Statistics South Africa briefing on 2006/7 Workprogramme

Meeting report

FINANCE PORTFOLIO COMMITTEE
24 May 2006
STATISTICS SOUTH AFRICA BRIEFING ON 2006/7 WORKPROGRAMME

Chairperson: Mr N Nene (ANC)

Documents handed out
Presentation on Statistics South Africa Work Programme 2006/7: Part 1 & 2
Statistics SA Work Programme 2006/7
Statistics South Africa Press Release on 2006/7 Work Programme

SUMMARY
In the Statistician-General’s presentation on the Statistics SA Work Programme for 2006/7 he outlined the institution’s legislative framework, its strategic overview and themes, its key priorities, the specific activities used to realise each of the priorities and the outputs, measures or indicators and targets of each priority. He stated that Statistics SA’s future plans were to provide accurate, relevant and reliable statistics, to ensure trust in official statistics and to enable a numerate society.

During the discussion the Committee sought clarity as to how Statistics SA would align its Work Programme with the Accelerated Shared Growth Initiative of South Africa, the reason for the incorrect calculation of the total amount of R41 billion in tax collected and the R102 million allocated for the analysis component of the community survey. Statistics SA had moved from a postal survey to a direct collection model in collecting data for the Consumer Price Index (CPIX), which allowed them greater control and cost much less. The basket of goods for the income and expenditure survey was not being changed, but was merely re-weighted to reflect the changing spending patterns and priorities of society. The Committee was concerned that a lack of a unified definition of poverty was a major failing in the accuracy of the statistics compiled by Statistics SA, but the Minister of Finance contended that the real reason was the practical problems experienced in collecting the data.

Statistics SA was now looking at collecting data on the informal sector, which was not currently being covered. It was doubted whether Statistics SA data could adequately inform the United Nations’ Millennium Development Goals because the sample was not large enough. Efforts would be focused on improving Statistics SA’s campaign to communicate its statistics to the general public. The sound and effective measures in place to ensure accurate and reliable data was captured in the field were explained. Statistics SA experienced problems in collecting data on causes of death, especially as it related to HIV/AIDS, because the information on the Department of Home Affairs’ database was not always accurate and clear.

MINUTES
Introduction by Chairperson
The Chair requested that the input from Statistics SA be brief so as to allow sufficient time for interaction with Members. He welcomed Mr P Lehohla, the Statistician-General, and his delegation: Dr J Arrow, Deputy Director-General (DDG): Quality and Integration; Ms N Mokoena, DDG: Corporate Services; Dr L Gavin, DDG: Statistical Support and Informatics; Ms T Latief, Chief Financial Officer; Mr R Maluleke, Executive Manager: Office of the Statistician-General; Ms C De Klerk, Manager: Strategic Planning; Mr T Oosterwyk, Manager: Communications and Dr R Cassiem, DDG: Economic Statistics.

He stated that the Minister of Finance would be joining the Committee later, as he was currently chairing a Cabinet meeting.

Briefing by Statistics South Africa
The Statistician-General thanked the Committee for the opportunity to address it. He outlined the institution’s legislative framework. It was explained that the Statistics SA strategic overview and themes included the provision of relevant information to meet the needs of users, enhancing the quality of products and services provided by Statistics SA, the developing and promoting of statistical co-ordination and partnerships and the building of human capacity within Statistics SA.

He explained that Statistics SA aimed to provide relevant information on its following key priorities: economic growth, price stability, employment and job creation, life circumstances and service delivery and on demographic profile and population dynamics. The activities used to realise each of the priorities were outlined and the outputs, measures or indicators and targets of each priority were detailed.

Statistics SA’s supporting initiatives were explained, namely improving the quality of its statistics, improving its quality dimensions, the co-ordination and partnerships engaged in, the building of capacity within Statistics SA and its governance and accountability. Statistics South Africa’s future plans were to provide accurate, relevant and reliable statistics, to ensure trust in official statistics and to enable a numerate society.

Discussion
The Chair thanked the Statistician-General for being brief. He requested Members’ inputs to be both brief and concise.

Mr K Moloto (ANC) sought clarity on the reason for the significant reduction in administration transfers and subsidies to provinces. He asked whether it was due to Statistics SA scaling down its personnel component.

Ms Latief responded that Statistics SA’s Programme 1 expenditure for property management, in its Estimates of National Expenditure (ENE), had increased substantially. The primary reason for the decrease in the allocation to local municipalities was that more regional, provincial and district offices have been established. She stated that the reduction was not related to Statistics SA’s human resource component, because its personnel budget has in fact increased.

Mr Moloto asked whether the postal survey was the best route to take when conducting the agricultural survey. The presentation indicated that direct collection was the best methodology for the collection of the Consumer Price Index (CPIX) data. He sought clarity on the weaknesses or limitations of using the postal survey.

The Statistician-General replied that the census would still be run on the basis of commercial agriculture. Statistics SA intended running a broad agricultural survey in 2008, and an agreement was reached with the Department of Agriculture on the matter. In the 2003/4 financial year, Statistics SA managed to secure the necessary responses from that sector at a slightly cheaper cost via the postal survey. In terms of there being a possible contradiction between the postal survey and the mention made of the benefits of direct price collection he stated that the agricultural census was quite different from asking a storekeeper to give price quotations for items in his store, whereas securing information from farmers on their stocks and foods was a much more complex exercise. The geographical distribution and location of storekeepers across the country was quite high and extensive, whereas farming communities were sparser and were therefore more suited to a postal survey.

Dr Arrow added that the agricultural survey was in fact a business survey. The data on the informal subsistence economy was collected via the household survey, whereas this agricultural survey collected data on the agricultural business sector in the same way as the data on the manufacturing sector was collected. The standard manner of conducting the business survey was via a postal method.

Mr I Davidson (DA) noted that Statistics SA was changing its method of collection for CPIX data to direct collection in order to align it with international best practice models. He asked Statistics SA to indicate the results revealed by the test sample, and whether Statistics SA was currently over- or under-measuring its CPIX figures.

The Statistician-General responded that its first test showed very little or no difference between the direct or postal survey method, but it did highlight differences at the sub item level. The question then was why Statistics SA chose the one method when there was in fact no real significant difference between the two. The fact was that the direct collection method allowed Statistics SA greater control, and it cost much less. He stated that Statistics SA had so far transitioned very smoothly from the postal to the direct collection method.

Mr Davidson stated that Statistics SA was changing the basket of goods for its income and expenditure survey, which had an effect on the trust people put in the statistical data produced by Statistics SA.

Dr Arrow replied that Statistics SA had not changed the basket. It merely substituted the storekeeper with a Statistics SA official in the collection process. The major difference will occur when Statistics SA reworks the weight allocation on various items, which did in fact redefine how current consumption patterns changed within society. This re-weighting will be conducted in 2008. The last reworking of the CPIX basket was in 2000. He stated that Statistics SA planned to move towards more frequent re-weighting of the CPIX because it was not appropriate to assume that patterns of consumption would remain standard over a five-year period, which was the current forecast period.

Mr Davidson stated that his question had not been answered. He explained that the presentation referred to the updating of the basket of goods for the income and expenditure survey. The reply from Statistics SA referred to a re-weighting and he sought clarity on when that re-weighting would take place, because it lay at the core of the CPIX calculation.

Dr Arrow responded that the reality was that there was a different ranking of goods and needs currently than those employed ten years ago. The income and expenditure survey was thus important because it allowed government to gain a sense of the spending patterns and priorities of citizens, and that was the kind of data government found valuable in informing it of the emerging priorities of society so that it could respond to them.

He informed the Committee that the weight attached to food in the current CPI was just 23%, which ranked second. Ten years ago fast food was uncommon, but the income and expenditure survey results illustrated that these days it was commonplace.

Mr Davidson sought clarity on the purpose of the additional business surveys.

Dr Arrow replied that the business survey measured typical characteristics that Statistics SA measured in other business surveys. It would thus focus on turnover and on various ratios in production and profitability measured in other areas. The additional business surveys would address the gaps and inaccuracies that occurred in past surveys, which was a result of less accurate coverage of the economy. The data collected from those business surveys would be used to inform the quarterly estimations of the Gross Domestic Product (GDP).

Mr Davidson sought clarity on the re-engineering of the labour force survey (LFS), as indicated in the presentation.

Dr Gavin responded that the process involved the collection of information via face-to-face interviews, processing that data and then translating it into data and indicators that described the current labour market. Re-engineering meant that Statistics SA would be revising those steps with a view to possibly refining them. The work programme detailed the specific activities Statistics SA would be engaging in in 2006 and 2007, which were aimed at revisiting and improving the quality of the statistics it provided. These included focusing on accuracy of statistics provided, redoing its questionnaire, the timeliness of results published, a rethink of its sample and the relevance of the data provided to ensure Statistics SA provided data its users would be interested in. To that end Statistics SA would be redesigning a very short quarterly questionnaire and various supplementary questionnaires aimed at securing feedback from users on the relevance and usefulness of the statistics provided.

Dr S Van Dyk (DA) stated that the 2006/7 Work Programme indicated how Statistics SA would align its programme of action with the Accelerated Shared Growth Initiative of South Africa (ASGISA). He asked how exactly Statistics SA’s strategy and planning differed from its previous work plans in order to align itself with ASGISA, and whether it would be surveying new topics or provide support in new areas to government departments.

The Statistician-General replied that the presentation emphasised the alignment of the Statistics SA programme of action with that of ASGISA. It would not be engaging in new issues but would be making improvements aimed at conducting its programmes in line with ASGISA.

Dr Van Dyk stated that it was reported that taxes in the amount of R41 billion extra had been collected. Incorrect calculations were made either by Statistics SA or by the South African Revenue Service (SARS), which had serious consequences for the economy. He sought an explanation of Statistics SA’s role in the miscalculation.

The Chair welcomed the Minister of Finance to the meeting.

The Statistician-General responded that on the one hand the budget deficit was based on the GDP figure and if the GDP figure seemed to be exceeded by revenue, with the result that the budget deficit decreased, the question arose whether the GDP estimate process by National Treasury was in fact correct. That was the problem that faced Statistics SA. In fact, Statistics SA was backstating its revenue estimates to arrive at the core of the matter.

Dr Arrow added that the R41 billion collected could be attributed to increased coverage by SARS, a broader tax base in the country, increased capacity within SARS to extract from its existing clients or even due to a growing economy. It could be argued that all those factors, or some more than others, influenced the total tax collected and thus increased the level of revenue. He believed National Treasury was better placed to provide that kind of analysis. Statistics SA has covered the economy as best it could to ensure a more accurate reflection.

Mr Y Bhamjee (ANC) welcomed the informative input. His main concern was the collection of data on the extent of poverty in South Africa. He asked whether there was one working definition of poverty used by all government departments when collating information on the matter. If no uniform approach was used, based on a clearly defined meaning of the word, he asked why government was still grappling with the definition in 2006.

Mr S Asiya (ANC) stated that government’s goal was to half the current rate of poverty by 2014. Statistics SA indicated that it would be introducing methodology to measure poverty levels, and asked how effective it would be in ensuring the achievement of that government goal.

The Statistician-General replied to both questions by assuring Members that Statistics SA did in fact measure the extent of poverty within South Africa. The definitional issues have been resolved, which Statistics SA participated in. However the data Statistics SA already produced on housing, access to water and other facilities did provide an indication or a measure of poverty. There was thus not a total vacuum of information on the extent of poverty.

Dr Gavin added that Statistics SA was currently looking at standardising the definition of poverty. The reality was that different government departments used different definitions, and National Treasury was currently heading the process to standardise it. Most would agree that poverty was a multi-dimensional phenomenon, and the information Statistics SA was collecting in terms of access to services was important in continuing to provide a complete picture.

The finalisation of the definition was taking so long because Statistics SA needed to identify standard measures it needed to provide data for. The practical difficulty experienced by Statistics SA was that it was very difficult to get people to disclose their income and expenditure. Statistics SA found people were sensitive to declaring their income. The Work Programme described some of the activities Statistics SA would be launching to address the problem. These included the usage and expenditure survey that was aimed at re-weighting the CPIX basket of goods in order to give a better picture of what household were spending on a monthly basis.

Mr Bhamjee asked whether Statistics SA had seriously investigated the skills shortages in South Africa, whether it had identified the sectors that were experiencing skills shortages and which country was best suited to provide the needed skills.

The Statistician-General agreed that there were certainly skills shortages. He believed that with careful targeting and mentoring the skills shortages could be overcome, especially in the field of statistics. The real shortage was in the field of official statistics. It was very difficult for South Africa to bring in skilled statisticians from other countries, and it instead had to build the capacity within South Africa.

Dr Arrow added that Statistics SA did not conduct the type of survey referred to by Mr Bhamjee. It was however a fact that there was a serious skills shortage in the area of statistics, and it was a national challenge to address the matter. He believed the matter must be addressed at the very beginning of the education career, by ensuring numeracy in schoolchildren. Statistics South Africa was involved in partnerships with universities to make mathematics, which was the foundation of statistics, more attractive to learners so as to ensure more mathematics graduates. There was no shortcut to skills development and there were no countries that had "statisticians on the street" that South Africa could simply snap up. Previous attempts to bring skilled foreigners to South Africa have failed, precisely because it was a scarce skill worldwide. South Africa thus needed to build that skill in its own backyard.

Ms J Fubbs (ANC) sought clarity on the R102 million allocated for the analysis component of the community survey. She asked for the actual allocation to the fieldwork component as well as identifying the specific methodology to be used.

Dr Gavin replied that some of the activities were aimed at building capacity towards Census 2011, as part of Statistics SA’s monitoring and evaluation system. Statistics SA had tested its Census and Survey Administration System (CSAS) activities, which it regard as the system of choice for Census 2011. It would contribute to the provision of quality and accurate statistics in Census 2011.

Dr Gavin stated that the budget was aligned to planned activities. In February 2007 Statistics SA would be collecting information from approximately 280 000 households, and thus the bulk of the allocation would be on data collection. The second largest allocation was related to preparing people in the field, for the design of the sample and for the redefinition of the designated fields. In the next financial year Statistics SA would process the data. It was hoped that the results would be published in November 2007. The spending thus related to the transition from paper questionnaires to the production of reports and electronic data sets.

Ms Fubbs stated that the presentation indicated that Statistics SA planned to develop a monitoring and evaluation framework by 2007 for the population. She asked Statistics SA to explain the framework that was being used prior to the introduction of that framework, and which framework it would be using to conduct the community survey.

Dr Gavin responded that Statistics SA would build on the experience of conducting the community survey. The focus was really on the refinement and continued improvement of those systems as it moved towards Census 2011.

The Minister of Finance indicated that he would like to touch on a few of the issues raised by Members. The first was the country’s ability to know where its 48 million citizens resided. The best measurement was one that allowed Statistics SA to go back to the same individuals and households, to the same area or even to "proxies" who would be able to verify their current whereabouts.

The impact of the statistics gathered on the quality of life was exceedingly important. It was learnt from the information gathered on households in Census 2001 that households were increasing at a tremendous rate, but that South African family units were smaller than those in other developing countries. It was also discovered that far too many South Africans did not have a physical address, and thus the ability to work from a questionnaire that was processed electronically proved quite difficult. Thus the constraint experienced by Statistics SA regarding the absence of physical addresses affected the measurement of the quality of life of South Africans.

The second feature was the rapid rate of urbanisation. At the moment there was no unique definition of an urban, peri-urban or rural area because the Municipal Systems Act created 283 local authorities without distinction between those three forms. The uncertainty could very well result in bad distortions when measuring employment, for example, via the LFS. Statistics SA’s re-engineering process will enable it to start off with a better set of assumptions on where, geographically, people who were actively seeking employment were actually located. If the distortions in that first set of assumptions were minimised, it would create more accurate statistics. It was however an enormously intricate exercise.

The third issue he wanted to touch on was poverty. The pertinent issue, as raised by the Committee, was what the measure of poverty was. As Dr Gavin stated earlier, poverty cannot be measured only in money-metric terms. A fair balance was needed between the various forms of poverty, namely access poverty, income poverty and capacity poverty. He believed that, as South Africans, the matter has not been properly addressed. Discussions needed to be held on resolving the matter.

Fourthly, Minister Manuel stated that the income and expenditure survey was important. Unless it was known who earned what and what they spent it on, even the best estimates of the measure of inflation would have to be wrong. Statistics SA was currently in the process of shifting to the direct collection method, which was a very important shift. He stated that he was struck by the quality of the response received via the direct collection method. It was however a work in progress and sufficient data had not yet been collected to comment on.

Mr Davidson expressed his concern that the presentation made no mention of the measurement of the informal sector. It was a sizeable portion of the economy, but there was a lacuna of information on that sector. He asked Statistics South Africa to explain the steps it had taken to cover that sector, as well as the nature and size of the contribution of that sector.

The Statistician-General responded that Statistics SA was looking at the informal sector in its current LFS, and a report on the matter would be tabled. The survey on employers and the self-employed was being used to try to get to the second economy. That survey was a sub-module of the LFS. Yet, as the Minister indicated, there were many complicating issues surrounding the collection of data on that sector.

Mr M Johnson (ANC) thanked the Statistician-General for the informative presentation. He sought clarity on the statement in the work programme that the "executive authority vests in the Statistician-General", whereas he was of the view that the executive authority was the Minister of Finance.

The Statistician-General announced that he had just received word of the CPI figure, which was down by 0,1% to 3,3%, and the CPIX was at 3,7%.

Mr Johnson sought clarity on the extent to which the day-to-day functioning of municipalities was aligned with the United Nations’ Millennium Development Goals (MDG) He asked whether Statistics SA believed South Africa would be able to reach its MDG at the current rate, especially with regard to municipalities.

The Statistician-General responded that the data was there but he doubted whether it would inform the MDGs adequately, particularly as it related to the continent of Africa. A problem was that some African countries only had one data point, whereas at least two data points of an adequate sample was needed to make some inferences, going forward. Consequently, Statistics SA made a recommendation to the Secretary General of the United Nations that capacity building in the area of statistics was critical. Statistics SA was working with Canada on a possible partnership model that would ensure the MDG matter was fully addressed. The picture did not look that good but at least the diagnosis has been made, which was a step in the right direction.

Ms Fubbs sought an indication of the extent to which the information gathered from the community survey would be used to achieve the MDGs.

Minister Manuel replied that Statistics SA had put significant work into the report on the MDGs. He was of the view that it was more important for Statistics SA to gather the information and make it available to South Africans than to write lengthy reports on it for the United Nations. The information will be made available, but the work would probably be done within the Cabinet cluster.

Mr Johnson sought clarity on the extent to which Statistics SA’s targets were aligned to those of other government departments.

The Statistician-General replied that he would have to provide an indirect answer. He informed the Committee that he was invited to a Provincial and Local Government Parliamentary Portfolio Committee meeting eight weeks ago, in which the Department of Provincial and Local Government indicated that the Limpopo Member of Executive Council (MEC) for Finance had put in place Integrated Development Plans (IDPs) which made a lot of sense and which assisted in understanding the reality of integration of various government departments. Statistics SA would be considering his approach in the hope that it would begin to shed light on this problem

The Limpopo MEC presented information on the different economic sectors within the province, the province’s expectations of those sectors, the rate of job creation and the millions of Rands spent by each of those sectors. The question that was then posed to Statistics SA in that meeting was how it would use that data to identify the gaps. The Limpopo MEC in fact identified the gaps within municipalities, which he tabled before Cabinet. He had managed to standardise the province’s IDPs. The difficulty Statistics SA now faced was how it would progress on the basis of that 2001 data.

The data was standardised but the difficulty was the interaction with Statistics SA and other government departments. He was not sure how to ensure integration of data in the manner achieved by Limpopo.

Ms Fubbs stated that the presentation indicated that this year Statistics SA would be making a concerted effort on improving the communication of its statistics to the general public, and not only to those regular users. She sought clarity on the actual strategy, as well as the budgetary allocation. What is the strategy and what is the allocation to that this year and following two years?

Dr Gavin responded that Statistics SA regarded the improvement of its communication strategy as very important. It was of the view that there was no use in producing statistics that would never get used. As a measure to improving its statistical releases, Statistics SA has introduced the following initiatives: a writing course for staff to encourage them to expressing things as clearly and as simply as possible; and it was working on standardising the format of its releases to aid interpretation, because currently they were not standardised. Part of that standardising process involved its metadata as well. Statistics SA welcomed any feedback from its users on its publications, which it used to improve the publication.

Ms Fubbs stated that the work programme indicated that it was not financially feasible for Statistics SA to publish some of its publications in languages that were more accessible to the general public. The Committee was however of the view that it would be much more helpful, especially at local government levels, if the publication was in accessible languages, not necessarily all eleven official languages. Many people did not have access to computers, and thus hard copies of the publications must be provided, as well as electronic copies.

Minister Manuel replied that that would be very difficult. A sense was needed of who precisely would require the information, and the forms in which they would be able to access it. He stated that the Chairperson of the Statistics Council chaired a body in the advertising industry, which made much use of Statistics SA’s publications which indicated Statistics SA’s statistics were never used broadly, but only by a select few.

Mr Moloto sought an explanation of Statistics South Africa’s draft statistical master plan and its benefits.

The Statistician-General responded that it assisted in at least providing a broad strategic view of the work that Statistics SA needed to perform. The danger in a master plan was however that Statistics SA could get carried away with it and forget the critical things that needed to be done. The Statistics SA Act allowed for the co-ordination of statistics across government departments, and the co-ordinating function rested with Statistics South Africa. The master plan provided for the possibility of co-ordination and the prioritisation of key areas such as physical addresses and the population and business registers.

Mr Moloto sought clarity on the timeframes within which Statistics SA planned to have resolved the problem with a lack of uniform definitions.

The Statistician-General replied that Statistics SA had devised a concepts and definitions document to co-ordinate concepts within Statistics SA, which every Statistics SA questionnaire had to comply with.

Dr Gavin added that the problem was that Statistics SA relied on external sources for some of its data, which proved to a hindrance to deliverable timelines.

The Chair requested that that document be made available to the Committee.

The Statistician-General agreed that it would be made available to Members. He stated that it was also available on www.statssa.gov.za.

Mr Bhamjee stated that it was important that the fixing of a uniform definition of poverty be placed firmly on the Statistics SA agenda, and urged that the Committee be kept informed of all developments.

Minister Manuel responded that Mr Bhamjee was not asking the correct question by focusing solely on the definition of poverty. The reality, and the practical difficult being experienced by Statistics SA, was that measuring some parts of the informal sector was virtually impossible, and emphasised that it was a very difficult issue. He informed the Committee that on a recent visit to rural areas he found that non-South Africans owned most spaza shops, many of whom were in South Africa illegally. Part of the difficulty experienced by Statistics SA was that it was almost impossible to extract information because people feared government agencies, more so if they were foreigners and especially if they were here illegally. Minister Manuel made it clear that he was not being xenophobic, but was merely raising real problems. Government was experiencing enormous difficulty in cracking this problem, which was a global problem as well.

Dr Arrow added that even if the definition were fixed Statistics SA could still find that people re-classified themselves when they provided the information, because people might not want to admit that they were in fact that poor.

Mr Bhamjee maintained that the definition of poverty must be clarified and fixed.

Dr Van Dyk stated that the Statistics SA’s previous Annual Report indicated that it used temporary staff to do field work and gather information. He asked whether any mechanisms were put in place to monitor their collection and to ensure that those staff collected information properly and accurately.

Minister Manuel replied that Statistics SA would never be able to produce figures that were 100% accurate, because the nature of statistics was to proceed from a series of assumptions and approximations. The institution can therefore not be faulted for that. In fact there were no fixed international standards on the collection of statistics. There were thus endeavours underway on international statistical commissions to codify collection standards, to enable measures of comparison. It was a very difficult area.

The Statistician-General explained that each statistical publication released by Statistics SA first had to receive a clearance, in the form of a clearance document. It served as a notification that the data in the publications were accurate, and each publication was signed off by the Statistician-General himself. Statistics SA also had an error log, which enabled it to review the major errors it encountered, so as to improve its capacity and quality of statistics in future.

Dr Gavin added that different levels of supervision were built into the field collection process. She stated that one of her colleagues was this morning evaluating data collected by field workers in the CPI project. In the household survey Statistics SA had defined processes for quality checks and had entrenched an efficient system of supervisors. There were naturally times when mistakes did happen, but Statistics SA was committed to constantly working to improve its systems.

She stated that Statistics SA was looking at using technology to assist in monitoring the fieldwork process. In the community survey pilot project it had attached Netstar tracking devices to track the movements of field workers through the areas. It was also exploring the possibility of using hand-held devices, but recognised that there was no simple solution to the issue.

Mr Johnson expressed concern with the inconsistent use of HIV/AIDS terminology in the Statistics SA literature, as it referred to the two synonymously. He asked how Statistics SA distinguished between the two, because people would not always declare their status. He asked how those statistics were gathered.

Minister Manuel responded that the most comprehensive report on the matter was probably the causes of death survey. It involved the processing of all available death certificates, and recorded both the primary and secondary causes of death. It collected the data from every single death certificate on Department of Home Affairs’ system. Statistics SA then processed the data based on the World Health Organisation (WHO) algorithm, and published the outcomes. The problem was that the death certificates were not entirely clear and accurate, with the result that Statistics SA had to use the best information available on those certificates. Statistics SA could not be expected to function as a doctor and interpret the vague and inaccurate cause of death listed on the Department of Home Affairs certificate, and then fix a medically accurate specific cause of death. That was simply not its area of expertise.

The Statistician-General informed the Committee that Statistics SA’s second report on the causes of death would be released soon. He stated that another reason for Statistics South Africa’s difficulty in recording the causes of death was the lack of notifiable diseases.

Dr Gavin added that she wished to confirm that Statistics SA used administrative sources of data and did not interpret the certificates when it recorded the causes of death. Instead it used the WHO classification ICD 10. Statistics SA used the same methodology as the WHO. It would shortly be releasing the results for the remainder of 2003, because it had now received the remainder of the 2003 death certificates from the Department of Home Affairs.

Mr Bhamjee stated that Statistics SA used field workers for each survey, and asked why it could not employ dedicated teams. This would ensure that Statistics SA had properly trained and experienced workers who could go out and collect data immediately, whenever the need arose.

Minister Manuel responded that one of the difficulties was that the nature of the fieldwork that needed to be conducted varied with time and circumstances. Statistics SA would properly train all those field workers, but they would then only be needed again for the next census. It was thus far too expensive to retain them as permanent staff, "twiddling their thumbs in the meantime". The key challenge for Statistics SA was to continue to invest in people and ensure it had sufficient residual people so that it could draw on those people when they were needed.

Mr Bhamjee made it clear that he was not suggesting that Statistics SA have "a whole department of field workers". He stated that that Dr Gavin gave the impression that Statistics SA had an in-house team who interpreted and processed data.

Dr Gavin replied that the real bottleneck was at the level of production, once Statistics SA had collected the raw data that had to be compiled into a release. Statistics SA was critically short of resources in that area, which were the actual statisticians. She did however believe that Statistics SA was making inroads via its internship programme, but it remained a long-term problem. The shortage of that skill will remain a constraint for a period of years.

Mr Bhamjee sought clarity on the re-engineering process, as it suggested manipulation or tinkering with data.

Minister Manuel responded that it could also be referred to as re-tooling. It therefore did not involve the manipulation of data, but was instead the re-ordering of the system by which Statistics SA arrived at its figures.

Mr Bhamjee did not believe that existing data could be re-engineered, because the methodology would not allow it. He was not convinced that his questions had been adequately answered. Secondly, he sought an indication of the impact of scarce skills pre-1994 as opposed to the present day.

Mr B Mnguni (ANC) expressed concern with the time lines within which Statistics SA publications were released. He stated that surveys that were conducted in the current year would only be released in the following year which meant that the data was already outdated, to some extent. Secondly, he noted that the Statistics SA figures did not consider the crude oil price as a factor in determining the rate of inflation, whereas the South African Reserve Bank (SARB) considered it a major factor.

Minister Manuel reminded Mr Mnguni that it was the rate of change over a period that was measured, and not the individual prices on every day.

Ms Fubbs stated that the Work Programme did not indicate whether the statistics provided by Statistics SA also provided the architecture of information for policy development.

Minister Manuel replied that that benefit was self-explanatory. He stated that evidence-based decision-making "peppered" the work of Statistics South Africa.

The Statistician-General added that Statistics South Africa’s master plan had a role to play here. The problem however was that at times the architecture was undermined by the practicalities on the ground.

The Chair informed the Minister, before he had to leave the meeting, that the challenges facing Statistics SA were gradually being overcome. He assured the Minister that the Committee was going to ensure that Statistics SA did not drift back into its past problems and, with the co-operation of Statistics SA itself, the Committee would ensure that it was equal to the challenges it faced.

Minister Manuel thanked the Committee for its support and appreciated the sentiments expressed by the Chair. .

Concluding remarks by Chairperson
The Chair thanked Statistics SA for the input.

Dr Arrow assured the Committee that Statistics SA was doing everything possible to build the needed capacity. He stated that well-known statistician Dr Carlton, from the United States administration, was at Statistics SA for two weeks. He assisted with the household surveys and mentored 43 of Statistics SA’s staff, as well as eight staff from Southern African Development Community (SADC) countries. Statistics SA was thus taking measures to strengthen its capacity. It was a slow process but Statistics SA was very focused and remained confident that over time it would be able to show the results of its consistent efforts to conduct its own surveys.

The Chair stated that the sentiments expressed by Dr Arrow were encouraging. He once again thanked the Statistics SA delegation for its briefing and looked forward to future valuable interaction.

The meeting was adjourned.