PARIS21 Annual Consortium Meeting, 4 and 5 October 2001, Paris
South African South-to-South Cooperation
J Akiiki Kahimbaara,
Statistics South Africa
South-to-South cooperation in this paper refers to the process of mutual support or interdependence regarding statistical development among developing countries in order to inform policy leading to especially positive socio-economic development. North-to-South cooperation has traditionally been the norm for development efforts in countries of the South. However such cooperation has largely not fostered interdependency between North and South, but dependency of South on North. Over time this asymmetrical relationship has tended to continually increase inequality between North and South. While South-to-South cooperation is not an alternative to North-to-South cooperation, it provides a supplementary strategy for initiating and sustaining development efforts in the South. The main benefits for a country in the South are:
Encouraging and strengthening regional development, which is a pre-requisite for success of any individual country;
Taking responsibility and accepting accountability for development initiatives and their sustenance;
Rationalising the use of human and financial resources; and
Sharing and learning from common experiences.
Against this background is an outline of the role Statistics South Africa1 has played to foster South-to-South cooperation, primarily in Africa, and to a lesser extent beyond Africa. The following issues are examined:
The role of Statistics South Africa in supporting south-to-south cooperation within South Africa herself; especially changes made to the statistical system to accommodate policy in SA;
Training of civil servants in statistics
South Africa’s support for regional activities in Africa, including exchange visits or study tours; and
Lessons for other countries and challenges facing south-to-south cooperation
Until 1994 today’s South Africa did not exist. The country was geographically fragmented into several so-called Republic of South Africa, four independent homelands (Bantustans) and six self-governing states. It was also demographically fragmented along colour lines into four so-called population groups – Whites, Coloureds, Indian/Asian and African, in that order of importance. Of significance to this paper is that the statistics function was equally fragmented. The erstwhile Republic of South Africa had different statistical institutions differently focussing on population groups. Statistics South Africa focused on the White, Coloured and Indian populations, while other institutions focused on the African population, on which very limited accurate statistics were ever generated. Each homeland had its own “national” statistics office.
After the advent of the new political dispensation in 1994, the national statistics office was transformed in 1995, integration of offices being completed 1996. The statistics service was characterised by
Geographical and demographic under-coverage;
Under-coverage and irrelevance of statistical products;
Statistics that were unreliable and lacked integrity;
A staff that was inexperienced, and with limited skills in collecting data within a Third World context; and
An organisation that was demographically unrepresentative.
3. South–to-south cooperation within South Africa
The foregoing scenario is characterised by a statistics function that is irrelevant to the needs of the new South Africa. Accordingly, this section addresses what was needed and was actually done in order to make Statistics South Africa relevant to the new South Africa. In essence Statistics South Africa had to engage in intra-national south-to-south cooperation for it to be relevant. In short, what was needed was for Statistics South Africa to change into an organisation would:
Produce statistics that were demanded by users, and not driven by producers; especially statistics that were relevant to government policy, emphasising growth and development;
Produce statistics that were reliable in terms of accuracy, integrity, timeliness and users’ perception; and
Be representative in terms of staff composition.
The main objective was to reorient the vision of Statistics South Africa to support users, particularly by engaging in a number of activities, as indicated below.
3.1 Advocacy for the use of statistics in policy-making and decision-making by government Advocacy for the use of statistics was equivalent with marketing Statistics South Africa. This was process in which Statistics South Africa had to tread delicately because it involved making “promises” some of which could not meet the expectations they raised.
3.2 Re-definition of existing products or introduction of new ones, including innovations, in order to be relevant to the needs; first, of government; and subsequently, of the private sector and the public at large. As a national statistical office, Statistics South Africa was known to produce only economic statistics.
As a response to the new government’s launch of the Reconstruction and Development Programme (RDP), which attempted to effect equity in the delivery of social services, good governance, capacity building and civil service transformation, Statistics South Africa introduced demographic and social statistics into its programme.
In 1996 Statistics South Africa run a population census that covered the entire country on a wide range of life circumstances in the country, and sensitive to respondents needs such as language
Annual household survey (came to be known as the October Household Survey), was extensively modified to inform the RDP, with emphasis on geographical and subject-matter coverage. In the course of events the annual household survey was supplemented by other surveys, such as the Income and Expenditure survey (IES) held in both 1995 and 2000, the Labour force survey (LBS), agricultural survey, rural survey, etc.
Narrative user-friendly reports were introduced to assist users in interpreting tables.
A gender statistics unit was established and published a comparative analysis on life circumstances of men and women in South Africa as well as how men and women use their time in South Africa.
Statistics South Africa acceded to the government request to establish Special Data Dissemination Standards (SDDS), which is an IMF requirement and condition of access to international money markets and loans.
Statistics South Africa also implemented the United Nations’ 1993 System of National Accounts (SNA-93), which was an SDDS requirement.
A move towards production of environmental statistics and accounting was made in 1998, arising from a World Bank-funded project.
Statistics South Africa also got involved in HIV/AIDS campaign and research, which is quite a challenge given the difficulty of obtaining information on infections.
Another response to relevance was the introduction of consulting services for clients prepared to pay for surveys. This occurred in the absence of adequate resources available to Statistics South Africa to undertake the survey. Some examples are:
A request from the Department of Transport: Administration for Statistics South Africa to continue the processing of road traffic collisions and monthly survey of price indices for the civil engineering industry
A request from the National Secretariat for Safety and Security, jointly with UNDP for Statistics South Africa to undertake a victims of crime survey; and
A request from the Department of Labour, in conjunction with ILO, for Statistics South Africa to undertake a survey on child labour in South Africa.
Statistics South Africa decided to be timely with the release of its products by setting up fixed schedules when products would be released, especially with regard to economic statistics releases.
Finally, there was initiated a move towards more frequent and timely survey results, such as the currently bi-annual LFS, which will later become quarterly.
3.3 Structural re-organisation in order to be more effective
It was necessary to institute a comprehensive restructuring and rationalisation of staff, especially with regard to incorporation of homeland offices.
Staff were re-organised into accountable work teams, and their performance was reviewed against key performance areas.
Branches of the national office in provinces to support provincial needs
A new structure incorporating demography, household surveys and vital statistics, finance and provisioning administration were introduced.
An analysis unit to produce user-friendly digests of published tables, as well a statistical consulting service were established.
A review of the Statistics Act to increase statutory independence was put in place.
A programme of technical assistance, supported by Statistics Sweden, the Australian Bureau of Statistics, UNFPA, Statistics Norway, Statistics Canada, the Swiss Agency for Development and Cooperation, wa put in place.
3.4 Re-organisation of processes for more effectiveness and efficiency
In order to be more effective and efficient, Statistics South Africa reorganised its organisational and work processes.
There was a deliberate effort to implement leadership development in order to encourage strategic planning.
Strategic planning necessitated institutional development – involvement of all staff.
Stakeholder participation and cooperation were harnessed in the form of advisory committees to major products for relevance and quality assurance. An example is consultations during survey design and during the various phases of census planning.
Cooperation with other government departments to advocate intra-state south-to-south cooperation. Here are some examples.
Tracking local government finance together with the South African Reserve Bank (SARB), Finance and Fiscal Commission (FFC), South African Development Bank (SADB), Departments of Finance and Constitutional Development.
A study of VAT returns, with the South African Revenue Service (SARS)
Implementation of SNA-93, with SARB.
A new business register based on VAT, with SARS, the Registrar of Companies and Department of Trade and industry
Vital Statistics - birth, death, migration statistics – with the Departments of Home Affairs and Health together with the Medical Research Council
Tourism statistics, with the Department of Environmental Affairs and Tourism and the South African Tourism Board
Electoral geography whereby Statistics South Africa provided the Independent Electoral Commission (IEC) with GIS-based enumerator area data of Census 96 to demarcate electoral wards and voting districts.
Efforts to move away from expensive collection of primary data (frequent censuses and surveys) to cheaper register-based data.
Development of a national statistics system (NSS) to integrate the interests of all stakeholders – users, suppliers and producers of statistics. Tis is a major undertaking currently in vogue at Statistics South Africa.
3.5 Re-engineering of products for relevance, effectiveness and efficiency
Re-engineering of products has occurred largely in economic statistics. It is intended to streamline economic indicators and to speed up turnaround time. For example, 17 disparate collections on employment were replaced with 3 complementary economy-wide surveys
3.6 Introduction new technology for efficiency and capacity-building
Statistics South Africa staff have basically moved away from mainframe computers to network-linked PCs.
PCs are available to every member of staff, which has resulted in building capacity in computer literacy.
A geographic information system (GIS) has been put in place to drive census and sample survey data collection. The use of this technology extends beyond the needs of the organisation. For example, it was instrumental in the production of electoral data. The enumerator area spatial database provides a sampling frame for many surveys nationally.
Statistics South Africa has developed an Internet website where most of its products may be found.
E-mail and fax technology is commonplace in the organisation.
Technology for the scanning of questionnaires during data capture has been in use for household surveys for some time. It will now be used in the processing of Census 2001. The advantage is timeliness (should reduce processing time by half) and reduction of error in data entry.
3.7 Capacity building programmes
In-house training, training by consultants as part of technical assistance and training at international institutions form part of capacity building at Statistics South Africa.
In-house training consists of short courses both in management and in professional areas. A relatively long-term course has been on complex survey sampling for statisticians. Statistics Sweden has continually conducted another course, Statistics in Action. It is a hands-on introduction to the survey research cycle. Understudying a consultant by a local staff member is also a common method of capacity building.
International institutions, such as the Institute of Social Research (ISR) at the University of Michigan, the University of Pennsylvania, the Institute of Statistics and Applied Economics at Makerere University (ISAE) and the Eastern Africa statistical Training Centre (EASTC) at the University of Dar es Salaam, have provided or are all providing short-term and long-term training.
3.7 Creation of a national statistics system
Establishing a national statistical system (NSS) is a currently a high level priority for Statistics South Africa because it is the most feasible way to maximise satisfaction of user needs and to cater for the interests of suppliers of data and producers of statistics. The NSS is intended to streamline production of statistics at the country level by integrating user needs with the activities of producers and suppliers of statistics. The main advantages of the NSS are to facilitate efficiency in national planning, to save costs by avoiding duplication of activities and to set and monitor standards in definitions and methodology for the production of harmonised and consistent statistics.
4. Regional south-to-south cooperation
This section outlines regional cooperation; first at the sub-regional level in southern Africa; and second, at the Africa (continental) level.
South Africa is a member of the Southern African Development Community (SADC). The region shares a history and a culture that necessitate cooperation in development programmes. Statistics South Africa has accordingly played a prominent role in building statistical capacity in the region.
4.1.1 The 2000/01 round of censuses of population and housing Statistics South Africa has been responsible for organising the collaborative effort for the round of censuses planned by SADC members to take place in 2000/01. Eight countries have scheduled censuses in the period 2000-2002. Other countries will run large demographic surveys. Cooperation has already occurred with regard to developing a common set of core items in the questionnaires.
Census-mapping workshops were held in Bloemfontein in May 2000 and in Pretoria in March 2001, and another one is scheduled for November 2001.
Statistics South Africa was also instrumental in convening workshops outside South Africa. One workshop, on sampling where short and long questionnaires are used, was held in Lusaka, Zambia, in February and March 2000; and another, on census data processing was held in Gabarone, Botswana, in October 2000.
A census managers’ meeting was held in March 2001 in Pretoria, to evaluate the joint SADC/UNSD/UNFPA programme.
A workshop on the harmonisation of census data was held in Pretoria in October 2000, and was attended by representatives from the African Census Analysis Project (ACAP) based at the University of Pennsylvania and SADC. This cooperation will certainly continue, and will include other census processes. Training at a regional level and sharing both of methodological practices and data will remain issues.
4.1.2 National accounts
There already exists a SADC Sub-committee on National Accounts involved with coordination of workshops especially for the implementation of the System of National Accounts of 1993 (SNA-93).
A meeting took place in South Africa last May.
The next meeting, to focus on measuring the informal sector, will take place in Malawi in May 2002. There is emerging a trend for member states to prepare quarterly GDP estimates or estimates of some quarterly main aggregates of GDP where staff compliments are small.
4.1.3 Harmonisation of the Consumer Price Index (CPI) and budgets of household surveys
A meeting took place in South Africa in June 2001 to review methodologies, product classifications and to prepare technical specifications for the harmonisation.
4.1.4 Poverty statistics Poverty is in many ways a uniting theme in the region. The focus is on developing both indicators and indices to measure poverty.
In December 2000 a SADC/PARIS21 workshop to identify statistical needs to address poverty in Zambia was held in Lusaka.
Statistics South Africa has decided to adopt the concept of Poverty Reduction Strategy Papers (PRSPs) as a guide to planning strategy to generate relevant statistics, although South Africa is not classified as highly indebted.
The main challenge with poverty statistics is training in poverty maps. There was a training workshop on some money metric methods to measure poverty in South Africa in April 2001.
Sustainable management of the quality of statistical products is a concern of SADC members. Consequently, a workshop on systematic quality management is scheduled for the middle of November 2001.
4.1.6 Observation visits
Statistics South Africa encourages learning by observation, and often affords its staff opportunities to be observers of certain activities of interest of sister organisations. The following are some of the visits:
In July 2000 two officials visited Namibia for a rapid assessment mission on their preparedness for Census 2000
Three officials visited Malawi to study the statistics system in Malawi in May 2001.
In May 2000 one official was an observer at the Zambian Census 2000 data processing rehearsal.
In October 2000 three officers were observers of field operations of the Zambian census
Three officials visited Mozambique to study the setup and operations of the National Institute of Statistics Mozambique in May 2001
CWIQ in Mozambique
Four officials visited Tanzania to observe implementation of the World Bank’s Core Welfare Indicators Questionnaire in January 2001
4.2 Africa region Statistics South Africa has also sought to encourage south-to-south cooperation skills transfer programmes in the Africa region.
Training of civil servants in official statistics.
The President of South Africa has expressed a desire for an economically literate society. Statistics South Africa believes that economic literacy subsumes statistical literacy, and has accordingly taken the lead to train for statistical literacy for the South African society, beginning with the training of its own officials from. This is in keeping with its vision of an NSS.
18.104.22.168 Training of Statistics South Africa officials The current priorities of Statistics South Africa, as a national statistics office, are to set up a NSS, to build capacity required to sustain the system and to improve the quality of its baseline products. There currently are no institutions that specialise in training in official statistics. Most statistical training at South African tertiary institutions emphasise actuarial sciences. However, training programmes in official statistics are available at two regional institutions in eastern Africa – the Eastern Africa Statistical Training Centre (EASTC) at the University of Dar-es-Salaam in Tanzania and the Institute of Statistics and Applied Economics (ISAE) at Makerere University in Uganda. There are programmes in place at these two institutions to provide major training to both staff from Statistics South Africa and other stakeholder institutions in the NSS. A main advantage of training at these institutions is relevance, arising from isomorphic data collection environments. There is, for example, less sense in seeking better training in data collection from informal settlements from institutions in the North than from institutions in the South.
Against this background that in September 2001 Statistics South Africa sent 12 officers to study for a Certificate in Applied Statistics and Economics at EASTC for a period of 10 months. These officers are mainly administrative clerks, including one transport officer and one assistant director. The objective is to provide basic literacy in official statistics. The main objective is to transform most of the staff into professionals.
A second group of 9 candidates was also sent in September 2001 to ISAE for varying lengths of time. Three are studying for a Master of Statistics; one for a Postgraduate Diploma in Statistics at the honours level; three for a Postgraduate Diploma in Demography; one for a Bachelor of Statistics; and one in a Bachelor of Science in Quantitative Economics.
These officers are legally bound to work for the organisation for as long as they will be absent during their training.
Statistics South Africa meets tuition and residence expenses. The officers depend on their own salaries for their upkeep.
On completion of their studies the officers will return to their previous positions at Statistics South Africa; they will have to compete for promotions like everyone else.
22.214.171.124 Training beyond Statistics South Africa A tentative agreement has been reached between Statistics South Africa and the South African Management Development Institute (SAMDI) on the need for training of civil servants in other government departments. It is envisaged that as many as 50 000 government officials could be trained.
4.2.1 Poverty mapping With support from the World Bank Statistics South Africa has embarked on a poverty mapping project designed to indicate the spatial incidence of poverty rates using small area data. The project has innovatively integrated population census data with sample survey data within a spatial framework provided by an enumerator area (EA) database. EAs can be aggregated into practically any administrative geography in South Africa. Mapping of poverty rates is important for poverty alleviation programmes. Poverty mapping has featured relatively significantly in south-to-south cooperation in the Africa region. The role played by the poverty mapping project is two-fold:
It involves sharing experience in poverty mapping, which is important for policy implementation; and
It encourages using poverty maps to communicate information about poverty conditions in any area.
Poverty mapping experience has been shared mainly through workshops and presentations.
SADC/PARIS21 Regional Workshop on poverty reduction strategies held in Lusaka, Zambia, in December 2000. Statistics South Africa presented a paper on the use of population census and sample survey data for census mapping in South Africa.
In January 2001 Statistics South Africa contributed to a workshop on poverty mapping for NGOs and other organisations (such as the International Livestock Research Institute and the African Economic Research Council) in Kenya.
In May 2001 a week-long training workshop in methodology and techniques of poverty mapping was held in Pretoria, South Africa. The training was done jointly by Statistics South Africa and the World Bank.
Plans for a regional training workshop in poverty mapping are under way for 2002.
5. Beyond Africa Statistics South Africa has also rendered professional support to countries beyond the African continent. Examples are:
Consulted for international agencies in Eastern Europe (to advise a conference on the use GIS) and Cambodia (to advise on preparedness for a population census).
6. Some lessons learnt
Growth and sustenance of a statistics organisation depends largely on satisfying user needs. The organisation has to produce what users want.
Advocacy is a tool a national statistics organisation has to have and constantly utilise, and it is best implemented from a user’s viewpoint. However, the organisation has to tread carefully between what it promises to deliver and user’s expectations.
Stakeholders – users, suppliers, other producers – should be treated as partners.
Working together as a group of statistics organisations facilitates advocacy and funding.
Every statistics organisation has something positive to offer another.
To be effective a statistics organisation has to have a clear, purposeful identity and a reasonable level of autonomy.
There is no opportune time for building sustainable capacity, as there will never be surplus staff to handle the workload of others going on long-term training. There must be a deliberate effort to train irrespective of workload.
Lobbying for resources is a necessity.
1 The statutory body responsible for the production of official statistics in South Africa was originally known as the Central Statistical Service. It was renamed Statistics South Africa in 1998 when it felt that the organisation had been sufficiently transformed to have acquired a new image and therefore a new name. For the sake of simplicity the name Statistics South Africa is used throughout the paper.