Overview of the Financing and Impact of Capital Expenditure

Document Sample
Overview of the Financing and Impact of Capital Expenditure Powered By Docstoc
					                         School of Economics
                        University of Cape Town

               CONFERENCE 2008
            The Regulatory Environment
            and its Impact on the Nature
              and Level of Economic
            Growth and Development in
                     South Africa

Overview of the Financing and Impact of
 Capital Expenditure by Provincial and
           Local Government

                        David Savage

 Employment Promotion Programme
         Overview of the financing and impact of capital
         expenditure by provincial and local government

  Final background paper prepared for the Presidency as part of the 15 year review

                                            David Savage
                                           March 17, 2008

 This background paper reviews trends in capital expenditure and, where possible, outputs of major sub-
national (provincial and local government) infrastructure sectors. It seeks to analyze the performance of
infrastructure investment and maintenance expenditure by sub-national governments, and the extent to
which this has been assisted or constrained by new or reformed institutions, policies and programmes. As
such it has a particular emphasis on the efficacy of the various financing instruments utilised.

1         Introduction ............................................................................................................................................ 2

2         The infrastructure investment needs of sub-national government ..................................................... 3

    2.1        Universalising access to sub-national infrastructure......................................................................... 3

    2.2        Supporting economic growth ............................................................................................................ 5

    2.3        Infrastructure maintenance .............................................................................................................. 6

    2.4        Key issues in the estimation of sub-national infrastructure investment needs ................................ 7

3         Institutional and fiscal arrangements for infrastructure delivery ....................................................... 8

    3.1        Institutional arrangements ................................................................................................................ 8

    3.2        Fiscal arrangements and financing instruments for sub-national infrastructure investment ........ 10

    3.3        Institutional and fiscal challenges ................................................................................................... 17

4         Capital expenditure programmes of provincial and local government .............................................. 23

    4.1        Trends in capital spending by provinces and municipalities ........................................................... 23

    4.2        Capital spending and service delivery ............................................................................................. 26

    4.3        Job creation and public works ......................................................................................................... 30

    4.4        Maintenance Expenditure ............................................................................................................... 30

    4.5        Underlying issues in capital and maintenance expenditures .......................................................... 31

5         Issues and emerging lessons in sub-national infrastructure delivery ................................................ 32

    5.1        Policy Uncertainties ......................................................................................................................... 33

    5.2        Information Systems........................................................................................................................ 33

    5.3        Institutional Capacity ....................................................................................................................... 34

    5.4        Fiscal Framework ............................................................................................................................. 34

    5.5        Confused Accountability .................................................................................................................. 35

References....................................................................................................................................................... 37

1    Introduction
This background paper reviews trends in capital expenditure and, where possible, outputs of major sub-
national (provincial and local government) infrastructure sectors. It seeks to analyze the performance of
infrastructure investment and maintenance expenditure by sub-national governments, and the extent to
which this has been assisted or constrained by new or reformed institutions, policies and programmes. As
such it has a particular emphasis on the efficacy of the various financing instruments utilised.

The main sectors that are included in this review are health, education, roads and housing at the provincial
level, and electricity distribution, water, sanitation and roads at the local government level. In general,
provincial governments deliver community infrastructure, while local governments provide economic
infrastructure. The transport functions of provincial governments and 2010 stadium building programmes
of local government are the obvious exceptions to this rule.

Regardless of the category of infrastructure programme, they all share (and must try to reconcile) the
objectives of meeting basic needs (poverty reduction) and enhancing the productive capacity of
communities. South Africa’s post-1994 sub-national infrastructure programmes initially focused almost
exclusively on poverty reduction objectives. They sought to universalize access to basic services, through
rolling our infrastructure networks to poor, historically marginalised communities as rapidly as possible.

More recently, government has also sought to utilise infrastructure programmes to stimulate higher levels
of economic growth and job creation. The positive contribution of infrastructure programmes to growth
and job creation is generally well-accepted. Hassen (2000:2) associates infrastructure investment with the
benefits of lowering transaction costs, improving productive capacities of firms and households, creating
economic linkages and boosting upstream demand, concentrating economic activity, allowing economic
sectors to respond to changing market conditions more rapidly, and creating jobs. These expanded
objectives of sub-national infrastructure programmes have been associated with a rapid rise in available
resources for investment.

As is evident from the brief discussion above, the objectives and potential impacts of infrastructure
programmes are heavily mediated by institutions and public policy. Esfahani and Ramirez (2002) model
these relationships and find that “the impact of infrastructure on GDP growth turns out to be substantial”,
but also that “institutional capabilities that lend credibility and effectiveness to government policy play
particularly important roles…” (2002:25). They continue:

        “the effects indicate that countries can gain a great deal by improving investment and
        performance in infrastructure sectors. But, the exercise also implies that achieving better
        outcomes requires institutional and organisational reforms that are more fundamental
        than simply designing infrastructure projects and spending money on them.” (2002:25)

This review largely supports this finding. Emerging issues and lessons from the last decade of infrastructure
investment by sub-national governments in South Africa demonstrates that national priorities and fiscal
systems can support a rapid achievement of core goals. However, it finds that policy uncertainties,
information issues, problems in the fiscal framework, and variable institutional capabilities can work to
undermine performance. Ultimately, the framework for institutional accountability will need to be further
strengthened for ongoing progress In sub-national infrastructure programmes.

Methodological notes

This paper follows a simple structure. Firstly, demand for infrastructure investment is analysed, and a
number of underlying issues discussed that affect the nature and quantification of demand. Secondly, the
framework for resourcing this need is discussed, and various underlying issues in resource availability and
targeting are discussed. Thirdly, actual expenditures and outcomes by sector are discussed. The paper then
concludes with some overarching themes.

This overview is restricted to direct spending by provincial and local governments in core sectors, namely
education, health, housing, transport, water services and electrification. This restriction implies that:
    • Capital expenditure by State Owned Enterprises (Eskom, Transnet, Water Boards & TCTA) and
        Special Funds (e.g RDP Fund, SPRA, uTshani, TNDT) is not the primary focus of evaluation here.
    • “Non-core” sectors, and those where capital expenditure is limited, are excluded. This includes
        social development.
    • Analysis of intergovernmental transfers is limited to the core sectors identified above.

The review has been conducted as a desktop exercise, building on existing evaluations and data sources.
Weaknesses in the accuracy and comparability of historical datasets and a general absence of mid-term or
end-of-programme evaluations reduce the scope of the evaluation.

Analysis of provincial and local government finance is made complex by three factors. Firstly, the financial
year of local governments differs by a quarter from that of national and provincial governments.
Municipalities begin their financial year on 1 July, whereas other spheres begin their years on 1 April. While
this has important managerial benefits, particularly related to the transmission of budget information
across spheres, it does complicate analysis across all sub-national entities. The paper uses data from the
national and provincial financial year for those spheres, and for all transfers between spheres. Municipal
spending is reported on the basis of their financial year. Secondly, data on local and provincial government
finance is unreliable. Historically poor recordkeeping and reporting by municipalities, long delays in
auditing, a major realignment of local government boundaries, as well as by the current process of reforms
to municipal financial management (including budgeting and accounting practices) have collectively
resulted in incomplete and not strictly comparable datasets. While the reform process has begun to deliver
significant improvements to the quality and quantity of municipal financial information, this is not always
comparable between years, between municipalities, or even between years for the same municipality. The
functional classification of expenditures is particularly problematic. This review uses a composite of
budgeted and reported actual figures for each year, drawn from National Treasury data sources. All data is
presented in nominal form unless otherwise indicated.

Data on physical progress with eliminating backlogs is similarly unreliable. Most data from government
departments conflates expenditure with sustained improvement in access to services, and thus steadily
diminishes backlogs as total expenditure rises, with adjustments for demographic change. This ignores two
critical factors. Firstly, the extent to which access is sustained over time, which can be reversed due to a
collapse in services. Secondly, the extent to which expenditures on access achieved their intended
developmental outcomes. For example, providing a pit latrine does not automatically result in people using
it. Unsafe sanitary practices can continue, with infrastructure being used for other purposes such as
agricultural storage.

2     The infrastructure investment needs of sub-national government
Sub-national governments must balance two, often competing pressures in planning their infrastructure
investments. On the one hand they must expand access to basic services to all citizens through building
new assets or extending them, while on the other hand they need to maintain, rehabilitate or replace
existing infrastructure assets so that their value (as an operational asset) is protected over time. Striking
this balance is made more difficult by distorted patterns of access to services that reflect an apartheid
social and economic legacy, the limited availability of resources, shifting needs associated with
demographic change and shortfalls in institutional capacity.

2.1    Universalising access to sub-national infrastructure
Since 1994, most sub-national infrastructure investment has been driven by the priority of universalising
access to basic services. Infrastructure investment patterns under apartheid had resulted in highly
differential levels of access to services and differentiated quality of service between racial group areas, and
across the country. However, this variation was also accompanied by a pattern of duplication and spatial
dislocation of investment. In racially segregated towns and cities bulk infrastructure (such as water

  treatment works) was often duplicated, while in former homelands high levels of service were occasionally
  installed to showcase the benefits of separate development.

  Initial sub-national infrastructure investment strategies focussed primarily on expanding access to services,
  rather than looking at the longer term sustainability of investments. In particular, the conundrum of over-
  specified but installed infrastructure, particularly in remote locations, was largely ignored. At the heart of
  this approach was a rights-based approach to infrastructure investment. This held that all citizens,
  regardless of their ability to pay for access, should receive services at a basic level. The calculation of
  investment needs was thus made on the basis of the number of individuals or households who did not have
  access to a basic level of service, as defined by sector norms and standards. This fundamental assumption
  continues to underpin discussions on sub-national investment needs even today.

  a) Economic Infrastructure

  Considerable effort has been expended on modelling service backlogs for municipal infrastructure services,
  as reflected in various Municipal Infrastructure Investment Framework (MIIF) documents over the years.
  These estimates are necessarily based on a variety of demographic, level of service, cost and capacity
  assumptions, which will vary between urban and rural areas and will also be subject to the preferences of
  local communities. Table 1 summarises estimated of backlogs for local economic infrastructure and housing
  over time. It demonstrates the significant progress that has been made in improving access to services.
  However, it should not be assumed that the same households (“the backlog”) remain unserved across all
  sectors. A household that lacks adequate water services may have an electrical connection, while those
  without adequate housing may still enjoy access to other services.

                      Table 1: Estimated backlogs in local economic infrastructure( 1994 to 2007)

                   1994 (RDP)                   1997                      2001                    2004                   2007
               Unserved                 Unserved                   Unserved                 Unserved               Unserved
              Population        %      Population        %        Population        %      Population      %      Population     %
Total Pop.    40,400,000               41,141,459                 44,819,866               47,845,153             48,897,554
Water         15,890,249    39.33%     15,782,494      38.36%     13,102,570     29.23%     9,431,499    19.71%    6,968,757    14.25%
Sanitation    21,000,000    51.98%     20,329,378      49.41%     19,594,970     43.72%    16,277,538    34.02%   14,329,889    29.31%
Electricity   13,500,000    33.42%                                13,464,151     30.04%    13,406,166    28.02%
(units)        1,200,000    13.40%                                                          1,800,000    13.50%
  Sources: RDP (1994), PDG estimates based on StatsSA census and survey data, NDoH, DWAF

  Basic access to water is defined as households with access to a piped water supply within 200m of their
  dwelling. The delivery of residential water infrastructure is generally the responsibility of local government.
  Backlogs in access have declined from almost 30% in 1994 to under 20% in 2004. The Stats SA Community
  Survey (2007) reports that 88% of households now have access to water. The nominal backlog figure is now
  reported to have stabilised at about 2.7 million households. The target for the eradication of backlogs is

  Basic access to sanitation is defined as access to a Ventilated Improved Pit Latrine, or comparable unit, on
  site. The backlog thus comprises households with unimproved pits (most of the backlog), bucket systems or
  no sanitation facility. Sanitation services are also typically the responsibility of local governments. An
  estimated 52% of households had no access to basic sanitation services in 1994, whereas 34% had no
  access to 2004. Subsequently (to 2007), overall access has continued to improve. The Stats SA Community
  Survey (2007) reports that 78% of households now have access to sanitation. However, about 300 000
  households still use a bucket collection system, and this has remained more or less constant over the last
  few years. The number of households using an unimproved pit latrine has increased from about 2.5 million
  to about 3 million over the period. The number of households with no sanitation facility has declined from
  about 1.5 million to about 800,000. The target for the eradication of backlogs is 2010.

Basic access to electricity is defined as a low voltage household connection. The responsibility for
infrastructure installation is split between local governments (who hold the constitutional authority) and
Eskom, which supplies electricity in all or part of many municipal jurisdictions. Backlogs in access to
electricity have declined from 33.4% in 1994 to 28% in 2004, although the rate of decrease has slowed. The
Stats SA Community Survey (2007) reports that household electricity access has increased to 59% for
heating, 67% for cooking and 80% for lighting. The target for the eradication of backlogs is 2011.

Basic access to housing is defined as a formal, weatherproof shelter. About 1,6 million housing
opportunities have been provided since 1997/98, and 500,000 families have had the opportunity to secure
titles of old public housing stock. An estimated 1,8 million housing opportunities are still required. Housing
delivery peaked in 1997, and the subsequent decline in delivery has led to rising backlogs associated with a
quicken pace of demographic change. The increasingly demand for housing must also translate into
increased demand for associated infrastructure connections, although this is not always reconciled across

Basic access to roads is typically defined as an improved (all-weather) road connecting to a settlement. A
higher service level would be a gravel or graded road top within 500m of a dwelling. Little data is available
on the extent of the road backlogs, and only limited data on the existing sub-national road network and its
condition, which is discussed later.

b) Community Infrastructure

Unfortunately, backlogs for community infrastructure (clinics, schools, etc) have been less easy to calculate.
This reflects the underlying needs of the sector, which has not always been focussed on the rapid
expansion of facilities, but rather of services.

Basic access to primary and secondary education requires adequate numbers of classrooms relative to the
number of learners, located within functioning schools. Provincial governments are responsible for the
construction of education infrastructure. The National Education Infrastructure Management System audit
(2007) reported that 6 331 schools, or 23% of all the schools in the country, had an average of 45 pupils a
classroom - 10 pupils more than the required learner-teacher ratio. The Department of Education notes
that this is due to a lack of classrooms rather than a shortage in teachers, and that 12,000 new classrooms
are required. Classroom overcrowding occurs mostly in the rural areas of Limpopo, KwaZulu-Natal, Eastern
Cape and to a lesser extent in Gauteng.

Health sector backlogs are poorly understood. While significant provincial investment in primary health
care occurred after 1994, funding responsibility has now been shifted to provincial budgets. Estimations of
infrastructure backlogs are driven by a combination of clinic catchment areas and average number of
patient visits per year, although no national data on this appears to be available. Similarly, estimations of
the adequacy of hospital infrastructure are not readily available. Expenditure on medical equipment is an
important dimension of capital investment, but is not regarded as infrastructure spending for the purposes
of this paper.

2.2   Supporting economic growth
Fixed investment by the private sector relies strongly on sub-national, and particularly municipal,
infrastructure investment. Municipalities are required to provide infrastructure services to support these
fixed investments, through road, electrical and water services connections. Inadequate provision of this
infrastructure can delay investments.

There is little data available on the extent of physical municipal infrastructure investment to support
economic growth, although Section 4 considers the extent of municipal capital expenditure relative to
building plans passed and building completed.

Data on building plans passed and completed (StatsSA, P5041.1) provides a useful proxy on private
investment intentions and performance over time. The total value of building plans passed has grown at an
average of 19.1% since 1998, with particularly strong growth from 2002 to 2006. The value of building

       completed has averaged 55.3% of plans passed over the period. The non-residential sector has grown at a
       slower pace (15.5%) than the average, although completion rates have been higher (60.2%). This reflects
       the particularly strong growth that has occurred in the residential sector, although this is now slackening.
       Another frequently citied reason for the relatively slower growth in non-residential investment (and its
       associated decline as a proportion of total investment) is that “bottlenecks” in the provision of municipal
       infrastructure regularly delay the expansion plans of firms.

             Table 2: Nominal value in R’000 of building plans passed and buildings completed (1998 to 2007)
                1998         1999         2000           2001     2002       2003     2004     2005      2006        2007      p.a
Plans passed 18,137,000 17,801,839 20,126,132 19,710,923 25,832,162 31,028,535 44,336,732 64,480,717 75,769,810   80,351,579 19.1%
Completed    11,688,000 10,241,312 10,775,642 12,590,750 14,034,387 16,616,738 22,577,891 30,110,132 38,834,985   45,561,022 17.2%
% completed 64.4%        57.5%        53.5%           63.9%     54.3%    53.6%      50.9%    46.7%    51.3%       56.7%     55.3%
% Non-res.
            63.2%        48.6%        61.3%           111.1%    60.1%    62.0%      58.6%    40.0%    38.7%       58.3%     60.2%
… of which
            29.6%        26.4%        29.1%           30.7%     18.8%    18.4%      16.7%    14.0%    15.0%       19.7%     21.8%
       Source: StatsSA, P5014.1 (1998 through 2007)

       2.3     Infrastructure maintenance
       The need for maintenance of infrastructure differs markedly between sectors and depends on the initial
       quality of infrastructure that is installed. As a general rule, the higher the quality of the initial investment,
       the longer the period before significant maintenance or rehabilitation will be required. However, this is not
       always the case as maintenance needs will also vary with the intensity of infrastructure use.

       a) Economic infrastructure

       Water and sanitation: Water services assets of local governments and water boards were estimated to
       have a replacement value of R180 billion in 2007 (CSIR estimate in cidb, 2007). Maintenance is reported to
       be a growing problem. In 2004, 37% of households had water distribution interrupted mainly for technical
       reasons, while 28% of households experienced failures in their sanitation systems. Over 50% of a sample of
       30 local governments reported leakages above international norms. In 2005, 63% of local governments
       could not confirm that they met water quality standards, although by 2007 it was estimated that 72% were
       now compliant.

       Roads: In 2001 (RIFSA) it was estimated that sub-national governments collectively owned 224,000 km of
       surfaced roads, and 522,000 km of other roads (from gravel roads to tracks). About 75% of surfaced roads
       and 42% of other roads were owned by local governments. The estimated total asset replacement value of
       this stock was R260 billion (CSIR estimate in cidb, 2007), or an average of R498,000 per kilometre. However,
       replacement costs vary dramatically by type of road. Total provincial road assets are estimated to be worth
       R200 billion (or R560,000 per km), while municipal assets were valued at R60 billion (R364,000 per km).
       Information on the condition of road assets is incomplete, but asset condition was reported to be adequate
       for surfaced roads, and inadequate for other roads. It is estimated that annual maintenance costs should be
       approximately 4% (and up to 10%) of total asset values.

       Electricity: The replacement value of electricity distribution assets has been estimated at R84 billion, of
       which R55 billion is held by local governments and the remainder with Eskom (CSIR estimate in cidb, 2007).
       Very little data is available on the age or condition of these assets. A NERSA (2007) audit of eleven
       electricity distributors reported the standard of data collection and reporting to be “well below
       international standards”. The audit estimated the weighted average of all distribution assets at between 25
       and 30 years, which is young by international standards.

       It should be noted that responsibility for housing maintenance falls to households, except in the case of
       public rental housing stock. Similarly, households are theoretically responsible for non-networked (on-site)
       sanitation system maintenance.

b) Community Infrastructure

Health and Education. The combined total replacement value of health assets is estimated to be R275
billion, although this may have risen steeply in the last 18 months in the case of health (Blecher, pers com).
Health and education assets are estimated to require maintenance expenditure of 4% of the total asset
value per annum. In the health sector this is required across over 400 hospitals and more than 1000 clinics.

              Table 3: Asset values and maintenance requirements for selected sub-national sectors.
                                     Asset values              Estimated % annual              Estimated annual
                                                            maintenance requirement*           maintenance cost

Water Services                         R180 bn                          4%                          R7.2 bn
Electricity                             R85 bn                          4%                          R3.4 bn
Municipal Roads                         R60 bn                          4%                          R2.4 bn
Subtotal: Local Govt                   R325 bn                                                      R13 bn
Provincial Roads                       R260 bn                          4%                           R8 bn
Health                                 R137.5 bn                        4%                          R5.5bn
Education                              R137.5 bn                        4%                          R5.5bn
Subtotal: Provinces                    R475 bn                                                      R19 bn

Total: Sub-national govt               R800 bn                                                      R32 bn
*Estimated annual maintenance requirements are drawn from cidb (2007 and refer to sector averages. Investment needs by type of
asset may thus vary from the average. Other estimates (e.g. SAICE, 2007) suggest that these may be occasionally understated in
some sectors.

2.4      Key issues in the estimation of sub-national infrastructure investment needs
Four issues emerge from this brief review of sub-national infrastructure investment needs.

Firstly, the information on servicing needs is highly variable between sectors and generally outdated.
Demographic trends, particularly the phenomenon of household decompression in urban areas (caused by
the drop in average household size from 4.5 people per household in 1996 to 3.8 in 2001), has dramatically
raised the number of consumer units that require economic infrastructure services. As a result, the
geographic distribution of investment needs has also changed dramatically. Information management
systems also tend to be fragmented and weak. National department’s typically construct database’s that
are at odds with official government information captured by StatsSA, both due to different survey
methods (and sample sizes) and more fundamental errors of confusing expenditure with actual
improvements in access to services. A recent roads survey found that only 36% of responding municipalities
had some form of management system on the extent and condition of road assets.

Secondly, planning processes have typically been weak. In addition to the methodological issues outlined
in section 1, national estimations of backlogs have tended to ignore actual community preferences for
investment (effective demand) as they pursue sector targets for improved access at centrally determined
service levels. This has been compounded by the provision of full subsidies for basic infrastructure services,
which tends to further obscure community preferences. Only the housing programmes has a formal
requirement for community contributions. Local planning tools (such as municipal IDPs) have been
methodologically too weak to accommodate extensive public participation and oversight of municipal

Thirdly, the norms and standards for investment in various sectors has been criticised on two grounds. On
the one hand it is argued that very basic levels of service do little to install an appropriate infrastructure
platform for local economic development. Low voltage electricity connections, for example, can hamper
the development of more energy intensive small businesses. On the other hand, these norms and
standards may be mis-specified relative to intended development outcomes. For example, wide road
reserves and large minimum plot sizes raise the cost of road infrastructure and other reticulated
infrastructure and leads to more sprawling, car-oriented settlements. Water borne sewerage systems may

be fiscally and technically unsustainable, yet are given preference over alternatives such as community
toilet blocks. In the community services sectors the absence of infrastructure norms and standards has
made it difficult to measure current future investment requirements.

Finally, the focus on backlogs has led to maintenance expenditures being accorded less priority than new
investments, often leading to the replacement of assets. This is reflected in data weaknesses on existing
assets, as well as the comparative ease with which new assets can be created or networks expanded,
relative to the more technically and politically more challenging task of planning, financing and
implementing asset maintenance programmes. Effective asset maintenance requires a strong institutional
capacity to undertake these tasks, which in turn presupposes a focus on service delivery outcomes rather
than connection targets.

These issues will be investigated further in the course of this paper.

3     Institutional and fiscal arrangements for infrastructure delivery at provincial
      and local government levels
The preceding section described estimates of physical investment needs by sector over time. In 1998, the
cost of municipal infrastructure investments required to address backlogs alone was estimated to be in the
order to R75 to R150 billion, and this figure was regularly increased. The nominal value of this figure is itself
of limited value. It is a function of a variety of assumptions on needs, investment norms and cost structures
that vary dramatically in the real world. The cost has always clearly been beyond the immediate reach of
public funding. It thus primarily serves to emphasize the importance of seeking additional resources
beyond the national fiscus to finance sub-national investment needs.

This section briefly describes institutional and fiscal arrangements for sub-national infrastructure delivery
and highlights key issues that have impacted on the effectiveness of related spending programmes.

3.1    Institutional arrangements
Institutional arrangements for infrastructure delivery encompass (a) the intergovernmental assignment of
functions, and (b) agency assignments and institutional procedures, rules, norms and standards within a
sphere of government.

a) Intergovernmental functional assignments

South Africa’s three “spheres” of government are constitutionally inter-dependent yet autonomous, and
established within the framework of a unitary state. Their inter-dependence is reflected in both functional
concurrency as well as a fairly broad set of financial relationships. Provincial governments are assigned
functions associated with community services (notably, health, education, welfare functions) that tend to
require a significant degree of inter-personal interaction, and are thus personnel intensive. Local
governments are assigned water, electricity distribution, sanitation and refuse functions, which require a
far lower degree of interaction with individuals, and are typically capital intensive. Importantly, housing,
transport and land use regulation functions (and thus finances) are concurrent mandates between
provincial and local governments, with housing and transport functions being more capital intensive in
nature. There are two important aspects of the current system of functional assignments that must be

Firstly, concurrency in functional assignments cuts directly through built environment functions, affecting
not only housing and transport, but also related infrastructure investments in water, sanitation, electricity
distribution and solid waste infrastructure. A significant degree of coordination is thus required between
functions to ensure that investments occur in an efficient manner. Two concerns have been raised on the
effectiveness of current functional assignments:

a) It creates considerable bureaucratic complexity, through failing to provide clarity on lead agencies for
   infrastructure investment. A decision to approve a new housing project, for example, requires

    complementary project approvals for infrastructure and roads both within the settlement and leading
    to it. Similarly, installation of new bulk and transmission infrastructure has a directive influence on
    future settlement patterns. The housing sector has typically regarded itself as the lead agency in these
    decisions, although local governments have considerable control over land use and infrastructure

b) It allows the costs of investment decisions to be externalised across sectors. In particular, the trend
   towards peripheral, low density housing development (itself driven by difficulties to unlocking well-
   located land in urban areas) results in long term costs for other infrastructure sectors. These other
   sectors are required to install and then subsidise the operating costs of associated services, while the
   housing sector can reduce its construction costs per unit.

Secondly, the system of functional assignments is symmetric within spheres of government, in that it
assumes that all local or provincial governments are equally suited to perform a function (commonly known
as a “one size fits all” approach). However, there is a considerable degree of variation between individual
provinces and municipalities in terms of investment needs, appropriate investment standards and
institutional capabilities. Local government functions in non-metropolitan areas are, however, divided
between District (Category C) and Local (Category B) municipalities based on determinations by the
national or provincial government for specific functions. This has resulted in provincial variations in intra-
sphere allocation of functions across provinces (and indeed within them. Typically, poorer provinces (KZN)
and / or those with weaker institutional capacity (Eastern Cape, Limpopo) have allocated more functions to
District Municipalities.

The symmetric allocation of functions has been criticised for failing to recognise the significant variation in
institutional capabilities between a metropolitan municipality, for example, and its poor rural counterpart.
Thus the City of Tshwane, for example, must subject itself to the same project-based approval process to
use infrastructure grant funds from national government despite it already managing self-funded capital
projects of a significantly greater size and complexity.

b) Intra-sphere task assignments

Responsibilities for capital investment within a sphere of government are located across a number of
agencies. Generically, revenue, budgeting and expenditure monitoring functions are located within
provincial or local treasuries (although this is not always the case where public entities have been
established). Strategic planning functions are typically located in the Office of the provincial Premier or
Municipal Manager. Project planning is located with relevant sector departments, while project
management functions are located either with sector departments or with special purposes departments,
most notably the provincial Departments of Public Works (DPWs). This assignment of roles is intended to
allow for coordination between sector investments (at a strategic level), allow a degree of role
specialisation to occur, and encourage economies of scale and skill in investment management.

Three aspects of this task assignment have proven problematic:

a) Planning systems have often proven to be weak: Considerable effort has been put into sub-national
   strategic planning systems since 2000. At provincial level, growth and development strategies (PGDS’s)
   have been developed, while municipalities have generated Integrated Development Plans (IDPs). This
   has been accompanied by extensive medium term budget reforms in both spheres. However, these
   plans have most often struggled to generate credible investment strategies for municipalities. This is
   partly due to methodological constraints associated with a linear (problem          vision      action)
   approach that assumes that public action have a greater degree of influence over local development
   trajectories than has turned out to be the case, and partly due to a focus on perceived needs rather
   than effective demand (as mediated by resource availability). Citizen engagement in the planning
   process has also been criticised as being weak or non-existent, reflecting perhaps an overly
   technocratic orientation in planning systems. More problematically for infrastructure investment,
   however, has been that strategic planning has often occurred at the expense of detailed business
   planning for individual capital investments.

b) The separation of budgeting and revenue management from expenditure management has reduced
   efficiency incentives: Budget reforms have increasingly decentralised budget authority to spending
   departments, alongside the rollout of medium term budget horizons. This is intended to provide
   spending departments with incentives to raise revenues for investment in their own service delivery
   processes. However, capital assets (particularly property) continue to be subject to rules and
   procedures that militate against efficient management of assets. For example, revenues from the sale
   of land are not retained by departments, giving them little incentive to release unneeded fixed assets.
   In addition, it continues to be substantially easier to obtain budget allocations for high profile new
   investments than for standard maintenance expenditure, which may promote the decay of installed
   assets (asset stripping).

c) The separation of project planning from implementation has led to the fragmentation of the delivery
   system: The historical separation of planning for implementation is premised on (a) DPWs having the
   necessary experience, skills and capacity to manage capital projects across sectors and (b) effective
   contracting arrangements between sector departments and DPWs. Experience has shown that neither
   of these are necessarily the case, leading to conflict, delays and cost escalation in capital projects.

In response to these concerns (and particularly the quality of business planning and implementation
arrangements), the National Treasury developed the Infrastructure Development Improvement Programme
(IDIP) from 2002. This has already begun to show considerable success, and is discussed in more detail in
section 4.

The focus on improving the effectiveness of the system of task assignments does tend to obscure another,
more difficult to discern trend in mechanisms of infrastructure investment. Most sub-national public
infrastructure investment projects adopt a “statist” approach to implementation, whereby provinces or
municipalities either manage projects in house, or procure engineering and construction support. The use
of non-state actors in the investment process appears to have declined, particularly in the case of small
scale infrastructure investments (such as community-based rural water supply schemes that were
extensively utilised by agencies such as Mvula Trust in the mid to late 90’s). Utilisation of public-private
partnerships has also been below anticipated levels. The net effect of this tendency is that sub-national
infrastructure investment programmes have not taken full advantage of community contributions (in cash
or kind) or private sector management expertise.

3.2   Fiscal arrangements and financing instruments for sub-national infrastructure investment
This sub-section identifies and quantifies the various sources of capital finance available to provincial and
local governments arising from own revenue, debt instruments and intergovernmental transfers.

a) The inter-governmental fiscal framework

An important development since 1994 has been the institutionalisation of a transparent, predictable and
equitable fiscal framework to guide relations between the three spheres of government. That this has
occurred in the context of fundamental restructuring, transformation and refocusing across the public
sector further underlines the scale of this achievement. All spheres of government now operate multi-year
budgeting systems that are informed by strategic plans. Fiscal relationships are conducted through a
coherent and cooperative system of intergovernmental budgeting, and reporting systems are increasingly
being aligned to allow comprehensive monitoring of expenditures.

Key to this process has been the articulation of a set of guiding principles for intergovernmental fiscal
relations, structured around the need to avoid moral hazard distorting accountability for decisions within
each sphere. Major fiscal reforms have included the introduction of equitable shares of nationally raised
revenues for both provincial and local government, both of which are distributed on a transparent,
formula-driven basis, the rationalisation of the number of conditional grants and other transfers to sub-
national governments, and the re-regulation of sub-national own revenue and debt raising instruments.

A transparent, predictable and equitable fiscal framework is an essential pre-requisite for effective
infrastructure investment. Infrastructure projects typically have longer lead times than other forms of

       spending, and thus require certainty over the resources that will be available prior to commissioning work.
       The availability of resources between sub-national jurisdictions needs to correlate with existing and
       emerging needs, and thus needs to be equitable distributed. Finally, infrastructure investment decisions
       need to account for the ongoing operational costs they will impose, while minimising negative externalities
       on other spheres of government or adjacent jurisdictions. The framework for intergovernmental relations
       in South Africa emphasizes the importance of transparent, predictable relationships between spheres of
       government, in order to ensure that each sphere is able to exercise its powers and functions, and be held
       accountable for the implications of associated decisions. However, the bulk of public revenues are raised by
       national government, which implies an extensive system of fiscal relationships between spheres.

       The key mechanisms through which government has institutionalised an effective intergovernmental fiscal
       framework have been the PFMA, MFMA and the annual Division of Revenue Acts.

       b) Financing instruments

       Provincial and local governments are constrained in financing their investment needs by the availability of
       resources and the regulatory environment for public financial management. Resources, or revenues, to
       finance capital expenditure are available from three generic sources, namely own revenues raised through
       taxes or user charges assigned to them, grants or other transfers from other spheres of government that
       are made either unconditionally or earmarked for particular functions or types of expenditures, and loans
       or equity financing arrangements that leverage resources from citizens and the private sector into public
       infrastructure investment.

     i.     Own revenues of provincial and local government

       Provincial governments have very few own revenue sources assigned to them, raising only R7.95 billion in
       own revenues in 2006/07 (see Table 5). This accounted for only 4.3% of total provincial receipts, and is
       projected to drop to 3.2% by 2009/10. Negative nominal growth in provincial revenues is also being
       forecast in the current MTEF. Provincial own revenues are predominately drawn from motor vehicle licence
       fees (45.3 per cent), sales of goods and services (18 per cent) and casino taxes (13.4 per cent). Other
       revenue sources include interest, dividends and rent on land, horse racing taxes, liquor licences, asset sales,
       and financial transactions. Although provincial own revenues are a negligible source of capital revenues,
       their decline in proportional and nominal terms will reduce space for discretionary capital investment by
       provinces, and potentially remove their (un-exercised) capacity to borrow from private capital markets.
       In contrast, local governments have had access to a range of own revenue instruments. Most notable
       among these are property rates accounting for 17.3% of total budgeted revenues in 2006/07 for
       metropolitan and local municipalities. User charges for trading services such as water and electricity
       account for 11.4% and 24% of total revenues respectively, while intergovernmental transfers account for
       24.6% of total revenues. Other revenue sources include license fees, fines, housing rentals, donations and a
       host of other smaller charges, some of which can be lucrative for municipalities, as shown in Table 4.
       Regional Services Council (RSC) levies used to account for around 7% of revenue for metropolitan and
       district municipalities, but this tax has been abolished, so the limited revenue shown in Table 4 is related to
       arrears only.
            Table 4: Comparative budgeted revenues of selected municipalities in 2006/07*

                                           Property                               Subsidy/                      Refuse                   TOTAL
                             Electricity                Water**     RSC Levies                 Sanitation**                   Other
                                            Rates                                  Grants                      Removal                   (R'000)
                               26.1%         20.0%       12.4%         0.2%        19.4%          2.2%           2.7%         17.0%           64,196,589
Local Municipalities           25.1%         15.9%       10.5%         0.0%        24.9%          4.4%           4.1%         15.1%           36,900,987
District Municipalities        0.1%          0.0%         7.3%         7.9%        69.0%          1.1%           0.1%         14.6%            7,284,246
National Average               24.0%         17.3%       11.4%         0.7%        24.6%          2.9%           3.0%         16.2%         108,381,821
            Source: NT Database. Notes: * Figures may not reconcile with other tables due to use of budgeted data and municipal estimates of grant
            income. ** Water & Sanitation charges treated differently across metro’s.

  There has been a structural change in the composition of municipal revenues. Firstly, municipal tax
  revenues have been remarkably buoyant, with property rates showing volatile but dramatic real growth
  since 2000 (and average increase of over 30% in real terms). This may reflect both administrative
  improvements and growth in house prices (in cases where valuation roles have been updated). RSC levy
  revenues have shown similar (though less volatile and dramatic) improvements, increasing by 28% in real
  terms since 1999/2000. As a result municipalities have become comparatively less dependent on user
  charges as a source of own revenues, whose contribution has declined from around 42% in the late 1990’s
  to 35% in 2005/06.

  Municipal own revenues are particularly important as a revenue stream to fund long term borrowing for
  capital investment. Conditional grants, in particular, cannot be pledged as a security for loans. The abolition
  of the RSC levy is thus likely to have a significant impact on metropolitan finances in particular. Its
  immediate replacement with a transfer from the national government does not guarantee long term access
  to a proven and productive source of income, or provide a readily available security for long term
  borrowing to finance capital expenditure. This is only partially count-weighted by the stabilisation in
  consumer debtors. The Auditor-General (2004, 2005) reports that the growth in total debt slowed from
  12% to 11% between 2003/04 and 2004/05, with municipalities provision for bad debt beginning to decline.
  The average debt collection period also fell from 322 days in 2003/04 to 136 days in 2004/05. Although this
  provides some (limited) hope, accumulated consumer debt remains large at an estimated R28 billion
  (National Treasury, 2005) and will place considerable pressure on municipal balance sheets for the
  foreseeable future.

ii.    Intergovernmental transfers to provincial and local government

  In 2006/07 national government transferred R26,1 billion to support the infrastructure investment
  programmes of provincial and local governments. These transfer programmes accounted for 12.1% of all
  transfers in that year, with this proportion growing rapidly since 2003/04. While growth in all transfers to
  provincial and local governments has averaged 11.4% between 1999/00 and 2009/10, infrastructure
  transfers have averaged 27% over the same period. The majority of this funding was allocated to provinces,
  who have received an average of 57.7% of conditional infrastructure transfers over the period. This reflects
  the respective functions and available alternative revenue sources of each sphere.
  Provincial governments are heavily reliant on intergovernmental transfers to finance their expenditures.
  Transfers to provinces account for 96% of all revenues (average 1999/00 to 2009/10) and have risen
  strongly over this period, averaging 10% growth annually. The primary source of transfer revenue is the
  provincial “equitable share”1 of nationally raised revenue. This is a constitutional entitlement and an
  unconditional transfer that makes up 81% of total revenues. It has grown strongly since its inception and is
  projected to continue rising as a proportion of total revenues over the medium term. The remainder of
  provincial revenues (14.7%) are made up by various conditional grants. Many of these are targeted towards
  supporting provincial capital expenditure (although the equitable share can also be used for this purpose).
  Major conditional grant programmes that are targeted to infrastructure amounted to R16,7 billion in
  2006/07 and account for an average of 6.2% of total transfers to provinces. The bulk of these resources are
  concentrated in the Infrastructure Grant for Provinces (IGP)2 and the Housing Subsidy, which together have
  made up and average of 79.3% of all conditional infrastructure transfers. These programmes are also the

    The provincial equitable share allocation formula consists of six components which capture the relative demand for services between provinces
  and take into account specific provincial circumstances. The components of the formula are neither indicative budgets nor guidelines as to how
  much should be spent on those functions. Rather, the education and health components are weighted broadly in line with expenditure patterns to
  provide an indication of relative need. Provinces have discretion regarding the determination of departmental allocations for each function, taking
  into account the priorities that underpin the division of revenue. The weights of the component are education (51 per cent) based on the size of the
  school-age population (ages 5-17) and the number of learners (Grade R to 12) enrolled in public ordinary schools, health (26 per cent) based on the
  proportion of the population with and without access to medical aid, a basic share (14 per cent) derived from each province’s share of the national
  population, an institutional component (5 per cent) divided equally between the provinces, an income poverty component (3 per cent) reinforcing
  the redistributive bias of the formula and an economic output component (1 per cent) based on GDP by region (GDP-R), although the latter is
  somewhat dated.
     The IGP is allocated to provinces based on their average proportion of the equitable share and a backlog component. Provinces allocate funds to
  sectors within broad national guidelines of education (40%), roads (40%), health (13%) and rural development (7%)

most rapidly expanding component of transfers to provinces, growing at an annual average of 26.7%
between 1999/00 and 2009/10, and are budgeted to account for 9.6% of all transfers in 2007/08.
On average, intergovernmental transfers account for 14% of total revenues of local governments. Again,
this figure has risen rapidly since 1999/00, but masks considerable variation between municipalities.
Transfers make up an average of 53% of the revenues of district municipalities, 31% of the revenues of local
municipalities, but only 12% for metropolitan municipalities, as the latter have the benefit of both property
taxes and, until recently, RSC levy revenues. Wide variations in grant dependence also exist within
categories of municipalities. The City of Johannesburg relies on transfers for less than 8% of its revenues,
while the City of Cape Town receives over 18% of its revenue from transfers. Infrastructure grant
programmes to local governments take the form of both cash and asset (“in kind”) transfers. On average,
these have accounted for 39.8% of all transfers to local governments and have grown at more or less the
same pace as other transfer programmes (on average 28.7% per year).
The rapid real growth in national transfers to provinces and local governments, and the associated increase
in conditional infrastructure transfers to provinces, occurred alongside the introduction of a new
generation of infrastructure grants. The IGP and MIG were introduced from 2002 as the flagship sub-
national infrastructure programmes. They replaced a set of sector focussed, project based programmes
that effectively earmarked resources to particular sectors, and required national approval of specific
projects in each sector. These programmes were typically poorly designed and highly bureaucratic, leading
to extensive delays in project approvals, and very limited real oversight of spending. They also displaced
accountability for investment decisions from sub-national to national government departments (who in
some cases were audited for individual project expenditures). The IGP and MIG were introduced to provide
greater transparency over the availability of resources (which were allocated over a three year horizon) as
well as greater discretion to sub-national governments on the allocation of funds between sectors and the
selection of projects. They were intended to act as supplements to the capital budgets of receiving entities;
thereby ensuring accountability for outcomes was placed firmly at that level. These reforms and the
associated growth in allocations have had three potentially countervailing effects:
Firstly, it has revealed capacity shortfalls. While these new instruments have directly and significantly
expanded the scope for sub-national capital spending this did not immediately translate into spending
increases. This was due to both shortfalls in spending capacity and the long lead times associated with
planning for infrastructure projects. More recent improvements in spending levels can be traced to the
streamlining of grant allocations. Greater predictability in allocations permits longer term planning. Greater
discretion removes unnecessary, fragmented and bureaucratic grant application, approval and reporting
requirements that absorb available sub-national capacity without doing much for spending outcomes.
However, the shift to this new approach has also revealed capacity constraints at sub-national level that
have had to be directly confronted. These issues are discussed in Section 4.

   Secondly, it has potentially weakened incentives for sub-national governments to seek funding from
   private capital markets. Grant increases in the context of limited spending capacity have resulted in local
   governments being less reliant on borrowing, which is in any event harder to arrange. Grants have typically
   not had matching requirements associated with them (which could be met through loan finance) and most
   local governments have postponed plans to take on additional borrowings.
   Thirdly, the increase in overall conditional infrastructure allocations has increased the scope for national
   government to intervene in infrastructure spending by sub-national governments. The MIG, in particular,
   still undertakes project–specific approvals of individual municipal projects and requires extensive reporting,
   regardless of the capacity of specific municipalities. Conditional infrastructure grants continue to exist
   outside of these two mechanisms, most of which tend to be focussed on managing inputs (finances, plans)
   rather than development outcomes, and contain no or few incentives for institutional strengthening
   (through requirements for grant matching, leveraging of private sector resources or community
   contributions). In all, little experimentation seems to have occurred on the role of these (or new) remaining
   conditional grants relative to the IGP and MIG.
iii.   Sub-national debt and equity partnerships

   Borrowing from private capital markets to finance infrastructure investment is generally regarded as an
   efficient way to finance capital programmes for reasons of the inherent lumpiness of capital spending,
   inter-generational equity and the governance benefits seen to result from having to negotiate funding from
   private parties. Debt finance needs to be repaid, so borrowing merely spreads the cost burden over time.
   The low level of provincial own revenues has, to date, precluded provinces from borrowing to finance
   capital expenditure. This has been due to concerns over debt costs: national government is able to raise
   debt more cheaply than provincial governments, and as national revenues (from grants) will ultimately fund
   provincial debt repayments this may result in rising public sector debt costs (assuming the efficiency of
   investments remains stable). Although provinces are now authorised to borrow none have done so to date.
   In contrast, local governments are permitted to borrow to finance capital expenditure, and have a long
   history of doing so. The municipal debt market has undergone a profound transformation since 1994.
   Government ended a policy of providing guarantees for municipal debt, removed prescribed asset
   requirements that ensured a steady interest in municipal debt, and put an end to a range of tax structuring
   packages and speculative borrowing practices that had become popular among lenders and municipalities,
   including foreign currency transactions. These steps forced lenders to evaluate municipalities on the
   strength of their own balance sheets, which effectively quashed investor appetite for any further lending
   until such time as balance sheets were both healthier and more transparent, and the rules of the game,
   particularly for debt defaults, more clearly articulated (see van Ryneveld, 2006). After considerable policy
   development, municipal borrowing is now regulated by the Municipal Finance Management Act, with the
   objective of re-establishing a vibrant market for municipal debt. Municipalities budgeted to fund over 17%
   of total capital spending, or R4,4 billion, from long term external lending in 2005/06. Metropolitan
   municipalities account for 73% of this scheduled lending, with the total external debt of metropolitan
   municipalities reported at R19,6 billion in December 2005 (National Treasury, 2005).
   External loans accounted for 3.9% (or R5.4 billion) of total budgeted local government income in 2006/07.
   External loans have risen marginally in nominal terms over the last few years, driven mainly by the City of
   Johannesburg’s debt restructuring (including bond issues refinancing existing debt). Cape Town has a
   comparatively low level of external debt, amounting to only R2 888 per household and 13.2 per cent of
   total annual revenues, as shown in Table 5. However, despite an investment grade credit rating it reports
   the highest average interest rates of the cities reviewed.

                                        Table 5: External debt levels as at 30 June 2007
                                        Total outstanding         No. of debt                  Debt per       External Debt as % of
                                          debt (R’000)           instruments                  Household          Total revenues
           City of Cape Town                2,247,534                22           13.5%         R 2,888               13.2%
           City of Johannesburg             6,150,132                40           12.5%         R 5,856               29.0%
           eThekwini Metro                 3,869,612                 10           10.9%         R 4,694               27.6%
          Source: NT database, StatsSA Census 2001

Most municipalities borrow from two specialist municipal lenders, the Infrastructure Finance Corporation
(INCA) and the Development Bank of Southern Africa (DBSA). Together they had an active municipal debt
portfolio of R12,4 billion in December 2005 (or 63% of total outstanding municipal debt), of which
metropolitan municipalities accounted for about 65% (National Treasury, 2005). Project-based financing
instruments are most commonly used by municipalities, although the recent bond issues by the City of
Johannesburg have highlighted the potential for the re-emergence of a municipal bond market.
Municipal debt has increased from 1997. The role played by the public sector has risen in relative
importance largely because of the expansion in lending by DBSA, which has become the single most
important provider of loan funds to municipalities. The relative importance of the private sector has
consequently declined, but has undergone important changes in its composition. Insurers and pension
funds have become relatively small providers of loan funds whereas banks have generally retained their
important role, particularly in project financing. INCA, which started operations in 1997, grew rapidly as a
source of funding to municipalities, partly through the swapping of the municipal investment portfolios of
insurers and pension funds for INCA bonds, indicating the important role that a well managed intermediary
of this kind can play in municipal finances.
The very important role that the issue of securities has played in the past as a means of attracting loan
capital has diminished to a level commensurate with annuity and other long-term loans. Short term loans
have however remained fairly stable during the four years, and seems to indicate a hard core of simply
revolving overdrafts which may be employed to finance capital expenditure.
It appears that municipal borrowing has begun to recover, without central guarantees, although slowly and
from a low base. The National Treasury (2005) reported a 3.3% increase in municipal lending between
March 2003 and 2004. It is not yet clear whether government will be able to declare a policy success in
attracting lenders back into the municipal debt market, despite the considerable emphasis this has received
in policy. This is partly due to limited effective demand from municipalities for debt finance3. Most
municipalities have been otherwise occupied with managing a difficult, prolonged transition from the
interim to final form of local government. The presentation of consolidated financial statements (across
formerly separate municipalities) and a strategic investment vision have taken time to emerge. The balance
sheets of secondary cities, in particular, have been weakened by the large rural hinterlands they have
absorbed. This has changed their underlying creditworthiness, and in some cases has required them to take
on large, poorly performing debt liabilities. While the re-demarcation process may have been a necessary
cost for local government transformation, it may be that the extent of municipal consolidation has
fundamentally altered the prospects for the municipal debt market in non-metropolitan areas.

  Demand side constraints to the municipal debt market include the complexities faced by municipalities in managing the local transition, capacity
constraints, the impact of demarcation, and poor financial and budget management. Supply side constraints refer to ineffective policy coordination
by central government, lack of appetite for risk by lenders and market distortions caused by DBSA.

                                          Main sources and real values of municipal loan funding
                           25 000

                           20 000
 R millions (2005 rands)

                           15 000

                           10 000

                            5 000
                                      1997      1998       1999        2000         2001         2002       2003       2004        2005
                                                                              Period (quarterly)
                           PIC   DBSA (Including LALF)   Local Auth & Public Enterp.     Insurers   Pension Funds   Banks   INCA     Others:

However, at least part of the constraints to a policy success may lie with poor policy coordination by
government itself, which is resulting in private players being crowded out of the municipal sector.
Firstly, the rapid growth in intergovernmental transfers has expanded the resources available to
municipalities for capital spending, which might otherwise have had to be borrowed. Many municipalities
may have difficulty in expanding capital spending beyond the pace at which they are currently doing so,
given their capacity constraints.
Secondly, the DBSA has continued to lend actively to creditworthy metropolitan municipalities and other
large cities – in some cases being reported to have outbid private sector lenders at the last moment with
offers of cheaper finance. The DBSA argues that it is able to do so due to its specialist focus, experience and
creditworthiness, and must continue to do so to allow it to cross-subsidize lending to weaker
municipalities. Private lenders have expressed dissatisfaction with what they regard as a subsidised public
agency providing competing unfairly against them, and acting in a predatory fashion by under-cutting at the
last moment bids that had taken considerable resources to prepare and negotiate. The National Treasury
(2003) reported that a higher proportion of DBSA lending was directed at metropolitan municipalities then
was the case with INCA, a private sector operator. This reflects in part INCA’s acquisition of the municipal
portfolios of Banks and insurance houses and more recently that of the PIC. This has created a situation
where the private sector is increasingly lending to less creditworthy municipalities then the DBSA
(regardless of whether it is originating or purchasing debt), which clearly cuts across governments’ policy
objectives of broadening municipal access to lending and reinvigorating the debt market. More recently,
this problem has been acknowledged by the DBSA, which has shifted its strategy to supporting market
development (DBSA, 2005), focussing both on lending to smaller municipalities and on supporting the
introduction of new lending instruments, such as through the provision of a partial credit guarantee for a
municipal bond issued by the City of Johannesburg.
Public Private Partnerships (PPPs) represent another mechanism that can be used to finance provincial and
local infrastructure investment. PPPs involve private parties performing provincial or local government
functions, including investing in associated infrastructure or acquiring the use of publically owned property.
Private parties are then compensated through collecting user fees or charges, or receiving a series of
payments from a province or municipality. Infrastructure related PPPs thus allow sub-national governments
to spread the burden of infrastructure financing over the life of an asset, and to share investment and

operational risk with private parties. Provincial PPPs have predominantly focussed on the health and
transport sectors. Besides a range of fleet management deals by provincial transport departments, the
major transport PPPs have been the Chapmans Peak toll road and Gautrain. The latter, which is the largest
deal to date, has capital value of R23 billion. There have been 6 hospital PPPs since 2001.
                                                 Table 6: Provincial PPPs
 Sector         Project           Province         PPP type;         Financing            Value to gvt   NPV of benefit
                                                   Contract                               (NPV of        to government
                                                   duration;                              unitary
                                                   Date of                                charge)
 TRANSPORT      Chapman’s         Western Cape     DF(part)BOT;      Debt: 44%                N/a        R 450 m
                Peak Drive toll                    30 years;         Equity: 10%                         in form of
                road                               May-03            Govt. Contrib: 46%                  capital works
                                                                                                         and operations
                Gautrain          Gauteng          DBFOT;            Capital Contrib:         N/A        Capital Value:
                Rapid Rail                         20 years;         87%                                 R23,09 bn
                Link                               Sep-06            Debt: 11%
                                                                     Equity: 2%
   HEALTH       Inkosi Albert     KZN              DFBOT;            Debt: 70%              R4.5 bn            N/a
                Luthuli                            15 years;         Equity: 20%
                Hospital                           Dec 01            Govt contrib: 10%
                Universitas       Free State       DFBOT;            Equity: 100%             N/a        R43 m cash plus
                and Pelonomi                       6,5 years;                                            R38 million in
                Hospitals co-                      Nov-02                                                form of
                location                                                                                 upgrade
                Humansdorp        Eastern Cape     DFBOT;            Equity: 90%            R18.9 m      R15 m upgrade
                District                           20 years;         Govt contrib: 10%                   plus R34 m cash
                Hospital                           Jun-03
                Western Cape      Western Cape     Facilities        Equity: 100%                        R 334 m
                Rehabilitation                     Management;
                Centre &                           Nov-06
                Polokwane         Limpopo          DBOT;             Equity: 100%          R88,35 m      N/a
                Hospital Renal                     10 years;
                Dialysis                           Dec-06
                Port Alfred &     Eastern Cape     DBFOT;            Debt:78%               R275m        Capital Value:
                Settlers                           17 years;         Equity:22%                          R168.6 m
                Hospital                           May-07
                                                                     Pure Equity:10%
                                                                     S/Holders Loan:90%
Source: PPP Unit, 2007

Municipal PPPs were initially focussed on long term asset leases in the water sector, but this focus has now
moved to management contracting for water services, and to solid waste management. Little information
on municipal PPPs is available since the closure of the Municipal Infrastructure Investment Unit (MIIU).
However, anecdotal evidence suggests a decline in the use of PPPs by municipalities due to the procedural
and technical complexities of negotiating them under the current legislative framework and declining
investor appetite in the global market.
Other forms of partnerships, such as public-community partnerships (which typically involve donations,
development charges or sweat equity by citizens or service users) are used by local governments, but are
not reviewed here.
3.3    Institutional and fiscal challenges
Three sets of challenges in current financing arrangements for sub-national infrastructure have become
evident since 2000.

a) Management and funding of the built environment

Functional concurrency and overlaps between provincial and local government in the areas of housing,
public transport land use and urban infrastructure functions has undermined effective management of
urban built environments and increased the costs of public sector investments. In particular:

•         A lack of policy coordination between investment programmes results in different agencies
          externalising the long term costs of their decisions to other entities. For example, investment in
          housing stock for poor communities on the urban periphery creates long term demands for transport
          subsidies as residents must commute to seek income opportunities. Low density housing may also
          raise the costs of infrastructure programmes, or not make efficient use of existing infrastructure stock
•         A lack of operational coordination between programmes can result in extensive delays in the
          completion of new investments. For example, new housing projects commissioned by the provincial
          government require both land use approvals from provincial and local governments as well as
          infrastructure investments by municipalities.
b) Intergovernmental issues

      Significant reforms have been introduced to the instruments of intergovernmental fiscal relations.
      These have brought greater equity, transparency and predictability to the fiscal relationship between
      spheres of government. This has laid the foundations for enhanced infrastructure investment through
      accommodating the longer lead times associated with capital projects, reducing bureaucracy and
      strengthening accountability for spending outcomes within spheres of government.

      However, three issues have emerged within this new framework for grant financing:

    i.       The introduction of the IGP and MIG has not been able to entirely address problems arising from
             the fragmentation of transfers. Large infrastructure transfer programmes remain outside of these
             two instruments. This is likely continue for the foreseeable future, as national government seeks to
             priorities certain sectors and aspects of sub-national infrastructure investment (from hospital
             revitalisation to stadium construction), and is not necessarily a problem in itself. However, this
             expands the challenges of coordinating investments between programmes across spheres, such as
             provincial housing and municipal infrastructure, to reduce the extent to which sectors are able to
             externalise costs of their investment decisions. Underlying this is the problem of the assignment of
             functions between spheres, particularly in the built environment.

    ii.      The IGP and particular MIG have not fully lived up to their potential of replacing project-based input
             controls on sub-national spending with output and outcome management of sub-national capital
             investments. While the relationship between plans and budgets has been strengthened, national
             monitoring remains fragmented. Grant programmes tend to focus their monitoring on projects
             they support, rather than on the entire capital budget of provinces or municipalities. In the case of
             MIG, individual projects are still effectively approved (by dplg) before spending can commence,
             while municipal budgets are evaluated (by the National Treasury) without specific reference to
             capital spending outcomes. This fragmentation in oversight is compounded by the absence of any
             rigorous evaluation of programme performance at a national level, and challenges experienced in
             strengthening mechanisms of community oversight of spending outcomes at sub-national level.

 iii.        The current framework for infrastructure grants does not fully address strategic priorities or
             emerging blockages in sub-national infrastructure investment. These include the need to leverage
             non-state contributions or ensure the release of strategic urban land parcels. The absence of
             requirements for grant matching (other than by beneficiaries in the housing programme) or
             authority to securitize loans against grant flows potentially reduces the incentives and capacity of
             sub-national entities to borrow. Input based reporting requirements (on receipts, expenditures and
             projects) inhibit the capacity of communities to have oversight of the performance of sub-national
             entities about changes in their quality of life that arise from investments. These issues point to
             potential design innovations required within existing instruments.

c) Own revenue issues

 The RSC levies on the turnover and payroll of businesses was a buoyant and productive source of revenue
for metropolitan municipalities in particular (although was highly imperfect from a national tax policy
perspective). It was intended to fund infrastructure investment by municipalities, either directly or through
being pledged as security for loans. In reality it was often used as a general revenue source.

Its withdrawal and replacement by a grant raises the prospect of erosion in both the fiscal autonomy of
cities and their ability to self finance capital expenditure over the longer term. Although national
government has made no attempt to impose conditions on the transfer of funds, and has allocated it in
accordance with historical collections by individual municipalities, the quantum of funds made available has
increased at a slower rate than actual collections are likely to have done, given past trends.

Until such time as a suitable own-revenue replacement for the RSC levies is found, the capacity of cities to
finance extensive infrastructure investment will be weakened, and their long term fiscal sustainability will
be in question. The National Treasury has published a discussion document on this matter, and the South
African Cities Network will shortly respond, so it remains likely that an alternative own revenue source(s)
will be introduced over the medium term. Options under consideration include, amongst others, a local
business tax, or revenue sharing arrangements on national taxes such at VAT and the fuel levy. The latter
now appears to be the preferred option.

                                      Table 7: Current revenues of national, provincial and local government (2002/03 to 2007/08)
               R millions                  2002/03           2003/04                    2004/05                      2005/06                 2006/07                  2007/08              AVERAGE
                                           Amount        Amount    Change           Amount    Change             Amount    Change        Amount    Change         Amount    Change

               Local                        65,408.9      75,071.6      15%          84,095.4      12%           95,101.4     13%       108,381.8      14%       128,335.7       18%
               Provincial                    5,624.0         6,132.7     9%           6,238.7       2%            7,379.1      18%         7,954.0      8%          7,673.1      -4%
               National                   278,507.7      299,431.2       8%        347,854.4       16%          411,747.9      18%      475,835.6      16%       544,601.6       14%
               Total                      349,540.6      380,635.5       9%        438,188.5       15%          514,228.4      17%      592,171.4      15%       680,610.4       15%

               Percentages of total
               Local                        18.7%            19.7%                   19.2%                        18.5%                   18.3%                    18.9%                      18.9%
               Provincial                   1.6%             1.6%                    1.4%                         1.4%                    1.3%                      1.1%                       1.4%
               National                     79.7%            78.7%                   79.4%                        80.1%                   80.4%                    80.0%                      79.7%

               Revenues as % GDP
               LG % GDP                                      6.0%                    6.1%                         6.2%                    6.2%                      6.6%                       6.2%
               Provincial % GDP                              0.5%                    0.4%                         0.5%                    0.5%                      0.4%                       0.5%
               Nat % GDP                                  23.8%                       25.1%                       27.0%                     27.1%                     28.1%                     26.2%
Source: Estimates of National Revenue 2004, 2005, 2006, 2007 (National Treasury), IGFR (National Treasury, 2003) Provincial Budgets and Expenditure Review (National Treasury, 2007) and LG Budget
database (National Treasury). All figures are nominal.

                                            Table 8: Nominal Vertical Division of Nationally Raised Revenue (2002/03 - 2007/08)
                                            R millions                              Actual       Actual      Actual         Revised    Budget        Budget      Average
                                                                                   2002/03      2003/04     2004/05         2005/06    2006/07       2007/08
                                            National government                    129 297      148 142     168 018         196 429    214 964       233 996
                                            Provincial government                  107 317      122 673     137 836         154 528    176 679       196 351
                                            Local                                     8 102      11 581         13 837       16 859     26 532         30 503
                                            Total                                   244716       282396         319691      367816      418175        460850
                                            National share (%)                        52.8         52.5           52.6        53.4        51.4          50.8          52.2
                                            Provincial (%)                             43.9         43.4          43.1         42.0        42.3          42.6         42.9
                                            Local (%)                                    3.3         4.1           4.3          4.6         6.3           6.6          4.9
                                            Total                                     100,0        100,0         100,0        100,0       100,0         100,0
                                            Growth in National share (%)                          14.6%         13.4%        16.9%        9.4%          8.9%       12.6%
                                            Growth in Provincial share (%)                        14.3%         12.4%        12.1%       14.3%         11.1%       12.8%
                                            Growth in Local share (%)                             42.9%         19.5%        21.8%       57.4%         15.0%       31.3%

                                                                      Table 9: Transfers to Provincial and Local Government
                                          1999/00       2000/01       2001/02        2002/03       2003/04       2004/05       2005/06       2006/07       2007/08        2008/09          2009/10       Average
                                                                                                                                              Prelim.                                                     over
                                                        Actual                      Est. actual                  Outcome                     outcome               Medium-term estimates                 period
Provincial Transfers                       103,615       113,754        126,340       142,560       128,805       144,075       161,161        186,115       210,438         237,397        262,974
of which
Equitable share                             89,095        98,398        107,460       123,457       107,538       120,885       135,292        150,753       171,271         193,474        215,784
Conditional grants                          10,482        10,825         13,938         13,479       15,134        16,951        18,490         27,408        31,494          35,822         38,660
… Infra Grant for Provinces (IGP)                   -             -             -              -      2,534         3,348         3,731          4,983         6,164           6,847          7,997
… Housing Subsidy                            2,571         2,945          3,132          3,731        4,355         4,589         4,868          6,404         8,238           9,853         11,531
… Hospital Revitalisation                        140         323           550             694           810          912         1,256          1,527         1,907           2,283          2,582
… Further Education & Training Recap.               -             -             -              -             -             -             -           470           595            795                -
… Mass Sport & Recreation                           -             -             -              -             -             9           24            119           194            290           402
… Gautrain                                          -             -             -              -             -             -             -       3,241         3,029           3,266          2,507
… Other                                      7,771         7,557         10,256          9,054        7,435         8,093         8,611         10,664        11,367          12,488         13,641
Subtotal: Selected Infra Grants              2,711         3,268          3,682          4,425        7,699         8,858         9,879         16,744        20,127          23,334         25,019
… as % of total                           2.6%          2.9%           2.9%           3.1%         6.0%          6.1%          6.1%          9.0%          9.6%           9.8%             9.5%           6.2%
% growth in provincial transfers                        9.8%          11.1%          12.8%         -9.6%         11.9%         11.9%         15.5%         13.1%          12.8%            10.8%         10.0%
% growth in Selected Infra Grants                       20.5%         12.7%          20.2%         74.0%         15.1%         11.5%         69.5%         20.2%          15.9%            7.2%          26.7%

Local Govt Transfers                         4,610         5,536          6,516          9,500       13,255        15,544        18,654         28,465        36,278          44,326         48,372
(a) of which Direct (cash)                   4,610         5,536          6,516          8,104       11,582        13,837        16,821         27,079        34,312          41,844         45,608
     Equitable share (including IGGs)        1,673         1,867          3,184          4,187        6,350         7,678         9,643         18,058        20,676          23,775         29,444
    Conditional grants                       2,937         3,669          3,332          3,917        5,232         6,159         7,178          9,021        13,636          18,069         16,164
    … CMIP (incl MIP, EMIP, BCIG)                696         851           996           1,723        2,285
    … Implement Water Services Projects          626         725           757             999        1,022           217              139
    … Community Based Public Works               211         209           357             260           262
    … Building for Sports & Recreation                                        36             84          122          134
    … LED Fund                                    10            78            87           143           117
    … Urban Transport Fund                        30            22            38             40              9
    … National Elec Programme                                                              225           245          251              297           355           468            596           897
    … Municipal Infra Grant (MIG)                                                                          41       4,481         5,436          6,756         7,549           8,053          9,130
    … Public Transport Infra. & Systems                                                                                                242           170       1,174           3,170          2,325
    … 2010 Stadiums                                                                                                                                  600       2,700           3,800          1,300
    … Neighbourhood Devt Partnership                                                                                                                 50            500         1,500          1,650
    … Other                                  1,364         1,784          1,061            443        1,129         1,076         1,064          1,090         1,245              950           862

 Subtotal: Selected Infra Grants                1,573        1,885         2,271         3,474       4,103       5,083        6,114         7,931        12,391       17,119      15,302
 (b) of which Indirect Infrastructure                 -             -             -      1,396       1,673       1,707        1,833         1,386         1,966        2,482       2,764
      … Bulk Water                                                                                                                                          300          450         450
      … Elec Backlogs clinics & Schools                                                                                                                      45           90         150
      … Water Backlogs clinics & schools                                                                                                                    105          210         350
      … Financial Management Grant                                                                       60         69             66            53          53           50
      … CWSS Operating                                                                     656          817        819            904           440         490          531         393
      … National Elec Programme (Eskom)                                                    740          796        819            863           893         973        1,151       1,421
       Subtotal: Selected Infra Grants &
 Indirect Transfers                           1,573         1,885         2,271         4,870       5,776       6,790      7,947         9,317        14,357       19,601        18,066
      .. as % of total                     34.1%           34.0%         34.9%         51.3%       43.6%       43.7%      42.6%         32.7%         39.6%        44.2%         37.3%      39.8%
 % growth in all LG transfers                             20.1%         17.7%         45.8%       39.5%       17.3%       20.0%         52.6%         27.4%        22.2%        9.1%        27.2%
 % growth in selected infra grants                        19.8%         20.5%         83.7%       18.4%       20.2%       18.4%         22.7%         55.1%        37.3%        -9.1%       28.7%

 Combined Prov & Local Transfers              108,225      119,290       132,856       152,060     142,060     159,619     179,815        214,580       246,716      281,723     311,346
 Total Selected Infrastructure Transfers      4,284           5,153         5,953         9,295      13,475      15,648      17,826        26,061         34,484       42,935      43,085
 … as % of all transfers                   4.0%           4.3%          4.5%          6.1%        9.5%         9.8%       9.9%          12.1%         14.0%        15.2%        13.8%        9.4%
 % growth in all LG transfers                             10.2%         11.4%         14.5%       -6.6%       12.4%       12.7%         19.3%         15.0%        14.2%        10.5%       11.4%
 % growth in selected infra grants                        20.3%         15.5%         56.1%       45.0%       16.1%       13.9%         46.2%         32.3%        24.5%         0.3%       27.0%
Sources: NT LG database, PBER 06; IGFR 2003

4     Capital expenditure programmes of provincial and local government
This section will review trends in budgeted and actual expenditure by provincial and local governments
across sectors. Where it identifies demand side constraints to capital expenditure by provinces and local

4.1     Trends in capital spending by provinces and municipalities
The consolidated capital expenditures of provinces and local governments, including housing transfers) was
over R51 billion in 2006/07. It has grown at over 17% per year (and has thus almost doubled to from
2002/03), with more rapid increases in recent years. However, it still accounts for only an average of 14% of
total spending by these spheres. This share of total sub-national spending is expanding at around 7% per
year. This reflects the robust growth in operational spending at provincial and local level. Of the two
spheres, local government has a marginally greater share of capital spending, amounting to an average of
54.2% since 2002/03. This contrasts with local governments receiving a lower proportion of national
infrastructure transfers, as discussed in Section 3.

Transport (predominantly roads) and housing dominate the consolidated capital spending of these two
spheres, accounting for an average of 21% each over the last three years to 2006/07. However, Health and
Electrification spending have seen the most dramatic increases, averaging 32.3% and 30.7% respectively
over the same period.

a) Trends in Provincial Capital Payments

Provincial capital payments4 have grown at 16.9% a year since 2002/03, and now accounts for 11.9% of
total provincial expenditure (2006/07). The strongest growth has occurred in health (25% p.a.) and housing

  For the purposes of this paper provincial capital payments include housing transfers which are technically regarded as transfers to households,
and thus not counted as capital expenditure.

(15.3% p.a.) Together with transport, these make up the majority of capital payments. Health, like
education, is not typically capital intensive (averaging only 7.5% of total sector spending), so this growth is
all the more remarkable. Education capital payments average only half of this (4.2%), reflecting the lower
effective priority accorded to capital spending in the education sector, and perhaps also the absence of a
dedicated capital grant from national government. Average growth in transport capital payments are only
4.4%, reflecting a real decline in capital spending since 2003/04.

                 Table 10: Capital & Housing payments by provinces (2003/04 to 2009/10)
 R’millions                       2003/04        2004/05        2005/06       2006/07   Average   2007/08    2008/09   2009/10
 Health                             2,428          2,693          3,844         4,685               5,496      5,971     6,662
 % change                                           10.9%         42.7%         21.9%    25.2%       17.3%      8.6%     11.6%
 as % of total sector exp               6.6%           6.6%        8.2%          8.7%      7.5%       9.3%      9.2%      9.4%
 as % of total Capital                17.4%         18.3%         21.6%         21.2%    19.6%       21.0%     20.1%     19.6%

 Education                            2,572         2,500          2,808        3,724                3,845     4,455     5,186
 % change                                           -2.8%         12.3%         32.6%    14.0%        3.3%     15.9%     16.4%
 as % of total sector exp               4.3%           3.9%        3.9%          4.7%      4.2%       4.3%      4.5%      4.8%
 as % of total Capital                18.4%         17.0%         15.8%         16.8%    17.0%       14.7%     15.0%     15.3%
 Provincial Transport                 3,714         4,016          4,847        5,758                6,949     7,749     8,904
 % change                                              0.1%        9.0%          4.1%      4.4%       9.2%      0.1%      5.6%
 as % of total sector exp             55.6%         55.6%         60.6%         63.0%    58.7%       68.9%     68.9%     72.7%
 as % of total Capital                26.6%         27.3%         27.2%         26.0%    26.8%       26.6%     26.1%     26.2%

 Housing                              4,331         4,520          5,052        6,557                8,177     9,715    11,383
 % change                                              4.4%       11.8%         29.8%    15.3%       24.7%     18.8%     17.2%
 as % of total other sector exp       61.5%         62.5%         63.8%         66.5%    63.6%       68.2%     70.3%     72.0%
 as % of total Capital                31.0%         30.7%         28.4%         29.6%    29.9%       31.3%     32.8%     33.6%

 Other                                  922              972       1,264        1,406                1,651     1,756     1,793
 % change                                              5.4%       30.1%         11.2%    15.5%       17.4%      6.4%      2.1%
 as % of total other sector exp         4.1%           4.1%        4.8%          4.5%      4.4%       4.3%      3.9%      3.6%
 as % of total Capital                  6.6%           6.6%        7.1%          6.4%      6.7%       6.3%      5.9%      5.3%

 Total Capital Payments            13,967          14,701         17,815       22,130               26,119    29,647    33,928
 % change                                              5.3%       21.2%         24.2%    16.9%       18.0%     13.5%     14.4%
 as % of total exp                    10.6%         10.4%         11.1%         11.9%    11.0%       12.4%     12.6%     13.2%
Source: PBER 2007

Of interest is the divergence in the composition of provincial capital payments from the guidelines provided
in the IGP. When housing transfers are excluded, health spending has consistently exceeded IGP guidelines,
due to the effect of the Hospital Revitalisation Grant, while education and roads expenditures have
consistently been de-prioritised. This situation is projected to be fairly stable over the medium term.

              Table 11: Annual allocation of provincial capital payments (excluding housing transfers)
                                              relative to IGP guidelines
                              2003/04          2004/05         2005/06        2006/07   Average   2007/08    2008/09    2009/10

 Health            13%        25.2%            26.5%           30.1%          30.1%      28.0%    30.6%      30.0%      29.6%
 Education         40%        26.7%            24.6%           22.0%          23.9%      24.3%    21.4%      22.4%      23.0%
 Transport      40%           38.5%            39.4%           38.0%          37.0%      38.2%    38.7%      38.9%      39.5%
Source: PBER 2007

b) Trends in Local Government Capital Expenditure

Local government capital expenditure has grown by 21.7% per year since 2002/03 and now accounts for
21.2% of total local expenditure. While over the longer term total municipal expenditures have increased in
real terms by an annual average of 17.4% (between 1996/97 and 2004/05 in 2005 rands), the early part of
this growth was largely driven by rising personnel costs, with capital spending remaining stagnant. Since
2002/03, capital spending has shown real annual average growth of over 21% since 2002/03, compared to
22.4% for operating expenditures.

A significant proportion of this capital spending is financed by infrastructure grants (56% and 71% for local
and district municipalities respectively, while only 28% for metropolitan municipalities), particularly the
newly introduced Municipal Infrastructure Grant (MIG). Although no doubt helped by improved reporting
and grant management systems, this trend represents a significant policy success for government in re-
focussing municipalities on their service delivery functions after a prolonged period of restructuring.

The strongest growth has been in District Municipalities (43.9% p.a.), but this is off a very low base and
reflects functional reassignments from Category B (Local) Municipalities, along with associated grant
funding, in certain provinces. It should not be confused with actual spending outcomes. The most
significant growth has been in metropolitan municipalities (17.5% p.a.) and secondary cities (24.3% p.a.)
who collectively account for the bulk of local capital spending. Other local municipalities have also
experienced strong growth, with capital spending accounting for an average of over 25% of their total

Sectorally, the greatest increases in capital spending have focussed on housing (35.5% p.a.) and
electrification (30.7% p.a.), although these make up smaller proportions of total capital spending than
water services and transport. Unlike provincial transport, local transport investments have experienced
strong growth.

                         Table 12: Local government capital expenditure by sector
                 R’000                 2004/05          2005/06          2006/07           Average
                 Local Transport          2 829 052        3 796 075        4 456 676
                 % change                                      34.2%           17.4%             25.8%
                 as % of total Capex         16.3%             16.0%           15.4%             15.9%
                 as % of total exp               2.8%             3.2%             3.3%              3.1%

                 Water & Sanitation       2 778 887        3 863 203        4 379 381
                 % change                                      39.0%           13.4%             26.2%
                 as % of total Capex         16.0%             16.3%           15.1%             15.8%
                 as % of total exp               2.7%             3.3%             3.2%              3.1%

                 Elecrification           1 848 198        2 420 740        3 157 313
                 % change                                      31.0%           30.4%             30.7%
                 as % of total Capex         10.6%             10.2%           10.9%             10.6%
                 as % of total exp               1.8%             2.1%             2.3%              2.1%

                 Housing                  2 319 598        4 015 697        3 930 162
                 % change                                      73.1%               -2.1%         35.5%
                 as % of total Capex         13.3%             17.0%           13.6%             14.6%
                 as % of total exp               2.3%             3.4%             2.9%              2.9%

                 Other                    7 609 306        9 572 515       13 034 235
                 % change                                      25.8%           36.2%             31.0%
                 as % of total Capex         43.8%             40.4%           45.0%             43.1%
                 as % of total exp               7.5%             8.1%             9.6%              8.4%

                 Total Capital Exp.      17 385 041       23 668 230       28 957 767
                 % change                                      36.1%           22.3%             29.2%
                 as % of total exp           17.2%             20.0%           21.3%             19.5%

                 Total Municipal Exp    101 096 201      118 058 442      136 264 218
                 % change                                      16.8%           15.4%             16.1%

4.2   Capital spending and service delivery

General data on progress with the elimination of backlogs by sector is provided in Section 1 (Table 1).

a) Progress with delivery at Provincial level

Limited data is available on physical progress with delivery of provincial infrastructure by sector. Most data
focuses on financial progress.

The Department of Education reports that the number of overcrowded schools has fallen from 51% in 1999
to 24% in 2006. Some 74% of schools are reported to be in a good or excellent condition. The number of
schools with electricity has increased from 11 174 in 1996 to 20 713 in 2006. The number of schools
without water has fallen from 8 823 in 1996 to 3 152 in 2006. The number of schools without on-site toilets
has fallen from 3 265 in 1996 to 1 532 in 2006. In addition, and impacting negatively on the quality of
education, only seven percent of schools have adequate libraries, and 10% of secondary schools have
functioning laboratories. 68% of schools have no computers

The housing sector has been most forthright in its estimations of progress and problems. Over the last 10
years, state-assisted housing investment of some R29,5 billion has provided 1.6 million housing
opportunities (at an implied average cost of R18,437 per unit) and has allowed 500,000 families the
opportunity to secure titles of old public housing stock. Housing subsidy grants increased from R2,7 billion
in 1996/97 to some R5 billion in 2006/07 and are budgeted to continue increasing strongly over the
medium term. These increases in housing development funding have largely gone towards funding the
increases in the quantum of the housing subsidy, which is now adjusted annually for inflation. The annual
delivery rate of housing units peaked in 1997 at some 323 000 units. The past few financial years have seen
a decline in the actual numbers of houses produced annually. Since 2000/01, annual rollovers have
increased to above 10% of the voted amounts (a 5% to 10% rollover is considered acceptable for major
capital programmes). Table 13 shows the rollovers for 2000/2001 to the 2003/04 financial years.

        Table 13: Housing rollovers as percentage of allocation
                                                                     2000/01   2001/02   2002/03   2003/04
      Total allocation of SA Housing Subsidy Grant (R’m)               2 998     3 226     3 801     5 245

      Total Provincial rollovers on SA Housing Subsidy grant (R'm)      519       458       885       560

      Percentage of total allocation                                    17%       14%       23%       11%
        Source: Department of Housing, Breaking New Ground

The slowdown in delivery and the under-expenditure of provincial budgets has been attributed to: (i) the
withdrawal of large construction companies from the sector and insufficient capacity among emerging
contractors to fill the gaps this has created in project and financial management, and subsidy
administration; (ii) Difficulties in securing private sector contributions in the low and medium income
housing market; (iii) Difficulties in obtaining the release of strategic, often state-owned land for
development; and (iv) Policy shifts, capacity constraints bureaucratic procedures that disrupt or delay
housing delivery. For example, national policy and provincial allocations have not always been able to
respond to the changing nature of housing demand arising from urbanisation.

In particular, the lack of affordable well located land for low cost housing, which has resulted in the housing
programme largely extending existing areas, often on the urban periphery and achieving limited
integration. Post-1994 extensions to settlements have generally lacked the qualities necessary to enable a
decent quality of life through becoming “valuable assets” in the hands of the poor. In addition to this the
inability of recipients of subsidy-housing to pay for municipal services and taxes has meant that such
housing projects have been viewed as liabilities to municipalities and have not assisted many of the
country’s major cities struggling to come to grips with rapid changes to economic conditions since South
Africa’s inclusion into the global economy. This is attributed largely to the lack of funding and poor
alignment of budgets and priorities between line function departments and municipalities responsible for
providing social facilities in new communities.

b) Progress with delivery at the Local Government level

The MIG is the primary instrument for addressing infrastructure backlogs at municipal level. To September
2006, and since its inception in April 2004, this programme has provided over R26 billion to municipalities
for infrastructure investment, supporting a total of 6195 projects, of which 2168 had been completed. The
grant directly funded 81.3% of project costs, with 17.4% being from municipal contributions and 1.3% from
municipal borrowings. The majority of expenditure has been on water (47%), sanitation (26%) and roads
(22%) infrastructure.

The MIG programme reported that in September 2006 it had provided 769,107 households with access to
water, 300,794 with access to sanitation and 512,000 with access to roads & stormwater. This would imply
an average cost per household for a full set these services at a basic level of R45,858.

                       Table 14: Average cost per household for MIG residential
                              infrastructure projects to September 2006
                                    Total expenditure      Total HH       Average Cost   MIG Guideline
                                       (R millions)        serviced           / HH         cost / HH
               Water                       R10,784.14          769,107        R 14,022            R 6,000
               Sanitation                   R6,145.88          300,794        R 20,432            R 3,650
               Roads & Stormwater              R5,839.1        512,009        R 11,404
               Total                       R22,769.12                         R 45,858

The major challenge facing municipal infrastructure investment is the slowdown in progress with
eliminating backlogs. From 2004, an annual reduction in backlogs of 4.5% for water supply and 5.3% for
sanitation was required in order to meet the targets of eliminating backlogs by 2008 and 2010 respectively.
However, rates of 2.6% for water supply and 2.7% for sanitation were achieved. As demonstrated in Table
15, it now seems highly unlikely that backlog targets will be met in their specified timeframes.

        Table 15: Rate of reduction in backlogs for local government services and housing
                                         %           %
                                      change      change        Backlog     Required      Current
                                       1994-       2001-         target      annual       annual
                                       2001        2004           year      reduction    reduction
                   Water              -0.50%      -9.50%          2008        4.50%       2.60%
                   Sanitation         -8.30%      -9.70%          2010        5.30%       2.70%
                   Electricity        -3.40%      -2.00%          2011        4.00%       1.90%
                  Housing (units)     -13.40%                     2014
               Sources: DWAF, DME, Housing

This slowdown may be explained by four factors:

a) Remaining under-served areas are more costly and difficult to reach, requiring innovative political and
   technological options to be explored.

b) Connection costs per household have risen. Table 14 demonstrates the extent to which average costs
   for water and sanitation far exceed official guidelines. This reduces the resources available to address
   the servicing needs of other households.

c) Weak maintenance practices have led to a rise in the number of network failures, particularly with
   older infrastructure. This can result in resources being diverted from planned new connections, or can
   result in serviced households losing access to services. For example, a recent audit of water services
   projects funded by the Municipal Infrastructure Grant (DWAF, 2007) found that approximately 40% of
   bulk water projects completed since 2003 had not been operational at some stage in the past year,
   with 13.4% having broken down three or more times; 9% of breakdowns took more than a month to
   repair. 21% of electrical pumps and 14% of diesel pumps were out of order, and a further 57% of diesel
   pumps lacked fuel. Technical problems reported include water pumps that were non-operational for
   various reasons, pipe leakages, illegal connections to pipelines and reservoirs that were non-
   operational. In terms of household supply, more than 16% of street and yard taps were not working,
   with a similar percentage damaged. The same study found that 16% of recently served people
   surveyed have now reverted to unprotected sources, generally more than 200m away.

d) Priorities for investment have shifted, most notably to the focus on eliminating the bucket system. In
   2005 a backlog study was undertaken based on the 2001 StatsSA census data and indicated that there
   were 252 254 bucket systems in formal and informal settlements. In 2005 an amount R1, 2 billion was
   allocated through MIG over the medium term. A further R400 million was made available for the
    2007/8 financial years specifically targeting the bucket eradication project. In July 2006 it was found
    that there were 165 192 buckets in use in formal areas. Between June 2006 and July 2007, 56 057
    buckets were removed in formal settlements and replaced with a basic sanitation service. The total
    number of buckets that were eradicated at the end of June 2007 stood at 109 855. Subsequently,
    buckets that do not “qualify” for removal have been subtracted from the estimates of the backlog by
    DWAF, reducing the figure to 49000 units.

Electrification is not funded through the MIG, due to uncertainties in the future of this function in
municipalities. In December 2000, the NER reported that a total of 3.14 million electricity connection had
been installed since 2000, 34% by local governments. The average cost per connection over the period was
reported to be R2871, with costs declining over the period. A total of almost R9 billion had been spent
(averaging R1,3 billion annually). The remaining backlog in December 2000 was estimated to be almost 4
million units, or 37% of households. The majority (58%) of this backlog were located in rural areas. To this
must be added the backlog on the electrification of schools and clinics, which was estimated at
approximately 17 000 schools and 1 000 clinics, distributed across most of rural South Africa. By May 2005,
the INEP had delivered connections to 232 287 households (at a cost of R582-million), 2 233 school
connections (R100-million), and 50 clinic connections (R118-million), though the majority of these were
likely installed by Eskom rather than municipalities. The target for the eradication of backlogs is 2011,
which required a 4% annual reduction in the backlog from 2004, but a rate of only 1.9% per year has been

Trends in the provision of municipal infrastructure to support economic activity are difficult to measure.
Correlating data on building plans passed and completed with municipal capital expenditure over time
provides a picture of the extent to which municipal investment is able to keep pace with that of private
fixed investment. The chart below demonstrates that that total capital expenditure of municipalities
exceeded or equalled the total value of buildings completed until 2003, but has since sunk significantly
below these levels. Of additional concern is that the value of plans passed has accelerated during the same

4.3   Job creation and public works
The MIG programmes reports that it has provided employment to 84,533 people, for an average length of
240 days per person or R300,000 of total expenditure per person. 464 MIG projects are EPWP compliant,
with a value of R2,76 billion.

4.4   Maintenance Expenditure

Investment in new infrastructure requires ongoing expenditure on maintenance. Annual maintenance
spending should amount to 4% of total asset values, according to the CSIR (quoted in SAICE, 2007).
Quantifying required and actual maintenance expenditure at sub-national level is a difficult exercise. Asset
registers are mostly incomplete (progress is being made in this regard) and accounting standards do not
capture maintenance expenditure, particularly in provinces.

Notwithstanding, Section 1 (Table 3) estimated the replacement costs of sub-national infrastructure assets
at R800 billion, with an estimated total required maintenance expenditure of R32 billion annually.
Provinces account for over 60% of installed assets, and should spend R19 billion annually on maintenance.
Informal estimates of maintenance spending by Provinces in 2006/07 (National Treasury) amounted to R4,5
billion (or 24% of total capital payments). Of this, 66% was spent on road maintenance and only 8% and 6%
in education and health respectively. The Department of Education reports that substantial problems still
remain in relation to the condition and standard of education facilities, due to the effects of vandalism,
neglect and inadequate maintenance. 14% of existing facilities were found to be in a poor condition and
12% very poor. In the Eastern Cape 40% of schools are assessed as being in a poor condition.

Municipalities account for R325 billion (40%) of existing assets, and should spend R13 billion annually on
maintenance. Table 16, below, demonstrates that they spent only R7.1 billion for this purpose in 2006/07.
While maintenance spending grew at an average of 10.5% over the last three years, this is happening at a
slower pace than other forms of spending (particularly capital investment). It is thus declining as a
proportion of total local government expenditure. This trend is evident across all categories of

             Table 16: Repairs & Maintenance Expenditure by Category of Municipality
          R’000                                          2004/05     2005/06     2006/07     Average
          Metro Muni’s                                   3 778 663   4 223 282   4 440 279
          % change                                                      11.8%        5.1%        8.5%
          % of total Metro Exp                               6.5%        6.2%        5.7%        6.2%

          Secondary cities                                 889 314     990 912   1 180 592
          % change                                                      11.4%       19.1%       15.3%
          % of total City Exp                                5.4%        5.3%        5.3%        5.3%

          Other Local Muni’s                               919 882     992 176   1 180 912
          % change                                                       7.9%       19.0%       13.4%
          % of total Other LG Exp                            5.2%        4.6%        4.5%        4.7%

          District Municipalities                          279 087     306 879     359 975
          % change                                                      10.0%       17.3%       13.6%
          % of total District Exp                            3.1%        3.1%        3.4%        3.2%

          Total                                          5 866 946   6 513 248   7 161 759
          % change                                                      11.0%       10.0%       10.5%
          % of total LG Exp                                  5.8%        5.5%        5.3%        5.5%

The electricity regulator, NERSA (2007) that maintenance spending backlogs in just 11 audited electricity
distributors amounted to R431 million in 2005, based on estimates of the age of distribution infrastructure.

Annual maintenance requirements for these distributors alone was estimated to amount to R422 million a
year. Spending on asset maintenance and refurbishment was identified as insufficient, partially as a result
from the very poor quality of data on assets.

NERSA now requires that municipalities spend at least 5% of income from electricity sales on repairs and
maintenance. This, together with a great focus on quality of service in the industry, has resulted in
significant increases in maintenance starting to be reported. Cape Town reports spending R185 million a
year on electricity maintenance. Johannesburg is allocating R1 billion per year for asset refurbishment and

But it is clear that the metropolitan municipalities and larger secondary cities have delayed or postponed
asset maintenance for some time, as a result of the local government transition and associated financial
pressure. This is likely to be having a negative impact on the quality of services. The Auditor-General (2004)
reported a 13% annual increase in water reticulation losses in the top 50 municipalities in 2003/04.

4.5     Underlying issues in capital and maintenance expenditures
There are three underlying issues affecting the expenditure performance of sub-national governments with
respect to infrastructure investment:

a) The capacity of sub-national governments to spend the resources available to them has been a
   perennial concern of government since a more expansionary fiscal stance was adopted from 2002.
   Following the introduction of the IDIP across all provinces spending has improved. However, this
   programme has identified fairly fundamental institutional problems underlying poor spending capacity
   and undermining the quality of outcomes. These include:

      (i) Problems in the infrastructure delivery process, from planning to budgeting and financial
          management. This leads to extensive delays as poorly planned projects cannot be budgeted for,
          budgets are often under-estimated, funds not made available timeously, and procurement
          activities not properly undertaken; and

      (ii) Weak institutional arrangements for infrastructure delivery, including inadequate contracting with
           DPWs, over-centralisation of project planning and approval, especially for small scale projects, and
           weak reporting and monitoring systems

      Underlying this has been a concern over the availability of skilled staff within the public sector. Planning
      systems and institutional arrangements are complex to manage, and all the more so for capital
      projects. On the one hand there are few, if any, tailored training programmes for public officials. On the
      other, high vacancy rates compromise the operational capacity of even the best designed systems.
      Table 17 provides data on vacancy rates in municipalities (albeit relative to preferred rather than
      efficient staffing levels) and indicates the extent and distribution of the problem in that sphere.

                            Table 17: Comparative personnel data for municipalities (2006)
                                                                Staff / 1000     Average annual     Vacancy   Manager /
                                            Total staff
                                                                Households     cost of employment    Rate       Staff
            Metropolitan Municipalities,
                                             100,203               23.3            R 165,111        25.7%       1.4%
            of which …
            Cape Town                         22,322               28.7            R 149,422         4.3%       0.0%
            Johannesburg                      26,274               25.0            R 146,865        12.7%       0.0%
            eThekwini                         18,067               21.9            R 177,279        48.5%       0.7%

            Local Municipalities              93,305               12.5            R 130,662        17.6%       2.7%

            District Municipalities           10,226                1.4            R 216,943        18.7%       6.8%
            National Total / Average         203,734               17.3            R 151,936        21.9%       2.3%
          Sources: StatsSA Census 2001 and P9115, NT database

b) Until recently the need for adequate maintenance expenditure has been largely ignored in the drive to
   eliminate backlogs. This resulted in existing assets being allowed to degrade to the point where
   substantial rehabilitation or outright replacement is required, significantly reducing the total value of
   installed assets. This trade off is likely to become less sustainable over time. Firstly, the asset values of
   existing networks will decline more rapidly as assets run down or are increasingly overloaded. This
   means that they will require replacement sooner than may otherwise be the case. Secondly, the total
   burden of operating and maintenance costs will rise as infrastructure stocks expand. The capacity of
   sub-national governments to finance ongoing maintenance and operating costs will thus be increasingly
   tested. The problem may be compounded by skills shortages, if expertise in running ageing assets is
   unavailable. Of particular concern is the limited information that is available on the condition, age and
   current and replacement values of assets.

    Uncertainty surrounding the restructuring of the electricity distribution industry has also had a
    negative effect on the maintenance of electrical infrastructure by local governments. Proposals for the
    restructuring of the EDI in a set of 6 Regional Electricity Distributors (REDs) have been around for
    almost 10 years, although little progress has been made with implementation. Although efforts were
    made to form the first RED in the Cape Town Functional Region, these were abandoned once the
    incumbent city administration assumed power. This extended inter-regnum has had a severe impact on
    the maintenance of EDI assets by licensed municipal distributors, who have faced the prospect of losing
    their licenses (and assets) to REDs, and have thus limited maintenance and new investments. Given the
    significance of electricity expenditure in total municipal spending, and its importance for economic
    activity, this is a significant trend in itself. However, should municipalities finally lose this function they
    may also lose: (i) a lucrative source of revenue gained for surcharges imposed on electricity tariffs; (ii)
    liquidity resulting from the scale of revenues raised by electricity; and (iii) the use of electricity
    disconnections as an important means of credit control for municipal account holders.

c) Insufficient attention has been paid to cost structures and cost inflation across sectors, despite its
   significance for the speed at which investment needs can be met. This has resulted in a high level of
   cost instability both within projects and across projects. A famous example from the mid-90’s had
   Umgeni Water installing a rural water supply scheme at an average cost of R100,000 per connection.
   Table 14 demonstrates the extent to which MIG projects are currently exceeding cost norms in the
   water and sanitation sector, with the latter doing so by over 450%. This reflects the installation of
   water-borne networked sanitation systems by municipalities. The implication of this is that fewer
   households can obtain access to basic services within a specific time, unless dramatic budget increases
   are allocated.

    Construction cost inflation is likely to have contributed to a subsequent rise in average costs per
    households across all sub-national infrastructure sectors. This has a knock-on implication for asset
    replacement values, and also required maintenance expenditure. Existing investments are likely to have
    significant and growing maintenance and operating costs over their lifetimes.

5    Issues and emerging lessons in sub-national infrastructure delivery
This paper has reviewed demand for sub-national infrastructure investment, the institutional and fiscal
framework for resourcing associated investment needs, and capital and maintenance expenditure patterns
among provinces and municipalities. Although it has focussed on a limited number of sectors, these are the
primary responsibilities of provincial and local governments. A number of issues and concerns have been
identified in each section of the review. This section presents an integrated summary of the findings of the
above sections and identifies cross-cutting issues related to the performance of provincial and local
government infrastructure investment programmes.

Of over-riding interest is the significant success that the sub-national infrastructure investment
programmes have had in addressing the lack of access to basic services since 1994. This is made more
remarkable in that it has occurred through a period fiscal restraint, significant restructuring of sub-national
governments, and changes to intergovernmental fiscal systems.
However, these associated reforms and a more recent slowdown in progress are useful in highlighting
lessons from the experience of sub-national infrastructure investment, and highlighting emerging areas of

5.1   Policy Uncertainties
A series of policy uncertainties associated with the assignment of functions and fiscal powers have come
into focus since 2000, and are likely to pose ongoing challenges to the effectiveness of sub-national
infrastructure delivery.

At an intergovernmental level, the current assignment of built environment functions is fragmented,
resulting in coordination challenges across sectors within urban systems. In the short term this has slowed
down delivery of basic services and housing, but also contributes significantly to sub-optimal built
environment outcomes, including urban sprawl and the peripheralisation of poverty. This is because sectors
are able to externalise the long term operating costs of their investment location decisions to other spheres
of government. The short and long term costs (arising from delays to costs entrenched in the pattern of
urban development, such as transport and services subsidies) are difficult to quantify, but are likely to be
significant and growing.

Local government faces two significant policy uncertainties that have and will continue to impact on its
capital expenditures until they are resolved. Firstly, the removal of RSC levies leaves large urban
municipalities in particular without an important source of revenue for direct fixed investment or security
for loans. This is a significant problem, which has been repeatedly highlighted by city governments (through
the SA Cities Network). These revenues have been replaced by a grant and may well be replaced by a
revenue sharing arrangement on fuel levies. It is too early to quantify the impacts of either of these
arrangements. However, even the disruption of a previously buoyant revenue stream will have already
negatively impacted on the scale, distribution and speed of municipal infrastructure investment. The RSC
levy replacement grant is unlikely to have grown as fast as RSC levy revenues would have, given their past
performance. The distribution of the actual revenues is also likely to have begun to diverge significantly
from the estimations contained in the grant. The absence of an own revenue source has also limited
municipal borrowing options, and thus slowed the pace of municipal lending. Collectively, this means that
municipalities are already likely to be spending less, with this trend more pronounced in areas of rapid
economic growth, and also borrowing less than they otherwise might have done. Secondly, ongoing
uncertainty over the future of the electricity function in municipalities has led to sustained under-
investment in asset maintenance and refurbishment by municipal distributors. This has already begun to
impact negatively on service quality.

5.2   Information Systems
Considerable efforts have been made since 1994 to measure backlogs in the delivery of basic services and
how these have altered as a result of public infrastructure investments. Given these efforts, the continued
low quality and limited availability of data is surprising. While multiple analyses and findings may well be
a desirable stimulus to policy improvements, information on some sectors and dimensions of infrastructure
investment programmes remains very difficult to come by. In particular, the absence of information of the
asset base of individual sectors is a key blockage to the development and monitoring of appropriate
maintenance expenditures.

However, there is another perspective that argues the focus on backlog data is increasingly inappropriate
as access to services expands. Backlog data captures only whether citizens have the necessary
infrastructure assets to receive a service. It implies that the primary focus on expenditure should be to
expand infrastructure networks. As a result it does not focus on the quality of services that are being
delivered or the developmental outcomes associated with them. These data sets, which measure technical
standards of delivery (such as the number of service interruptions) and citizen satisfaction with services,
are only at an embryonic stage of development in South Africa. Their mainstreaming into infrastructure
policy would be likely to have significant impacts on the orientation of sub-national expenditures.

5.3   Institutional Capacity
The capacity of sub-national governments to spend resources allocated to them has rightly become a key
focus of national government. The IDIP was specifically introduced to reverse the trend of persistent under-
spending in provinces, and rightly focussed on core systems of planning, procurement and project
management. The success of this approach is evident in significant declines in under-expenditure by
provinces. Given its success, the expansion of the programme to include large municipalities and issues of
maintenance expenditure (i.e the entire asset management cycle) may be desirable.

However, highly symmetrical intergovernmental arrangements may also constrain expenditure at
municipal level. The autonomy of large city municipalities in particular is eroded by the imposition of
standard procedures in national grant programmes, and limited authority over important aspects of the
built environment. These structural impediments to expenditure performance have become increasingly
inappropriate as the institutional capacity of large city municipalities improves. Negotiating greater
discretion for these entities, as part of a shift to a more outcomes driven approach to infrastructure
investment, is likely to become a significant issue moving forward.

5.4   Fiscal Framework
There have been significant reforms to the fiscal framework for sub-national infrastructure investment.
These reforms have created the architecture necessary for national government to dramatically expand
transfers to these entities, and so significantly boost levels of infrastructure investment. However, three
sets of issues have been raised by the installation of this framework.

Firstly, it has significantly reduced the ability of national sector departments to establish and control
expenditures in their respective sectors. Multi-sector infrastructure grants inherently provide a greater
degree of discretion to benefiting authorities over the allocation of expenditures. Some national
departments argue that the removal of these specific purpose transfers has undermined the ability of
national government to establish and fund development priorities across the public sector. This perspective
is flawed on four levels:

a) The past performance of specific purpose grants was generally dismal. Departments failed to invest
   adequately in the design or management of their programmes. Programmes were often overly
   centralised and bureaucratic in orientation, yet were often unable to report on even basic aspects of
   progress on a regular or timely basis. In general, they assumed accountability for expenditures (such as
   retaining invoices for audit purposes) despite provincial or local governments holding ownership of
   assets; and

b) It assumes that controlling programme inputs will increase the ability of departments to leverage
   developmental outcomes. Achieving these outcomes requires careful integration of choices across
   sectors, and financial commitments over time. Unless these departments propose to control the
   allocation of operating expenditures a system of input controls will remain a mirage and may well
   create liabilities for the national fiscus;

c) It ignores existing mechanisms of national budget oversight that provide considerable scope to exert
   influence over outcomes of sub-national spending. Performance-based adjustments to allocations, for
   example, remain possible within the current framework but have not yet been fully utilised. Indeed, the
   counterargument is that the problem is rather one of insufficient discretion being granted, particularly
   to municipalities in the case of MIG. The symmetric allocation of powers to local government has
   resulted in all municipalities being subject to the same, detailed, input-based oversight mechanisms
   regardless of their capabilities. This significantly delays and distorts investment decisions by more
   capable municipalities, and restricts the ability of national government to monitor developmental

d) Specific purpose transfers have continued to remain a feature of the intergovernmental system. The
   number of conditional grants was initialled rationalised through the introduction of multi-purpose
   infrastructure grants. However, this created scope for a new generation of outcome-focussed grants to
      be introduced. Large new infrastructure grant programmes include those for 2010 stadiums, public
      transport and neighbourhood development. This suggests that a balance is already being struck, and
      that further proliferation of specific purpose transfers may confuse sub-national priorities.

Secondly, the rapid expansion of transfers to sub-national governments may have unintended
consequences. There is some evidence that municipal borrowing from the private sector has not grown to
its full potential due to capacity constraints in spending available grant funds, as well as the aggressive
expansion of DBSA lending to creditworthy municipalities. The effect of this is to expand the fiscal burden
on national government. It may also have the unintended consequence of masking effective municipal
capital expenditure demands, as grant funds are allocated based largely on national estimations of
investment needs.

Effective demand is best expressed through self-funded contributions towards infrastructure, backed by
appropriate strategic and operational business plans. The revenue performance of sub-national
governments has been weak. Own revenue makes a small and shrinking contribution to provincial
revenues, while municipalities have struggled to collect revenues due to them and have large outstanding
debtors’ balances. The capital investment planning process in sub-national governments also demonstrates
some weaknesses. The absence of measures to ensure investment projects reflect sub-national demands
may thus be a major weakness in the overall approach to national funding of local infrastructure. Common
policy options to address this include a requirement for the allocation of matching funds by sub-national
authorities, which can be raised either from own revenues or the capital markets.

5.5     Confused Accountability
A number of factors mentioned in this review point to an underlying problem of confused lines of
accountability for sub-national infrastructure investment. The current assignment of built environment
functions encourages the externalisation of long term costs between sectors and spheres of government.
Most funding programmes continue to rely on input-based controls that reduce both discretion and
accountability. Few positive or negative incentives are provided for performance in meeting key priorities
or for adopting innovations that could overcome these constraints. In general, this makes it difficult for
national government to hold sub-national authorities to account for their infrastructure investment

However, at the heart of the notion of accountability is the role of citizens in being able to hold their
governments to account for performance. Central oversight of sub-national performance can support, but
can seldom replace, this direct form of accountability. Yet municipal infrastructure programmes in
particular tend to be highly centralised in orientation, supported by the rapid growth of full grant funding.
Incentives and mechanisms to strengthen citizens capacity for oversight of local infrastructure delivery thus
appear to offer a fruitful line of future enquiry.

                              Table 18: Local Government Capital Expenditure by Category of Municipality
R’000                             2002/03      2003/04      2004/05      2005/06      2006/07     Average    2007/08      2008/09      2009/10

Metropolitan Municipalities       7 005 650   7 611 502    8 669 749    11 046 101   13 249 259             19 329 595   19 697 819   15 055 904
% Change                                         8.6%        13.9%        27.4%        19.9%      17.5%       45.9%         1.9%        -23.6%
% LG Cap Exp                       52.1%        46.1%        50.4%        46.7%        45.8%      48.2%       48.7%        50.4%         49.6%
% tot exp                          15.0%        14.4%        15.0%        16.3%        17.1%      15.6%       20.0%        19.3%         15.3%
Secondary cities                  2 003 337   2 473 054    2 874 329     3 643 843   4 761 313              6 116 653    5 786 277    3 829 099
% Change                                        23.4%        16.2%         26.8%       30.7%      24.3%       28.5%        -5.4%       -33.8%
% LG Cap Exp                       14.9%        15.0%        16.7%         15.4%       16.5%      15.7%       15.4%        14.8%        12.6%
% tot exp                          15.9%        17.3%        17.3%         19.6%       21.5%      18.3%       23.8%        22.9%        15.9%
Other Local Municipalities        3 266 034   3 976 526    4 002 356     5 997 264   7 435 831              9 201 900    9 337 573    7 198 996
% Change                                        21.8%         0.6%         49.8%       24.0%      24.1%       23.8%         1.5%       -22.9%
% LG Cap Exp                       24.3%        24.1%        23.3%         25.3%       25.7%      24.5%       23.2%        23.9%        23.7%
% tot exp                          24.0%        24.5%        22.4%         27.5%       28.5%      25.4%       30.0%        29.8%        23.5%
District Municipalities           1 160 099   2 447 332    1 657 617     2 979 708   3 485 734              5 077 013    4 243 642    4 293 564
% Change                                       111.0%       -32.3%         79.8%       17.0%      43.9%       45.7%       -16.4%         1.2%
% LG Cap Exp                       8.6%         14.8%         9.6%         12.6%       12.0%      11.5%       12.8%        10.9%        14.1%
% tot exp                          20.5%        30.5%        18.5%         29.8%       32.6%      26.4%       36.0%        34.8%        34.2%
Total                            13 435 121   16 508 413   17 204 051   23 666 916   28 932 138             39 725 162   39 065 312   30 377 563
% Change                                        22.9%         4.2%        37.6%        22.2%      21.7%       37.3%        -1.7%        -22.2%
% tot exp                          17.1%        18.1%        17.0%        20.0%        21.2%      18.7%       23.8%        22.9%         18.3%


Auditor-General of South Africa, 2004. Report on the audit outcomes of local government for the year
ending June 2004. http//www.agsa.co.za

Construction Industry Development Board (cidb), with the Department of Public Works and CSIR, 2007:
“The National Infrastructure Maintenance Strategy”. July 2007.

Development Bank of Southern Africa, 2005. Interim Report of the 2005/06 Financial Year. Midrand: DBSA.

Department of Housing, 2004, Breaking New Ground. Department of Housing: Pretoria

Department of Transport, 2007: “Assessment of the Municipal Roads Network”. Final draft (version 7),
March 2007.

Department of Water Affairs and Forestry (DWAF) 2007 “Summative evaluation of the Masibambane II
Programme. Project Evaluation Report”. August 2007.

Esfahani, H.S. and Ramirez, M.T 2002 “Institutions, Infrastructure and Economic Growth”, Repec Working
Paper (downloaded online)

Hassen, Ebrahim-Khalil, 2000 “Infrastructure Delivery and Employment Creation in South Africa”, Paper
presented at Sept. 2000 Annual Forum of Trade and Industrial Policy Secretariat, Glenburn Lodge,

National Electricity Regulator of South Africa (NERSA), 2007: “Independent technical audit on eleven
electricity distributors’ networks: Executive summary of the consolidated audit report”.

National Treasury, 2003: Intergovernmental             Fiscal   Review.    Pretoria:   National    Treasury.

Van Ryneveld, P, 2006 “The development of policy on the financing of municipalities”, ” in Pillay, U.,
Tomlinson, R. and du Toit, J. (eds.) Democracy and Delivery: Urban Policy in South Africa. Pretoria: HSRC

South African Institution of Civil Engineering (SAICE), 2007: “The SAICE Infrastructure Report Card for South
Africa: 2006”. SAICE, Midrand