Statistics SA Annual Report: briefing

Meeting report

FINANCE PORTFOLIO COMMITTEE
11 October 2005
STATISTICS SA ANNUAL REPORT: BRIEFING

Statistics SA Powerpoint Presentation
Statistics SA Annual Report [2004/5]
Statistics South Africa Strategic Plan 2005/6 – 2009/10
Draft Committee Programme

Statistics SA delegation: Mr P Lehohla, Statistician-General; Dr R Hirschowitz, Deputy Director-General: Quality and Integration; Dr L Gavin, Deputy Director-General: Statistical Support and Informatics; Mr R Maluleke, Executive Manager: Statistician-General Office Support; Mr T Oosterwyk, Manager: Communication; Mr Z Nkosiyane, Acting CFO and Ms C De Klerk, Manager: Strategic Planning.

SUMMARY
Statistics SA reported on their response to the Portfolio Committee recommendations concerning its 2003-4 Annual Report. This was followed by an overview of Statistics SA’s operations and performance this financial year. Statistics SA had decided not to conduct a Census in 2006 because it had failed to reach agreement with the Statistics Council on the methodology that would be used. A pilot Community Survey would be hosted in February 2009, with the full-fledged survey being conducted in February 2011.

The Committee raised concerns with Statistics SA’s failure to produce many of its statistical publications, the ongoing problem with the public perception of the quality of statistics produced by Statistics SA and the impression that there was "a real world and a statistical world". Clarity was sought on the rationale behind dropping the broad definition of unemployment in the Labour Force Survey (LFS), as the strict definition did not accommodate the large informal and self-employed sector, which needed to be measured. Statistics SA was instructed to provide the Committee with a full report on its suspended officials, as well as the state of its financial management system. A progress report was sought on the spatial infrastructure needed for both the household surveys as well as the business register. Also questioned was the R158 million that Parliament's Standing Committee on Public Accounts (SCOPA) had recommended be approved as unauthorised expenditure.

MINUTES
Introduction by Chairperson
The Chair informed the Statistics SA delegation that the Committee had examined the Statistics SA 2004/5 Annual Report, and thus urged Statistics SA to limit its presentation to main issues in order to allow sufficient time for interaction on substantive matters. This would also prevent any rehashing of the content of the Annual Report.

Statistics SA briefing
The Statistician-General reported on their response to recommendations made by the Portfolio Committee based on Statistics SA 2003-4 Annual Report and provided an overview of Statistics SA’s operations. He provided a breakdown of Statistics SA’s performance with regard to its economic and social statistics, population statistics, quality integration, statistical support and informatics programmes. He outlined the financial performance of Statistics SA over the last financial year and highlighted its challenges and future plans. Statistics SA had received an unqualified audit report from the Auditor-General, and its corporate governance structures were improving. A new Statistics Council had been appointed in May 2005 and Statistics SA and the Council were actively promoting good relations. A Community Survey would be conducted in 2007. There had been an improvement in the quality of Statistics SA’s statistical information through improvement of its methodology and collection methods, and its human capacity was developed via collaborative training projects.

With regard to financial information, the Statistician-General reported that Statistics SA had been allocated a total of R500 million for the 2004/5 financial year but it had only spent R349 million, which resulted in an under-expenditure of R151 million. He explained that approximately R90 million earmarked for the community survey had not been spent due to a delay in reaching agreement between Statistics SA and the Statistics Council on the survey.
.
The challenges that faced Statistics SA included its efforts to meet stakeholders’ needs, the aligning of its priorities to government priorities and international frameworks, the implementation of the National Statistics System (NSS), the implementation and monitoring of its strategies and plans, continued compliance and proper risk management systems.

In conclusion, he informed the Committee that this would be the last time that Dr Hirschowitz would be addressing the Committee as she would be retiring in January 2006. He thanked her for her contribution to Statistics SA and stated that it had been a pleasure to work with her. She would not be remembered for the gap she left when bidding farewell to Statistics SA, but would instead be remembered for the capacity she had built.

The Chair echoed the Statistician-General’s sentiments in thanking Dr Hirschowitz for the immense contribution she had made to Statistics SA, and wished her luck in future.

Discussion
The Chair stated that questions would be posed to Statistics SA in an organised fashion per programme. He commended Statistics SA on its very reader-friendly report, as well as the obvious attempt to address all elements included in its strategic themes in the Strategic Plan. This was good practice and must be continued.

Programme 1: Economic and social statistics
Mr K Moloto (ANC) stated that he was impressed with the improvements made by Statistics SA in a number of areas mentioned in the Strategic Plan. The Annual Report indicated that Statistics SA had planned to publish 12 statistical releases on products manufactured in the manufacturing sector, yet none had been published due to "a poor response rate and quality problems". The same was reported for the land freight transport sector and for restaurants, bars and canteens. He stated that even if the R90 million that was earmarked for the community survey were to be deducted from the overall under-expenditure of R151 million it would still have left Statistics SA with an amount totaling R61 million, which it could have used to finance these publications. He stated that he failed to understand why the publications were not produced when Statistics SA clearly had the necessary resources within the organisation. He asked Statistics SA to explain the measures it would take to address this problem.

Dr Hirschowitz responded that one of the main reasons why Statistics SA did not produce statistics on manufacturing products was a quality issue. Statistics SA’s list of products on which it was collecting information was incomplete and not very up to date, and one of the tasks for better quality improvement was to develop a new list of products and test them out in the last scale sample survey of manufacturing to improve the quality of the products as well as to update and modernise the products. Thus, the difficult decision had been taken to not produce the publications on manufacturing products because Statistics SA was concerned about the quality.

With regard to the land freight transport publications, Statistics SA had discontinued the publication because it was looking at better ways to measure the whole transport industry. It was felt that it would not be very useful to produce statistics on land freight only as the sole form of transport that Statistics SA actually measured, when it was not measuring the taxi industry or air freight for example. Statistics SA would thus be revising and looking at the transport industry much more comprehensively, as the statistics that it was currently collecting were not as accurate or as relevant. The decision was taken to start from scratch and to look at new ways of collecting this information. There were also capacity problems in terms of getting people appointed who actually understood the statistical issues to start a new series at that stage.

Statistics SA understood that the gathering of data on restaurants and bars was a big problem. The aim was to first gather the accommodation statistics correctly, which was in itself a major undertaking. In the past only the hotel industry had been measured, yet the bed and breakfast facilities, game parks etc. was a massive industry that had perhaps overtaken the hotel industry in many ways. Statistics SA believed that it could not conduct the accommodation and restaurant survey at the same time given its lack of capacity, and thus "chose to get the one right first". Statistics SA had published the accommodation survey and was currently in the process of sending out the questionnaires, having drawn samples of restaurants, bars and canteens.

She stated that Statistics SA was conducting a large scale sample survey on manufacturing to identify the products that it should be measuring. This also applied to the retail trade survey to some extent, because Statistics SA needed better product information from this sector, especially for the Consumer Price Index (CPI). Statistics SA was currently conducting the two large scale production sample surveys of wholesale and retail trade as well as manufacturing in the present financial year which would continue into the next financial year.

Mr I Davidson (DA) thanked Statistics SA for the good report and indicated that it was in many ways an improvement on the report submitted last year. He stated that the quality and accuracy of the information produced by Statistics SA had been called into question last year, and he asked Statistics SA to explain the steps it had taken to address this problem and to rectify the public perception that the statistics produced by it were not credible. This was very important because, for example, the Department of Health’s estimation of the current HIV/AIDS infection rate was actually two million more than Statistics SA’s figure. He was not sure which figure was to be believed.

Dr Hirschowitz replied that, through its standards and methodology section, Statistics SA had looked at ways of improving its sampling methods. Statistics SA also had three consultants of good international repute who had been helping Statistics SA to improve the quality of all its economic statistics, and to help with the planning needed to improve the quality. Statistics SA firmly believed that it had improved the quality of the information it produced.

She stated that Statistics SA had also introduced a systematic quality management system which emphasised that quality of information produced was every Statistics SA official’s responsibility, from the person who collected the data through the edit programme, and to the final signing off. Every single phase of the process was broken down into the basic steps and analysed for areas in which the quality could be improved. Furthermore, every release that the Statistician-General signed off on was accompanied by a document that explained exactly why the statistics were as they were, the date on which the harmonisation had been done and the "confrontation" done with other information sources. Dr Hirschowitz stated that she had personally been checking every single edit programme.

The Statistician-General added that one of the quality measures in all Statistics SA reports, but which had not been drawn to the Committee’s attention, was how variable the estimate would be. This was measured through the standard error, and for each indicator that was produced Statistics SA included an estimate which would be accurate within a given range. There were very few producers of statistics who provided that kind of quality check.

There was also a measure of quality in relation to the production of metadata, which was data about data. This point was however not captured in the Annual Report nor in today’s presentation. All in all, Statistics SA believed in being transparent with regard to admitting the limitations inherent in providing quality statistics, and it was for that reason that Statistics SA would decide not to publish data if it was not up to standard. Reasons would also be provided as to why the data could not be published.

Dr Gavin replied that, with regard to the HIV/AIDS statistics, Statistics SA proceeded from a basis of imperfect knowledge about the country. Statistics SA was not able to measure HIV/AIDS prevalence by testing everybody, and as such measurements would be made on the basis on certain modeled assumptions, Statistics SA would make a prediction or would develop an understanding of what it believed would be the pattern of HIV/AIDS prevalence and HIV/AIDS deaths.

The reasons for the discrepancy between the two figures lay in the assumptions that had been made, as well as in the sources of data used. These assumptions were spelt out very clearly in both the Department of Health and Statistics SA reports, which enabled users and readers to make an assessment as to whether the figures were reasonable as time went by. The Department of Health, for example, used the data gathered from the measurement of attendees of publicly funded antenatal clinics, and made the assumption that the HIV/AIDS prevalence rate obtained from those surveys were actually the same rate for all women in the country. This was not however an assumption that Statistics SA would make, because it regarded that as somewhat of a biased sample in terms of the HIV/AIDS measurements. Instead the HIV/AIDS prevalence rates that Statistics SA believed were consistent with the mortality data which the Statistician-General mentioned Statistics SA published in February 2005, and which would be less than the rates yielded by the assumptions used by the Department of Health, which in turn were based on methodology used by UN AIDS. Statistics SA was however working with the Department of Health and there was ongoing collaboration regarding statistics relating to health, and HIV/AIDS in particular.

Mr Davidson sought clarity on Statistics SA’s rationale behind dropping the broad definition of unemployment in the Labour Force Survey (LFS). The 13^th International Conference of Labour Statisticians suggested that the broad definition of unemployment "may be appropriate in situations where the conventional means of seeking work are of limited relevance, where the labour market is largely unorganised or of limited scope, where labour absorption is at the time inadequate or where the labour force is largely self-employed", and that situation was very applicable to South Africa.

Mr N Van Dyk (DA) contended that the unemployment rate produced by Statistics SA was very important for the South African Reserve Bank (SARB) in determining its monetary policy, for the Minister of Finance in compiling the National Budget, as well as for the Department of Labour and the labour market. He asked Statistics SA to indicate how it planned to assist these stakeholders, and what Statistics SA’ s approach would be to the unemployment statistics.

Dr Hirschowitz responded to these two questions by stating that South Africa was not dropping the expanded definition of unemployment as such. Instead it was defining an extra category of people that it wanted to describe in much more detail: the ‘discouraged work seekers’. These were South Africans who were not economically active, according to the official definition, but were unemployed by the expanded definition. The aim was to identify the characteristics of this group, what stopped them from going out and looking for work and how Statistics SA could be made to understand them in greater detail. Thus the aim was not to diminish the information published by Statistics SA but instead to expand it by identifying the specific group of people who needed better exploration. The International Monetary Fund (IMF) had reviewed Statistics SA’s labour statistics, and provided certain recommendations which Statistics SA had taken forward.

Dr Gavin replied that, as Dr Hirschowitz had stated, the IMF conducted a review of Statistics SA’s LFS earlier this year. The IMF report was studied thoroughly and on the basis of it, as well as other input, Statistics SA had actually put together a project that would, over a period of two years, look at the re-engineering of the Statistics SA LFS. The aim of the project would to be able to provide credible, reliable and much more timely information on the LFS. It would be a far-reaching project that would cover issues of methodology, how Statistics SA could produce short-term information more rapidly and on a quarterly basis as well as providing results and data that Statistics SA users actually required, both on a quarterly and annual basis.

She stated that Mr Davidson’s concern with the definition of unemployment to some extent revolved around a presentation issue. She believed the manner in which Statistics SA presented the information was causing great confusion, and people were comparing South Africa’s broad definition with other countries’ strict definition. As Dr Hirschowitz had stated the intention of Statistics SA was not to collect any less data, but it was aimed at presenting information in a way that would be more useful to users.

Mr Davidson stated that he was not sure whether he was encouraged by the response given by Statistics South Africa, because it created the perception that there was "a real world and a statistical world". The real world would come through from the Department of Health via its HIV/AIDS statistics, for example. The sample might not necessarily be accurate but was nevertheless the information provided by "the real world", as opposed to a statistical world which was plagued with perceptions of inaccuracy.

He welcomed Statistics SA’s intention to provide an additional category of people seeking employment, but argued that that still did not provide the much needed information on the difference between the broad and the strict definition of unemployment. The broad category of unemployment must be retained and captured as it would reflect the real jobless rate.

Mr Davidson contended that the measurement of GDP would be primarily biased towards growth in the formal sector. South Africa had a burgeoning informal sector, and at the previous meeting with Statistics SA he had asked it to explain the progress it had made in measuring the informal sector. The growth in the informal sector must also be measured as accurately as possible.

The Statistician-General responded that it was true that how data was measured determined the result.

Mr Van Dyk stated that the Annual Report reflected the Auditor-General’s concern that "leave forms were not always on file to substantiate leave taken, whilst in some instances, the actual leave taken was incorrectly captured". He asked Statistics SA to explain the steps it had taken to address the matter.

The Statistician-General responded that this was an important issue and Statistics SA was currently improving on it by evolving its leave management systems, including the introduction of a computerised leave form system. A concerted effort was being made to improve this area.

The Chair stated that the Committee would definitely follow up on these issues raised.

Programme 2: Population statistics & Programme 3: Quality and Integration
Ms J Fubbs (ANC) sought clarity on the high under-expenditure in this area, particularly as a number of the reports had not been published in time.

The Statistician-General replied that there was high under-expenditure but the one lump sum could be attributed to Statistics SA’s not conducing the census, as well as the disagreement with the Statistics Council. The other reason was the recruitment of personnel, and approximately half of the R50 million was due to the fact that Statistics SA was unable to secure adequate personnel.

Ms Fubbs noted that the Annual Report indicated that much training was provided around the gathering of population statistics, and that the raw data provided by the Departments of Health and Home Affairs was also used by Statistics SA in this regard. She asked whether Statistics SA was involved in training with those government departments, and whether they were all linked up to each other electronically in order to provide the information timeously. She stated that she was not sure which was actually in charge of this area within Statistics SA.

Mr T Vezi (IFP) stated that in his constituency, most of the death certificates of people who were married at the time of their death, stated that they were unmarried. He asked how this would impact on the accuracy of Statistics SA’s statistics.

The Statistician-General responded to these two questions by stating that Statistics SA generally used the cause of death, as well as the age and sex of the person who died. Other ancillary information was important, but not much could be done with that information by the time it reached Statistics South Africa. Ms Fubbs’ question thus became relevant because the issue was to what extent the information would be integrated with all the other data sets once the incident took place, so that Statistics SA would not have to ask whether the person was married at the time of death or not. Statistics SA had a session with the Director-General of the Department of Home Affairs on the SmartCard, in order to decide on the kinds of basic information that Statistics SA required that could be incorporated on the SmartCard. A solution was thus in sight, but at the moment Statistics SA would have to collate the different sources of information wherever they existed. He agreed that there was information on the death certificates that did not make much sense, including the cause of death, and that required much attention.

Dr Hirschowitz replied to Ms Fubbs’ question by stating that Statistics SA had been in contact with the Departments of Education, Health and Home Affairs. The aim was to look at their databases and how they collected their data, and for Statistics SA to propose ways in which those government departments could improve their methods so that they could become official statistics under the National Statistics System (NSS). This project was strongly supported by the World bank and it has sent out consultants who were currently examining those databases and assisting Statistics SA to identify criteria for evaluating the databases around those issues.

The Statistician-General informed Ms Fubbs that Dr Gavin was currently in charge of the population statistics, in an acting capacity. As the Committee might be aware Statistics SA had suspended its DDG for population statistics a while ago, and he was still under suspension.

Mr A Asiya (ANC) stated that it would be proper for Statistics SA to provide the Committee with a full report on the suspension of the that official, but it does not have to do so at this meeting.

The Statistician-General answered in the affirmative.

Ms Fubbs requested that Statistics SA also provide the Committee with a detailed report on its trend of under-expenditure over the last two Annual Reports, and how it planned to address this matter.

Mr B Mnguni (ANC) noted that the Annual Report indicated that Statistics SA had published a Draft Report of Census 2011 and of the Community Survey, and asked whether these draft reports were available for public comment.

Ms Fubbs stated that Statistics SA’s Strategic Plan placed much more focus on improving the metholodogies of the community and the population census than the Annual Report did. She asked whether it was correct to assume that Statistics SA officials were currently "out in the field" working on the pilot community survey.

Dr Gavin responded to these two questions by stating that the documents mentioned in the Annual Report were largely internal planning documents. In the case of Census 2011 Statistics SA did have a very long time horizon, although it had already sketched out very broadly the planning for the survey. A pilot would be conducted in 2009 in order to avoid a very high profile event in 2010. Many of the people involved with the previous Census would be involved with running the community survey, and this was a deliberate decision by Statistics SA aimed at capacitating people to be able to conduct the 2011 Census.

A second reason for the draft reports was partly due to the slippage of dates. The Strategic Plan indicated that Statistics SA planned to conduct the pilot Community Survey in October 2005, and the main survey involving approximately 200 000 households was to be conducted in October 2006. However, as mentioned in the presentation, before Statistics SA could progress with the Community Survey it first needed to gain the support of the Statistics Council. Statistics SA began discussions with the Statistics Council in October 2004 and really only reached an agreement on the broad methodological approach a few months ago. The result of the agreement reached Statistics SA was quite a significant change in Statistics SA’s original plans for the community survey, and it thus necessitated quite substantial replanning. Both Statistics SA and the Statistics Council were of the view that it would be better to delay the Community Survey, but to ensure that the thorough planning had been completed. The revised Statistics SA plans also allowed for a full year between the pilot survey in February 2006 and the main survey in February 2007, so that the lessons learned from the pilot survey would allow Statistics SA sufficient time to incorporate them into the actual Community Survey.

Dr Hirscowitz stated that she would like to provide information on the way in which Statistics SA published reports. Apart from the official statistical releases Statistics SA also had a forum for discussion papers on its website, aimed at encouraging discussions on Statistics SA activities which Statistics SA was not yet ready to declare as official statistics. A distinction was thus made between discussion documents and discussion papers.

Mr Moloto noted that Statistics SA had only produced on mid-year population estimates report and three draft analytical reports under its demographic analysis sub-programme, when its stated target was to have published four reports. He sought clarity on this.

Dr Hirschowitz replied that Statistics SA produced only one report because it had insufficient personnel and capacity. Furthermore a conscious decision was taken to redeploy people into the survey areas themselves in order to improve the quality of the survey areas, rather than to focus on report-writing. She stated that one of the most important lessons that she has learnt was that the amount of time needed to properly plan, test and pilot a new survey was under-estimated, as there was so much background and research that needed to be conducted to understand the issue. While Statistics SA did receive the funds it requested it was not however always able to spend the funds in that year because it had not yet considered properly how long it would take to conduct the research and understand the background and complexities of the issue. She believed that this had been Statistics SA’s biggest weakness in its budgeting and planning system.

Ms Fubbs sought clarity on the indicators used.

Dr Hirschowitz replied that Statistics SA had worked with the Presidency during the current financial year on refining and defining the indicators for the next ten year review, and how progress would be measured. The indicators to be used were identified through the NSS, in collaboration with the Presidency, as well as the information that still needed to be collected in order to measure the indicators. This was an ongoing exercise that would take quite a significant amount of time.

Programme 4: Statistical Support and Informatics
Mr M Johnson (ANC) noted that the Annual Report indicated that Statistics SA had done no work on the spatial infrastructure needed for both the household surveys as well as the business register, and sought clarity on the progress made in this areas.

Secondly, he noted that the Annual Report indicated that Statistics SA had planned to provide the sample for the collection of economic statistics in June 2005, whereas it had only produced the sample in August 2005. He sought clarity on the reasons for the delay, as well as its implications on the statistics that needed to be provided.

The Statistician-General responded to these question by stating that, with regard to the business register, Members would recall that Statistics SA had identified the business register as the one core area that was central to the improvement of statistics. The concern raised by Mr Johnson was apt. There were two levels at which Statistics SA was working on the business register. The first was its collaborations with SARS and the Department of Trade and Industry (DTI), in order to ensure that the raw data provided to Statistics SA was sufficiently authentic and could be worked with. That process had taken place and was the basis for improvement. The second aspect was that when the register was within Statistics SA it needed to be converted into a statistical register, and there was much work for sustaining that relationship between SARS and the DTI that required attention to ensure that the register did not degenerate. He stated that he, the SARS Commissioner and the Minister of the DTI were attending to this to ensure the register did not deteriorate.

He stated that Statistics SA was fairly satisfied with the progress it was making with regard to the classifying of the data received, but the delays were a cause for concern and were being addressed.

Dr Hirschowitz replied that there may be a delay in drawing samples for the business register. The plan was basically for Statistics SA to look at the classification of large businesses and complex businesses and to conduct a profiling exercise. Before Statistics SA could actually implement a new sample it needed to compare the old and new samples around the kinds of activities that have been performed by the different businesses and how those activities have changed in terms of classification by industry and by activity. The sample might also need to be redrafted after the profiling exercise because some businesses might have moved and were now out of scope, and needed to be reclassified into another sector. Thus the harmonisation and comparison process was a very intricate exercise, and Statistics SA was conducting this exercise much more rapidly. Statistics SA was working very closely with SARS and the DTI to gather the original data but, as the Statistician-General indicated, Statistics SA still needed to reclassify them into statistical units.

Dr Gavin responded that the Community Survey involved approximately 17 000 enumeration areas, whereas the master sample used for the general household survey was about 3 000 areas. Thus because the Community Survey involved a much larger coverage of the country, much focus was placed on updating information across the country. For instance during the previous financial year Statistics SA worked with three metro’s to share information and obtain their spatial databses, which were incorporated into Statistics SA’s. During the current financial year a further three metro’s were being worked with. Statistics SA was also working on updating imagery, and had conducted an audit on the imagery of the country it currently had available by working with the satellite application centre. Statistics SA was also working with the Chief Directorate of Surveys and Mapping within the Department of Land Affairs which also produced photo maps but, due to difficulties with their funding, there was a delay in the processing of that data.

She stated that Statistics SA was also involved in partnerships with other parastatals such as Eskom, who have indicated that they would be refreshing their information on the Eastern Cape and Statistics SA would be working closely with them so that both could benefit from the collection of that kind of information. This was thus the broad preparation towards the Community Survey.

Moreover a total of 760 enumeration areas would be covered by the pilot Community Survey in February 2006, and it was actually true that all Statistics SA collection officials would be in field as of next week busy with compiling the list of every structure and dwelling unit in each of the enumeration area. The national and provincial level training was taking place this week.

As far as the spatial infrastructure for the business register was concerned, Statistics SA was able to localise all the entities on the list to a municipal level. The difficulty was that, due to a lack of standardisation of place names, it was very difficult to refine the spatial referencing. The other problem was the Statistics SA’s geography division would take the data provided at face value, even though some of the addresses might be linked to enterprises or merely to branches of businesses. Additional information was thus needed to fully and properly interpret that data.

Mr Johnson sought clarity on the kind of practical support that Statistics SA stated it could provide to the small municipal towns in his constituency, as many of the small municipalities were struggling due to a lack of capacity.

Secondly, he stated that although Statistics SA was able to gauge the number of people that used its site and the documentation provided thereon, but asked whether it was also able to track what exactly those users did with that information.

Mr Risenge replied that Statistics SA had gone into the field with its user satisfaction survey, and many interesting issues were raised. Statistics SA had information officers in every province who were supported by the information clerks, and they worked with statisticians in the province and the NSS support staff. The staff members would go out into the provinces to the users in an outreach programme and would train them, wither the information was downloaded from the Statistics SA website or whether it was supplied directly to the users. Statistics SA also provided installations on the machines of users if they so requrested, and no installation would be performed unless it was followed by training. The provincial offices were always open to provide any support needed by the users. He stated that Statistics SA did therefore have an active outreach programme targeted at the local level and, although it might not yet have reached all local municipalities yet, efforts were being focused to reach as many as possible. A further reason why all municipalities have not yet been reached was that Statistics SA provided such information and training on a request basis, and it would thus spend its time with those municipalities who requested the data..

Dr Hirschowitz responded that there were two ways in which Statistics SA collected information on local areas. The first was via a population census and household survey, where the demand side was accessed. Secondly Statistics SA also gathered non-financial information from the municipalities themselves on the monies they have collected over the time period, how the money was spent and the kinds of services they provided with those funds. At the moment Statistics SA has trained nine economists, one in each province, to go round to the municipalities and collect the information for both the financial and non-financial cencus at municipalities. Unfortunately that information was not published at a municipal level, because its Act stipulated that it was not allowed to publish information collected from the supplier of that particular service. She stated that Statistics SA had sought a legal opinion on this matter from its legal division. Thus that information could not be published at a local level unless the permission of the municipality was first secured. One of the areas in which Stastics SA could be assisted would be to get municipalities to publish information at that level.

Programme 5: Administration and Human Resource Management
The Chair noted that the Committee was running out of time, and allowed Statistics SA to respond only to the substantive issues raised. Statistics SA would be allowed to respond to the remainder of the questions that would require a more detailed response in writing.

The Statistician-General replied that his organisation would prefer to come back to the Committee with written document on many of the issues, especially the suspension of the officials referred to.

Mr Davidson wished to once again voice his concern with the huge under-expenditure within Statistics SA as a whole. This was also the case during the previous financial year. It necessitated the question whether there were adequate financial planning within Statistics SA, as it was.one of the core problems within the organisation. Funds that were not used by Statistics SA were monies that could instead have been given to other government departments. The audit report of the Auditor-General also raised substantive concerns with the financial management system currently in place within the organisation, such as the uncertain status and functioning of its audit committee. He contended that that must be read together with the firing of three Statistics SA officials for raising concerns around corrupt practises within the organisation, and the accusations remained in the air. Furthermore a DDG was also fired for allegedly referring issues of mismanagement within the organisation to the Auditor-General. The Committee needed answers as to what was going on within the organisation as well as the state of its financial management framework.

The Chair reminded Members that, as raised earlier by Mr Asiya, the Committee would be getting a full report on the suspended officials.

The Statistician-General responded that a forensic audit had been conducted by the Public Service Commission, which concluded that those who alleged corrupt practices within Statistics SA were lying. He stated that he had acted on those individuals, referred to by Mr Davidson. The Statistician-General informed Members that he was not against whistle-blowing, and stated that he was in fact in favour of it. It was however clear that the officials in that instance were misguided and made those false allegations in pursuance of their own agendas. They had to go through a disciplinary procedure and none of them were fired, as the Statistician-General had pardoned them and suspended them for a period of three months after which they would be allowed to return to work. Two of the officials had resigned subsequently, and the person who perpetrated the allegations was still on suspension. Statistics SA thus followed the necessary procedures that were cherished by this country in law, and would continue in that manner.

He stated that he would provide the necessary documentation to the Committee on this matter, and informed Members that he had apprised both the Minister and Deputy Minister on these matters. Anyone who attempted to undermine the operations of the institution would be disciplined.

The problem raised by the under-expenditure, without exonerating Statistics South Africa, was that Statistics SA had organised its functions and activities around the Census which is had planned to conduct in 2006. However when it was advised otherwise by the Statistics Council, Statistics SA then had to change its plans. The change of course had a very disruptive effect on the financial expenditure and plans of Statistics South Africa, and it was that disruptive effect that Statistics SA was trying to lay to rest at the moment. The figures indicated that there was no under-expenditure on the non-Census activities of Statistics SA. He stated that the under-expenditure reported in 2004 was due to the problems faced by Statistics SA with the planning for Census 2006, and that under-expenditure had spilt over into 2005. It affected various issues within Statistics SA, such as the possibility of recruitment etc.

Statistics SA was addressing the problem by getting dedicated project support, in the form of an official at the level of a director, who would recruit people for projects undertaken by Statistics SA. This measure was paying off. He assured the Committee that were would be no future cases of under-expenditure..

Mr Moloto noted that the Auditor-General's report highlighted irregular expenditure of R16,1 million and R12,3 million for the 2003 and 2004 years respectively. However Parliament's Standing Committee on Public Accounts (SCOPA) had recommended that R158 million be approved as unauthorised expenditure by Statistics SA. he asked whether these two figures referred to the same matter.

Mr Asiya asked whether Statistics SA would have the R158 million in its coffers when Parliament approved the amount, or whether it would come from the national fiscus.

The Statistician-General replied that the two figures were separate. He stated that all these issues emanated from the Census which Statistics SA had decided not to run in 2006, and everything that has followed since has been as an effect of that decision. This was not an indication that the organisation had not improved, as it had in fact improved by far, but it was the last remnants of the decision not to hold the Census in 2006.

The amounts Statistics SA had requested be authorised have already been discussed with the Tender Board and was condoned.

Mr Nkosiyane responded that a total of R36 million was the result of overspending on the 2006 Census which was no longer taking place, and R122 million was a result of overspending in 2001. This overspending was done consciously by management, because it "could not hold the census ro ransom not to succeed". The decision was taken consciously by Statistics SA management to ensure that the country benefited from the results of the 1996 and 2001 Census.

He stated that Statistics SA had requested that Parliament approve the amount of R158 million so that National Treasury could return the funds to Statistics South Africa, because Statistics SA’s bank statements were currently overdrawn in respect of the 1996 and 2001 cencus.

In 2004 a total of R90 million was not spent in terms of the community survey, because of the delay in the agreement reached between Statistics SA and the Statistics Council. Amounts of R18 million and R20 million were due to capital expenditure and goods and services incurred during 2004, which were committed in 2004 but were never used because Statistics SA did not receive the goods timeously. A total of R22 million was attributed to vacancies which were never filled. He reminded the Committee that Statistics SA was still dealing with the legacy of the past as it did not have sufficient adequately skilled statisticians and economists, with the result that Statistics SA had to use foreign consultants. Statistics SA ensured that those consultants trained Statistics SA staff to be able to produce results that the institution could be proud of.

The Chair stated that the use of consultants was a problem, both from a capacity point of view and expenditure concerns, and the Committee will look into the matter.

Mr Mnguni noted that Statistics SA had planned to assess the statistical capacity of three government departments, whereas it exceeded that target and managed to assess eight departments. He sought clarity on the reasons for that.

Secondly, he noted that Statistics SA had planned to identify 25 development indicators, yet the Annual Report did not mention exactly how many it had actually identified.

Thirdly, he noted that Statistics SA had planned to certify three statistics as official, yet it was unable to certify even one. Its failure to do so was a critical issue as it eroded the credibility of the work done by Statistics SA in the public's eye. He sought clarity on the reasons for failing to meet the target.

Fourthly, he noted that the targets under the Office of the Auditor-General subprogramme had been met, and asked whether that had improved the core operations.

Fifthly, he noted that Statistics SA had planned to conduct an HIV/AIDS awareness and support programme in all of the nine provinces, yet it had succeeded in reaching only seven with a total of 119 participants. he sought clarity as to the number of participants Statistics SA had initially expected and why the remaining two provinces were not reached.

Finally, Mr Mnguni asked Statistics to explain why it had only conducted eight employee assistance programme workshops, when it had set a target of ten such workshops. He also sought clarity on the number of participants initially expected.

Mr Moloto asked whether the Statistics SA audit committee was currently fully functional, because concerns were raised by the Auditor-General's report.

Secondly, Mr Moloto sought assurance from Statistics SA that the concerns raised by the Auditor-General's report regarding internal controls within the organisation were being attended to.

Thirdly, Mr Moloto noted that the Auditor-General's audit report raised concerns with the expenditure incurred by Statistics SA in contravention of tender procedures, and sought clarity on the nature of the problem.

Mr Van Dyk sought clarity on the problems with the administration of the internal audit controls, and the steps taken to address the problem.

Mr Asiya was of the view that the Annual Report did not speak directly to the Strategic Plan, and proposed that the Annual Report be made more reader friendly in future by ensuring that thee two documents actually spoke to each other. That would enable Members to properly exercise their oversight function.

Secondly, Mr Asiya stated that it was a serious problem if Statistics SA's audit committee had not sat for an entire year, as noted by the Auditor-General's report. The matter must be raised with the Minister.

The Statistician-General replied that he preferred to first speak to the Auditor-General about the above questions posed by Mr Mnguni, Mr Moloto, Mr Asiya and Mr Van Dyk regarding the internal controls and financial management of Statistics SA. He would then provide the Committee with a full report.

Mr Asiya requested that the Committee also be provided with the management letter submitted by Statistics SA to the Office of the Auditor-General on these matters. He stated that further that he understood that letter to contain all these substantive issues, and was thus not sure why the Statistician-General requested a further opportunity to discuss the matters again with the Auditor-General.

Mr Davidson agreed with Mr Asiya. He stated that he understood the Statistician-General’s request, but contended that these issues should indeed have been clarified between Statistics SA and the Auditor-General prior to the printing of the Annual Report. He failed to understand why another meeting had to take place between the two institutions before the Committee could get clarity on matters which the Auditor-General had already reported on. He believed that the Committee had a right to inquire as to the financial management within Statistics SA..

The Statistician-General responded that he had no doubt that the Committee was entitled to such information, but pleaded that he be allowed to discuss the matters with the Auditor-General. He could forward to the Committee the letter he had written to the Auditor-General on the issues raised.

The Chair noted that the Committee would allow the Statistician-General’s request, but that all the documentation and information Members had requested must be provided by Statistics SA at a later date.

Mr Van Dyk sought clarity on the resubmission of the Statistics SA's financial statements to the Auditor-General.

Mr Nkosiyane responded that the Office of the Auditor-General commenced its audit in April 2005, and the audit was supposed to run the conventional time frame of three months. However at the end of that period in June 2005, while Statistics SA was waiting for the management letter from the Office of the Auditor-General, Statistics South Africa discovered that that institution had not yet provided Statistics SA with the findings of the audit they had conducted. Statistics SA then wrote a letter to the Office of the Auditor-General stating that the audit had not been conducted in accordance with the audit plans. They replied that their audit manager had been ill for more than six weeks, and for that reason that had to commence the audit using an external private company. It was for that reason that Statistics SA then had to resubmit its financial statements to the new auditing firm..

Ms Fubbs stated that training at municipal level was very important, as lack of capacity was one of the causes for the shortcomings experienced in that sphere of government.

Secondly, she noted that Statistics SA was perennially indicating that it would prioritise the improvement the quality of the statistics if produced, yet each year there was no clear indication that it was making headway in that regard. She sought clarity on the problem, and questioned whether it was due to a gap in human resources or a lack of capacity within the institution.

The Statistician-General replied to these question by stating that Statistics SA would be meeting with the universities on 3 November 2005 in Grahamstown at the South African Statistical Association, and one of the issues on the agenda would be the measures that needed to be put in place to ensure proper training and a satisfactory uptake of statisticians.

Dr Hirschowitz responded that Statistics SA had forwarded to the Minister its plan on establishing a statistical training institute. Training programmes were also being set up with tertiary institutions as well as with other SADC countries.

Concluding remarks
The Chair stated that the questions that were not answered would have to be replied to by Statistics SA in writing.

The meeting was adjourned.