Public Expenditure Review: National Treasury briefing

Standing Committee on Appropriations

22 November 2016
Chairperson: Ms Y Phosa (ANC)
Share this page:

Meeting Summary

The Government Technical Assistance Unit (GTAC) in the National Treasury (NT) presented to the Standing Committee. Performance and expenditure reviews (PERs) had been started because the Presidency and the NT needed detailed and reliable information. It was a joint project between the NT and the Department of Planning, Monitoring and Evaluation (DPME). Findings were presented under the headings of urban space, higher education and training, economic development and administrative services.

The GTAC reported there were 50 colleges and 26 universities, with 1.7 million students. There had been rapid enrolment growth. Spending would have to rise from 2.3% of the gross domestic product (GDP) to 4.4% to meet the goals of the National Development Plan (NDP). The average value of the National Student Financial Aid Sceme (NSFAS) awards was not enough to meet full university costs. The training of artisans was costly, at R450 000 per artisan. The throughput rate was 56%. Curricula would have to improve to meet the demands of industry.

In a sample of 1 000 leased office accommodations for national government, it had been found that 60% of the leases were 45% above market rates. It was concluded that policies sometimes were designed and implemented with unrealistic expectations of the availability of funding.

In discussion, higher education funding and artisan training were most intensively interrogated. Members found it unacceptable that there were 13 routes to artisan training, and that the cost of training one artisan was estimated at R450 000. There were other critical remarks and questions about the inadequacy of artisan curricula for the demands of industry. The adequacy and impact of the National Student Financial Aid Scheme (NSFAS) was questioned, especially within the context of rising university costs. There were other remarks and questions about the relationship between road and rail infrastructure; the withdrawal of NT support to the Micro Agricultural Financial Institutions of South Africa (Mafisa) because loans were not being paid back; consultant costs incurred by the PERs; the progress of legal transfers with respect to land restitution; nutrition programmes; skills development; and the PER information-gathering methodology

Meeting report

Chairperson’s introductory remarks
The Chairperson said the performance and expenditure reviews were a joint initiative of the Treasury and the Department of Performance Monitoring and Evaluation (DPME). They assisted with the effectiveness of public policy, fiscal consolidation and reprioritisation in all spheres of government. The Standing Committee (SC) would tell all who appeared before it, that the impact of approved budgets had to be visible. There had to be sustained progress towards the National Development Plan (NDP) vision. A new approach was needed, and a stronger sense of urgency. The briefing would assist the SC to rethink the allocation of funds. It would use the findings about allocations in the Medium Term Budget Policy Statement (MTBPS) report.

Government Technical Assistance Unit (GTAC), National Treasury: Briefing
Ms Ronette Engela, Head: Public Expenditure and Policy Analysis, GTAC, said Performance and Expenditure Reviews (PERs) had been started because the Presidency and the National Treasury (NT) needed rich, detailed and reliable information. It had been initiated as a joint project between the NT and the DPME in the Presidency. Work was done in close collaboration with departments. Information was readily available in existing government systems. Findings were presented under the headings of urban space; higher education and training; economic development; and administrative services.

Between 2010 and 2012, 370 000 Reconstruction and Development Programme (RDP) housing units had been delivered at a cost of R45 billion. Fragmented management of public transport had reduced efficiency.

There were 50 colleges and 26 universities, with 1.7 million students enrolled, following rapid enrolment growth in the preceding five years. Spending would have to rise from 2.3% of the gross domestic product (GDP), to 4.4% to meet the NDP/White Paper goals. The average value of the National Student Financial Aid Scheme (NSFAS) awards was lower than the full cost of study. Producing artisans was costly, at R450 000 per artisan. The throughput rate was 56%. Curricula needed to improve to meet demands and needs of industry.

In a sample of 1 000 leased office accommodations for national government, 60% of the leases were 45% above market rates. It was concluded that policies were sometimes designed and implemented with unrealistic expectations regarding the availability of funding.

Discussion
Mr N Gcwabaza (ANC) said housing delivery was reported to have been somewhat lower than previously reported, and asked what had been found regarding resources that could increase delivery. He asked if the rise in university fees had been booked against the NSFAS, as the rise in university costs could eat up the NSFAS funds. Would the benefits of the NSFAS scheme be reduced? Was it possible that fewer students could benefit? It had been stated that there were 18 nutrition programmes across government, most of them, but not solely, for children. One would have thought that these programmes properly should be located in the Department of Basic Education (DBE), for social improvement, and he asked for an explanation.

He asked if GTAC intended to look at the broad national road and rail infrastructure. Rail was under Transnet, which could remove destructive elements like trucks from the roads by encouraging the use of rail transport. He asked if there had been any progress at all, as this was an issue that went back to 2010. The intention was to release pressure on the roads so that there would be longer life and lower maintenance costs.

Dr M Figg (DA) said the GTAC methodology included participation by departments, and he asked at what level that was -- whether it was clerks or senior management. It seemed that government gave loans without concern for when they would be paid back. The NT was waiting for repayment of outstanding loans. He asked how long it waited, as it was a predetermined system and people had signed conditions of repayment. It was a reckless way of doing things.

He asked when the new window period for land claims would be finalised. It was not helping people to get land they did not own. They could not qualify for funding. It was a contradiction.

He noted that expenditure on NSFAS for 2014 had been R9 billion. What did the figure have to be to satisfy student needs? It had been stated that it cost R450 000 to train an artisan, and that there were 13 training routes. How was that possible? Previously one would get a job at a company, and the government and the company each gave a portion towards training costs. If it cost so much to be an artisan, he would become one. R3 500 per month for three years came to R126 000, and if the government paid 60%, that would be R75 600. He asked how the R450 000 figure had been arrived at. It had been said that there was a 56% pass rate. Under the old system, an artisan trained for two and a half to five years. A trade test could be taken after two and a half years, if a level of competence had been reached, which at that stage was called the National Technical Certificate (NTC) 3. If the apprentice passed it, he would get a red ticket and would be qualified. If the test was not passed, a blue ticket was awarded, and the five years had to be completed. He asked if the State Education and Training Authorities (SETAs) were the problem. Going back to the old system had to be considered.

Mr A McLoughlin (DA) commented that GTAC had said that the Presidency and the NT needed rich information. However, that information was gathered by the Presidency and the NT themselves, and it could be an attempt to keep information from the rest of government, which needed it in terms of the Constitutional mandate.

The presentation had referred to a body of work done by specialised consultants, such as sector experts and modellers. He asked about the cost of consultants. Expenditure review projects had been completed in five to six months, although the international norm was 18 months, and he asked if corners were being cut by speeding up the process. What caused the NT to believe that it could achieve in six months what was elsewhere only achieved in 18 months?

He had noted that the cost of RDP houses delivered between 2010 and 2012 had been R45 billion. He asked what the current cost of an RDP housing unit was. It had been R121 000 in 2010/11. It had been said that there was a strain on local government because the unit cost of RDP houses had become greater than the value of the subsidy. However, housing was a provincial matter -- local government provided only the land. It had been stated that housing delivery was somewhat lower than previously reported -- how much lower?

The presentation had indicated that 55% of subsidies went to the Passenger Rail Agency of SA (PRASA). He asked how much of the 55% translated into effective rail use. It had also been stated that rapid bus transport systems were more expensive than anticipated -- how much more?

He asked for the statement that intermediaries carried the risk of default on Micro Agricultural Financial Institutions of South Africa (Mafisa) loans to be explained. Money went from the government to intermediaries, who distributed it. The Treasury had withdrawn Mafisa funds while waiting for repayment of outstanding loans. He asked how much was owed.

Mr McLoughlin referred to land restitution, observing that R29.3 billion was expected to settle 85% of the approximately 60 000 claims lodged. That would work out at R4510 per claim. 397000 new claims would cost way more than the Department of Rural Development and Land Reform (DRDLR) cost estimate of R230 to R260 billion. It commented that under the two components of the land restitution backlog, legal transfers had not been finalised, and he asked for an explanation.

It had been stated that higher education spending would have to rise from 2.3% of GDP, to 4.4%. He asked if that was an example of the statement that policies were sometimes designed and implemented with unrealistic expectations of the availability of funding. It was unacceptable that the amount had to be doubled.

Ms D Senokoanyane (ANC) asked why RDP housing supposedly placed a burden on local government. She referred to upgrading programmes for informal settlements. Informal settlements were often the result of land invasions. Whenever settlers were relocated, new ones took their place. She had observed that often. She referred to the role of the private sector, saying it was more concerned with profit than with social responsibility.

She referred to Mafisa and intermediaries. The eight percent interest rate on loans was high. Loans were small-scale. She asked if people were not able to repay because they were not making enough profit.
With regard to land restitution, she had seen on the news that a number of new claims had been made since 2014. She asked if there was still an opportunity to make land claims.

She asked if TVET colleges were cost-effective. The NSFAS expenditure of R9.1 billion was a small amount in relation to the number of higher education beneficiaries. It was not adequate. Something had to be done. She asked about NSFAS’s impact, and whether there was value for money. She referred to the student dropout rate, and asked if the NSFAS was doing what it was supposed to do, with regard to its product.

Did the government have criteria and prescribed amounts for office accommodation, because people seemed to go where they liked?

Ms M Manana (ANC) said the government was pumping lots of money into higher education and training. The Committee had received a report about higher education and training which indicated that in some Technical and Vocational Education and Training (TVET) colleges, only one student out of ten qualified. She asked GTAC to elaborate on the “missing middle” in the funding strategy, which had been cited as a key NSFAS issue.

It had been stated that malnutrition was high in South Africa, relative to comparable countries, and she asked which countries were referred to.

National Treasury’s response
Mr Dondo Mogajane, Deputy Director General (DDG): Public Finance, NT, responded that if it was understood where resources went, Treasury would know where to cut. Expenditure reviews touched only the tip of the iceberg in terms of government programmes, including those at the provincial and municipal level. In terms of available resources, it was the first GTAC major exercise. The NT had told the SC before that under budget conditions of fiscal consolidation and an expenditure ceiling, savings made could be used to expand other programmes. Some programmes would have to be stopped. On the GTAC website there were details of each area. The SC could use that information to call the Department of Higher Education and Training to talk only about NSFAS, for instance. GTAC information could be used in different ways. Parliament could choose to do expenditure reviews in various areas of interest. GTAC worked with departments. For instance, the accommodation costs for the DPW were 45% more than the market rate. The SC could call the DPW about that, but it was also part of the NT oversight role.

Ms Engela replied about general methodology questions. There was participation by departments apart from the Director Generals’ office. GTAC would write a letter to the department, asking it to become part of the process, and the department would send medium to high level presentations relevant to provinces. There was a new tranche of identified topics. It also worked together with provincial treasuries. There was a process. A list of areas appropriate to the year would be sent. Provinces were trained in the methodology. GTAC saw fit to publish what had been reviewed up until then, as it was a large body of work. There would be ongoing publication.

She could not provide the precise cost of consultants used. On average, it was R800 000. GTAC did not cut corners. Top international experts were pulled in. Existing information was made use of. There was a shift in what was possible to be done, with improved data systems. It was done in-house because the systems were understood. It was a consultative process, and could be done more quickly than through an international body.

She was hesitant about giving detailed policy reasons related to RDP housing. Information gathered was not claimed to replace current notions about the causes of slow delivery in the RDP environment. Built environment processes were complex with respect to the approval, planning, coordination and flow of funds. There was concern about tendering processes. Local government acted as implementing agents (IAs), who received funds from national and provincial treasuries. GTAC costed the top structures. The difference between what subsidy flowed down to implementing agents and what could be seen as top structures, was often carried by the people responsible for infrastructure. It was difficult to distinguish per project between two funding streams. The subsidy that was coming down did not cover the full cost of the top structure.

A list of the 18 nutrition programmes across government was available in GTAC studies. In the school nutrition programme, GTAC looked at children younger than five. Most programmes in the area of social development led to expenditure on health, especially malnourished children.

GTAC did not look at the possibility of rail taking pressure away from roads, as it was outside of its ambit.

She answered Dr Figg and Ms Senokoanyane about the Mafisa process. Money from Parliament was paid over to the Land Bank and sent to intermediaries, who managed the loans. Some intermediaries could pay the loans back to the Land Bank, and some could not. The Department of Agriculture, Forestry and Fisheries (DAFF) had been alerted about the matter, and could take it further.

Mr Mogojane added that the NT acknowledged the challenges of the Mafisa package, such as poor management and corruption. Repackaging had to be done, with the support of the Land Bank and the Department of Rural Development and Land Reform (DRDLR). He referred to competition between grant funds and loans. The Treasury and the Land Bank would look at the package. The new approach was informed by the GTAC study.

Ms Engela said that land claims were awarded in terms of an administrative process of research and agreement, before awarding it to a specific party. The transfer of legal deeds was called settled, but it was not finalised. A study of land claims was undertaken before new window estimates were made. Unit costs were complicated. One could not generalize, as the cost of land varied between districts.

NSFAS projected numbers had been calculated and were included in the technical report. It was true that there were 13 routes to artisan training, which led to confusion and double costs. The model could be engaged with. Subsidies were calculated. SETAs were engaged with. Artisans were not paid a minimum wage. They needed maths and science. The throughput rate was 36%. Prior learning was recognised. It was a complicated process.

She said that the current cost of an RDP housing unit was R153 000, and the subsidy was R90 000. The cost was carried by local government. PRASA subsidies were complex. Higher capital assets eventually lowered costs. The value depended on how many people were carried, and what the costs were. The model was complicated. Implications had to be looked at. It was in the GTAC technical report.

The risk of default with respect to Mafisa was carried by intermediaries, as they were doing the lending. Large transfers were in the technical report, which showed the transfers to intermediaries.

The cost implications of the rise in higher education spending from 2.3% of the GDP to 4.4% were being addressed.

She replied to Ms Senokoanyane that an informal settlements study had looked at what were declared informal settlements in 2010. Informal settlement backlogs were being addressed.

With regard to the NSFAS, the number of higher education dropouts was heartbreaking. Dropouts were caught in a double bind -- they had no higher education and were saddled with a burden of debt. There was a growing dropout trend. The NSFAS environment had to be reprogrammed.

There were criteria for office accommodation. There were A, B and C grades. Government officials could not use A grade accommodation. Costs were driven by landlords who wanted regular payments. It was a technical process, and government was not as skilled as the private sector.

She answered Ms Manana that the very low throughput rate at TVET colleges was a definitional issue. There were complexities around the calculation. Some colleges had a pass rate of below 10%. The “missing middle” referred to the NSFAS low cutoff for middle income families. Salary earners were struggling to get their children through university.

Comparable countries, with respect to malnutrition, were other middle income countries like Brazil, Indonesia, Malaysia and Chile.

With respect to calculations for foreign missions, international law on the cost of living adjustments was based on New York. If the city was more or less expensive, adjustments were made. The basic salaries remained the same. A city like Oslo was very expensive, and adjustment had to be made upwards. In South Africa, instead of putting initial calculations as a percentage, it had been decided to put down a hard number. Adjustments up and down were very high.

Mr Mogojane replied to Mr Gcwabaza about NSFAS. Higher education enrolment had shot up two to threefold. The study showed that government had spread itself too wide in trying to take on everybody. NSFAS could not sufficiently fund those in the middle income bracket. University fees, enrolment and the NSFAS subsidy had all increased. There would be full engagement on the cost structure of universities. Government did not determine this. If all factors were considered in terms of current allocations, there were shortfalls here and there. If the White Paper and the recent demands of students were taken into account, R120 billion would be needed over and above current subsidy levels to address the issue. Study costs at Wits were R120 000, and R70 000 at Venda. The NSFAS threshold was R75 000. If everyone who qualified was taken on, every modification would require R20 billion. To supply students with free food, books and transport would require R20 billion. It was a thorny issue that involved funding, subsidies and the NSFAS. When one looked at the quantum, subsidies to universities were decreasing, but allocations to NSFAS and TVETs were increasing.

Ms Engela said that the NSFAS had grown by 23% since 2011. University costs had grown by 11%.

Dr Madlopha commented that the information received could help oversight. She referred to the cost of artisan training. Government had paid R6.5 billion in 2014/15, but it was not mentioned how many artisans had been funded. There were challenges with the curricula. She asked if there had been improvement. The artisan system was unnecessarily complex. She suggested that a system be developed that would allow more people to register.

She referred to the sample of 1 000 leased office accommodations which had indicated that 60% of leases were 45% above market rates. She asked which financial year this referred to. The DPW had embarked on a turnaround strategy with respect to accommodation in 2012. She asked what the challenge currently looked like.

Ms Engela said that there were 14 000 to 15 000 artisans. Efforts were being made to push the number up by 1 000 every year.

Dr Madlopha said that if GTAC was not sure about any of the figures, these should be put in writing.

Ms Engela replied about gaps in the artisan curricula. The feedback from industry was that the training was inadequate. Training consisted of theory and laboratory training, and in-service training. The Higher Education Department had acknowledged that the training was inadequate. The 13 routes made it unnecessarily complex. A reform process had been started around National Certificate Vocational (NCV) certification. It was supposed to replace the old system, but was not successful. Different approaches were still possible. Previous approaches had not been taken off the books. Parallel processes were available. Artisan training had moved into Higher Education and Training. It had to be seen as a real alternative to academic courses. There were people who completed academic studies and then opted for artisan training.

With respect to leases, the 2013/14 sample had been the only one that had square meterage and the grading of buildings related to office accommodation. The Department of Public Works (DPW) database was inadequate. If government had more information, it would have more power to negotiate with the private sector. The DPW did not keep the necessary records of what leases cost and who were being paid. Details were needed around IT, security, gardening and extra costs. GTAC proposed a different route for the renegotiation of leases, without automatic escalation. Leases had to be renegotiated back to market rates, on expiry.

Dr Madlopha asked if leases above the market rate had been signed before the DPW turnaround in 2013. The R6.5 billion invested in artisans had not resulted in value for money. She asked if GTAC was saying that artisans were not getting the desired product in terms of training.

Ms Engela replied that 2013/14 was when the sample was taken. There were NT regulations about capital costs. She was interested to know which DPW turnaround had been referred to. The DPW had admitted that it could not turn around on leases. There were artisan centres of excellence, and areas where there was value for money. Higher Education could pronounce on the value to industry.

Mr Gcwabaza commented that the government was pumping in money to drive relevant skills development, in order to make skills available to the economy. It did not seem to have enhanced skills development. R6.5 billion had been invested. The private sector seemed to invest, but there was no movement. He requested that the subject be looked at as part of the review, so that it could be known what was happening. It would not be appropriate to put money into skills development without determining the skills that the economy required. The skills required had to be specified. There had to be a list of the most demanded skills. Government and the private sector had to come together to say what it wanted to achieve.

The Chairperson asked GTAC to elaborate about its information-gathering methodology. She asked when the next round of reviews would be published, and what they would focus on.

Ms Engela replied that existing information was used. GTAC relied on transversal knowledge which did not represent anyone’s opinion. It was augmented by intensive interviews. The basic accounting system (BAS) was underrated. It was possible to go down to expenditure in a hospital, school or police station, or area spending in district municipalities. Information could be obtained about a housing project in section 4, for instance, which was not readily available to government managers. The system was not sufficiently used, and the review would make it more visible. GTAC made use of a new body of work around the Personnel Administration System (Persal).  Data was triangulated. It was possible to say to whom money was transferred, and who got paid for projects. Information was augmented. The next review would begin in the first term of the following year, and at the end of the last term of the financial year. The next tranche would be longer, as the review would deal with information not available at the national level. Information gathering at the provincial level was slower and more expensive. It would be completed by the end of the following year, in October, and published.

The Chairperson said that written questions would be submitted to the Treasury.

Mr Mogojane said that GTAC was beginning to look at the budget. It would dig further to be more informative. There was a three-month training course for budget analysis. People were trained to look at data differently. More usable reports would be published. It could empower Parliament to say where more could be done.

The Chairperson told the Treasury that the report was appreciated. To respond to the demands of the NDP, impementation had to more innovative. The PERs were critical to the Committee to intensify its oversight work. Government impact could be increased. It was clear that departments had to spend more effciciently. There had to be recommendations for the redirection of resources for more impact. Budget decision makers could be helped. There were growing demands on the fiscus. More could be done with less. Higher education could not tolerate programmes that did not work. There had to be better alignment and more resources for positive impact. More higher education resources had to be directed to deal with the low throughput. There had to be a turnaround, with a clear picture of the desired quality of students and artisans.

Spending on infrastructure had to lead to better skills development, and a lower dropout rate. The NT and DPME expenditure reviews were appreciated. It had to be seen where resources went, and what impact they made. A sense of urgency to implement the NDP was needed. The NT findings would be considered and incorporated into the MTBPS report. She was looking forward to the next meeting with GTAC.

Committee business
The minutes of 15, 16 and 18 November were adopted with minor technical corrections. It was resolved that henceforth the meeting agenda had to be adopted, and should also be minuted.

The Committee had a brief discussion of the Committee programme.

The Chairperson referred to a request for an input tabled by Dr Figg. She advised that he should table it as a Committee Member, and not as a member of the DA.
 
The Chairperson adjourned the meeting.
 

Share this page: