Statistics South Africa: Strategic Plan 2009/10: briefing

Meeting report

Statistics South Africa (StatsSA): Work programme for 2009-10 financial year.
Mr Pali Lehohla, Statistician-General: Statistics South Africa, presented the work programme for Statistics South Africa (StatsSA) 2009/10 financial year.

He introduced the Isibala programme. Isibala was the Zulu word for a mathematical sum and this programme, focusing on statistical literacy in schools, was to be the guiding principal for StatsSA 2009/10 financial year.

There was a short review of the challenges posed by the former Minister of Finance, Hon Trevor Manuel. The strategic overview was contained in the strategic importance of statistics and StatsSA's mandate. The ten binding principles for the Statistics South Africa Act were presented. These principles were based on the United Nations (UN) Fundamental Principles of Official Statistics. Mr Lehohla made particular mention of the importance of the first principle - a democratic society. Without this in place all the other principles would be compromised.
Mr Lehohla pointed out the difference in approach to statistics between StatsSA and academia, specifically that StatsSA's main goal was the utilisation of information. The need for the independence of a knowledge institution (StatsSA) from government and policy makers was noted.

The legislative mandate was outlined according to the roles of the Minister, the Statistics Council and the Statistician-General in Sections 5, 7, 13 and 14 of the Statistics Act No. 6 of 1999 (the Act).

The key competencies sought by StatsSA were described as intellectual, technological, political, logistical and administrative.

The recent negative media coverage received by StatsSA focused specifically on disputes with fieldworkers, the better-than-expected unemployment figures released in 2008 and the feasibility of the goals set by President Zuma in his State of the Nation Address.

The key strategic priorities were listed as: the benchmarking and rebasing of the Gross Domestic Product (GDP) measure, reviewing the Producer Price Index (PPI), conducting the Living Conditions Survey, piloting the Census 2011, compiling a geo-referenced dwelling frame, developing a National Strategy for Development of Statistics, and hosting the International Statistics Institute Conference in 2009. He also detailed the activities to contribute to South Africa's strategic priorities of economic growth, price stability, job creation, improvement of service delivery, poverty reduction and producing accurate information on the demographic profile and population dynamics of South Africa (see attached presentation).

The strategic risks and challenges faced by StatsSA were described as the development of:  a single integrated business registration system, increasing statistical skills in the country, implementing an effective operating model for field operations, rolling-out the South African Statistical Qualifications Framework (SASQAF) and establishing a corporate data processing centre.

Beyond 2009, the planned initiatives were the development of the future state for statistical production in the country, and reviewing the baseline information in the statistical system. Strategic shifts would be required in terms of a national planning requirement for structural issues and for markets and their failure, the research agenda and globally coordinated statistical responses. Enabling sections of the Statistics Act for the shift beyond 2009 were listed as subsections 2(c) and (d); 3(1)(a),(b) and (c); 4(3); 7(2)(f); 7(2)(g)(ii); 7(2)(j)(i) and (ii).
The full implementation of the Act was to happen, chiefly, through the National Statistical System (NSS). The mechanisms presented to achieve the desired state of statistical capacity in South Africa were User- Producer Partnerships, Certification and Capacity Building.

Discussion
Dr D George (DA) noted that the presenter had referred to measuring poverty in multiple dimensions and the past talk of developing a poverty line index (PLI). He asked if measuring poverty in multiple dimensions would include the poverty line index and how it would be composed.

Ms Yandiswa Mpetsheni, Acting Deputy Director-General:Population and Social Statistic, : StatsSA, replied that StatsSA was updating the PLI to be in line with inflation. This would be released in June along with the dimensions for the Living Conditions Survey. When it came to evaluating the issue of what constituted decent work, there were questions asked on job security. There were however, no questions on social protection and the information gathered on fair income needed improvement.

Dr George pointed out the confusion that often existed around understanding statistical methodology, and asked if there was a way to translate these complex concepts into simpler language for the media and public in general.

The Chairperson referred to the approximation of R 30 billion as the value of the informal sector to the economy and asked what tools were needed to unlock potential in this sector.

Mr Lehohla responded that StatsSA had to improve the communications approach. This was a point that the Chairperson of the Statistics Council had raised. He referred to the furore surrounding the previously released labour statistics, and he added that although the statistics were correct, the reporting on those statistics had caused many problems.

Dr George referred the planned re-basing of the GDP and asked what plans there were to measure the level of economic activity in the informal sector. He was of the view that this was likely to be substantial, but the impact of the informal sector was unknown, as it was not measured.

Mr Lehohla responded that this was not a sector that StatsSA directly addressed. The third survey of employers and the self employed was to take place this year. The data collected so far told StatsSA several things: that women dominated in the informal sector, that incomes were small and collectively, the tradeable income was around R30 billion, that the environment in which they operated was not the best but where there was shelter and electricity, the self-employed were able to perform better.  Many started their businesses on money from relatives or their own capital. This showed that the small business initiatives so often talked about did not reach these people. The positioning of these schemes had to be better if it the informal sector was to flourish.

Dr Rashad Cassim, Deputy Director-General: Economic Statistics, StatsSA responded that the measurement of the GDP was often a trade of between timelines and quality. There was more detail on the quarterly figure of the GDP and the detail on the annual aggregation of the GDP was not as good. The annual GDP figure was more focused on tracking the turning points in the economic cycles, especially from a fiscal perspective, for analysis by policy makers. GDP was driven by the value-added approach. Change in the GDP was driven by the formal sector. At the best of times, the figures available on the informal sector were riddled with problems. How to measure the levels of variable versus the rates of change of those levels was the main issue in determining the turning points in the economy. The structure of economic statistics was based on the system of national accounts. These measures were essentially macroeconomic. StatsSA’s efforts were more geared toward measuring the sophisticated formal sector and it relied heavily on data from South African Revenue Services (SARS) to do that. There was probably the need to introduce an economic census, but this was still up for debate.

Dr George asked if in general the measurements were quantitative or qualitative.

Mr Lehohla responded that there was a need to look at issues of structure. StatsSA’s job was to monitor and report but it had not provided adequate information for Integrated Development Plans (IDPs). Producing these statistics had to be the core business, as this information was the basis for business decisions in government. StatsSA did not have adequate capacity and this was the aim of its capacity building plans. This strategic shift was both quantitative and qualitative.

Ms N Sibhidla (ANC) wondered how StatsSA interacted with government departments and institutions. She said the information produced by StatsSA was crucial to the planning process in government.

Prof Akiiki Kahimbaara, Executive Manager, StatsSA, responded that the NSS was aimed at this kind of co-ordination but it would only produce about 10 % of the data needed. The other 90% would have to come from the government departments. StatsSA had to interact with these departments to address this issue.

There were problems with data availability and data quality. South Africa also needed capacity in the form of more trained statisticians. Even when government departments had statisticians on their staff, these statisticians often did not enjoy importance at the decision making level. Interaction with departments took place through a well thought-through legal framework but in practice, StatsSA lacked teeth. Empowering the Statistician General could be achieved through legislation but this would take time. With this in mind, an interim political arrangement would be necessary. Although StatsSA did have good interactions with several government departments, this was strictly on a voluntary basis. There was a need to encourage compliance from departments with StatsSA.

Mr Lehohla responded that StatsSA had introduced the Isibala programme as an outreach programme in schools to enthuse children about statistics.

Ms Sibhidla queried the possibility of a briefing on the public information initiatives on statistics and how this might help to deal with the problem of undercounting and false information. She queried the level of preparedness StatsSA had for the 2011 census, especially regarding the arrangements around fieldworkers and other logistical problems.

Mr Lehohla replied that the undercount was a big concern. There was no solution as yet. This appeared to be a structural issue. StatsSA had tested the theories that the problems were money and time. These theories were disproved. He was of the opinion that there was a need for a system similar to that of the USA. This involved leaving the census form in the household, to be filled in for collection at a later date. One of the obstacles to this approach was the large number of South Africans who did not have an address. He concluded that unless there was structural change the undercount would remain in the results of the 2011 Census.

Ms Sibhidla asked if StatsSA had a strategy to counteract the negative publicity.

Mr Risenga Maluleke, Deputy Director-General: Corporate Relations, StatsSA, responded that this was sensationalist reporting from the media. The reality was that the media operated in a self-regulatory environment. This was beyond StatsSA's power. This was an issue of the information that was in the public domain versus inaccurate reporting. StatsSA did do some training with journalists on the methods used to produce the figures released. This was ongoing and there was also recognition by StatsSA that it needed to be improved.

Mr Trevor Oosterwyk, Manager: Communications, StatsSA, responded that there was a schools outreach programmes. One of these was the "Census at school" project. This project had a two-fold purpose. It was an instrument of generating interest in statistics among young children.

He added that the media coverage often comprised the views of economists, analysts and academics that did not agree with the statistics released, and raised these issues with journalists. StatsSA had undertaken user discussion groups with these academics to reduce misunderstandings in the press.

Ms Sibhidla asked what was being done to generate interest in the general public - beyond school children. She pointed specifically to the reluctance in many communities to comply with surveys and the Census, and its impact on the undercounting problem.

Mr Oosterwyk replied that StatsSA had appointed publicity officers who would go door-to-door to prepare communities for fieldworkers. StatsSA planned to launch a newsletter and was developing the idea of media liaison officers in communities. The media liaison officers would be people who lived and worked in the community and would eventually report back on the results of the survey and census to their communites. The aim here was to improve StatsSA's relationship with respondents.

Mr Maluleke noted that a census was the highest public mobilisation, other than war and elections. He said that the Government might want to find a way, in the context of Africa Statistics Day (on 18 November) to engage policy makers in advocacy for statistics development, particularly in elevating the status of the census. StatsSA needed support to reach society.

The Chairperson commented that there was a lot of interest in the first census in South Africa. Interest had since decreased, along with interest in other surveys and voter education. Awareness programmes had to be intensified. As the South African economy became more sophisticated and immigration continued, challenges to statistical measurement were likely to increase. The awareness programme would mainly comprise educating the population about the importance of the census.

Ms Z Dlamini-Dubazana (ANC) asked StatsSA to differentiate between the concepts of impartial compilation of statistics and estimates. She further queried the process used to arrive at estimates.

Mr Lehohla responded that statistics was the science of estimation. This involved quantifying and making categorical statements about the level of uncertainty on the figures that would be released. These were technically called confidence intervals. This was a transparent scientific method.

Dr Cassim responded that estimates were usually based on consultations. StatsSA did provide metadata on the methodology and inputs used to arrive at the estimates.

Ms Dlamini-Dubazana pointed out that she did not see a programme that measured the levels of competencies in the Strategic Plan. She noted the paradigm shift in the global environment and asked what had been done to ensure the consistency of these competencies.

Ms Dlamini-Dubazana asked if StatsSA provided advice to government sectors and departments and if so, how often.

Ms Dlamini-Dubazana noted the relatively small equitable share assigned to KwaZulu-Natal and asked what demographic factors were used to arrive at this equitable share. She felt there was no correlation between the equitable share and the demographic reality of Kwazulu-Natal.

Mr Howard Gabriels, Chairperson: Statistics Council, replied that it was important to remain vigilant on where the separation lay between providing information and formulating policy. StatsSA's job was to provide the right information and it was not its job to make policy, as this involved making value judgements. The equitable share issue was a good example of this. Statistically, Gauteng and the Western Cape were growing faster than Kwazulu-Natal. The implication was that this then affected policy, as the figures were used by the Financial and Fiscal Commission (FFC) in its formulas, to determine the provincial equitable share.

Ms Dlamini-Dubazana referred to the five key deliverables that government had outlined and asked how StatsSA was advising government to aid the achievement of these deliverables. She further asked if the programmes that government planned for achieving the deliverables were achievable, or simply amounted to a wish list.

Mr Lehohla replied that the next strategic document to be presented by StatsSA would focus on this. With regard to the deliverable of employment, the Labour Force Survey had looked at that. With regard to rural development, there were not good statistics on the levels of subsistence agriculture. StatsSA was working with University of Fort Hare on this problem. Similarly, StatsSA worked with the University of Stellenbosch on territorial statistics, specifically urban and rural development.

Mr Gabriels added that there were significant shifts in the national priorities from the programmes of the last electoral term. There was a new five-year strategy. It was critical for the Committee to consider where StatsSA fitted in with the national priorities. In regard to Census 2011, there was a need to seriously address the issue of undercounting. If this was not achieved, then it would be impossible to accurately isolate the trends in demography, and this was a critical to providing the right information to the National Planning Commission. This data would also inform on planning for provincial and local government.

StatsSA, together with the Department of Trade and Industry (dti) and the South African Revenue Services (SARS) planned to produce an integrated business register. He suggested that the Committee interrogate the progress made on this. This register was part of the infrastructure for a good statistics system and would facilitate other forms of sampling.

The Dwelling Frame was necessary for better sampling in household surveys.

He noted that the Committee had broadly expressed dissatisfaction with StatsSA's approach to communications. He pointed that this was a matter that was also raised by the Council. He suggested that StatsSA also be called back to report on this matter.

Ms Sibhidla asked if StatsSA was of the view that policy makers were using the information effectively. She asked what steps were necessary to ensure that information was used effectively by policy makers.

Mr Gabriels reiterated that StatsSA was not and should not be involved in policy making. It was important to protect the independence of the statistics produced. The statistics agency was accountable to all in the country - business, unions and citizens. StatsSA should not have to defend the numbers. Often StatsSA had a wealth of information that was not utilised. However, the doors to access this information should be opened through better distribution, so that the information was put into the hands of policy makers. StatsSA would discuss this further with the Committee once the benchmarking was completed.

Ms Sibhidla asked what the strategy was to deal with building capacity and filling vacancies.

Ms Thandeka Mxengwe, StatsSA replied that this was contained in the human capacity development document. StatsSA had started the implementation two years ago. It expected long-term gains to be realised from these projects. The strategy had three phases. At school level, StatsSA had initiatives to develop interest in careers in statistics. At university level, it would concentrated on intake and development of university recruits. The problem at the third professional level was that there was a lack of official statistics on training in South Africa. In response to this, StatsSA had programmes in place, to use other African countries to do this.

Ms Dlamini-Dubazana suggested that StatsSA recognise the registered schools in Kwazulu-Natal and use the teachers to assist in the survey and the census. She felt this would be very cost effective.

Mr Lehohla replied that it was increasingly difficult to use teachers to complete these surveys, given the pressure currently on teachers. Moreover, surveys needed to be conducted outside of office hours, when people were home. This may also prove problematic for teachers. StatsSA was ready to re-engage with the Department of Basic Education on the possibility of using teachers for surveys, as they were in many respects ideally placed.

Mr Ashwell Jenneker, Deputy Director-General: Statistical Support and Informatics, StatsSA responded that the basic element of the dwelling frame was in developing good quality statistics. StatsSA was working on getting all dwellings geo-referenced and use of unique numbers. 70 % of this project was completed. It had planned to have finished the project already, but there had been reprioritisation due to changes in resources. This would help to reduce the undercount for Census 2011.

The Chairperson noted that the StatsSA budget was not allocated within the National Treasury. StatsSA essentially had the status of an independent government department and had authority over its allocated resources. He suggested that StatsSA return for a more intense discussion on the strategic plan and budget, where the officials would present measurable objectives and programmes.

Prof Kahimbara replied that he agreed with the suggestion.

The Chairperson noted that the issue of capacity went beyond the control of StatsSA.
He asked how the ISI could address the problem of undercounting. This undercount of about 17% had a serious impact, especially on the rural areas. He also enquired as to the role South Africa played in the preparations for the conference.

Dr Jaire Arrow, Deputy Director-General: Methodology and Standards, StatsSA replied that participation in scientific dialogue from the African continent had been minimal. The numbers for attendance looked good. Of the 1 500 registered delegates, 600 were from the African continent. The largest proportion of these 600 delegates were from South Africa, followed by Nigeria and Uganda. StatsSA had raised funding for gifted statisticians to present papers at the conference and interact with professionals. This had positive impact on capacity building. He added that there was a need to find a way to foster appreciation for the work statisticians did and how this related to more centralised planning through the National Planning Commission (NPC).

The Chairperson noted that there was a lack of mathematics and statistics researchers in the country and pointed out the importance of PhDs in future. There was an ageing population of researchers at this level. There was no emphasis on the stimulation of intellectualism in South Africa and this was a factor that could be used to guide policy makers in future. He pointed to the relatively low spending on Research and Development (R&D) in developing countries versus the relatively higher spending on this area in developed countries.

Mr Gabriels replied that the Council convened a workshop to look specifically at building capacity for statistics training at universities. This was necessary as universities would soon be unable to provide high level statistics training and the knowledge capacity needed to support Masters and PhD (Doctoral) students.

The Chairperson referred to the target of creating 500 000 jobs by December 2009 set in the State of the Nation Address. He asked if there was integration of the three departments, dti, StatsSA and SARS, to facilitate businesses and ensure that there was not too much lag between the initiation of these projects and results.

Dr Arrow responded that this related to the NSS. StatsSA was putting together a project plan. This would be an important source of information to inform on government priorities.

The meeting was adjourned.