162

Brian Dollery is Professor of  Economics & Director of  the UNE Centre for Local Government, School of  * 
Business, Economics and Public Policy University of  New, England Armidale NSW 2351 Australia.   Email: 
bdollery@une.edu.au. Gert van der Westhuizen is Professor of  Economics, School of  Economic Sciences, 
North-West University, Vaal Campus, Vanderbijlpark, South Africa. E-mail: gvdwesthuizen@absamail.co.za

Efficiency measurement of basic service delivery at 
South African district and local municipalities 

G van der Westhuizen and B Dollery*

Keywords: Efficiency; local government; local service delivery; South Africa
Disciplines: Economics, Public Management and Administration, Management Sciences.

Introduction1. 

Local government systems across the world have come under intense scrutiny over the past two decades. 
A good deal of concern by policy makers has been expressed over various aspects of local government, 
including its operational efficiency. This has seen a wave of reform in many local government jurisdictions 
which has encompassed almost all dimensions of local government structure and function (Dollery, et al., 
2008). South African local government has not been immune to these international trends and it too has 
been subjected to vigorous change. Indeed, few local government systems have undergone a greater degree 
of transformation (Cameron, 2001; Visser, 2001). 

The dramatic nature of local government reform in South Africa has its roots in the apartheid legacy that 

Abstract. South Africa has experienced immense changes in the post-
apartheid era and coordinated local public policy has sought to expand and 
improve the level of  basic services provided to previously disadvantaged 
people. Local government has played a pivotal role in this process and 
has been subjected to intense reform in an effort to enhance its effective-
ness and broaden its range of  activities. While a number of  scholars have 
examined the administrative, political and social dimensions of  the local 
government reform program, little attention has focused on the economic 
efficiency of  service delivery. This paper seeks to remedy this neglect by 
evaluating the productive efficiency with which municipal councils have 
delivered electricity, domestic waste removal, sanitation and water in line 
with their new responsibilities using Data Envelopment Analysis (DEA) 
techniques applied to cross-sectional data covering the period 2006/2007 
for 231 local municipalities and 46 district municipalities.



Efficiency SA municipalities

163                                                           TD The Journal for Transdisciplinary Research in Southern Africa, 5(2), December 2009, pp. 162-174.

confronted policy makers in the first post-apartheid administration. This legacy included a high degree 
of economic, spatial and social inequality between the different population groups, together with a ‘dual 
economy’ divided along racial lines, and segregated urban areas. Post-apartheid policy makers responded 
with a range of measures aimed at addressing these inequalities across the entire gamut of economic 
and social indicators. The most important of these measures consisted of the 1994 Reconstruction and 
Development Programme (RDP), which sought to coordinate the efforts of central, provincial and local 
governments into an integrated national whole (Lester, et al, 2000).

The RDP placed great emphasis on the role of local government in this process based on the premise 
that

the democratic government will reduce the burden of implementation which falls 
upon its shoulders through the appropriate allocation of powers and responsibilities 
to lower levels of government (African National Congress, 1994, p.140). 

This emphasis was given sharper focus in the White Paper on Local Government, which held that local 
government must co-operate with local communities

to find sustainable ways to meet their needs and improve the quality of their lives 
(Republic of South Africa, 1998, p.17). 

The notion of ‘developmental local government’ thus grew out of the perception that the role of local 
government in South Africa should expand decisively from its traditional narrow concentration on ‘services 
to property’ to embrace local economic development and local economic growth as well as the conditions 
that stimulate these elements of national development (Nel and Binns, 2002). 

The White Paper on Local Government (Republic of South Africa, 1998) set out four key aims for 
‘developmental local government. In the first place, the provision of a basic level of household services, 
especially electricity, sewerage and water, to households without these services, should take priority. 
Secondly, municipalities should seek to ameliorate the ‘spatial legacy of apartheid separation’ through 
the integration of previously segregated urban areas. Thirdly, local economic development should be 
stimulated through local economic growth and local job creation. Finally, ‘community empowerment and 
redistribution’ should be addressed.

While these objectives are certainly laudable in the context of post-apartheid South African society, striven 
by striking racial inequalities, they placed a massive burden on a municipal system that had previously 
confined its activities to a narrow range of local services. This burden was exacerbated by an acute lack 
of local government capacity, especially in terms of administrative and technical skills (Buthelezi and 
Dollery, 2004; Dollery, et al, 2005), as well as very weak financial compliance management (Dollery 
and Graves, 2009) and chronic funding shortages (Bahl and Smoke, 2003) - problems also apparent at 
higher tiers of government in South Africa (Dollery, 2004; Dollery and Snowball, 2003). In essence, the 
problem of local government incapacity seems to derive mainly from a lack of skilled human resources in 
South African local government to perform its newly expanded responsibilities. Against this background, 
there is an obvious need to assess the progress made by local government to achieve its new objectives. 
Some scholarly attention has already been at this important aspect of the reform process (see, for instance, 



Van der Westhuizen and Dollery

164

Macdonald and Pape, 2002 and Miraftab, 2004). However, to date there has been no work on evaluating 
econometrically the efficiency of local government service delivery. The present paper thus seeks to address 
this neglected question by focusing on the first of the quadrilateral suit of policy objectives set out in 
White Paper on Local Government (Republic of South Africa, 1998). Accordingly, we assess the economic 
efficiency with which municipal councils have delivered electricity, domestic waste removal, sanitation 
and water in line with their RDP responsibilities using DEA techniques applied to cross-sectional data 
covering the period 2006/2007 for 231 local municipalities and 46 district municipalities.

The paper is divided into four main parts. Section 2 provides a synoptic description of the service delivery 
responsibilities of district municipalities and local municipalities in the South African local government 
milieu. Section 3 briefly outlines the theory of local government efficiency analysis and the empirical 
literature that has developed around it. Section 4 sets out the data sources and models employed in the 
analysis. Section 5 considers the results obtained from the estimations. The paper ends with some brief 
concluding comments in section 6.

2. South African local government service delivery

In common with its counterparts in Australia and New Zealand, South African local government has 
traditionally provided a rather narrow range of ‘services to property’ (Graves and Dollery, 2009). However, 
in line with the White Paper on Local Government and the RDP aims, the role of municipalities was rapidly 
expanded to include a much broader array of service objectives. 

In terms of its structure, South African local government is sub-divided into three basic categories; 
metropolitan municipalities, district municipalities and local municipalities. Metropolitan municipalities 
have municipal executive and legislative authority in their respective local government areas which 
embrace the major urban concentrations of Cape Town, Durban, East Rand, Johannesburg, Pretoria and 
Port Elizabeth. By contrast, district municipalities have municipal executive and legislative authority over 
significant spatial areas, with primary responsibility for district-wide planning and capacity-building. 
Within the local jurisdiction of each district municipality there are typically several individual local 
municipalities which share their municipal authority with the district municipality. In essence, district 
municipalities administer and make rules for a district, which includes more than one local municipality. 
South Africa has a total of 284 municipalities in these three categories combined (Atkinson, 2002).

The intended purpose of district municipalities and local municipalities sharing responsibility for local 
government in their given areas is to ensure that all communities, particularly historically disadvantaged 
communities, have equal access to resources and services. Since district municipalities usually cover 
both relatively affluent and relatively poor concentrations of people, consequent upon the apartheid 
legacy, this facilitates ‘cross-subsidisation’, enabling local municipalities without adequate administrative, 
financial and technical capacity to provide basic services to their historically disadvantaged communities. 
In addition, the system provides scope for shared services between local municipalities which can generate 
economics of scale and scope and attendant cost reductions.

The functions of district municipalities are manifold. From the perspective of local service delivery, district 



Efficiency SA municipalities

165                                                                                                                                                                     TD, 5(2), December 2009, pp. 162-174.

municipalities inter alia must plan for development for the district municipality as a whole; provide bulk 
supply of water that affects a large proportion of the municipalities in the district; supply bulk electricity; 
provide municipal health services for the whole municipalities in the district; ensure bulk sewerage 
purification works and main sewerage disposal; provide waste disposal sites for the whole district; provide 
municipal roads and storm water drainage for the district municipality area; provide municipal public 
works; ensure street lighting; and provide municipal parks and recreation facilities.

In the present context, financial management represents an important aspect of the overall South African 
local government reform process. In this regard, the National Treasury has played a pivotal role in the 
introduction of financial management reforms across local government since 1996. The legislative 
cornerstone of the municipal reform process has been the Municipal Finance Management Act, 56 of 2003 
(MFMA), supported by the annual Division of Revenue Act. This legislation has been aligned with other 
important local government enactments, such as the Structures Act, Systems Act, Property Rates Act, to form 
a coherent package.

The prime objective of the National Treasury in the local government realm has been to secure sound and 
sustainable management of the financial affairs of local government. This has required the development 
of a coherent approach to improve the delivery of services to local communities. The National Treasury 
has implemented a strategy of financial and technical support for local government based around the 
MFMA, including conditional grants, subsidies, technical guidelines, policy advice and the placement 
of international advisors with various municipalities (Dollery and Graves, 2009). This strategy took 
into account the vastly different levels of capacity of the different municipalities for implementing the 
reforms. It also prescribed requirements for institutional strengthening, municipal capacity building 
and improving municipal consultation, reporting, transparency and accountability. In sum, the MFMA 
aims to modernise budget, accounting and financial management practices by placing local government 
finances on a sustainable footing in order to maximise the capacity of municipalities to deliver services to 
communities.

3. Local government efficiency measurement

Ongoing local government reform around the world has led increasing sophisticated attempts to measure 
the effects of reform programs. This has given rise to a growing theoretical and empirical literature on local 
government efficiency measurement (Worthington and Dollery, 2000). Economic efficiency has various 
dimensions, including allocative efficiency, productive efficiency, dynamic efficiency, scale efficiency and 
scope efficiency. Technical or productive efficiency refers to the use of resources in the technologically 
most efficient manner in order to obtain the maximum possible output(s) from a given set of inputs. 
When productive efficiency is determined in monetary terms, it is sometimes known as cost efficiency. 

Several features of local government service delivery have made it difficult to develop accurate measures 
of productive efficiency, especially for benchmarking and comparative performance measurement. These 
features include (a) multiple inputs and outputs in service delivery (b) problems in estimating the costs 
of service delivery (c) the existence of different stakeholders with competing needs impedes efficiency 
improvement, and (c) ‘non-discretionary’ factors beyond the control of local government (Worthington 
and Dollery, 2002).



Van der Westhuizen and Dollery

166

Despite these difficulties, economists have applied five different approaches to the analysis of local public 
sector efficiency (Worthington, 2001): Least squares econometric production models; the deterministic 
frontier approach; the stochastic frontier approach; the fee-disposal hull approach; and DEA. DEA is 
often used to measure the relative efficiency (or productivity) of organisations in the same industry, such 
as municipalities. DEA is typically the preferred measure of relative efficiency for complex organizations 
in complex environments because it readily lends itself to the analysis of multiple output organisations, 
especially where binding constraints affect the behaviour of the organisations in question. In essence, 
DEA combines all the input and output information on the local municipality into a single measure of 
productive efficiency that lies between zero (i.e. a completely inefficient municipality) and unity (i.e. a 
completely efficient municipality). DEA is thus an application of linear programming that can be used to 
measure the relative efficiency of organisations with the same goals and objectives.

The empirical measurement of allocative, cost, productive and scale efficiencies involves the estimation 
of production frontiers. DEA effectively estimates the frontier by finding a set of linear estimates that 
bound (or envelop) the observed data. Measuring productive efficiencies using DEA requires data on 
output and input quantities whereas measuring allocative and cost efficiencies also needs data on input 
prices. Outputs measures used in DEA studies of local government services have typically used only 
quantitative measures, and often ‘non-discretionary’ quantitative output measures like the number of 
residents receiving garbage collection. Worthington and Dollery (2000) provide a detailed discussion 
of DEA and other frontier techniques for the measurement of economic efficiency in local government, 
whereas Worthington (2001) has summarised the empirical literature on local government.

4. Data and models

Securing adequate and satisfactory data on South African local government represents a formidable 
challenge to all researchers in the area (Bahl and Smoke, 2003). After repeated unsuccessful requests to 
individual local municipalities and individual district municipalities to supply relevant data that could 
be used to estimate the efficiency of various local services, to which only two municipalities responded 
positively, we were obliged to use data published by the Demarcation Board of South Africa (2006/2007). 
This data comprised inter alia total number of households, RDP water, RDP sanitation, RDP electricity, 
RDP refuse removal, the number of staff, various types of income that can be aggregated as total operating 
income and staff costs. The number of households, RDP water, sanitation electricity and refuse removal 
are derived from the 2001 national census. This was adjusted to represent number of households in 
2006/2007 using the population growth rate in proportion to the estimates of the population per province 
(Statistics SA, 2008). The total number of households was checked using a regression equation with an 
independent variable given as the total number of households for 2006/2007 and the dependent variable 
the number of households serviced for each type of service rendered.

The data used for the analysis in this paper represent budget data for the relevant variables used. In 
common with all efficiency measurement exercises, the quality of the data obviously determines the 
quality of the results obtained from the analysis. This raises various potential problems in the present 
context. For instance, an improved municipal budgeting process may change efficiency results without 
any corresponding real change in local services actually delivered. However, there is some evidence which 
suggests that this may not be a significant problem in the time period under review (Dollery and Graves, 



Efficiency SA municipalities

167                                                                                                                                                                     TD, 5(2), December 2009, pp. 162-174.

2009). In any event, cross-sectional data covering the fiscal year 2006/2007 were employed, which 
included 231 local municipalities and 46 district municipalities.

A DEA was performed using as output variables the total number of households receiving RDP water, 
RDP sanitation RDP electricity and RDP refuse removal. For input variables, the Rand value of staff costs 
and total operational income were used. Total operational income was employed as an input since this 
represent the aggregate funds to deliver the various local services under review; it incorporates the rand 
value of rates income, services income and government grants.

5. Discussion of results

The software package DEAP Version 2.1 developed by Coelli (1996) is ‘purpose-built’ to solve the DEA 
problem in efficiency estimation procedures and typically forms the methodological basis for many local 
government efficiency measurement exercises. Accordingly, it has been used for the estimations reported 
in this paper to generate estimates of technical efficiency and scale efficiency. The efficiency estimates are 
executed under constant returns to scale (CRSTE) and variable returns to scale (VRSTE); they embrace 
output-orientated as well as input-orientated approaches. The output-orientated approach applies to a 
situation in which a given district municipality seeks to maximise the output with a given set of inputs, 
whereas an input-orientated approach applies to the situation in which the district municipality seeks 
to deliver the desired output with the minimum input. In the present context, discussion of our results 
will primarily concentrate on the efficiency estimates under variable returns to scale (VRS). Following 
Coelli et al (1998:150), the use of the constant returns to scale (CRS) specification when not all firms 
are operating at the optimal scale results in measures of relative efficiency (TE) which are confounded by 
scale efficiencies (SE). The use of VRS specification thus permits the calculation of TE devoid of these SE 
effects.

The relative efficiency estimates for the district municipalities in the country are depicted in Table 1. 

Table 1: Relative Efficiency Estimates for District Municipalities

Output-orientated Input-orientated
District CRSTE VRSTE SCALE RETURN CRSTE VRSTE SCALE RETURN

1 0.071 0.370 0.193 drs 0.071 0.108 0.665 irs
2 0.118 0.743 0.158 drs 0.118 0.126 0.936 drs
3 0.076 0.301 0.254 drs 0.076 0.158 0.483 irs
4 0.095 0.609 0.155 drs 0.095 0.107 0.883 drs
5 0.043 0.087 0.492 drs 0.043 0.266 0.160 irs
6 0.102 0.135 0.759 drs 0.102 0.352 0.291 irs
7 0.106 0.185 0.574 drs 0.106 0.348 0.306 irs
8 0.223 0.365 0.610 drs 0.223 0.265 0.840 irs
9 0.221 0.429 0.516 drs 0.221 0.451 0.491 irs

10 0.108 0.411 0.262 drs 0.108 0.212 0.509 irs



Van der Westhuizen and Dollery

168

Output-orientated Input-orientated
District Crste Vrste Scale Return Crste Vrste Scale Return

11 0.222 1.000 0.222 drs 0.222 1.000 0.222 drs
12 0.134 0.449 0.299 drs 0.134 0.174 0.773 irs
13 0.045 0.185 0.245 drs 0.045 0.128 0.355 irs
14 0.075 0.399 0.187 drs 0.075 0.110 0.680 irs
15 1.000 1.000 1.000 - 1.000 1.000 1.000 -
16 0.602 0.913 0.660 drs 0.602 0.684 0.881 drs
17 0.614 0.825 0.745 drs 0.614 0.696 0.882 irs
18 0.776 0.840 0.924 drs 0.776 0.803 0.967 drs
19 0.648 0.765 0.847 irs 0.648 0.963 0.673 irs
20 0.045 0.342 0.132 drs 0.045 0.081 0.561 irs
21 0.279 0.840 0.332 drs 0.279 0.332 0.841 drs
22 0.087 0.363 0.239 drs 0.087 0.093 0.929 irs
23 0.230 1.000 0.230 irs 0.230 1.000 0.230 irs
24 0.506 1.000 0.506 irs 0.506 1.000 0.506 irs
25 0.089 0.318 0.280 drs 0.089 0.223 0.400 irs
26 0.084 0.188 0.445 drs 0.084 0.348 0.240 irs
27 0.141 0.462 0.305 drs 0.141 0.223 0.631 irs
28 0.071 0.309 0.228 drs 0.071 0.102 0.691 irs
29 0.615 0.848 0.725 drs 0.615 0.675 0.911 irs
30 0.835 1.000 0.835 drs 0.835 1.000 0.835 drs
31 0.475 0.796 0.597 drs 0.475 0.503 0.945 irs
32 0.872 1.000 0.872 drs 0.872 1.000 0.872 drs
33 0.422 0.838 0.504 drs 0.422 0.567 0.744 drs
34 0.163 0.752 0.217 drs 0.163 0.230 0.710 drs
35 0.582 0.608 0.956 drs 0.582 0.877 0.663 irs
36 0.651 1.000 0.651 drs 0.651 1.000 0.651 drs
37 0.199 0.563 0.353 drs 0.199 0.234 0.849 irs
38 0.288 0.397 0.725 drs 0.288 0.582 0.495 irs
39 1.000 1.000 1.000 - 1.000 1.000 1.000 -
40 0.197 1.000 0.197 drs 0.197 1.000 0.197 drs
41 0.080 0.202 0.394 drs 0.080 0.234 0.340 irs
42 0.085 0.184 0.464 drs 0.085 0.203 0.421 irs
43 0.021 0.149 0.140 drs 0.021 0.026 0.796 irs
44 0.154 0.209 0.739 drs 0.154 0.473 0.326 irs
45 0.423 0.586 0.722 drs 0.423 0.508 0.834 irs
46 0.135 0.700 0.193 drs 0.135 0.156 0.865 irs

Mean 0.305 0.580 0.480 0.305 0.470 0.641
Min 0.021 0.087 0.132 0.021 0.026 0.160
Max 1.000 1.000 1.000 1.000 1.000 1.000



Efficiency SA municipalities

169                                                                                                                                                                     TD, 5(2), December 2009, pp. 162-174.

From the results in Table 1, it can be seen that under the output-orientated approach the district 
municipalities were, on average, only 30.5 per cent technically efficient under constant returns to scale, 
58 per cent technically efficient under variable returns to scale, and 48.0 per cent scale efficient. According 
to the return to scale estimates (under the ‘RETURN’ column), all but two municipalities (i.e. district 
municipality numbers 15 and 39) were operating at decreasing returns to scale (DRS), which means that 
they were operating at a scale that was too large in efficiency terms.

Under the input-orientated approach, the district municipalities were, on average, 47.0 per cent technical 
efficient in the case of variable returns to scale, and 64.1 per cent scale efficient. With regard to the returns 
to scale, 32 municipalities were operating under increasing returns to scale (IRS), which necessarily implies 
that they were operating on a scale that was too small in efficiency terms. Only two district municipalities 
were operating at the optimal scale (i.e. district municipalities numbers 15 and 39), which means that they 
at the optimal size in efficiency terms. The remaining district municipalities were operating at decreasing 
returns to scale.

In the combined case of both the input-orientated and output-orientated approaches, 9 district 
municipalities (i.e. district municipalities numbers 11, 15, 23, 24, 30, 32, 36, 39 and 40) were fully 
technically efficient under variable returns to scale. This implies that these district municipalities were 
using their inputs optimally. Thus any reduction in the inputs employed by these municipalities will 
result in reduced output. The remaining district municipalities can either reduce their inputs without any 
reduction in output or expand their output without increasing their inputs.

We now consider the relative efficiency estimates obtained for individual local municipalities under the 
auspices of district municipalities. Given space constraints, only summary results from these estimations 
are provided in this paper. Table 2 displays the average efficiency as well as the minimum efficiency and 
maximum efficiency estimates for the best performing district municipality per province, according to 
the performance. 

Table 2: Best-Performing District Municipality per Province

Province/District 
municipality

Output-orientated Input-orientated
CRSTE VRSTE CRSTE VRSTE

Western Cape/5 Average 0.618 0.690 0.618 0.736
Min 0.507 0.525 0.507 0.511
Max 0.775 0.914 0.775 0.934

Northern Cape/4 Average 0.744 0.753 0.744 0.756
Min 0.445 0.445 0.445 0.475
Max 1.000 1.000 1.000 1.000

East Cape/6 Average 0.698 0.732 0.698 0.741
Min 0.459 0.464 0.459 0.481
Max 0.937 1.000 0.937 1.000

Free State/1 Average 0.763 0.780 0.763 0.768
Min 0.659 0.660 0.659 0.659
Max 0.962 0.972 0.962 0.974



Van der Westhuizen and Dollery

170

Province/District 
municipality

Output-orientated Input-orientated
CRSTE VRSTE CRSTE VRSTE

Kwazulu/Natal/9 Average 0.707 0.759 0.707 0.786
Min 0.431 0.461 0.431 0.548
Max 1.000 1.000 1.000 1.000

Mpumalanga/3 Average 0.543 0.797 0.543 0.740
Min 0.404 0.618 0.404 0.511
Max 0.703 1.000 0.703 1.000

Limpopo/3 Average 0.795 0.878 0.795 0.856
Min 0.519 0.722 0.519 0.657
Max 1.000 1.000 1.000 1.000

Northwest/1 Average 0.536 0.801 0.536 0.748
Min 0.455 0.667 0.455 0.540
Max 0.660 1.000 0.660 1.000

Gauteng/3 Average 0.780 0.905 0.780 0.887
Min 0.598 0.853 0.598 0.827
Max 1.000 1.000 1.000 1.000

In Table 2, district municipality 5 is the best performing local government in the Western Cape Province 
with an average efficiency estimate of 69.0 per cent (VRSTE) under the output-orientated approach and 
an average efficiency estimate of 73.6 per cent (VRSTE) under the input-orientated approach. More 
generally, efficiency estimates ranged between 52.5 per cent and 91.4 per cent under the output-orientated 
approach and between 51.1% per cent and 93.4 per cent under the input-orientated approach. None of 
the local municipalities operating under the auspices of this district municipality was fully technically 
efficient.

The Free State is the only other province in which its most efficient district municipality had none of its 
local municipalities operating at full technical efficiency. The average technical efficiency estimate under 
the output-orientated approach was 78.0 per cent and 76.8 per cent under the input-orientated approach. 
In terms of the output-orientated approach, the efficiency estimates ranged between 66.0 per cent and 
97.2 per cent. By contrast, under the input-orientated approach, estimates ranged between 65.9 per cent 
and 97.4 per cent.

In the remaining seven South African provinces, there were a number of district municipalities which 
encompassed local municipalities which were fully technical efficient. District Municipality 3 in the 
Gauteng Province exhibited the highest average technical efficiency estimate (i.e. 90.5 per cent under the 
output-orientated approach and 88.7 per cent under the input-orientated approach).

Table 3 exhibits the worst performing district municipalities per province. District Municipality 5 in the 
Northern Cape had the lowest average efficiency estimate. Under the output-orientated approach, its 
average efficiency estimate was 40.7 per cent and under the input-orientated approach it was 36 per cent. 
In general, the efficiency estimate for this district municipality ranged between 16.6 per cent and 60.0 
per cent.



Efficiency SA municipalities

171                                                                                                                                                                     TD, 5(2), December 2009, pp. 162-174.

Table 3: Worst-Performing District Municipality per Province

Province/District 
municipality

Output-orientated Input-orientated

CRSTE VRSTE CRSTE VRSTE

Western Cape/4
Average 0.456 0.517 0.468 0.489

Min 0.354 0.380 0.361 0.372

Max 0.672 0.689 0.617 0.665
Northern Cape/5

Average 0.368 0.447 0.368 0.412

Min 0.136 0.245 0.136 0.211

Max 0.695 0.726 0.695 0.753
East Cape/5

Average 0.328 0.407 0.328 0.362

Min 0.154 0.166 0.154 0.173

Max 0.538 0.600 0.538 0.542
Free State/5

Average 0.587 0.694 0.587 0.654

Min 0.527 0.620 0.527 0.577

Max 0.676 0.804 0.676 0.759
Kwazulu/Natal/10

Average 0.591 0.609 0.591 0.623

Min 0.388 0.399 0.388 0.485

Max 0.838 0.849 0.838 0.854
Mpumalanga/1

Average 0.529 0.643 0.529 0.586

Min 0.396 0.480 0.396 0.397

Max 0.644 0.812 0.644 0.752
Limpopo/4

Average 0.519 0.597 0.519 0.560

Min 0.315 0.387 0.315 0.315

Max 0.709 0.827 0.709 0.768
Northwest/4

Average 0.594 0.706 0.594 0.695

Min 0.407 0.459 0.407 0.463

Max 0.795 1.000 0.795 1.000
Gauteng/1

Average 0.510 0.670 0.510 0.620

Min 0.373 0.454 0.373 0.373

Max 0.669 1.000 0.669 1.000



Van der Westhuizen and Dollery

172

In Table 3, it is interesting to note is that although District 4 in the North West Province and District 
1 in Gauteng Province were included in the list of worst performers, both had local municipalities that 
were fully technically efficient. This means that there were local municipalities that were using their inputs 
optimally even in the worst-performing district municipalities. Table 4 displays the average efficiency 
estimates of all the local municipalities per province.

Table 4: Average Efficiency Estimates of  Local Municipalities per Province

Province/District 
municipality

Output-orientated Input-orientated
CRSTE VRSTE CRSTE VRSTE

Western Cape Average 0.469 0.552 0.469 0.508
Min 0.354 0.380 0.354 0.355
Max 0.775 0.914 0.775 0.934

Province/District 
municipality

Output-orientated Input-orientated
CRSTE VRSTE CRSTE VRSTE

Northern Cape Average 0.589 0.624 0.589 0.646
Min 0.136 0.245 0.136 0.211
Max 1.000 1.000 1.000 1.000

Eastern Cape Average 0.512 0.569 0.512 0.554
Min 0.154 0.166 0.154 0.173
Max 0.937 1.000 0.937 1.000

Free State Average 0.624 0.724 0.624 0.701
Min 0.435 0.450 0.435 0.436
Max 0.962 1.000 0.962 1.000

Kwazulu/Natal Average 0.619 0.673 0.619 0.676
Min 0.315 0.358 0.315 0.361
Max 1.000 1.000 1.000 1.000

Mpumalanga Average 0.558 0.722 0.558 0.678
Min 0.396 0.464 0.396 0.397
Max 0.773 1.000 0.773 1.000

Limpopo Average 0.624 0.750 0.624 0.714
Min 0.300 0.387 0.300 0.315
Max 1.000 1.000 1.000 1.000

Northwest Average 0.607 0.755 0.607 0.721
Min 0.406 0.459 0.406 0.406
Max 1.000 1.000 1.000 1.000

Gauteng Average 0.677 0.794 0.677 0.767
Min 0.373 0.454 0.373 0.373
Max 1.000 1.000 1.000 1.000

Table 4 shows that eight out of the nine South African provinces had one or more local municipality that 
was fully technically efficient. Only the Western Cape had no local municipality that was fully technically 
efficient. The Gauteng Province had the highest average efficiency estimate of 79.4 per cent under the 
output-orientated approach and 76.7 per cent under the input-orientated approach. The Western Cape 
was the worst-performing province with an average efficiency estimate of 55.25 per cent under the output-
orientated approach and 50.8 per cent under the input-orientated approach. There are thus noteworthy 
inter-provincial differences in the relative efficiencies of local municipalities.



Efficiency SA municipalities

173                                                                                                                                                                     TD, 5(2), December 2009, pp. 162-174.

6. Concluding remarks

The results obtained from this first preliminary attempt at measuring relative efficiency in South African 
local government should be regarded with caution mainly because the data from which they were derived 
are questionable. It would have been preferable to employ data provided directly from a large sample of 
district municipalities and local municipalities. However, since this proved impossible, we were obliged 
to use the only available data. Future research into the efficiency measurement of South African local 
government should seek to construct data from reliable sources.

Nonetheless, because there is no compelling a priori reason to assume that systematic differences in data 
reliability between provinces or between municipalities, from a policy perspective it is still possible to 
draw at least some useful tentative conclusions from our preliminary results. In the first place, as we have 
seen from the results in Table 4, considerable variation exists in the average efficiency scores between some 
provinces. This suggests that the financial and technical support for local government provided by the 
National Treasury to local government, which includes policy advice and the placement of international 
advisors in municipalities, should be more carefully targeted between the different provinces, and biased 
towards the worst-performing provinces. Secondly, because significant variation in efficiency also occurs 
within the constellation of municipalities in any given province, a closely related policy option would be 
to direct available resources to the worst-performing municipalities within each province.

References

African National Congress (1994), The Reconstruction and Development Programme: A Policy Framework, 
Umanyano Publications, Johannesburg.

Atkinson, D. (2002), Local Government, Local Governance and Sustainable Development, HSRC Press, 
Pretoria.

Bahl, R. and Smoke, P. (eds.) (2003), Restructuring Local Government Finance in Developing Countries: 
Lessons from South Africa, Edward Elgar Publishers, Cheltenham. 

Buthelezi, A. and Dollery, B. E. (2004), “An Exploratory Analysis of Local Government Failure in South 
Africa”, Studies in Economics and Econometrics, 28(2), 91-105.

Cameron, R. (2001), “The Upliftment of South African Local Government”, Local Government Studies, 
27(3), 97-118.

Coelli, T. (1996), “A Guide to DEAP Version 2.1 Data Envelopment Analysis (Computer) Program”, 
Centre for Productivity Analysis Working Paper 96/07, Department of Econometrics, University of 
New England, Armidale.

Coelli, T., Rao, D.S.P. and Battese, G. (1998), An Introduction to Efficiency and Productivity Analysis, 
Kluwer Academic Publishers, Boston.

Dollery, B. E. (2003), “A History of Inequality in South Africa: Review Note”, South African Journal of 
Economics, 71(3), 595-610.



Van der Westhuizen and Dollery

174

Dollery, B. E., Crase, L. and Marshall, N. A. (2005), “The Economic Case for Virtual Local Government 
in South Africa”, Studies in Economics and Econometrics, 29(1), 47-60.

Dollery, B. E., Garcea, J. and LeSage, E. (2008), Local Government Reform, Edward Elgar Publishers, 
Cheltenham.

Dollery, B. E. and Graves, N. (2009), “Funding Compliance Measurement to Understand Best-Practice 
Financial Management Technical Assistance to South African Local Government”, Public 
Administration and Development, 29(1), 101-116.

Dollery, B. E. and Snowball, J. (2003), “Government Failure and State Incapacity: The South African Public 
Sector in the 1990s”, South African Journal of Economic History, Special Issue, 18(1/2), 310-331.

Graves, N. and Dollery, B. E. (2009), “Local Government Reform in South Africa: An Analysis of Financial 
Management Legislative Compliance by Municipalities”, Public Administration and Development, 
29(5), 387-414.

Lester, A., Nel, E. and Binns, T. (2000), South Africa: Past, Present and Future, Longman, Harlow.

McDonald, D. A. and Pape, J. (eds.), (2002), Cost Recovery and the Crisis of Service Delivery in South 
Africa, HSRC Press, Pretoria.

Miraftab, F. (2004), “Neoliberalism and Casualization of Public Services: Waste Collection Services in 
Cape Town, South Africa”, International Journal of Urban and Regional Research, 28(4), 874-
892.

Municipal Demarcation Board.  2007, Capacity Assessments 2006/2007,                                                                                                                
http://www demarcation.org.za (Date accessed – various occasions).

Nel, E. and Binns, T. (2002), “Devolving Development: Integrated Local Planning and Developmental 
Local Government in post-Apartheid South Africa”, Regional Studies, 36(8) 921-945.

Republic of South Africa (1998), White Paper on Local Government, Department of Constitutional 
Development, Pretoria.

Stats SA. 2008, Mid-year population estimates 2008 – Series P0302,                                                                                                                                           
http://www. statssa.gov.za/publications/populationstats.asp (Date accessed: various occasions).

Visser, G. (2001), “Social Justice, Integrated Development Planning and Post-Apartheid urban 
Reconstruction”, Urban Studies, 38(10), 673-699.

Worthington, A. C. (2001), “The Empirical Measurement of Local Government Efficiency”, in B. E. Dollery 
and J. L. Wallis, The Political Economy of Local Government, Edward Elgar Publishers, Cheltenham. 

Worthington, A. C. and Dollery, B. E. (2000), “Measuring Efficiency in Local Government”, Public 
Productivity and Management Review, 23(4), 469-485.

Worthington, A. C. and Dollery, B. E. (2002), “Incorporating Contextual Information in Public Sector 
Efficiency Analysis: A Comparative Study of NSW Local Government”, Applied Economics, 34(3), 
453-464.