UBs), non-responses (NRs) and no activity (NA) households and sights. The 2000 Baseline Survey instruments are similar to those used in other GEMINI-style surveys conducted in the region. With the participation of the NSO, and local consultants, the questionnaire was adapted to a certain extent to suit national conditions. Field experience suggested that for future surveys some further adaptation would improve the efficiency in enumeration. The main questionnaire is eight pages in length, and covers questions about the distribution of MSEs by size, geographically by type of ownership it also contains questions that help estimate contribution to employment and incomes, business economic performance (births and deaths, costs, sales and profits). It also includes questions on current and start up constraints, as well as services (including credit) that are available to the MSE owners. In addition the questionnaire includes 4 indirect questions that help estimate the likelihood that a business is affected by HIV/AIDS and a direct question on whether the respondent believes the business is affected. This section has previously not been included in similar studies done in the region.
Data collection
| Start | End |
|---|---|
| 2000-10-23 | 2000-12-08 |
While in the field, all teams received regular and intensive supervision and support from senior researchers and survey managers from NSO, Wadonda Consult and ECI. For example, each of the four field teams was visited at least once in each of the five phases and in most cases once every week. During these visits the senior supervisors used a structured and commonly agreed supervision guide which covered both administrative and technical aspects of the survey. The supervision reports were sent back to the Principal Investigator after each visit in order to resolve any on-going problems in the course of the survey. This guide is included in a training manual also developed during the design and training phase of the survey. In addition, NSO staff brought invaluable experience on field conduct and dealing with resistance and difficult field situations. Wadonda Consult primarily did the data entry, cleaning and processing. This started in the first week of November and continued through to mid December. Data analysis started in November and continued to the end of January.
The data collection was undertaken in the seven weeks from October 23rd to December 8th. This coincided with a busy agricultural period (mainly planting), which meant that a certain number of potential survey respondents were out in the fields during enumerator visits, and consequently an underestimation of the number of existing MSE especially in the rural areas may have occurred.
The field survey was organized in 5 phases starting with Zomba area and the southern and problematic districts. The annual rains being around November / December in the southern regions of the country and thus the initiation of the survey here. To enhance data quality, 46 candidates received an intensive one-week training followed by a competitive process and intensive evaluation exercises to select the 28 staff to undertake the fieldwork. Then remuneration for all staff in the field and local survey managers was based on basic pay plus an additional bonus based on performance- a strategy that worked extremely well and to be recommended for future surveys of this kind.
Phase I of the fieldwork started in the Zomba area and allowed senior researches to provide further field based intensive supervision and guidance to the field teams. After that the teams started phase II further south in Nsanje and in difficult locations (due to accessibility during the rainy season) such as Makanjira. The team progressed north and finished an initial 71 Enumeration Areas just north of Mzuzu around 3rd of December. An additional 10 EAs spread throughout the country and mainly from the rural strata (see second tier EA in the attached EA summary sheet) were completed in the following five days.
Data Access
| Name | Affiliation | URL | |
|---|---|---|---|
| National Statistical Office of Malawi | Government of Malawi | www.nsomalawi.mw | enquiries@statistics.gov.mw |
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including country, acronym and year of implementation)
- the survey reference number
- the source and date of download
Disclaimer and copyrights
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.
Contacts
| Name | Affiliation | URL | |
|---|---|---|---|
| National Statistical Office of Malawi | Government of Malawi | enquiries@statistics.gov.mw | www.nsomalawi.mw |
Metadata production
DDI_MWI_2000_MSE_v01_M
| Name | Affiliation | Role |
|---|---|---|
| Development Economics Data Group | The World Bank | Documentation of the DDI |
2013-02-12
Metadata version
Version 01 (February 2013)