General Data Dissemination System (GDDS) Project - Phase 2 Socio-Demographic Statistics Project for Anglophone Africa

Transcription

1 General Data Dissemination System (GDDS) Project - Phase 2 Socio-Demographic Statistics Project for Anglophone Africa Report of provision of technical assistance during the first mission as a lead expert for the Labor Statistics Module to the National Statistical Office Zomba, Malawi January 14-25, 2008 Prepared by: Rachel Krantz-Kent

2 Main Issues Methods for conducting a labor force survey (LFS) require development Main Outcomes Reviewed and discussed draft proposal Assisted staff with development of labor force concepts Provided staff with resources for developing questionnaire, looked at and discussed questionnaires used by other countries Met with the Ministry of Labor to discuss their data needs and concerns Provided staff with feedback on proposed method of sample selection Developed work plans for conducting the survey Lack of funding for LFS Met with Ministry of Labor to discuss need for funding Discussed potential sources of funds and alternative ways to collect information about the labor force No sample frame available for an informal sector survey (ISS) Discussed questions that can be used to identify informal sector enterprises Discussed attaching questions to the 2008 Population and Housing Census or the LFS, to identify households with informal sector enterprises 1

3 Executive Summary At the request of the National Statistical Office (NSO) in Malawi, and with the support of the World Bank, a technical assistance mission visited Zomba January 14-25, 2008, to assist with the development of a labor force survey (LFS), an informal sector survey (ISS), and an establishment survey on wages and salaries. Twenty-five years have past since the last LFS was conducted in Malawi in The results of this survey were compiled and analyzed but because of political circumstances were never published. Basic labor force information is collected annually on the Welfare Monitoring Survey, every five years on the Integrated Household Survey, and every ten years on the Population and Housing Census. Information about children s work and work conditions was collected in 2002 in a special Child Labor Report. Malawi s NSO would like to conduct a LFS to gain a more comprehensive understanding of Malawi s labor force. Malawi s NSO has begun preparations for a LFS. They have drafted a proposal that outlines their plans for conducting the survey. They also have met with the Ministry of Labor to discuss the Ministry s data needs, plans for the LFS, and the need for funding to conduct the survey. Further development of the proposal, concepts, questions, and survey methods is needed, as well as a source of funds to conduct the survey. Malawi s NSO also would like to conduct an ISS and an establishment survey of wages and salaries. Both of these surveys are receiving some assistance from the Norwegian government, which currently is assisting Malawi on a project to improve its national accounts. As part of the national accounts project, the Norwegian government is sponsoring a census of large establishments, to update Malawi s business registry; this census went into the field in December The Norwegian government also has funds to collect information from informal sector enterprises, but first an appropriate sampling frame is needed. On its 2008 Population and Housing Census, the Malawi government may include questions to identify informal sector enterprises. Overview of visit I visited Malawi s National Statistical Office (NSO) in Zomba for two weeks, from January 14-25, During the visit, I met with several persons at the National Statistical Office, as well as with the Chief Economist at the Ministry of Labor in Lilongwe. While in Malawi, I met with the following persons: Charles Machinjiri, Commissioner Mercy Kanyuka, Deputy Commissioner and GDDS Manager Jumeson Ndawala, Assistant Commissioner Derrice Zanera, Chief Statistician 2

4 Chris Matemba Brian Ngoma, Chief Economist at Malawi s Ministry of Labor Each day I meet with Medson Makwemba, a Statistician, and Frank Matumula, an Assistant Statistician, at the NSO. The purpose of the assignment was to complete the first technical assistance mission to assess all work completed to date and to develop detailed work plans to guide the authorities in the development of a household LFS, an ISS, and an establishment survey on wages and salaries. The visit focused on three priorities: 1. Development of a household LFS (60%) 2. Development of an ISS (20%) 3. Development of an establishment survey on wages and salaries (20%) Most of the visit focused on development of a household LFS because this was identified by the NSO as their main priority. The NSO is receiving assistance from Norway on a project to improve Malawi s national accounts; work on this project overlaps with priorities 2 and 3, listed above. Because of this other assistance (from Norway) and the ongoing nature of that project, the NSO was primarily interested in working on the LFS. We discussed the development of a sample frame for an ISS and the work they currently are doing to update their business registry and collect establishment data. Main findings of the visit: Currently, basic information about Malawi s labor force is collected in short modules of questions that are attached to regularly scheduled surveys. Malawi would like to have a better understanding of its labor force and the NSO is committed to generating this information, but funding is a major obstacle. The NSO hopes to conduct a stand-alone LFS but if they are unable to find funding to do this, they are considering other options. One option involves featuring the labor force module on the 2009 Integrated Household Survey. If the labor force module is featured, they will be able to expand the number of questions they ask on the topic. In 2008, Malawi will conduct a decennial Population and Housing Census. If the LFS is conducted in a timely fashion after completion of the census, the household listing and enumeration areas can be re-used to select the LFS sample. The NSO has developed a draft proposal for conducting a LFS. They have met with the Ministry of Labor and the Ministry of Trade and Industry and they plan to meet with additional stakeholders. Ongoing development of the proposal and additional meetings with stakeholders are needed. 3

5 As part of a larger project to improve Malawi s national accounts, the Norwegian government is assisting the Malawian government in updating their business registry. In December 2007, they began a census of large establishments. In the census, they are collecting information from private companies, government establishments, and non-government organizations about their main activities and the number of persons they employ. As part of a larger project to improve Malawi s national accounts, the Norwegian government is planning to collect information about informal sector enterprises. The Norwegians currently are trying to get questions added to the 2008 Population and Housing Census that will identify households with informal sector enterprises. If they are unable to do this, they are hoping that this information can be collected on the 2009 Integrated Household Survey. Priority 1: Develop Household Labor Force Survey A. Availability of labor force information in Malawi Malawi s last LFS was conducted in These data were analyzed but political circumstances prevented their publication. While Malawi has not had a LFS in 25 years, they do collect basic information about their labor force via modules of questions that are attached to regularly-scheduled surveys. The annual Welfare Monitoring Survey includes eight questions related to employment that are asked of persons age five and older. These questions are used to identify persons who are economically active (one week reference period) and to generate employment rates, unemployment rates, and labor force participation rates for numerous demographic and geographic groups. For employed persons, information is collected about the status of their employer, the industry in which they work, the number of jobs they hold, and how they are paid. Persons who are not in the labor force are asked to provide a reason. The Integrated Household Survey is conducted every five years (most recently in ); persons age 15 and older are asked questions about their labor force participation, employment, and time use. There are 26 questions on these topics. In addition to collecting information to generate employment rates, unemployment rates, and labor force participation rates, the survey includes questions about the industry, occupation, earnings, and hours worked of employed persons; participation in casual labor; hours spent doing income-generating activities; and socio-demographic information. Every 10 years, Malawi conducts a Population and Housing Census. Malawi s NSO will conduct its next census this year, in The census form includes eight employmentrelated questions to be asked of all persons age 6 and older. From the responses to these questions, the NSO will be able to measure the employment rate, unemployment rate, and labor force participation rate; the economically active population (one week reference 4

6 period); the occupation, industry, and type of employer for employed persons; and the main reason persons who are not in the labor force do not work. B. Malawi would like to have more information about its labor force Malawi would like to produce additional information about its labor force and the NSO has begun developing a LFS for this purpose. The NSO has drafted a proposal that outlines their objectives and plans. Their goal is to generate reliable information on employment, unemployment, and other labor force characteristics of the population. They would like to learn more about underemployment, especially among farmers; about the socio-economic and demographic characteristics of persons, including children and youth, who are in the labor force; about the characteristics of persons who are not economically active; and about informal employment. C. Review of Proposal The NSO developed a draft proposal for a LFS that they asked me to review (see Appendix C). I reviewed the proposal and provided feedback on it; some of the main points of our discussion follow. Detailed notes from our discussions appear in the appendices. The 3-month data collection period may not adequately account for seasonality; because the majority of Malawians are farmers, seasonality of agriculture must be considered when designing the survey. Other, smaller industries such as construction also exhibit seasonality. We discussed different ways to account for seasonality in the survey design. We talked about including both long and short reference periods for questions about economic activity; spreading data collection over an entire year, if possible; and, choosing a typical month or period for data collection if the fieldwork cannot be extended over the year. Malawi would like to collect data about child labor, but the proposed universe does not include all economically active children. The 2008 census includes questions about economic activity for all persons age 6 and older; we talked about using the same age cutoff for the LFS. While asking the same questions of children and adults may lead to comprehension issues, this was not much of a concern because in Malawi parents usually respond on behalf of their young children. The proposal lists several ways that Malawi would like to use the labor force data. Some of these uses necessitate that the data be produced at regular, frequent intervals (for example, to produce employment projections or in collective bargaining and wage contracts), yet it is proposed that the data are produced every five years. We talked about ways in which Malawi could use infrequently produced labor force data. 5

7 The proposed method of sample selection is for a two-stage, self-weighting sample. It also is stated that data about urban and rural areas are desired. To ensure the sample is sufficient to produce urban and rural estimates, the NSO will need to over-sample urban areas in the first stage of sample selection; by doing this, the sample will not be self-weighting. If you choose to over-sample urban areas, record the probabilities that are used to select the sample; this information is needed to design appropriate weights for the data. Cost should always be a concern when designing a survey. We talked about whether it is cost effective to use laptops to collect the data. We discussed the pros and cons of using laptops versus paper questionnaires to field the LFS and which option made the most sense for Malawi. The data analysis appearing in the proposal requires more thought and careful consideration. Labor force results from past surveys and censuses are discussed but the presented results and analysis do not always make sense. We discussed economic and non-economic reasons for differences in measurements over time including whether the data were collected and generated using comparable methods and the need to explain large changes. According to the proposal, the NSO would like to collect information about occupational hazards in the LFS. Upon discussion, I learned that the NSO was interested in learning about on-the-job deaths and injuries and whether persons take preventative measures (for example, whether they wear protective gear) to reduce their chances of injury or death while on the job. We talked about whether a household survey was the best tool for collecting information about occupational hazards. I expressed concern that the NSO is trying to measure too many concepts on the LFS and that they should consider narrowing their focus on this initial iteration. D. Developing concepts and questions for the labor force survey Employment and unemployment are defined in the proposal. We discussed these concepts and also talked about whether it made sense to include non-market work activities that fall within the United Nation s System of National Accounts in determining whether or not a person is employed. We talked about how employment status had been determined in past surveys in Malawi and about which non-market activities to include. We looked at questions that have been used by Malawi and other countries to collect this type of information in the past. We talked about measuring underemployment. The NSO is particularly interested in measuring underemployment of farmers during their off-season. We looked at questions that have been used by other countries and what the International Labor Organization (ILO) recommends. We also talked about how the timing of data collection could impact these results. 6

8 We discussed informal sector employment and different criteria for determining whether a job is in the formal or informal sector. We looked at questions that other countries use to determine whether a job falls within the formal or informal sector. We looked at LFS questionnaires used by several countries. We also looked at questions appearing in the ILO manual, Employment, Unemployment, and Underemployment, and information appearing on the ILO s website and that of Statistics South Africa. See the appendices for more information. E. Meeting with stakeholders We talked about the importance of soliciting input from stakeholders during the survey design process, to ensure the data will meet their needs. Prior to my visit, the NSO had met with two of their stakeholders; they had met with the Ministry of Labor and the Ministry of Trade and Industry to discuss their plans for a LFS. Mr. Makwemba, Mr. Matumula, and I traveled to Lilongwe to meet with Mr. Brian Ngoma, the Chief Economist at the Ministry of Labor (detailed notes from this meeting appear in Appendix I). In our discussion with Mr. Ngoma, he suggested that the NSO hold a stakeholder meeting with other Ministries, trade unions, and business organizations, at the Ministry of Labor in Lilongwe. Malawi s NSO also plans to meet with the Chamber of Commerce and Industry in Blantyre. F. Selecting the Sample The NSO is conducting a Population and Housing Census in The 2008 census enumeration areas and household listing can be used to select the sample for the LFS, as long as the LFS is conducted soon after the census. Using this information from the 2008 census will reduce the survey s cost and improve data quality. To ensure the data are sufficient to analyze Malawi s urban labor force, urban areas need to be over-sampled. This can be done in the first stage of sample selection, by doubling the sizes of urban enumeration areas. Note that if urban areas are over-sampled then the data will not be self-weighting and thus you must be certain to document the probabilities used in the sample selection. This information is needed to generate appropriate weights for the data. See the appendices for more information. G. Work plan Malawi s NSO currently does not have funding to conduct the LFS and thus they are considering different options for producing these data. They would prefer to conduct a stand-alone LFS but if this is not financially feasible, they would like to feature the labor force module on the 2009 Integrated Household Survey (IHS). If the labor force module is designated as the featured module, it will appear in an extended form on the 2009 survey. 7

9 Because of this uncertainty, we produced two work plans for producing labor force information. Appendix J is a work plan based on the assumption that the LFS is conducted as a stand-alone survey. Appendix K is a work plan that was developed based on the assumption that the labor force questions are designated as the featured module of the 2009 IHS. Priority 2: Development of Informal Sector Survey A. Background information The informal sector is believed to constitute a large and growing part of Malawi s economy, yet little is known about it. The NSO would like to conduct an ISS in Malawi to learn more about this sector. B. Developing a sample frame for an informal sector survey There is no sample frame currently available to conduct an ISS. A sampling frame could be generated by adding questions to the LFS or another household-based survey or census. We talked about criteria that are used to identify informal sector enterprises and questions that the NSO could include in their LFS to identify whether a household enterprise meets the criteria. We also reviewed Ralf Hussmans recommendations on this subject, in his paper, Defining and measuring informal sector employment. See appendices for more information. C. Other donors involved in this project As part of a larger project to improve Malawi s national accounts, the Norwegian government is planning to collect information from informal sector enterprises. Their ability to do this depends on whether an appropriate sampling frame is developed. The national accounts project currently is trying to get questions added to the 2008 Population and Housing Census to identify households with informal sector enterprises. If they are unable to do this, they are hoping that this information can be collected on the 2009 Integrated Household Survey. D. Work plan Because they currently have no funding for an ISS and because their plans to conduct the survey overlap with assistance they already are receiving, the NSO was not interested in developing a work plan at this time. Priority 3: Development of Establishment Survey on Wages and Salaries A. Background information 8

10 Malawi would like to conduct an establishment survey to collect more information about employment, wages, and salaries. They collected establishment data in the 1980s, but have not done so since that time. B. Developing a sampling frame The NSO is conducting a census of large establishments to update its business registry. Fieldwork began in December 2007 and was ongoing at the time of my visit. The NSO is collecting information from large companies, non-government organizations, and government establishments about their main activities and the number of permanent and temporary persons they employ, by sex. The NSO has developed two collection forms one for businesses and another for non-government organizations and governments that enumerators are using to solicit the information (copies of the forms appear in appendix Q). The NSO plans to use the updated business registry as a sampling frame for establishment surveys. C. Concepts Mr. Matemba and I discussed concepts that can be measured on an establishment survey. We talked about the concepts measured on the Current Employment Statistics survey, the monthly establishment survey in the United States. We reviewed and discussed data collection forms used by this survey. We also reviewed the monthly news release as well as charts and analysis generated from these data. We discussed uses of establishment data. See appendices for more information. C. Other donors involved in this project As part of a larger project to improve Malawi s national accounts, the Norwegian government is assisting the NSO with the census of large establishments. D. Work plan Because of the ongoing nature of the establishment census, funding and technical assistance being provided by the Norwegians, and uncertainties about how the NSO s objectives and those of the national accounts project dovetail on this project, it does not make sense to develop an establishment survey work plan at this time. Summary of Recommendations Labor force survey Consider the impact of seasonality, especially in agriculture, when designing the survey. If possible, field the data over an entire year. If data are not fielded over a one-year period, carefully select the period when data will be collected and be certain to include questions with short and long reference periods. 9

11 Focus on producing information about employment, unemployment, underemployment, earnings, hours, informal sector employment, other core labor force concepts, and related socio-economic and demographic information. Delay collecting information about occupational injuries and deaths. Instead of asking separate questions for adults and children, expand the survey s universe to include all economically active children and ask all survey respondents the same basic questions. Generate separate estimates and analyses for children, youths, and adults. In the first stage of sample selection, use enumeration areas from the 2008 Population and Housing Census rather than those from the 1998 census. By delaying the survey a few months, the data will be of a better quality and you will be able to save money by re-using the Census household listing. If you decide to over-sample urban enumeration areas, record the probabilities that are used for the sample selection. This information is needed to generate appropriate weights for the data. Give careful consideration to how you plan to use the data you collect. Be realistic about how data that are not produced on a regular, frequent basis can be used. Consider cost, especially as you determine how the data will be collected. Weigh the costs and benefits of using laptops versus paper questionnaires for data collection. Organize additional stakeholder meetings to solicit input now, as you design the survey. Hold one at the Ministry of Labor in Lilongwe and another with the Chamber of Commerce and Industry in Blantyre. Include questions about non-market activities that fall within the production boundary for the United Nation s System of National Accounts. Solicit information from countries in your region about how they define labor force concepts and the questions they have successfully used on their LFSs. Ask Statistics South Africa for a copy of their questionnaire. Include questions to identify households with informal sector enterprises; this information can be used as a sampling frame for an ISS. If you are unable to identify funding sources to conduct a stand-alone LFS, try to get the labor force module designated as the featured module on the 2009 Integrated Household Survey. 10

12 Informal sector survey Include questions on the 2008 Population and Housing Census, 2009 Integrated Household Survey, or the LFS to identify households with informal sector enterprises. This step is necessary to generate a sampling frame for an ISS. Learn more about the plans the national accounts project has for collecting information about the informal sector. Explore the idea of working together to produce the informal sector information you and they need. Establishment survey. Learn more about the plans the national accounts project has for collecting information from establishments. Explore the idea of working together to produce the information you and they need. 11