ATC071109:Report on Annual Report 2006/2007 of Statistics South Africa.

Finance Standing Committee

Report the Portfolio Committee on Finance on the Annual Report 2006/2007 of Statistics South Africa (Stats SA), dated 9 November 2007

1. Introduction

The Portfolio Committee on Finance met on 17 October 2007 to consider the Annual Report 2006/2007 of Statistics South Africa (StatsSA). 

The Annual Report was presented by the Statistician-General. After the presentation discussions followed which are recorded below. 

Recommendations based on the engagement of the Committee with Stats SA are provided in the final section. 

2. Fruitless and wasteful expenditure

The Committee requested that StatsSA provide a more detailed explanation of the fruitless and wasteful expenditure documented in the Annual Report, particularly referring to services procured and paid for but not utilised, in excess of R2,4 million (p 136). 

StatsSA indicated that this expenditure was primarily related to field operators appointed to conduct the Community Survey. The Statistician-General explained that the particular company appointed had opted to accommodate fieldworkers at the Sheraton Hotel in Cape Town, at an excessive cost to StatSA. An immediate decision was made to dismiss the company, but expenditure had already been made. However, StatsSAcontended that despite the financial loss, a clear and important signal had been sent about misuse of funds, and maintaining reasonable standards affordable to the department. 

3. Debts written off

The Committee queried debts written off in the amount of R704,000 (p 135), as well as the source and circumstances around the non-recovery of funds.

StatsSA responded that the funds written off primarily refer to debts from the 1996 and 2001 censuses. StatsSA applies the law of prescription, meaning that debts will be written off if they are more than three years old and have little positive feasibility in terms of recovering funds. Most of these are small debts, in which the cost of recovery would be higher than the debt itself, and therefore would not constitute efficient use of public funds.  

4. Re-weighting and updating of the CPI and PPI

The Committee requested further information on the reported processes of re-weighting and updating of the CPI and PPI.

StatsSA responded by emphasising the organisational priority placed on Economic Statistics. The CPI is re-weighted approximately every five years or so based on the results of the Income and Expenditure Survey (IES), although this year the re-weighting is taking place slightly outside of the five year period. Currently, there are just over 1,020 products in the CPI basket, which is relatively high in comparison with other countries. Internationally, in both developed and developing countries, CPI baskets generally include between 200 and 800 products. 

The current CPI basket is based on products that represent 0.01% or more of total national household expenditure. This relatively low threshold means that too many items are reflected in the basket. StatsSA also indicated that the CPI basket has been criticised as a “rich man’s basket” that is overly reflective of expensive luxury goods, rather than products used regularly by “normal day-to-day people”. StatsSA reported that a new system of identifying products for the basket would address this issue. A provincial approach has also been adopted, whereby a basket is developed for each province based on both household expenditure, and the number of households that have bought a particular product. This will to some extent exclude items that are expensive, but are bought by very few households, as a minimum number of households will have to have purchased the product in order for it to appear in the basket.

A document detailing this approach was posted on the StatsSA website last month, and has received good media coverage. It outlines a total of 400 products which should appear in the new CPI basket, 300 of which were found to be common across all provinces. StatsSA now needs to begin collection of the items not currently in the basket, so that when the CPI is published in January of 2009 it will be based on the new basket of goods and will have new weights from the IES, which will be published in November of this year. 

StatsSA contends that this approach will mean a more “streamlined” CPI, which includes items bought by more people. Although the CPI will still very much be determined by expenditure, and therefore to some extent dominated by the spending habits of wealthier households, StatsSAsuggests that it will nonetheless still be more reflective of the products average households buy, and the impacts of these on inflation. 

5. StatsSA website 

The Committee requested more information on use of the StatsSA website, particularly related to the visitors accessing the site, and any feedback obtained. The Annual Report indicated an increase of up to 15% in visitors to the website. 

StatsSA responded that indeed the issue of how many people visited the website was less relevant than others, including: the specific hits received; how often links are followed from search engines; whether experiences in use of the website are positive or negative; whether queries can be dealt with automatically; whether the website is accessible; and who is accessing the website, and whether they are statisticians or other interested parties. 

Although the website has limitations in terms of providing information to people who are not connected to the internet, it remains and efficient and cost-effective means of providing a large amount of information to users.

At one level, StatsSA receives hits from “tourists”, who look at the front page, sometimes for access to the indicators that appear in the scrolling banner (e.g. CPI, PPI, GDP, etc). StatsSA keeps data on a monthly basis about what pages are accessed by visitors; incidentally, the vacancies page is one of the most frequently visited.

At a second level, StatsSA looks at “harvesters” of information from the website, particularly in terms of publications that are downloaded. For example, on the day of the CPI release, numbers of hits to the website are very high. The most popular publication downloaded between July 2006 and July of 2007 was the Labour Force Survey.

StatsSA also looks at “data miners”, as the website offers the ability to tabulate data online. Specific programmes are in places to facilitate access to time series data, including the household surveys and censuses. StatsSA also requests that users subscribe to data publications although these are free, in order to obtain information on who is interested in the data and how they use it. About one-third of subscribers are located in the private sector. 

StatsSA also runs a user satisfaction survey, and contends that over time satisfaction among external users is increasing, but this data is not released publicly.

6. Security of statistical information

The Committee enquired about the security of statistical information, given that this is very sensitive.

StatsSA responded that there are systems in place to protect the security of data, but that in relation to the website in particular, these are admittedly weak. The website has not been hacked successfully for several years, despite frequent automated attempts. But there is room for more vigilance. 

Also, StatsSA suggested that there are inevitably always some security issues with regards to personnel. StatsSA depends on staff within the organisation to maintain data security, and staff are required to sign an oath of confidentiality. StatsSA is looking at standardising confidentiality clauses for anyone working with statistical information.  

However, staff shortages and overlapping functions also threaten the security of data to a certain extent. For example, some staff are able to access statistical information systems, and also access logs that track entries in and out, and it would be good practice to have these separated. However, many staff members have had to assume multiple responsibilities due to HR shortages. 

StatsSA is, however, attempting to build internal IT capacity through training internal systems developers. Some StatsSA interns are working in the Data Management and Information Delivery programme, and are being mentored by a contracted external company. StatsSA is monitoring progress in the development of these skills, and the programme may serve as a model for further growing IT skills within the organisation. 

7. Financial performance and under-spending 

The Committee raised several questions related to financial performance and under-spending in the past financial year. Members questioned whether under-spending in some programmes was a result of poor planning, or an inability or lack of capacity to spend the allocated budget.

StatsSA responded that overall, 94.3% of 2005/06 budget was spent, which represented a significant increase from the previous year, in which only 86% of the budget was spent. One of the major sources of under-spending was in relation to the Community Survey conducted with more than 200,000 participants across the country. Some payments towards the Community Survey were not completed by the end of the financial year, resulting in an under-spending outcome.

However, StatsSA also indicated that total departmental expenditure on a “modified cash basis”, was 94%. There was a departmental savings of R65 million, of which R56 million has been requested from National Treasury for roll-overs, particularly for payments in the Community Survey project (p 102). On an accrual accounting basis, StatsSA would have spent 99% of the final appropriation allocated. 

StatsSA also contended that the organisation had fared well in terms of under-spending in comparison to other departments in the cluster. The Chief Financial Officer suggested that StatsSA should take a lead in terms of accrual accounting. She indicated that rollovers should also be taken into account, because those funds are now available to StatsSA in the current financial year. 

In another issue related to financial performance, Committee Members asked about the reasons behind high expenditure on consultants and contractors. 

Again, StatsSA responded that use of consultants in the department is relatively low when compared with other government departments. However, StatsSA also indicated that the reported number of consultants and contractors used also reflected the large number of temporary fieldworkers appointed to work on surveys, who were paid out of this line item. StatsSA cannot afford to employ fieldworkers on a permanent basis when they are not working on a particular survey. However, the organisation is developing an integrated fieldwork strategy that would result in different surveys running continuously, and would therefore allow for more permanent fieldwork staff. 

In terms of expenditure on fieldworkers over the past year, about 6,000 persons were employed temporarily for the Community Survey. Over the same period, about 2000 cars were hired. Rollovers related to the Community Survey explained poor spending performance in Programme 3. 

Also, again for the Community Survey, regional offices were opened up to accommodate logistical operations and provide a base for the 6,000 fieldworkers employed. There were also significant costs related to travel and subsistence. One of the areas of most significant expenditure was in hired cars, because the core business of StatsSA requires moving to different parts of the country. There were also high petrol costs, and issues around cash flow for petrol have been overcome by introducing petrol cards. StatsSA also hired a lot of venues for training fieldworkers in the various areas and regions, and on airfare to transport trainers to those venues. This explains some expenditure trends. 

8. Virements

The Committee questioned StatsSA on the extent of virements appearing in the Annual Report, and the effect that the shifting of funds may have had on particular programmes. Members asked whether these virements ultimately helped StatsSA to achieve programmatic objectives, or whether any programmes and outcomes were sacrificed. The explanations offered in the Annual Report were thought to be inadequate, given that Administration gained in excess of R5 million, and the Statistical Support and Informatics Programme gained in excess of R10 million. At the same time, Economic Statistics lost more than R7 million, Population and Social Statistics lost in excess of R5 million, and Quality and Integration lost R4 million. 

StatsSA responded that the virements conducted were within the limits of Section 43 of the PFMA. Some funds were taken from the compensation line, in congruence with the high numbers of vacancies recorded within the organisation. StatsSA indicated that savings in various programmes were used to correct budgeting imbalances, and highlighted the issue of not always receiving the full amount of funds requested from Treasury. 



StatsSA also indicated that all programmes are assessed in terms of achievements towards targets set in the work programme. The workprogramme is linked to MTF requests, which are also linked to the allocation of financial resources. Any virements or funding shifts are only made based on savings realised in the previous six months, and after careful consideration of the impacts and effects of such shifts. No virements take place where projects or programmes have not achieved their goals. StatsSA contended that no programmes have suffered as a result of virements. 

9. Human resources and capacity

Issues around human resource constraints were raised several times during the course of the Hearing. Members were concerned about high levels of reported vacancies, skills shortages, and the absence of team leaders, as well as staff withdrawing from training courses, or failing to competethem altogether. Members suggested that these issues could have serious consequences in terms of the overall performance of the organisation. 

StatsSA indicated that in many places in the Annual Report, the term “resource constraints” was used to explain a lack of adequate staff, rather than insufficient funding. 

StatsSA also indicated that the organisation faces a number of challenges related to Human Resources, including a shortage of specialised skills in South Africa and the region. Areas of acute shortage are in IT skills, statistical computation abilities, and statistical analysis. This is to some extent related to the ongoing problem of poor performance in maths among learners for historical reasons.

Vacancy rates within StatsSA remain very high. The organisation reported difficulties in attracting highly skilled workers into temporary contract posts. Therefore, StatsSA is looking to create more permanent positions. HR management, together with the Human Capacity Developmentprogramme, is working to identify internal capacity development programmes that will produce an internal pool of staff for recruitment to specialisedpositions. 

StatsSA has also acknowledged that the current advertising strategy, where vacancies appear in papers such as the Sunday Times, has not resulted in attracting the skills required. Despite high numbers of applicants, only between 10-20% are generally sufficiently qualified to be short-listed. StatsSA is investigating other forms of media to use for advertising, and also making use of a head-hunting policy. 

Resignations remain a problem, and have highlighted the problem of a lack of career progression options within the organisation. Many competingorganisations are able to offer better career progression options and benefits, and StatsSA staff tend to be skilled and highly mobile. Human Resources Management will be tabling a Retention Strategy, because ultimately it is not effective to direct efforts within the organisation towards skills development, and then lose newly-trained staff. 

StatsSA also discussed the need for more effective managers within the organisation. A new programme on leadership for managers has been introduced. Resignations are often related to the immediate working environment, including both physical working conditions and relationships with managers. The programme will focus on improving management skills and leadership. 

Human capacity development programmes have also been put in place to build awareness among learners about careers in statistics, which should be beneficial in the long-term. Bursaries will also be made available from StatsSA both inside and outside of South Africa

Members also queried high expenditure on consulting, contractors and special services, given high vacancy rates within StatsSA. Members asked how long this use of consultants, contractors and special services was likely to continue.

In response, as discussed above, StatsSA responded that high levels of expenditure on consulting, contractors and special services, was generally related to the use of temporary fieldworkers in the surveys conducted. StatsSA currently cannot afford to employ fieldworkers on a permanent basis, while they are not working on specific surveys.

Members also expressed concern about the lack of a learnership programme, on order to promote better access to the skills required by StatsSA.

StatsSA reported that an internship programme is in place, and has been supported by the East African Statistical Organisation. The programmecurrently has twice the number of interns than last year, and the number is expected to double again by next year. However, many more skilled employees are needed still, and StatsSA contends that it needs to open a Training Centre, which will be pan-African but will largely train people within the country.

A learnership programme will also be initiated in the next financial year, and a plan and draft strategy are already in place. Given the organisation’sskills shortage problem, when the Department of Public Service and Administration requested that Departments pilot internal skills audits, StatsSAwas one of the organisations that volunteered. Phase one of the pilot has been completed, and two phases remain. StatsSA has also embarked on a skills audit programmatic exercise as well, which focuses on management and the recruitment of suitable staff.

Also, StatsSA will begin offering training to teachers, related to learning outcome 4 in the FET grouping, for grades 9-12. It is hoped that this will eventually be implemented across 27,000 schools, whereby at least one teacher in each will receive training at the FET level. This will also support awareness and entrants into the bursary programme, and other initiatives to enthuse learners about careers in statistics. StatsSA then plans to expand training to include the GET level. A national coordinator for this initiative has been appointed, as well as a coordinator in each of the nine provinces. 

In response to a query around StatsSA’s relationship with tertiary institutions, these relationships exist within at least three programmes. Oneprogramme is specifically looking at course content related to social statistics. Progress has also been made with one specific university, while others are still being developed. This is particularly important in relation to the statistics training institute that is being developed. 

Members also queried why certain HR processes, including HR audits, were not conducted as planned. 

In response, StatsSA explained that a specific Team Leader who was not available, and therefore as a consequence, an internal audit in human resources functions conducted by the Internal Audit Unit could not be completed. This again highlights issues of staff shortages and capacity constraints. The Audit has been included in next year’s financial plans.  

10. Employment of staff with disabilities

Members also queried the low percentage of persons with disabilities employed by StatsSA.

StatsSA responded that recruitment for data processors in particular has been skewed in an attempt to attract more applicants with disabilities. Percentages of persons with disabilities who are employed within StatsSA may also increase as the organisation moves towards the creation of more permanent positions. A specific unit has also been established to address issues of disability within the organisation, including recruitment strategies. A DDG within the department has agreed to champion of this programme. StatsSA indicated that the percentage of staff with disabilities is unlikely to increase by year-end, or possibly even next year, but will rise in the medium- and long-term. 

11. Management and planning

Committee Members raised some general issues related to overall planning within StatsSA, as reflected by under-spending, virements and lack of delivery in some survey products.

StatsSA responded that in some cases, it is difficult to set priorities in the face of competing demands, but that a Programme and Projects office dealing with integrated planning had been established in an attempt to better address this issue. In some cases, StatsSA acknowledged that the organisation had “bit off more than it could chew” in terms of delivery.

12. Poverty line 

Committee Members queried whether progress had been made in the establishment of a national poverty line, as announced in the Budget in February.

StatsSA responded that a report had been launched with National Treasury in February detailing the methodology for the construction of a proposed poverty line. Since then, talks have been underway with Treasury and consultation has been invited from the public. A progress report will be provided in February of 2008.

The development of a Poverty Survey is also underway, and StatsSA anticipates that the survey will be have been piloted by end-2007. The survey includes a number of non-money metric measures of “subjective poverty” alongside money metric measures, which together will provide a multi-dimensional picture of poverty in the country. StatsSA anticipates that the results of the pilot will be available in 2008.

13. Delivery of survey products

Committee Members queried a number of survey products that had not been completed as planned, including data on Health and Vital Statistics, and Tourism and Migration annual reports. Resource constraints were cited as reasons for non-completion and failure to reach targets. 

StatsSA once again indicated that resource constraints generally referred to skills and staff shortages, rather than insufficient funding.

In terms of the Health and Vital Statistics, and Tourism and Migration reports, these were delayed in part due to the introduction of newprogrammes for editing data. Further, in both of these specific areas, StatsSA depends on other departments for data. In the case of Tourism and Migration, StatsSA is dependent on the Department of Home Affairs, and has worked with the department of issues of data quality. 

Last year, StatsSA reported on similar problems in releasing data on Health and Vital Statistics, and there is a need to improve HR specifically in the division.  StatsSA is now working more closely with the Department of Health in order to ensure that in the next year, a demographic and health survey can be conducted, and that the quality of thematic reports can be improved. 

In terms of the work programme, StatsSA has committed to undertaking a transport survey, and needs to ensure that that commitment is followed through. StatsSA has developed a comprehensive strategy for economic statistics over the next three to five year, based on the international framework system of national accounts. All economic statistics feed into the system of national accounts so that the GDP can be calculated. The Economic Statistics Strategy identifies ten sectors to be measured in GDP calculations, as well as quality and data gaps that currently exist. This is how the need for a transport survey and business survey were identified. One of StatsSA’s priorities is to improve measures of economic growth, so these surveys will remain in place.

StatsSA indicated that each year, the organisation undergoes a rigorous business planning process, whereby targets are re-examined, and feature in the next work programme. StatsSA contends that close follow-ups are conducted on targets committed to in the previous year. 

In terms of future survey products, postal surveys will continue as they are relatively successful. StatsSA is planning an agricultural census in the next few years, and one is underway already. This will feature a combination of postal surveys, phone follow-ups and evidence from the audited books of farms. There will be a poverty survey conducted in 2008, a tourism survey in 2008. The Statistics Training Institute is planned for 2009, and the Census will be conducted in 2011. 

14. Business register

Members asked a number of questions related to the Business Register, including how accurate and up to date the register would be, and whether it would give a good indication of the different kinds of businesses operating in South Africa. 

StatsSA responded that to some extent, the accuracy of the register is based on assumptions. Businesses that are registered for taxation through SARS appear in the register, and those which are unregistered do not. Classification is based on standards used by DTI and the Small Business Act, and these were revised in 2003. The classifications are quite detailed in terms of industry and turnover, but there are large numbers of businesses which do not provide financial information, and thus cannot be classified in terms of the Act. A summary of this information is available on the StatsSA website. 

The Committee queried why StatsSA does not simply use data from SARS for the Business register. One aspect of this is that VAT and revenue collection grows at a different rate to that of the GDP. In general, these should be going in the same direction, but they may not be of the same magnitude due to the specific ways in which calculations are done. The business frame that StatsSA has deduced from SARS data indicates the existence of businesses producing goods and services, as well as the size of these businesses, numbers of employees, and turnover. However, this does not allow for businesses to be located in specific sectors, which is a basic classification in terms of the system of national accounts. The data required beyond what is available from SARS requires specialised economic research. 

In terms of the question of completeness, in a closed economy in which national accounts are compiled, it is ideal to be able to measure the totality of all entities involved in the production of all goods and services for the marketplace. There are equivalent questions about people, and where they live. The best way to measure this within the population is to assign unique identity numbers, and most people know the advantages and disadvantages of an ID number. It is negative if the number is associated with a high number of traffic offences. The population register and system of assigning ID numbers has certain limitations, but is useful overall. 

Similarly, it is useful to assign unique numbers to businesses, and in the same way, a business ID number associated with tax evasion is a business you would not want to identify with. The project concept behind the Business Register is that we need something as unique as the ID’s used in the population register. If citizens or businesses perform well in terms of the law, they should be rewarded with clean certificates and permitted to do business. If they err, they should be held accountable and punished.

The Business Register also offers possibilities of tracking the nature of ownership and empowerment, and changes in the distribution of wealth. By associating a person’s ID number and the number of the businesses in which they are registered as a director, it is possible to map out how much a particular group is associated with ownership and directorships, if that is an indicator of empowerment. It also allows for a closer look into ownership patterns by sectors of the economy. 

Members also queried the health of the Companies and Intellectual Property Registration Office (CIPRO), housed within DTI. Without commenting on the health of a sister department, it is important to note that the Board is mandated by law to register and manage certain entities and companies, and to ensure that the register is updated so that entities that are created are timeously registered, and liquidations are timeouslyregistered as well. Timeous registrations are particularly important so that there is an equation related to the net stock of people and the net stock of businesses that are registered. 

In terms of the health of the business database, rather than CIPRO itself, it can be difficult to ascertain whether or not companies are actually active. The register only therefore provides information on aspirations, rather than economic activity. This is limiting, and it would be valuable to move towards a register that separates active and dormant businesses, or even more preferably, one which only includes active businesses, rather than indicating economic intention. 

StatsSA indicated that a national registration system would be advantageous. Economic trusts are not registered under CIRPO law or through SARS. StatsSA suggested that a streamlined, national registration process would be useful, in that businesses would be uniquely registered and would also be registered for tax. Businesses would therefore not be allowed to transact in the economy without SARS registration. Businesses ceasing to transact would also be taken out of the accounting system. Some progress in this direction has been made already – Treasury, DTI andStatsSA have agreed to form a project team, and each department has nominated 2 senior official s to represent them on a Project Steering Committee

15. Internal and external auditing processes

Members specifically queried references to a BAUD asset register that was reported to be unavailable at the time of the audit (p 64).

StatsSA responded that the BAUD is a basic asset register system, but that it had not been available due to timing. The Internal Audit conducted coincided with the External Audit, and the external auditors were likely to keep the BAUD register for a longer time, due to lack of familiarity with systems and controls. 

In response to issues raised about attendance to the Audit Committee, the Chair of the Audit Committee was absent due to ill health, and this was also explained to the external auditor. However, the Chair provided feedback and support throughout the audit process, although was not compliant with attendance. The Chair has now returned and is in good health, but his absence also highlighted the need to address strategic and risk issues that could potentially threaten the health of the organisation.  

The Auditor-General identified some management and control weaknesses within StatsSA, but these were not of sufficient concern to lead to a matter of emphasis or an audit qualification. StatsSA contends that most of the issues have been addressed on an ongoing basis, as soon as feedback is received.  

16. Dwelling frame

Members queried some aspects of the Dwelling Frame used by StatsSA in household surveys. 

In some survey exercises, StatsSA has assisted in providing addresses in communities, and this has been an important restorative process. 

StatsSA explained the processes behind the referencing of residences, addresses and built structures. The first step is to obtain imagery of survey areas, either from satellite pictures or from the Department of Land Affairs. Built structures, including dwellings, are then identified within these particular areas. In areas where no other information is available, satellite imagery is used. Where information is held by other partners and stakeholders - such as the Post Office, which often assigns numbers to houses in rural areas – StatsSA also uses this information. StatsSA also attempts to work with municipalities, particularly in validating information obtained through imagery. Some municipalities have strong databases, particularly for service delivery, but there are often pockets of informal settlements that do not feature in these databases. In these cases, StatsSAworks through databases and points out gaps in information, working in partnership with municipalities to ensure that databases are as complete as possible. 

In some areas, information held by municipalities is very poor or outdated, and this is reflected in an assessment conducted by StatsSA and the Department of Provincial and Local Government (DPLG) of Geographic Information Systems (GIS) capacity. StatsSA is looking to partner with local government in order to produce coverage for the whole country in terms of dwelling structures, but often information and infrastructure in municipalities is poor, and generally capacity to manage information exists only in better-resourced areas where more people pay rates. 

In areas where no information exists beyond satellite imagery, StatsSA goes into the field and attempts to identify and better understand information gleaned from the photo or satellite image. In the past, StatsSA has outsourced this fieldwork to the private sector, although information comes back to StatsSA for quality assurance and vetting before any payments are made. StatsSA is also moving towards building capacity through an internal fieldwork team that would be able to conduct information capturing and verification on dwelling units, but this will take time due to recruitment and training processes. Some small teams are already in place, but these will need to be upscaled dramatically in order to prepare for the Census in 2011.

17. Building public trust in StatSA

The Statistician-General discussed the need to change the public face of StatsSA, and to build higher levels of confidence in the institution. This process is being progressively achieved, and has been supported through closer links with the South African Statistics Council.  

Amongst other interventions, StatsSA has operated a programme called Census in Schools, and will repeat this process. StatsSA indicated that statistics are increasingly being used within the country, and encourages debate and engagement with the numbers produced. 

18. Recommendations

18.1. The Committee recommends that StatsSA provide more detail in future on the reasons for underspending when they occur.

18.2. The Committee recommends that StatsSA take a more innovative and active approach to recruit the staff it needs in order to address the challenges of human resource constraints, which in some cases appear to have hampered its ability to attain its objectives.

18.3. The Committee recommends that StatsSA make more of an effort to employ disabled people in functions across the organisation in compliance with the Employment Equity Act.