Anda di halaman 1dari 161


Pro-Poor Economic Growth Models

for South Africa
Research Reports prepared for Oxfam GB

Halving Poverty and Unemployment in South

Africa: Choices for the next ten years
Asghar Adelzadeh Director Applied Development
Research Solutions, USA

What is Pro Poor Growth? What are some of

the things that hinder its achievement in South
Charles Meth- Honorary Research Fellow School of
Development Studies University of Kwa Zulu Natal and
Research Associate South African Labour and
Development Research Unit, University of Cape Town

Pro Poor Trade and Industry Policies:

Considerations for South Africa
Mohammed Jahed Associate Professor Graduate School Of
Public and Development Management, University of The

26 May 2007
Contents Page No.

Halving Poverty and Unemployment in South Africa:

Choices for the next ten years 1
Asghar Adelzadeh Director Applied Development Research
Solutions, USA

What is Pro Poor Growth? What are some of the things

that hinder its achievement in South Africa? 60
Charles Meth- Honorary Research Fellow School of Development
Studies University of Kwa Zulu Natal and Research Associate
South African Labour and Development Research Unit, University
of Cape Town

Pro Poor Trade and Industry Policies: Considerations

for South Africa 133
Mohammed Jahed Associate Professor Graduate School Of
Public and Development Management, University of The

Oxfam Great Britain May 2007

These papers were commissioned by the Oxfam GB South Africa Programme.
The views expressed are those of the researchers. Oxfam GB acknowledges the
support to its partners in the production of the reports through funding, serving
on the reference group and commenting on the TORs and drafts. The text may be
used free of charge for the purposes of advocacy, campaigning, education, and
research, provided that the source is acknowledged in full. The copyright holder
requests that all such use be registered with them for impact assessment
purposes. For copying in any other circumstances, or for re-use in other
publications, or for translation or adaptation, permission must be secured and a fee
may be charged.
For further information on the issues raised in this paper please e-mail or call 2711 6429283
Halving Poverty and
Unemployment in South Africa:
Choices for the next 10 years

Asghar Adelzadeh
Applied Development Research Solutions (ADRS)

Prepared for

Oxfam GB South Africa

24 May 2007


Three years ago, South Africa embarked on its second decade of democracy, announcing
objectives that included halving both the unemployment rate and the poverty rate by
2015. Meeting these objectives poses major policy challenges. This paper formally
establishes key requirements for an accelerated poverty reduction process, which is also
similar to the findings of pro-poor growth literature (Kakwani et al 2003). The proposed
framework integrates the role of employment and social policy as a critical nexus
between growth and poverty reduction. A micro-simulation model of tax and transfers,
with links to macroeconomic performance, is used to compare and contrast the
effectiveness of ten policy scenarios to halve poverty and unemployment by 2015. Among
the findings are: (a) both reductions in income inequality and high growth rates are
necessary pre-requisites for an accelerated poverty reduction path; (b) to realise the
unemployment and poverty goals, the labour market needs to be pro-poor in terms of
employment allocation and income; additionally, the employment elasticity of growth
should reach 0.60, and (c) the social security system needs to be substantially expanded.
These findings inform the papers presentation of the basic tenets of a pro-poor economic
policy framework. Overall, the paper suggests an active pro-poor role for the state.


ABSTRACT....................................................................................................................... 2
1. INTRODUCTION........................................................................................................ 4
2. FROM MUTUAL BENEFITS TO PRO-POOR GROWTH................................... 4
A. MUTUAL BENEFITS: A RISING TIDE WILL RAISE ALL BOATS .................................... 5
B. HUMAN DEVELOPMENT IS THE END .......................................................................... 6
3. PRO-POOR ECONOMIC GROWTH AND POLICIES......................................... 8
A. THE DEBATE ON PRO-POOR GROWTH ...................................................................... 9
B. CHANNELS TO PRO-POOR OUTCOMES.................................................................... 11
C. GROWTH-EMPLOYMENT NEXUS ............................................................................ 19
IN SOUTH AFRICA....................................................................................................... 20
A. THE APPROACH ........................................................................................................ 21
B. HOW CAN THE UNEMPLOYMENT RATE BE HALVED BY 2015? ................................. 22
A. PRO-POOR MACROECONOMIC POLICIES ................................................................ 38
B. EMPLOYMENT GENERATION POLICIES ................................................................... 43
C. TRADE AND INDUSTRIAL POLICY ........................................................................... 46
D. COMPREHENSIVE SOCIAL SECURITY SYSTEM ........................................................ 48
6. CONCLUSIONS ........................................................................................................ 49
7. REFERENCES........................................................................................................... 50


Three years ago, South Africa embarked on its second decade of democracy, announcing
objectives that included halving both the unemployment rate and the poverty rate by
2015. Meeting these objectives poses major policy challenges. Halving the current
unemployment rate requires that the economy generate about 3.7 million new
employment opportunities during the next nine years. This is one and a half times the
number of jobs that were created between 1995 and 2004.1 At the same time, halving the
current poverty rate requires serious economic restructuring and commitment of
resources. The aims of this study are to show how the unemployment rate and the poverty
rate can be halved by 2015, and to present an outline of a supportive policy framework.

The paper is organised as follows. Section 2 presents a summary of the economic

development debate that has led to the emergence of the concept of pro-poor growth.
Section 3 summarises the debate on pro-poor growth and provides a formal presentation
of the channels through which growth and poverty are linked. It develops a measure of
the contribution of each channel to changes in poverty, and an aggregate measure of their
overall effect on poverty. It then uses the findings to develop the necessary conditions for
a pro-poor growth path. Section 4 applies the methodology that was developed to
examine the implications of diverse economic scenarios for halving poverty and
unemployment rates during the next nine years. We use a micro-simulation model of
South Africa to simulate the impacts of different scenarios and compare their potential
outcomes. Finally, Section 5 uses the findings of previous sections to present an
economic policy approach that internalises the requirements for fostering growth that is
pro-poor in South Africa.


Development theory and practice have gone through important changes during the last 25
years, with the emergence of such concepts as human development, sustainable
development and pro-poor growth. New thinking in development theory and practice

According to the October Household Survey (Statistics South Africa 1995), the official employment stood
at 9,186,000. The September Labour Force Survey 2004 reported total employment at 11,643,000.

has gradually shifted the focus towards the welfare of current and future generations. In
this context, increasing attention is paid to the analytical, empirical and policy links
between growth, income distribution and poverty. This section provides a brief overview
of the historical context of the new thinking on development.

A. Mutual Benefits: A Rising Tide Will Raise All Boats2

During the 1950s, the development debate centred on industrialisation, informed by the
catching up (or modernisation) perspective of development, whose underlying notion
was mutual benefit. The issue of how to industrialise led to an important debate on the
roles of market and planning, the desirable degree of openness to trade and investment,
the correct blend of capital intensive and labour intensive technology, whether to
prioritise industry or agriculture, and whether to move through an unbalanced or balanced
growth trajectory. Hotly debated issues included the relative role of the state, balanced or
unbalanced growth, and whether the path to industrialisation should be gradual or would
depend on a big push.

The development debate gradually shifted to deeper questions during the 1960s. There
were those who argued that, within the current socio-economic system, a poor countrys
path to development was inherently in conflict with the interests of developed countries;
moreover, that the underdevelopment of a large number of countries was part of the
historical processes of development of developed countries. Thus, the future
development of poor countries is achievable, but necessitates a degree of separation
from the dominant world economic system (Wallerstein 1979, Amin 1988).3 On the other
hand, the mainstream side of the debate believed that the underlying interdependencies
between developed and underdeveloped countries did not amount to a systemic
impediment to the development of the latter. They believed in mutual benefit, and their
mantra was that a rising tide would raise all boats.

According to Sutcliffe (1995), even though there were still differences in perspectives on
development, the different sides shared more or less similar views. That is, (1) as in

This section is based on Sutcliffe 1995.
These positions are known as world systems theory, dependency theory, neo-Marxism, structuralism, etc.

developed countries, development would be characterised by industrialisation, the use of
productive technology, high average standards of living, efficiency, etc; (2) there is a
close or even automatic relationship between the economic aspects of development (e.g.,
growth and rise in productivity) and meeting basic needs and human welfare; (3)
development involves social, economic and political (domestic or international) concerns,
but not necessarily natural ones such as the physical environment; (5) development would
implicitly lead to equalisation between countries through levelling up and not through
redistribution, and (6) development would be a permanent state.

However, the above understanding of the prospect and processes of development,

notwithstanding differences in views, found it difficult to explain the multifaceted
development experiences of the 1960s and 1970s, which included some successes and
many failures. More and more empirical data on country experiences showed persistence
and even a worsening of inequality and poverty. This began to highlight the conceptual
gap between the contemporary understanding of the benefits of development and the
human welfare benefits. This gradually gave way to the existing critical approach and the
emergence of new thinking on development.4

B. Human Development is the End

A new debate began to emerge, with different views on how the goal of development
could be reached. Streeten (1981) and Stewart (1985) advocated the attainment of basic
human needs as a primary rather than a secondary objective of development; others
focused on the problems of individual and group poverty rather than the
underdevelopment of nations (e.g. Lipton (1977). The common characteristic of these
approaches was to view development in terms of what happens to people rather than to
abstractions like nations5, and to define human well being as the purpose and goal of
development. This reverses the earlier idea that welfare is a by-product or necessary
outcome of economic development, as defined earlier. It argues instead that the need for
economic development should be justified by its contribution to welfare.
A series of articles by Dudley Seers (1977, 1979) were an important part of this process.
Sutcliffe 1995, p.239.

A substantial body of literature has emerged during the last twenty years, generating
lively debate on different aspects of human development (HD) among researchers and
policymakers. The Human Development Report of the UNDP, published annually since
1990, has been a major contributor and driving force behind the new thinking on
development.6 It has contributed to the debate by demonstrating that development
involves consideration of much more than just economics, and by focusing on human
poverty as distinct from income poverty (HDR, 1997); capabilities and choices (HDR,
1995); public policies to make new technologies work for human development (HDR,
2001); globalisation and rising new insecurities in peoples lives (HDR, 1999); the needs
of under-consumers against over-consumers (HDR, 1998); human security rather than
the security of national borders (HDR, 1994), and links between growth and human
development (HDR, 1996).

There have, at the same time, been major coordinated international efforts to define the
issues related to HD, and to urge national and international commitments to the
realisation of its goals. Important amongst these was the 1992 United Nations Conference
on Environment and Development (UNCED) in Brazil, which produced what has come
to be known as the Earth Summit Agenda 21. Sustainable development became the
organising principle for Agenda 21, attempting, as it did, to cover many areas of
sustainability and development.

In the mid-1990s, the World Summit on Social Development (WSSD) adopted a

Declaration and Programme of Action that pledged to make the conquest of poverty, the
goal of full employment and the fostering of stable, safe and just societies their
overriding objectives.7 Among the agreements of the WSSD were those dedicated to the
eradication of poverty, support for full employment as a basic policy goal and the
achievement of equality and equity between men and women.8

S. Fukuda-Parr et al (2002) provides an excellent review of contributions of HDR in the area of human
United Nations, Copenhagen Declaration and Programme of Action, the World Summit for Social
Development, 6-12 March 1995, United Nations Department of Public Information, p. vii.
Ibid, p. vii.

At both these and other major conferences, poverty eradication was made a priority and
became one of the major objectives. At the end of the last decade of the 20th century, the
worlds political leaders committed themselves to a set of goals at the Millennium
General Assembly in 2000. The principal objective of what became the Millennium
Development Goals (MDG) was to bring about a significant reduction in human poverty.
Therefore, a hallmark of the 1990s was to highlight the importance of eradicating

Despite its success in redefining the principle objective of development, the usefulness of
HD as a developmental approach depends on analysis that explains how the current
socio-economic system can be steered to respond effectively to human needs and provide
for human wellbeing.

Thus, the recent emergence of a massive literature on pro-poor growth and pro-poor
economic policies can be viewed as a response to calls for policy frameworks that
internalise the realisation of HD objectives.

3. Pro-Poor Economic Growth and Policies

The historical context that underlies the literature on pro-poor growth and policies
includes the widely recognised increased poverty and inequality outcomes of the trickle-
down development approach and the policy practices of structural adjustment
programmes over the last three decades (Box1). However, the new literature represents a
highly heterogeneous effort by both mainstream and non-mainstream analysts to define
what is meant by pro-poor, and to incorporate the analysis of poverty in the traditional
discourse on macroeconomic issues and policies.

Structural Adjustment Program
Structural adjustment policies encompass a set of short-run stabilisation measures and longer run
adjustment policies that are implemented together or sequentially. The stabilisation policies include
following measures:

fiscal policies aimed at reducing the public budget deficit;

monetary policies aimed at reducing the inflation rate or the money supply;
wage and price policy to control inflation in support of the above policies;
exchange rate policies to reduce the balance of payment deficits.

The adjustment policies comprise policies designed to alter the working of the product and factor markets
by removing price controls and subsidies through a process of liberalization and deregulation. Structural
adjustment programme (SAP) policies include trade liberalization to strive for freer trade; deregulation to
make enterprises more profitable, and privatisation of publicly owned enterprises. Needless to say, some
aspects of adjustment policies (e.g., privatisation of state enterprises) complement parts of stabilization
policies (e.g., deficit reduction).

Historically, the stabilisation and adjustment policies of SAPs have, in many countries, produced low rates
of economic growth, environmental degradation, together with increased unemployment rates, income
inequality and poverty. These results have been linked to the dynamic outcomes of the overall framework
and policies of SAPs. This includes, for example:

setting monetary and fiscal objectives and targets that result in the undercutting of public services,
reduce internal demand, aggravate unemployment, and frustrate efforts to increase investment in
infrastructure, SMME development and human development;
using tariff reductions as de facto industrial policy. Accelerated import tariff reductions often result in
a flood of imports that undermine domestic industrial and agricultural producers;
reducing labour protections and benefits by adopting labour market liberalisation measures.

A. The Debate on Pro-Poor Growth

Pro-poor growth refers to the interrelationship between growth, inequality and poverty
(Kakwani, Khandker and Son 2003:4). There are at least two main views on what is
meant by pro-poor growth. These are distinguished by the ways they utilise the absolute
or relative notions of pro-poor growth. The former considers only the incomes of the
poor and the rate at which absolute poverty is reduced. The latter refers to relative
changes in the incomes of the poor and their impact on poverty.

Therefore, the position adopted by one group (advanced by the UNDP among others)
argues that redistribution needs to take its place firmly beside growth as the means to
address poverty. Kakwani and Pernia (2000) argue that, if the trickle-down development
approach of the neo-liberal policies results in growth (rise in average per capita income),

it will also reduce poverty, but the rate of poverty reduction will be much slower. Rapid
poverty reduction necessitates enhancing growth that delivers proportionally greater
benefits to the poor than to the rich, i.e. a reduction in inequality (Kakwani et al 2003,
HDR 2005). More recently, the Human Development Report also uses the concept of
progressive growth to define a growth pattern in which average incomes are growing,
but the incomes of poor people are growing even faster. Therefore, progressive growth is
viewed as a positive-sum process in which nobody loses and the poor gain proportionally
more (HDR 2005).

The second position, which is frequently advanced by World Bank economists and the
International Monetary Fund (IMF), is that any growth is pro-poor. Proponents of this
position argue that increasing inequality is not a problem as long as the poor gain in the
process.9 In other words, any increase in the incomes of the poor is evidence of pro-poor
growth (regardless of their gains relative to the income gains of the non-poor income
gains). An influential paper in this group is by Dollar and Kraay (2000), published by the
World Bank. Its central message is captured in its title, Growth is good for the poor.
Even though the paper does not claim that growth is the only factor that matters for
poverty reduction, it centres on the primacy of growth as a means of addressing
poverty.10 This is consistent with the traditional neo-liberal view of income inequality,
which is not really treated as a concern since, in well functioning markets, individuals
earn rewards equal to their contribution to output, i.e., their marginal product. So the rich
are rich because they are more productive, and the opposite holds for the poor.11

An example of this type of thinking is found in a World Bank sponsored comparative analysis of 21 case
studies (Lal and Myint, 1996).
If the Dollar and Kraay paper had been a just a research exercise, Oxfam observes, it could have been
dismissed as weak and largely irrelevant. Sending the message that it did (leave it to the markets and
let the benefits of growth trickle-down), and bearing as it did, the imprimatur of the World Bank,
(Oxfam, 2000) it aroused a storm of protest (Dagdeviren et al, 2000; Lbker et al, 2000; Oxfam, 2000;
Wade, 2001).
The above two positions also underlie an intense ideological battle. For example, as Wade (2001) shows,
the controversy that followed the publication of the article by Dollar and Kraay (2000) reflected an
ideological struggle within the Bank to push equity considerations well into the second place behind
growth, even when research at the Bank was beginning to show that hoped-for impact of Bank lending on
poverty was not being achieved (Easterly, 2001), and that inequality matters to the pace of poverty
reduction that is achieved at any given rate of growth. (Ravallion, 2001, p.17).

- 10 -
The World Bank, which started with the weaker definition of pro-poor growth (using
the absolute pro-poor growth approach) has, in recent years, gradually come to
recognise the importance of reducing inequality as a means of fighting poverty. This is
evident from the poverty-growth-inequality triangle described by Bourguignon (2004),
which points out that changes in poverty are a function of growth in mean income,
changes in the distribution of income and initial distribution. Overall, as Klasen (2003)
notes, the literature suggests that pro-poor growth policy should put a significant
emphasis on reducing inequalities in order to accelerate poverty reduction, increase
growth, and enhance the poverty impact of such growth.

The next section develops a formal presentation of a different nexus between growth and
poverty, which is used in section 4 to analyse the implications of diverse growth
scenarios for halving poverty and unemployment rates in South Africa by 2015.

B. Channels to Pro-Poor Outcomes

What are the channels through which growth and inequality impact on poverty? To
answer this question, we need to be able to identify the relevant channels, develop a
measure of the contribution of each channel to changes in poverty and, if possible,
provide an aggregate measure of their overall effect on poverty. It is also desirable to go
one step further and use the findings to develop the necessary conditions for a pro-poor
growth path.

As discussed in the previous section, poverty reductions, especially rapid reductions in

poverty, depend on two important factors. First, the size of the economic growth rate: that
is, the higher (lower) the growth rate, the larger (smaller) the poverty reduction will be.
Second, changes in inequality that generally accompany economic growth: that is, a rise
(fall) in inequality decreases (increases) the impact of growth on poverty reduction.
Kakwani et al (2003) derives a measure of total elasticity of poverty that captures the net
effect of the above two factors on the overall poverty index, and uses it to define the
formal conditions for pro-poor growth. In the following pages, we adopt a different
formal approach to derive a similar measure of total elasticity of poverty and conditions

- 11 -
for pro-poor growth.12 The approach is informed by the need to include explicitly the
links between poverty and both the labour market and governments poverty related
expenditures. Osmani (2002) defines these channels as the personal income channel and
social provisioning channel. The former refers to the growth of the economy, which
through employment, translates into higher personal income amongst the poor. The latter
refers to the resources generated by growth (e.g. taxes) that can potentially be used by a
society to provide services to the poor.

Following Kakwani (1993, 2003), the general class of poverty measures (v) can be fully
represented by a poverty line (z) that captures the countrys minimum standard of living,
the mean household income (y) and the Gini index (G) as a measure of inequality.

v = v( z , y, G ) with v/z>0; v/y<0; v/G>0 [1]

In equation 1, the mean household income (y) is defined as the average household
income after tax and includes current transfers from general government.

The next equation expresses mean household income as a function of employment, E,

and governments cash and in-kind transfers to households, S. Therefore, we can write:

y = y( E, S ) with y/E>0; y/S>0 [2]

In equation 2, variable E captures the impact of the level of employment in the economy
and the associated rate of compensation (i.e. total compensation of employees) on the
mean household income. It, therefore, reflects both the quantity and quality of
employment in the economy.13 Our assumption is that the average household income is a
positive function of both the level of employment in the society and its quality.
Therefore, for example, it is assumed that, ceteris paribus (all things being equal), the

Kakwani et al (2003) uses a measure of the average deprivation in the society for which he uses f(x) as
probability density function of individual income (x) that is a random variable.
The quality of employment refers to those attributes of employment that have potential impacts on
returns on labour, such as formal versus informal employment, union membership, and sectoral and
occupational differences.

- 12 -
average household income increases (decreases) when the share of total employment in
formal, manufacturing or unionised sectors of the economy increases (declines).14

In equation 2, variable S captures the impact of cash and non-cash government transfers
to households on the mean household income.15 It is closely related to the concept of the
social provisioning channel used by Osmani (2002 and 2004). According to equation 2,
an overall rise in employment ceteris paribus is expected to be accompanied by an
overall increase in mean household income. Similarly, mean household income is
expected to increase ceteris paribus as a result of the states cash and in-kind transfers
to households.

The next equation expresses total employment (E) as a function of the size of the
economy, which is represented by the real gross domestic product (GDP). It is assumed
that employment will increase (decrease) with the rise (fall) of GDP and suggests that
economic growth is central to any employment, and as we will see later, poverty
reduction focused strategy.16 Thus:

E = E (GDP ) with E/GDP>0 [3]

Finally, the last equation of the model captures one of the basic empirical relationships
between the Gini-coefficient measure of inequality (G) and mean household income.
Since mean household income is directly related to the size of the economy, studies have
found that a positive or negative relationship between inequality and growth implies a
parallel relationship between inequality and mean household income (Kuznets, 1955;
Blank, 1989; 17 Arnand and Kanbur, 1984; and Deininger and Squire, 1998). Therefore,

In 2006, compensation of employees constituted a little less than three quarters of household current
income (Reserve Bank, March 2007, S-128).
In 2005, governments total cash and in-kind transfers to households amounted to about R220 billion.
The cash transfer was almost R80 billion, or one third of the total amount (Reserve Bank, June 2006, S-
Osmani (2002) points out, without economic growth, there can be no sustained expansion in
employment. Only a growth-induced shift in employment will make it possible for the poor to enjoy rising
income either through reduced unemployment/ underemployment or through higher returns on labour.
The only way employment can be expanded in a stagnant economy is either by depressing the returns to
labour or by increasing the rate of unemployment. Neither route is good for the poor.
For reference see Chuhay (2004).

- 13 -
equation 4 postulates that income distribution in a country is a function of mean
household income:

G = G( y) with G/y <>0 [4]

As will be discussed later, it is not possible to express a priori the sign of the relationship
between mean household income and inequality.

The total differentials of the above equation system yield:

v v v
dv = dz + dy + dG
z y G
y v
dy = dE + dS
dE = dGDP
dG = dy

Assuming that the poverty line (z) is kept constant in real terms, the above system can be
reduced to one central expression that captures the channels through which changes in the
real growth rate of the economy impact on the poverty index:

dv / v
= = ( yv + Gv yG ) Ey GDP
+ ( yv + Gv yG ) Sy GDP
+ Gv GDP


= total elasticity of poverty as a measure of the overall rate of decline in the poverty
index, due to a 1 percent real increase in GDP.

yv = poverty elasticity (partial elasticity) of mean household income as a measure of the

rate of decline (increase) in the poverty index, as a result of a small increase (decrease)
in mean household income. It is expected to be negative, (v/v)/ (y/y)<0.

- 14 -
Gv = poverty elasticity of the inequality index, using Gini index. (v/v)/(G/G)>0 is
expected to be positive, suggesting that the poverty index declines (increases) with
reductions (rises) in inequality.

yG = inequality elasticity of mean income can be either positive or negative.

Ey = income elasticity of employment is expected to be positive (equation 2).

= employment elasticity of growth is assumed to be positive (equation 3).

Sy = income elasticity of social provisioning is assumed to be positive (equation 3).

= social provisioning elasticity of growth can be either positive or negative.

= inequality elasticity of growth which can be either positive or negative.

Equation 6 can be expressed in a more compact form as:

= + + [7]


represents a combination of elasticities related to the employment nexus between

growth and poverty. It is a measure of how much a small increase in GDP reduces
poverty through employment. It captures both the income and the inequality
effects. In the rest of this paper, we refer to as the poverty elasticity of

represents a combination of elasticities related to the social provisioning nexus

between growth and poverty. It is a measure of how much a small increase in
GDP reduces the poverty rate through the social provisioning channel. It captures
both the income and the inequality effects. In the rest of this paper, we refer to
as the poverty elasticity of social provisioning.

- 15 -
measures the direct effect of 1 percent increase in GDP on inequality (Gini index).

measures the increase in the poverty index as a result of a small increase in the
inequality index.

Equation 7 shows clearly how a 1 percent increase in GDP is channelled through

employment and social provisioning to impact on mean household income and inequality,
which in turn impacts on the total poverty rate.

In Equation 7, captures whether growth is inherently accompanied by rising or

declining inequality. Kuznets (1955) suggests that inequality initially worsens as
economic development takes off; but, in the later stage of development, inequality begins
to improve. Recent empirical studies have questioned this proposition (Anand and
Kanbur 1984; Deininger and Squire 1998) and have found no support for Kuznets
inverted U-shaped pattern of income inequality. Recent studies (Kakwani and Son 2002,
Bourguignon 2004) have argued that it is not possible to state a priori the sign and
magnitude of , since how changes depends on a countrys initial level of economic
development, inequality and policies.

In order to separate the channels through which changes in inequality impact on poverty,
equation system [5] is summarized differently to derive the following equations:

dv / v
= = yv ( Ey GDP
+ Sy GDP
) + Gv ( yG Ey GDP
+ yG Sy GDP
) + Gv GDP


= + [9]


= yv ( Ey GDP
+ Sy GDP
) represents the growth elasticity of poverty (Kakwani, 1993),

which is the percentage change in poverty due to a 1 percent increase in economic

- 16 -
growth, provided that the growth process does not change inequality (i.e. the benefits of
growth are distributed equally among everyone in the country).

= Gv ( yG Ey GDP
+ yG Sy GDP
) + Gv GDP
represents the inequality elasticity of poverty. It is

an aggregate measure of all the channels through which a 1 percent increase in GDP
impacts on poverty through its impact on inequality. In other words, it is a measure of
how much the changes in the total poverty index relate to changes in inequality, given
a 1 percent increase in GDP.

Equation 9 shows that the total poverty index is equal to the sum of two combinations of
elasticities. The first () is an extension of Kakwanis concept of growth elasticity of
poverty (Kakwani, 1993). It measures percentage change in the poverty index that results
from the impact of a 1 percent increase in GDP on employment and social provisioning,
provided that the growth process does not change inequality.

The second combination of elasticities () a measure of the inequality elasticity of

poverty is the sum of different channels through which a 1 percent increase in GDP
impacts on poverty through its net effects on inequality. captures three channels
through which growth impacts on inequality. The first and second measures capture the
impacts of growth on inequality through employment and social provisioning channels
( yG Ey GDP
, yG Sy GDP
). The third measures the direct effect of growth on inequality


Kakwani, Khandher and Son (2003) show that economic growth is pro-poor (pro-rich) if
the change in inequality that accompanies growth reduces (increases) total poverty. In the
above system, this implies the growth is pro-poor if both total elasticity of poverty ()
and the growth elasticity of poverty () are negative and ||>||. These two criteria for
pro-poor growth are satisfied under the following conditions and assuming that the mean
household income elasticity of growth is positive ( GDP
> 0 )18:

This implies that total household income in the economy is expected to grow at a faster rate than the
growth rate of the number of households in the country.

- 17 -
a) if GDP
> 0 and GDP
< 0 (therefore yG < 0 ), both and will be negative, which

implies that ||>||. This implies that the pro-poor growth conditions are met if
governments spending on social provisioning increases with economic growth, and if, at
the same time, economic growth is accompanied by a decline in inequality. Relative to
other possibilities, under these conditions, ceteris paribus, the pro-poorness channels of
the growth process will be the most extensive.

b) if GDP
> 0 but GDP
> 0 (therefore yG > 0 ), both and will be positive, which

implies that >>0. This implies that the growth process is not pro-poor (is pro-non poor)
because poverty is expected to worsen through economic growth, both directly and
through the increase in inequality.

c) if GDP
< 0 but GDP
> 0 (therefore yG > 0 ), most probably, will be positive, which

implies that a direct effect of a 1 percent increase in GDP is to worsen (increase)

inequality, i.e. growth is not pro-poor. The growth elasticity of poverty () is also
weakened due to a decline in social provisioning as the economy grows by 1 percent.
Overall, the growth path is not pro-poor.

d) if GDP
< 0 and GDP
< 0 (therefore yG < 0 ), will be negative. The negative social

spending elasticity of growth means that one component of will be positive and the
second part will have a negative value. The overall value of will, most likely, be
negative, thus satisfying one of the conditions for pro-poor growth. This combined with
negative GDP
implies that the second condition (||>||) is also satisfied. Therefore, even

without a negative correlation between changes in social provisioning and economic

growth, as long as the employment intensity of growth is relatively high, the economic
growth path will be pro-poor. In this case, a strong poverty elasticity of employment
ensures that the poverty elasticity of growth is negative (i.e., reduces poverty) and the
negative inequality elasticity of growth ensures that economic growth, by reducing
inequality further, facilitates reductions in the poverty index.

- 18 -
C. Growth-Employment Nexus

The third equation in the equation system 5 reflects the total change (the total differential)
of the employment equation 3. It can be written as:19

GDP [10]

where E and GDP represent the growth rates of employment and the economy,
respectively; and GDP
is the employment elasticity of growth.

Equation 10 suggests that the growth rate of employment is proportional to the real
growth rate of the economy, where its proportionality is determined by the size of the
employment elasticity of growth. Therefore, for a given rate of economic growth, the
higher the employment elasticity of growth, the larger will be the rate of growth of

Using equation 10, it is also possible to derive the relationship between the
unemployment rate and economic growth. Let:

LF represent the labour force,

equal the employment rate (E/LF)

equal the unemployment rate, (LF-E)/LF

Using the definitions of employment and unemployment rates and equation 10, we derive
equation 11 for the rate of change of the employment rate:

GDP LF 0 [11]


Equation 10 is derived by dividing both sides of the expression dE in equation 5 by E, multiplying the
right side by GDP/GDP, and defining GDP
= (E / GDP )(GDP / E ) .

- 19 -

= 0 0 = E /( LF E ) [12]

that is, as the employment rate increases (decreases), the unemployment rate
proportionally declines (increases).

Substituting equation 12 in equation 11 provides:

= 0 ( LF 0 GDP
GDP) [13]

Equation 13 implies that, for a given growth rate of the labour force, the unemployment
rate declines (increases), with a higher rate of economic growth and higher employment
elasticity of growth. It also implies that the rate of change of the unemployment rate is
affected by the initial ratio of employment to unemployment (). In other words, for a
given rate of economic growth, employment elasticity of growth, the growth rate of the
labour force, if initially is high (low), the unemployment rate declines faster (slower).

Our analysis of channels through which growth and inequality impact on poverty
highlights the complexity of the underlying relationships and the conditions for pro-poor
growth. Moreover, equation 13 captures the relationship between economic growth and
the evolution of the unemployment rate. In the next section, we apply the above
methodology to examine the implication of diverse growth scenarios for the evolution of
poverty and unemployment in South Africa.



The South African government is committed to halving unemployment and poverty rates
by the year 2015.20 The presentation in section 3 lays the analytical foundation for an
empirical examination of the potential of different economic paths to achieve these
objectives. In this section, we define four distinct groups of economic scenarios that are

ANC 2007:3.

- 20 -
designed to open and widen the poverty reduction channels identified in section 3 and
estimate their effects on unemployment and poverty rates.

A. The Approach

We use a micro-simulation model of South Africa to simulate the impacts of different

scenarios and compare their potential outcomes. The criteria we use are: (a) lowering the
poverty rate, the faster the better; (b) reducing the unemployment rate, the faster the
better, and (c) decreasing income inequality, the more the better. Moreover, the
acceptable paths are expected to (d) possess positive per capita growth, and (e) follow a
sensible fiscal position. The model, South African Tax and Transfer Simulation Model
(SATTSIM-Plus), was built by Applied Development Research Solutions (ADRS) for the
quantitative assessment of the impacts of alternative macroeconomic, tax and transfer
scenarios on poverty, distribution and budget trends in South Africa. It includes a
macroeconomic performance module, two tax modules (income tax and value added tax),
and six social security modules (child support grant, old age pension, disability grant,
care dependency grant, basic income grant, and care giver grant). We have used this
model to generate projections for the period 2007 to 2015.

The model allows us to specify values for a number of exogenous variables to define and
run a new policy scenario. These can be divided into the variables that are subject to
policy control and those that fall outside such control. For a given set of future values for
a particular macroeconomic trend and policy variable, the model predicts the time paths
(within the next ten years) of its endogenous variables. These include: the poverty rate
and the poverty gap; income inequality, represented by the Gini-coefficient; the
distribution of employment and income among poor and non-poor households; the
allocation of social security grants among eligible groups within the population; the
budgetary implications of current and alternative grant programmes; government
revenues from income tax and value added tax, and the employment and unemployment

- 21 -
rates. In short, the model links the welfare outcomes of alternative tax and social security
policies with different assumptions about the countrys macroeconomic performance. 21

B. How Can the Unemployment Rate Be Halved by 2015?

The South African population is expected to grow by about 3.5 million between 2007 and
2015, or by an average annual rate of less than 1 percent (0.8%).22 During the same
period, it is estimated that the labour force will grow by about 1.65 million, assuming that
the labour force participation rate remains at the 2006 rate of 56 percent.

Given these basic demographic trends, and assuming that the economy will grow at an
average annual real rate of 5 percent between 2007 and 2015, 23 how can the economy
generate enough jobs to gradually reduce the unemployment rate by half by 2015?
Secondly, what are the poverty and inequality effects of different employment creating

To answer these questions, Group A scenarios are designed, using different assumptions
about the employment intensity of the growth paths ( GDP

Scenario A1: Low employment intensive growth path

The specifics of scenario A1 are as follows:

The economy will grow in real terms at an annual rate of 5 percent.

The average annual inflation rate will be 5 percent.

The employment elasticity of growth, gdp

, will be 0.20 between 2007 and 2015. This

means that a 1 percent increase in the real growth of the economy is expected to
generate a 0.2 percent increase in total employment.

For more information about the model used for this paper, please contact the author at
We have used the population projections of Statistics South Africa for 2007 to 2011. For the period 2012
to 2015, we have used own population projection using StatsSA data.
The governments growth projections for the next 9 years, including 2007, are as follow: 5 percent
growth until 2009 followed by 6 percent growth rates for the period 2010 to 2015. To simplify, we have
adopted 5 percent average annual real growth rate between 2007 and 2015.

- 22 -
The poverty line of R669 per month per adult equivalent for 2006 will be annually
adjusted to the inflation rate.24

The current distribution of income between poor and non-poor households and
governments current income tax, value added tax and social welfare programmes
remain unchanged over the next nine years.

The real value of different grants and their related means tests remain constant over
the next nine years. This implies that the grant amounts and the means test are
adjusted annually to the inflation rate.

The annual additions to total employment are allocated between the poor and the non-
poor, using their current share of employment about 15 percent and 85 percent

The differences between the average salaries of the employed poor and the employed
non-poor remain constant in real terms. Salaries of both groups are assumed to
increase annually in real terms by 3.5% percent.

Scenario A2: Medium employment-intensive growth path

Scenario A2 is the same as scenario A1, with only one exception. The growth process is
assumed to be twice as job creating as in scenario A1. Specifically, the employment
elasticity of growth is assumed to be 0.40.

Scenario A3: High employment-intensive growth path

The specifics of scenario A3 are the same as for scenario A1, except that the growth path
of the economy is associated with much higher job creation. For scenario A3, the

The estimation of the poverty line is based on (a) the calculation of the minimum dietary requirements of adult
individuals, (b) the estimation of the share of food in total expenditure of the bottom 20 per cent of the South African
population, and (c) the application of the Orshansky (1965) method of estimating poverty thresholds by dividing the
cost of the minimum dietary requirement for a person by the above share.

- 23 -
employment intensity of growth is set at 0.60, which is considered high by international

Findings on the impact of Group A scenarios on employment:

The first scenario (A1) will generate about 1.1 million new employment opportunities
between 2007 and 2015. This is about half a million fewer than the expected increase
in the size of the labour force in the same period. Therefore, under this scenario, the
unemployment rate is expected to increase by 0.6 percent by 2015, which is an
undesirable outcome.

By doubling the employment-intensity of growth from 0.20 to 0.40, simulation results

show that scenario A2 will be able to create more employment opportunities than the
expected increase in the labour force in the next nine years. As a result, the
unemployment rate is expected to decline from 25.8 percent in 2006 to 19.7 by 2015
(Table 1). Therefore, scenario A2 will also be unable to reduce the unemployment
rate by half by 2015.

It is only when the growth process is associated with a much higher rate of
employment every year that the economy will be able to create enough jobs to halve
the unemployment rate by 2015. Scenario A3 reflects a growth path that is capable of
generating about 3.7 million additional employment between 2007 and 2015, and
reducing the unemployment rate from 25.8 percent to 12.4 percent during this period
(Table 1). Therefore, with scenario A3, the economy will generate enough
employment annually to gradually reduce the unemployment rate by half by 2015.

Findings on impact of Group A scenarios on poverty

The simulation results show that:

The doubling and tripling of the employment intensity of growth in Group A

scenarios partially expands employment opportunities for the unemployed poor.
However, since the pattern of allocation of new employment is unchanged, the

For a cross country comparison of employment elasticity of growth see Islam 2004.

- 24 -
economic paths that underlie Group A scenarios benefit the non-poor proportionally
more than the poor. Consequently, only 15 percent of new employment opportunities
and their related income are allocated annually to poor families.

Higher employment intensity of growth results in higher mean household income

(equation 2), which in turn directly and indirectly helps increase the pace of poverty
reduction (equation 1) (Table 2). The direct effect is through poverty elasticity of
income ( yv ), which is relatively higher (in absolute value terms) in scenario A3; the

indirect effect is through income inequality, which is also relatively higher (in
absolute value terms) in scenario A3 (Table 3).

Higher employment-intensive growth is also accompanied by the greater

effectiveness of growth in reducing poverty through the social provisioning channel.
This is mainly due to the decline in the unemployment rate. As a larger number of
jobs are created and members of poor households obtain employment, the depth of
poverty declines, increasing the ability of social welfare programmes to move poor
households out of poverty. Therefore, simulation results show that the poverty
elasticity of social provisioning (), which in this case captures the ability of social
grants to move poor households out of poverty, is higher for the economic path that is
more job creating (Table 3).

The economy in scenario A3 is also more effective in reducing inequality than the
low employment-intensive growth path (scenario A1). Therefore, the Gini-coefficient
will be lower with scenario A3 in 2015 than with scenarios A1 or A2 (Table 2). This
is mainly due to a relatively larger increase in household mean income as the
economy generates a high level of new employment opportunities every year under
scenario A3.

Scenario A3 is more effective as it reduces the overall poverty index through a

reduction in income inequality. This is reflected in scenario A3s higher (in absolute value
term) inequality elasticity of poverty () (Table 4).

- 25 -
Overall, the economic path with the highest employment intensity of growth (scenario
A3) is almost twice as effective as scenario A1 in reducing poverty (Table 2);26 but it is
also quite insufficient to halve the poverty rate by 2015. The three scenarios in group A
demonstrate that, given the current structure of the labour market, halving the unemployment rate
by creating large numbers of employment opportunities (about 3.7 million between 2007 and
2015) will be insufficient for halving the poverty rate at the same time.

Moreover, the economic path of scenario A3 may be unsustainable, since about three-
quarters of the annual new employment opportunities are expected to go to non-poor
households already experiencing a much lower rate of unemployment. This means that
the dynamic of the current scenarios is predicted to generate further expansion of the
wage gap and worsening inequality and poverty. The solution is to make the employment
creation process more pro-poor in order to (a) avert increases in the excess demand for
employment in the non-poor households, and (b) accelerate the pace of poverty reduction.

C. From Halving Unemployment to Halving the Poverty Rate

The focus of the next set of scenarios (Group B) is to explore further the potential of
using labour market reforms to achieve our twin goals. Specifically, the question is: if the
job creation process becomes significantly pro-poor, will that be sufficient to halve both
the unemployment and poverty rates by 2015?

In section 3, we showed that the evolution of the poverty index relates, in a complex
manner, to the dynamic of inequality. Due to its underlying assumptions, Group A
scenarios have a relatively low impact on reducing inequality. Their impact on poverty is
mainly the result of employment-induced poverty elasticity of growth ().

The Group B scenarios build on scenario A3 and add two types of pro-poor measures to
the labour market. The measures are designed to make both the annual allocation of new
employment opportunities and the wage rates associated with different types of
employment gradually pro-poor. This would change the thrust of the economic path so

The relevant total elasticity of poverty are 1.095 and 1.923.

- 26 -
that the poor begin to benefit proportionally more than the non-poor from the expansion
of the labour market (Table 5).

Scenario B1:

Scenario A3, with an annual 1 percent increase in the allocation of new employment
opportunities to the unemployed from poor families, and an annual 0.5 percent increase
in the average real salary of new employment opportunities for the unemployed poor.
Therefore, the poor households share of total new employment increases from 15
percent in 2006 to 24 percent in 2015 (Table 5).

Scenario B2:

Scenario A3, with an annual 2 percent increase in the allocation of new employment
opportunities for the unemployed from poor families, shows an annual 1 percent increase
in the average real salary of new employment opportunities for the unemployed poor.
Therefore, the poor households share of total new employment increases from 15
percent in 2006 to 33 percent in 2015 (Table 5).

Analysis of Findings

Since Group B scenarios are built on scenario A3, they meet the goal of halving the
unemployment rate by 2015. At the same time, relative to scenario A3, the pro-poor
measures in Group B help increase the effectiveness of the three poverty channels:

The poverty elasticity of employment () increases (in absolute value terms) from
-1.009 in scenario A3 to -1.125 in scenario B2. This implies that the increased
pro-poor allocation of new employment and the annual real increase in the
average salary of the employed from poor families make the job-creating aspect
of the growth process more poverty reducing (Table 3).

The increase in the poverty elasticity of social provisioning () implies that the
social security system will become relatively more effective in moving recipient

- 27 -
families out of poverty, as more and better employment opportunity for the poor
help reduce the depth of poverty.27

Finally, relative to scenario A3, Group B scenarios, especially scenario B2,

decrease income inequality due to their higher mean household income and
gradual reductions in the wage gaps between the poor and the non-poor.
Consequently, Group B scenarios are relatively more effective in reducing
poverty through the inequality channel. This implies relatively higher inequality-
elasticity of poverty () and = (Tables 3 and 4).

Overall, the results of Group B scenarios show that the economic path that includes a
more pro-poor labour market measures embodies a stronger employment, social
provisioning and inequality nexus between poverty and growth (, , ). Therefore,
relative to scenarios A3 and B1, the growth process in scenario B2 is more effective in
reducing poverty ( B 2 > B1 > A3 ). However, even though poverty rate declines from 44

percent to 30.4 percent in scenario B2, the pace of poverty reduction is not sufficient to
halve the poverty rate by 2015 (Tables 2 and 3).

The results of Group B scenarios suggest that we need to go beyond labour market
reform and examine the possibilities of releasing the capacity of social provisioning to
help directly (through raising the income of the poor, i.e., increasing the growth elasticity
of poverty, ) and indirectly (through reduction in inequality, i.e., increasing the
inequality elasticity of poverty, ) increase the pace of poverty reduction.28

D. Effective Ways to Halve Both Unemployment and Poverty Rates

In South Africa, there is an ongoing debate about the need to establish a comprehensive
social security system that expands current government programmes.29 The social
provisioning component of the model presented in section 3 captures the links between
social welfare programmes and reductions in inequality and poverty. Group C scenarios
Poverty gap remains basically unchanged in scenario A3. However, in scenario B2, poverty gap declines
by 1.3 percent.
Please note that changes in the elasticities are in terms of their absolute values.
Department of Social Development 2002.

- 28 -
explore the implications of expanding the current social protection system in South
Africa, in terms of both coverage and financial support. Since the current South African
social security system has been a successful policy tool in reducing poverty, our working
assumption is that the expansion of the current social security system has the potential to
help achieve an economic path capable of achieving the goal of halving the poverty rate.
Therefore, the key question is whether reasonable extensions of the social security system
can help halve the poverty rate by 2015.

Scenario C1:

The same as scenario B2, except that the age requirement for the child support grant is
extended to include children younger than 18 years old, starting in 2007. The amount and
adjustment of all grants remain the same as in scenario B2 (Table 5).

Scenario C2:

Scenario C2 adds to scenario C1 a new grant programme for unemployed adults (i.e.
adult grant). For the purposes of simulation, scenario C2 adopts the adult grant
programme from 2007, setting the amount of the grant at R500 a month. The annual
adjustment of the adult grant will be the same as for the other grants (Table 5).

Scenario C3:

The same as scenario C2, except that all grant amounts are allowed to increase by 2
percent above the inflation rate every year, starting in 2007 (Table 5).

Analysis of Simulation Results (Scenario C1)

Scenario C1 brings an additional 1.9 million children from poor families under the social
grant umbrella. Consequently, the budget for the child support grant increases by about
R5 billion a year. However, the extension of the child support grant, which also increases
the disposable income of poor families, is estimated to raise value added tax revenue by
about R0.5 billion a year in the earlier periods, and by about R1 billion a year in the later
periods. Therefore, the net budgetary impact of scenario C1 is between R4 billion and
R4.5 billion a year.

- 29 -
Given the pro-poor bias of the child support grant, scenario C1 helps increase the average
income of poor families by about R1,000 a year, and reduces overall income inequality,
the poverty rate and the poverty gap by 2015. Therefore, relative to scenario B2, the
poverty rate will be 0.7 percent lower in 2015 (Table 2).

Overall, the economic path that reflects scenario C1 is a more effective route to halving
the current poverty rate than are the earlier scenarios, which did not include the
expansion of the child support grant. This is reflected in a higher, in absolute value terms,
total elasticity of poverty (=-2.313) than the corresponding elasticity for scenario B2 (
= -2.145). The main contributions to this outcome are the increases in the poverty
elasticity of social provisioning ()30, and the inequality-linked poverty elasticity of
growth (=).31 Relative to scenario B2, the absolute value of the former elasticity
increased from 0.457 to 0.493; in the latter case, it increased from 0.563 to 0.607.
Overall, the poverty rate declines from 44 percent in 2006 to 29.3 percent in 2015 (Tables
2 and 3).

Scenario C1 highlights the relative importance and potential of social security to combat
poverty. At the same time, it shows that the extension of the scope of child support is not
a sufficient measure to help halve the current rate of poverty by 2015. Therefore, scenario
C1 paves the way for the next scenario.

Analysis of Simulation Results (Scenario C2)

The debate on the need for a comprehensive social security system in South Africa
includes the call for the expansion of the current grant system to provide support to poor
or unemployed adults between 18 years old and the retirement age. The purpose of
scenario C2 is to add an adult grant programme to the current social security system and
assess its impacts. Therefore, in scenario C2, in addition to the extension of the child
support grant (scenario C1), we introduce an adult grant programme for unemployed men
and women. The grant is designed to help broadly defined unemployed adults and to

which captures the social provisioning nexus between poverty and growth.
The referred increases in the elasticities are in terms of their absolute values.

- 30 -
cover men aged between 18 to 64 and women aged between 18 and 59.32 The grant
amount starts at R500 a month in 2007 and is expected to adjust annually to the inflation
rate, like the other grants (Table 5).

For 2007, the model estimates that there are almost 5.6 million unemployed adults.33
Over time, however, the net effect of the evolution of the population and the rise in
employment that underlies scenario C2 show a gradual decline in the overall number of
those eligible for the adult grant. By 2015, the number of those eligible for the adult grant
is estimated to have decreased by about 2 million persons.

The adult grant requires an initial disbursement of R33.6 billion in the first year of its
introduction. However, even though the nominal value of the grant adjusts annually to the
inflation rate, the total cost of the grant is expected to decline gradually, due to the
gradual fall in the number of those eligible for the grant. Therefore, in 2015, the cost of
the adult grant is estimated at R32.8 billion. At the same time, the disbursement of the
grant generates an estimated annual increase of between R2.5 and R3 billion in total
revenue from income tax and value added tax.

The adult grant helps establish an economic path with a faster rate of poverty reduction.
As a result, by 2015, the poverty rate is expected to be 1.6 percent lower than the result
from scenario C1. The poverty gap also declines substantially (by 3.1 percent relative to
scenario C1) between 2007 and 2015 (Table 2).

Overall, the economic path of scenario C2 results in lower poverty rates than previous
scenarios, since it embodies multiple improvements in poverty reduction channels. In this
case, all the three composite elasticities associated with the employment, social
provisioning and inequality nexus between growth and poverty (,, and ) improve
with the introduction of the adult grant. Consequently, with scenario C2, the total
elasticity of poverty () will be higher than for previous scenarios and poverty rate
declines from 44 percent to 27.7 percent by 2015. Despite its desired effects, the

The broad definition was used since the objective is to reduce poverty. At the same time, the scenario
can be redefined in terms of who should receive such a grant.
The estimated number of unemployed adults does not include the disabled adults.

- 31 -
simulation results show that the introduction of an adult grant programme in scenario C2
is not sufficient to lower the poverty rate sufficiently to reach the objective of halving the
poverty rate by 2015.

Analysis of Simulation Results (Scenario C3)

The simulation results of scenario C2 show that, despite the expansion of the social
security system and the annual full adjustment of all grants to the inflation rate, the
magnitude of social grants relative to GDP declines gradually from its peak of 5.7 percent
in the first year of implementing the measures (2007) to 3.7 percent of GDP in 2015.34
Similarly, the share of total grants to the sum of income tax and value added tax declines
substantially. The reason for the relative decline in the cost of social grants is that
scenario C2 (and all the previous scenarios) keeps the real value of grants constant, even
though the economy is assumed to grow in real terms at an average annual rate of 5

Scenario C3 takes note of this finding and expands on scenario C2 by allowing the
amount of all grants, including the adult grant, to adjust partially to the real rate of growth
of the economy. The partial adjustment is set at 2 percent annually. This is considered a
financially acceptable measure.

The simulation results of scenario C3 include:

A relatively larger increase in the average income of poor families and their share of
total household income.35 As a result, the income inequality index declines from
0.584 to 0.575. The growth elasticity of poverty () and specially the inequality
elasticity of poverty () are higher with scenario C3, indicating the increased pro-
poorness of the scenario (Table 4).

A sharp rise in the cost of total grants relative to GDP in the first year of
implementing scenario C3. However, the cost of the grants relative to GDP declines

The total amount of all grants increases from 3.8 percent of GDP in 2006 to 5.7 percent of GDP in 2007
since we have assumed that the above measures will all be implemented from 2007.
This is relative to results of scenario C2.

- 32 -
gradually to about 4.4 percent of GDP in 2015, which, compared to 2006, is only 0.7
percent higher. Therefore, the economic path that supports scenario C3 annually
provides larger resources for social provisioning through social welfare programmes,
which is reflected in a relatively higher social provisioning elasticity of
growth ( GDP

As a result of the increase in the real value of grants, and relative to scenario C2, the
poverty rate declines by 2.2 percent from 27.7 percent under scenario C2 to 25.4
percent. The depth of poverty, measured by the poverty gap, also declines by a little
more than 2 percent.

Overall, relative to scenario C2, the economic path of scenario C3 is more effective in
reducing poverty. This is reflected in its relatively higher overall total elasticity of
poverty (), which implies that, compared to scenario C2, the same rate of economic
growth is expected to reduce the poverty index by a larger amount
( (| C 3 |= 2.917 >| C 2 |= 2.566) . This is due to scenario C3s relatively stronger

employment, social provisioning and inequality nexus between poverty and growth,
which are reflected in the greater magnitude of the relevant elasticities (, , and )
(Table 3).

Scenario C3 reflects an economic path that realises the objective of halving the
unemployment rate and, at the same time, brings the economy closer to meeting the
objective of halving the poverty rate. However, scenario C3s projection of the poverty
rate for 2015 (estimated at 25.4 percent) still falls short of the expected target of 22
percent (Table 2). Following are two final scenarios that incorporate additional measures
to scenario C3 in order to meet the two goals.

Final Scenarios

If the amounts of the different grants are allowed to grow annually at the rate of 4 percent
in real terms, this revised version of scenario C3 will be sufficient to reduce the poverty
rate to 22.6 percent, which is almost half of the estimated current poverty rate of 44

- 33 -
percent (Table 2). The measure initially adds a little more than one billion rand to the cost
of the social security system. Overtime, however, as the amounts of grants grow, the
overall social security budget for this scenario will be R30 billion higher than scenario C3
by 2015.

Another approach would be to augment scenario C3 with non-cash transfers to poor

households. The Group D scenario examines the impacts of in-kind social transfers on

In equation 1, the variable mean household income (y) was defined to include income
from employment and both the cash and in kind government transfers to households.36
Social transfers in-kind consist of individual goods and services that are provided as
transfers in-kind to individual households by government entities. They can be divided
into two types: those where beneficiary households actually purchase the goods or
services themselves and are then reimbursed, and those that the relevant services provide
directly to the beneficiaries. The appeal to policy makers of transfers in-kind over
transfers in cash is that the resources transferred can be targeted towards meeting specific
needs, such as education or health, and must be used in the ways that their providers
intend. The South African government spent R150 billion on social transfers in-kind to
households in 2006.37

In equation 2, we used the concept of social provisioning (S) as one of the determinants
of the household mean income. Group C scenarios used one category of social
provisioning (i.e. cash transfers) to examine the effectiveness of South Africas current
grant system and its extension to help halve the poverty rate by 2015. For the next
scenario (scenario D), we examine the impact of expanded in-kind social transfers to poor
households as part of social provisioning.

To measure the value of goods and services provided to households as transfers in kind, the
governments final consumption expenditure is divided into collective consumption expenditure (defence,
justice, etc) and individual consumption expenditure (education, health, etc.).
This is recorded in the Current Income and Expenditure of General Government and also in the Current
Income and Expenditure of Households (Reserve Bank, Quarterly Bulletin).

- 34 -
Scenario D:

Scenario D is the same as scenario C3, plus an annual lump sum addition to the current
social transfers in-kind to fund specific government targeted (near-cash) basic services to
mainly poor households. The scenario includes an initial injection of R20 billion (1
percent of estimated nominal GDP for 2007) into the governments in-kind transfer
budget to expand current government services and introduce new services to poor
households. The amount is expected to increase by 2 percent in real terms every
subsequent year. This injection is expected to reach R34.4 billion (0.86 percent of
nominal GDP) by 2015. We have assumed that the adjusted disposable income of the
poor will benefit by an equivalent of 80 percent of the above addition to the
governments in-kind expenditure (Table 5).

In terms of areas of support to poor families, the above injection to the in-kind budget is
expected to be used as a transportation subsidy or transportation voucher for the
unemployed, and to include other types of in-kind support to poor households that
indirectly help increase their disposable income, such as subsidies for medicine, food aid
(school feeding), and expanded free essential services. 38

Analysis of Simulation Results (Scenario D)

Relative to scenario C3, the simulation results of scenario D show:

A greater decline in the inequality measure (Gini index) and the poverty gap. Relative
to scenario C3 (with a Gini coefficient of 0.575 and a 14 percent poverty gap), Gini
coefficient declines to 0.560 and the poverty gap declines to 8.3 percent (Table 2).

An increased contribution of the social provision channel to poverty reduction, which

is evident from relatively higher poverty elasticity of social provisioning, . It
increases (in absolute value terms) from 0.621 in scenario C3 to 1.492 in scenario D
(Table 3).

See Meth 2007 for a detailed analysis of important gaps in current services to the poor.

- 35 -
A stronger redistributive thrust, which underlies the increased effectiveness of the
scenario in reducing poverty. The growth elasticity of poverty for this scenario is -
1.320 while the distribution elasticity of poverty is -2.258. This suggests that the
inequality reducing channels of the scenario have greater effects on reducing the total
poverty index than do the increasing mean household income channels of the scenario
(Table 4). To express this differently, in scenario D the impact of the decline in
income inequality on poverty is 1.7 times the poverty effect of the increase in mean
household income.

An increased effectiveness of the employment creation to reduce poverty. This is

important since it shows how scenario D, that has significantly larger impacts on
reducing the poverty gap, makes it possible for a larger number of poor families to
move out of poverty as they find employment. This is captured by the relative
increase in poverty elasticity of employment (), compared to scenario C3.39

Overall, the rate of poverty declines from 44 percent in 2006 to 21.2 percent in 2015
(Table 2). This means that the economic path that underlies scenario D has the potential
to reduce both the unemployment and poverty rates by half by 2015. As a reminder,
scenario D is based on and requires (Table 5):

(a) an annual real growth rate of 5 percent between 2007 and 2015;

(b) employment elasticity of growth of at least 0.60 to generate 3.7 million new jobs
during the period;

(c) transformation of the economy so that the labour market benefits the poor
proportionally more than the non-poor, in terms of distribution of new
employment and employment related income;

(d) extension of the child support grant to poor children between the ages of 14 and

Poverty elasticity of employment increased (in absolute value terms) from 1.531 in scenario C3 to 2.985
in scenario D.

- 36 -
(e) introduction of an adult grant programme for unemployed adults, with an initial
grant amount of R500 per month;

(f) a real annual increase in the value of all grants by 2 percent;

(g) expansion of the current government in-kind transfers to poor households in the
form of transportation subsidies, food aid, subsidy for essential services, etc. The
proposal is to increase the budget for in-kind social transfers by R20 billion in the
first year, followed by an annual real increase of 2 percent.

The implementation of scenario D has important implications for macroeconomic,

industrial and social security policies, and for the allocation of resources. For the
economy to grow in real terms at 5 percent a year, to generate a high rate of employment
(about 3.7 million new jobs) that is pro-poor, and to provide for a major expansion of the
social security system, the government will need to play an active role in designing and
implementing growth, employment and social security policies that are strongly pro-poor.
For example, financing the expansion of the social security system along the lines of
scenario D requires the adoption of a combination of measures that include deficit
financing (e.g. equivalent to about 2 percent of the GDP), budget reprioritisation,
elimination or major curtailing of programmes that effectively subsidise the income of
non-poor households, and changes in income, wealth and value-added taxes.

In this context, the next section is devoted to important features of a pro-poor policy
framework with the potential to implement policies such as scenario D policies to
achieve the twin goals of halving the poverty and unemployment rates by 2015.


Sections 3 and 4 showed that, on the one hand, the combined goals of halving the
unemployment and poverty rates depend on both the magnitude of the growth rate (the
higher the growth rate, the larger the decline in poverty) and the distribution of the
benefits of growth (poverty declines faster when more growth benefits go to the poor than

- 37 -
to the non-poor). Therefore, achieving high growth rates does not necessarily translate
into a high reduction in poverty. On the other hand, the scenarios clearly highlight the
relative importance of policy interventions to foster pro-poor growth and outcomes.

The aim of this section is to present an economic policy approach that can be considered
a pro-poor approach, since it internalises the requirements for fostering growth that
delivers proportionally greater benefits to the poor than to the non-poor. This is important
as many discussions on pro-poor policies effectively reflect the orthodox macroeconomic
policy prescriptions. In these presentations, it is usually the requirements of the orthodox
macroeconomic framework that are prioritised and allowed to define a poverty reduction
agenda. Moreover, in many instances they also ignore or insufficiently integrate policy
issues related to high employment creation.

We need, therefore, to overcome the intellectual hurdles of mainstream views and policy
prescriptions if we are to open up the space to design a developmentally suitable policy
framework. Our aim is to present a basic guideline that will inform the specifics of a pro-
poor set of policies. To the extent that halving the poverty and unemployment rates
depend on the macroeconomic policy framework, the approach to industrial development,
labour market reforms and the establishment of comprehensive social security system,
our basic policy guideline needs to reflect the coordination of these policies.

A. Pro-Poor Macroeconomic Policies

In South Africa, macroeconomic policy was significantly biased towards stabilisation and
liberalisation from 1996 to 2001. In recent years, however, fiscal policy has become
relatively less restrictive. The importance of growth and redistribution for employment
creation and poverty reduction underscores the pertinent role of a macroeconomic policy
framework to foster high rates of growth and channel a disproportionate share of its
benefits to the poor.

Pro-Poor Fiscal Policy: The central aim of a pro-poor fiscal policy should be to help
establish a growth path that embodies simultaneous progress towards minimising income
inequality and achieving full employment. Its necessary elements are (a) counter-cyclical

- 38 -
policies, (b) a strong public investment programme, and (c) measures to re-orient the
private sector.

(a) Counter-cyclical policies: A major challenge is to avoid pro-cyclical fiscal policy,

either to cushion the impact of adverse shocks, or to adapt to government revenues that
are pro-cyclical, drop during recessions and increase during expansions.

In this context, how can the government establish a counter-cyclical fiscal policy? On the
one hand, this requires that the monetary authorities support the pro-poor fiscal policy
goal and, together with the fiscal authorities, design a coordinated counter cyclical
macroeconomic policy framework for the country. On the other hand, it is necessary to
establish the needed fiscal space to sustain strong public investment and expanded
social security programmes over economic cycles (see below). One suggestion is to
create a fiscal stabilisation fund that could be used to smooth public spending over
economic cycles (Islam 2003). This can be financed from excess revenues that are
collected by the government during a given year (e.g. SARSs overshooting of its
projected tax collection) or during periods of growth (e.g. current fiscal surplus) or during
commodity price boom (e.g. rise in gold prices).40

(b) strong public investment: another defining element of a pro-poor fiscal policy is a
strong public investment programme. Given the historical under-investment in new
infrastructure and the low maintenance of existing social infrastructure, many of the
capital inputs required education, training, communications, transportation and other
elements of social and physical infrastructure have a significant public good character
that precludes their being adequately supplied by private firms. Only the public sector (at
national, provincial and local levels) can realistically be expected to take responsibility
for the necessary social and infrastructure investment.

Public sector investment can become even more pro-poor by, for example, ensuring that
such investment is relatively less capital intensive than private sector investment. This
helps increase the employment-creating thrust of the growth process. To foster such an

This is similar to holding international reserves in liquid foreign assets to safeguard the economy against
fluctuations in capital flow (Islam 2003).

- 39 -
outcome, public investment needs to include: a) public works projects that directly hire
the poor; b) the creation of infrastructure assets that give the poor access to markets and
lower their production costs, and c) social sector assets such as schools and health clinics
that increase the productiveness of the poor.41

It is within this overall context that the UN Millennium Project (2005) emphasises the
need for a big push strategy in public investment to help countries break out of their
poverty traps, meet the MDG goals and set economies on a sustainable growth path. The
Project identifies seven main areas for public investment: rural development; urban
development; health systems; education; gender equality; environment and science, and
technology and innovation.

Given the importance of strong public investment for pro-poor fiscal policy, it is
necessary to complement the idea of creating a fiscal stabilisation fund with additional
measures to generate the required fiscal space. This can be made possible through one or
more of the following two options: raising the revenue share in GDP and/or using the
fiscal deficit.

Increasing the revenue share in GDP is the most effective way to create fiscal space and,
at the same time, reduce inequality. The very high income and wealth inequality in South
Africa makes it possible to press for a strong progressive income and company tax, value
added tax and wealth tax systems. The high-income section of the population, who are
expected to increase their contribution to the fiscal space for pro-poor public investment,
will be considerably compensated in a number of ways. As the main owners of small and
larger private sector enterprises, they will be important beneficiaries of public investment
in many of the capital inputs required for efficient productive activity in a modern
economy. Such investments, along with a counter cyclical policy approach, will help
their sources of income through increased capacity utilisation, capital and labour
productivities and profitability.

Finally, using the fiscal deficit to finance social and infrastructural investments to create
the necessary financial space for pro-poor fiscal policies is currently a viable option in
Roy et al 2005.

- 40 -
South Africa. An important guiding principle is that increasing expenditure on social
sectors, especially in terms of a relative increase in allocations to these sectors, is almost
always pro-poor. At the same time, it helps increase both the rate and sustainability of the
countrys long-term growth.

(c) Measures to reorient the private sector: The third element of a pro-poor fiscal policy
includes government procurement policies and investment incentive measures to increase
the volume of private sector investment, influence the composition of investment, and
increase the chances of future growth patterns being redistributive and job creating.

Pro-poor investment needs to go beyond the public sector. There is a need to reward
investment by businesses whose activities reflect support for a more broadly based
transformation of ownership, improved income distribution and a reduction in the
unemployment rate. Specific criteria can be developed to determine the eligibility and
extent to which businesses can be given differential tax incentives, access to subsidies
and access to government procurement.42

Monetary Policy: If the above fiscal policy measures are to produce the desired
outcomes, monetary policy authorities need to balance their concern for inflation with a
concern for economic growth.43 Ensuring the monetary policy commitment to foster
growth is of prime importance.44 This is because, as a major macroeconomic instrument,
monetary policy has limited influence in terms making growth pro-poor. However, in
support of an expansionary fiscal policy, it can indirectly foster growth that is pro-poor.
As a general rule, if inflationary pressures are weak, this support should take the form of

The South Africa Human Development Report (2003) proposes that businesses can be given above
benefits depending on the extent to which (a) they have restructured, or are going through an ownership
restructuring process that is resulting in a broad-based distribution of company ownership amongst their
employees and the public as a whole; (b) their capital productivity growth rates exceed increases in their
after-tax rates of profit (i.e., as a measure of improvement in income distribution), and (c) their real output
grows faster than their labour productivity growth rates (i.e., as a measure of significant employment
Similarly, Weeks et al (2005) suggests the following guideline for the monetary policy, foster growth
and counter inflation when necessary.
In scenarios that were developed in section 4, we assumed 5 percent real annual growth rate during 2007
to 2015. Reaching such a rate and sustaining it over years require monetary policy commitment to directly
promote growth and policy coordination with the Treasury.

- 41 -
low interest rates and an expanding monetary supply, which is also the basis for
supporting a counter-cyclical macroeconomic policy45.

For a growth accommodating monetary policy, the interest rate should be used as a long-
term investment instrument rather than for short-term stabilisation. More specifically, to
sustain economic growth, the long-term interest rate should not be greater than the
maximum sustainable rate of growth of per capita income.

In this context, nominal GDP targeting, for example, offers an important improvement
over inflation targeting or money supply targeting. The advantage is that this gives
weight both to inflation and to real GDP growth. This paves the way towards reorienting
monetary policy that supports economic growth.

In addition to its potential impacts on growth, relatively lower interest rates under
nominal GDP targeting have further positive developmental effects. For example, to the
extent that government bonds are held by the relatively prosperous portion of the
population, lowering interest rates can improve income distribution; at the same time,
lower rates reduce the size of domestic debt service in the budget, thus helping establish
additional fiscal space for pro-poor government expenditure.

In addition to the above guideline for monetary policy, government can increase the
portion of deficit that is money financed. This is considered the least anti-poor method of
financing deficits, since with bond financing the deficit redistributes public revenue from
the population as a whole to the wealthy in the form of interest payments. At the same
time, to deal with balance of payments constraints, monetary authorities need to consider
the adoption of a policy of managed currency depreciation in order to establish a stable
and moderately undervalued real exchange rate that helps generate extra exports.

Finally, the above macroeconomic policy framework for pro-poor growth requires
adopting a guarded approach to capital mobility. Domestic and international supporters of
the Washington Consensus have long advocated deep economic integration, including
both trade and capital account liberalisation in South Africa. However, it is well known

National Human Development Report of South Africa (2003) and Weeks et al (2005).

- 42 -
that short-term capital mobility increases the risk of financial crisis and economic
recessions.46 At the same time, capital account liberalisation has, in many countries
including South Africa, allowed the interests of domestic and global financial capital to
influence a countrys macroeconomic policy in favour of low inflation and fiscal
conservatism. Therefore, during the past decade, there has been a growing international
debate on how to safeguard the economy from the potential damages of short-term capital
mobility. Such controls would entail a combination of taxes and quantitative restraints.47

The above fiscal and monetary policy measures will enhance the pro-poor growth
potential of the economy, ensure that fiscal and monetary policies are coordinated to
respond positively to the need to sustain high growth rates, give necessary consideration
to the monetary authoritys concern about inflation, and safeguard the economy from the
risk of financial crisis and pressures to forego pro-poor policies.

B. Employment Generation Policies

The findings of the economic scenarios in section 4 make it clear that, in order to halve
both the unemployment and poverty rates by 2015, the labour market needs to become
pro-poor (in terms of the allocation of new jobs and the evolution of wages) and the
employment intensity of economic growth needs to increase substantially. Using policies
to transform the economy so as to generate about 3.7 million new jobs (that are also pro-
poor in terms of allocation and rewards) during the next nine years presents a significant
challenge to policymakers. The following policy suggestions are offered in order to help
develop a country specific package of measures to achieve the above objectives:48

Adopting policies to increase the labour intensity of production:49 Central to a successful

re-orientation of the economy is the need to ensure that the general production processes

William Easterly et al (1999:43) claims that countries with more open capital accounts are more
likely to go into recessions.
Chang (2005:162) suggests that policymakers should adopt a conservative stance on liberalization. That
is, liberalise only when the benefits are unambiguous and cannot be reasonably achieved through other
means. He argues that any benefits associated with financial liberalisation must be carefully weighted
against its costs (namely, increased levels of systemic risk, volatility and short-termism, all of which
increase financial fragility, and thus the chance of financial crisis).
Additional resources include Islam (2004) and ILO (2003).
For further analysis and empirical findings see, Islam (2004).

- 43 -
become significantly more labour intensive. Examples of policy measures in this area

(a) The promotion of employment-intensive social and infrastructural investments.

ILO (2000) proposes a public investment policy that stresses labour-based rather
than equipment-based production methods.50

(b) The identification of targeted, time-structured and conditional incentives to

promote labour-intensive production in the private sector.

(c) The withdrawal of explicit or implicit subsidies that systematically favour capital-
intensive and/or large-scale enterprises.

(d) The adoption of tax and incentive measures to engender structural transformation
of employment towards manufacturing and other non-farm sectors.

(e) Agricultural interventions to promote increased acreage under labour-intensive


(f) The promotion of alternative ownership and production arrangements, such as co-
operatives and equity arrangements that use labour-intensive methods of

In a number of these areas, the restructuring of diverse public enterprises can lead the
way. Moreover, government procurement policy, at both national and provincial levels,
should put greater stress on procuring goods and services from companies that seek to use
labour more extensively in their areas of activity. Overall, commitment to halving the
unemployment and poverty rates should reward sectors of the economy on the basis of
their contributions to these national developmental goals and not on the basis of their
industrial competitiveness on the global market.

Adopting measures to enhance the economys capacity to utilise more labour: It is

necessary to develop sector strategies that aim directly at promoting labour absorption.

See ILO (1999) for a case study of using labour-based production method.

- 44 -
Such sector strategies should not be seen as separate from an employment strategy, but as
vehicles for the realisation of the employment promotion objective. Sectoral strategies
need to be internally coherent, as well as in terms of how they link together nationally
and regionally. Supportive policies in this regard generally focus on the following:
enabling polices; pricing; inputs; innovation and technology; information; infrastructure;
the development of regulatory regimes and institutions; incentive structures; the
identification of specific activities or groups to be targeted, and resource mobilisation.

Examples of relevant policy measures in this regard are:

The facilitation of access to credit by disadvantaged individuals and communities

through the reform of financial institutions, a lowered interest rate and/or the
selective promotion of prescribed asset and/or investment policies.

Land reform through restitution, tenure reform and redistribution.

Design of projects that target poor labour tenants, farm workers, women and
emerging farmers, giving them access to land for residential and productive use.

Targeted provision of subsidised credit and extension services for labour-absorbing

large, medium and small-scale activities that enhance employment creation.

Use of an employment subsidy programme to encourage the private sector to

increase its employment of first time jobseekers, and the retraining and/or hiring of
retrenched employees.

Promoting sectoral programmes with high level of employment multiplier: Employment

opportunities can be expanded through the promotion and consolidation of value chains
or sectoral linkages as new economic activities are promoted. Pro-active policy measures

Support for the development of activities that fill gaps in the value chain
(distribution, marketing, financial intermediation, input provision, research and

- 45 -
development, technological capabilities, etc.) for select activities in agriculture,
manufacturing and services.

Tight coordination of public investment projects to maximise their impact on

mutually re-enforcing activities in the private sector.

Development of regional clusters of economic activities, especially in depressed

areas of the economy.

Special employment and public works programmes for those who cannot
immediately be absorbed into productive employment. Such public works should
focus on carefully chosen activities of a public good nature, which will enhance
social infrastructures that support private sector productivity.

C. Trade and Industrial Policy

Further liberalisation of the economy in terms of flows of goods, services and capital
must be carefully managed and sequenced. There is a need for critical assessment of the
underlying tenets of the argument for liberalisation, and a need to monitor carefully its
speed and extent in the context of creating jobs and achieving a higher standard of living
for the country's disadvantaged majority. Such analysis is vitally important, particularly
given the experience of countries that have, in the last two decades, adopted full-scale
economic liberalisation.

In this context, there is a good case for protecting or supporting sectors that generate large
positive externalities such as technological spillovers. By encouraging import substitutes,
protection can expand the domestic traded goods sector. The means of expansion operates
by reducing the propensity to import, thereby reducing leakage from the domestic
economy. The objective of protection in an economy with underemployment is to reduce
the propensity to import competitive goods, not to reduce the actual volume of imports.51
If the policy is successful, the rise in domestic incomes should encourage more imports of
complementary and subsequently competitive goods.

See also Jahed 2007.

- 46 -
In the direct and indirect formulation of industrial policy, the following needs to be given
the highest priority: (a) meeting basic needs; (b) generation of employment; (c)
education and training; (d) sectoral policy; (e) infrastructural provision and measures to
ensure spin-offs; (f) reform of the financial system to secure finance for industry; (g)
monitoring and control of foreign investment flows, particularly outward investments by
conglomerates of South African origin; (h) minimum labour standards and the narrowing
of wage differentials; (i) macroeconomic policy; (j) regional integration; (k)
restructuring of state assets, and (l), the reform of institutions that make industrial policy,
so that the allocation and coordination of responsibilities across government departments
is rationalised and coherent.52

Chang (2002) suggests specific measures related to some of the above set of priorities.
His recommendations can be summarised in the form of a guideline for developing pro-
poor trade and industrial policy:

Develop policies and measures to overcome the following challenges facing the
development of SMMEs: access to finance; marketing of products and services;
access to adequate and appropriate infrastructure; access to title deeds;
insufficient skills in manufacturing, and access to technological support and

Provide protection to certain strategic industries. In order to prevent de-

industrialisation associated with free trade policy recommendations, and establish
the long-term foundation for a vibrant industrial sector, strategic industries need
to be identified (in terms of employment creation, export generation, etc.), and
measures such as tariffs, quotas and subsidies used for their development.53 These
measures are also suggested by Jahed (2007) and Carim (2005).

Coordinate other policies with the industrial strategy. The objective is to ensure
that the governments overall policy framework provides short and long-term

For more details see Fine 1997.
For more on the short term and long term costs and benefits of this policy approach see Chang 2002 and
Helleiner 1994.

- 47 -
supports for industrial development. Government policies related to education,
infrastructure, health, public investment and technology policy need to support
industrial policy.54

Specify clear performance targets and incentives: Defined realistic performance

targets, rewards and penalties are essential in order to minimize the possibility of
implementation failure.55

Clarify the mandates of state owned enterprises and hold their management
accountable to those mandates.

Use FDI as a strategic means to promote technology and knowledge transfers and
stimulate innovation by domestic researchers. Moreover, tie the FDI policy to the
national development and/or industrial policy plans.

As much as possible, reduce reliance on foreign bank loans and adopt policies that
discourage rather than prohibit the use of foreign loans as a source of finance.

D. Comprehensive Social Security System

In March 2002, the report of the Committee of Inquiry into a Comprehensive System of
Social Security for South Africa was delivered to Cabinet. It provides a comprehensive
attempt to bring together the different elements of a fragmented social security system in
order to provide recommendations that could ultimately lead to comprehensive social
protection. The simulation results of our policy scenarios in section 4 support the main
conclusions and recommendations of the Committees report, particularly the
Committees position that an appropriate social security system for South Africa must
prioritise the needs of people without any income or with insufficient income, and must
encompass those engaged in informal sector. This, for example, implies an extension of
the child support programme to poor children aged between 14 and 18 years, who are not
currently covered by the grant. It also implies an extension of the social security system

For more on this topic see Helleiner (1994) and Jahed (2007).
Chang (2002) provides examples for Japan, Korea and Taiwan.

- 48 -
to provide income support to poor adults or to poor unemployed adults aged between 18
and the retirement age for households where no-one is employed.

In addition to expanding the scope for the eligibility for social grants, the real value of
social grants (after adjustment for inflation) also needs to increase in the medium term.
The combination of both measures is needed to ensure that the governments social grant
system contributes effectively to the fight against poverty.


The aims of this study were to show how the unemployment rate and the poverty rate can
be halved by 2015, and to present an outline of a related supportive policy framework.
Our findings support the argument that a combination of labour market and social
provisioning measures and outcomes are necessary to achieving the above twin goals.
Moreover, these measures and outcomes need to be pro-poor in the sense that policies
should increasingly target the poor, and that the poor should benefit disproportionally
from the countrys economic path. In this context, our findings confirm the international
consensus that, for an economy with high rates of poverty and inequality, unless the
growth process embodies reductions in inequality, it will be unfeasible to achieve
accelerated poverty reductions.

It was also argued that the above findings should inform the economic policy framework
in the sense that the countrys agenda for accelerated poverty and unemployment
reductions should not be compromised by the requirements of an orthodox policy
framework. Therefore, the paper presents the basic tenets of a pro-poor economic policy
framework that also includes the coordination of policies on which halving poverty and
unemployment depend. These include the macroeconomic policy framework, the
approach to industrial development, the labour market and the social security system.
Overall, the findings suggest an active pro-poor role for the state.

- 49 -

Amin, S. (1989). Eurocentrism, New York: Monthly Review Press.

ANC (African National Congress), (2007). Economic Transformation for a National

Democratic Society, Policy Discussion Document, pp.1-22

Anand, S.and Kanbur, S.M.R. (1984). The Kuznets Process and the Inequality
Development Relationship, Journal of Development Economics, Vol. 40, pp. 25-52.

Blank, R.M. (1989). Disaggregating the Effect of the Business Cycle on the Distribution
on Income, Econometrica, Vol.56, pp.141-163.

Bourguignon, F. (2004). The Poverty-Growth-Inequality Triangle, Paper Presented at

the Indian Council for Research on International Economic Relations, New Delhi,

Chang, H. (2002). Kicking Away the Ladder-Development Strategy in Historical

Perspective. Anthem Press, London.

Chuhay, R. (2004). The Business Cycle Behavior of GINI Coefficient of the Income
Distribution. A Thesis Submitted to the National University Kyiv-Mohyla Academy,

Cornia, G.A. (2002). Poverty and Inequality in the Era of Liberalization and
Globalization, In H.Van Ginkel et al. Human Development and the Environment:
Challenges for the United Nations in the New Millennium. New York: United Nations
University Press.

Dagdeviren, H.,Van der Hoeven, Rolph and J. Weeks, (2000). Redistribution Matters:
Growth for Poverty reduction, Working Paper Series No. 99, Department of Economics,
School of Oriental and African Studies, University of London, October.

Deininger, K. and L. Squire, (1998). New Ways of Looking at Old Issues, Journal of
Development Economics, Vol. 57, pp. 259-87.

Department of Social Development, Pretoria, (2002). Transforming the Present-

Protecting the Future: Report of the Committee of Inquiry into a Comprehensive System
of Social Security for South Africa, (Consolidated Report).

Dollar, D. and A. Kraay, (2002). Growth is Good for the Poor, World Bank,
Development Research Group.

Easterly, W. (2001). The Effect of International Monetary Fund and World Bank
Programs on Poverty, World Bank Policy Research Working Paper 2517.

- 50 -
Easterly, W., R. Islam and J. Stiglitz, (1999). Shaken and Stirred: Volatility and
Macroeconomic Paradigms for Rich and Poor Nations, XII World Congress of the IEA,
Buenos Aires, August.

Fine, B. (1997). Industrial Policy and South Africa: A Strategic View, NIEP
Occasional Paper Series, No. 5, April.

Fukuda-Parr, S., C. Lopes and K. Malik, (2002). Capacity for Development: New
Solutions to Old Problems, (eds.), Earthscan Publications, pp.1-21.

Helleiner, G. (1994). Trade Policy and Industrialization in Turbulent Times, Routledge.

International Labour Office (1999). Decent Work, Geneva.

International Labour Office (2000a). Employment-intensive Investment in Infrastructure:

Jobs to Build Society, Geneva: ILO.

International Labour Office (2000b). World Labour Report: Income Security and Social
Protection in a Changing World, Geneva: ILO.

International Labour Office (ILO) (2003). Working out of Poverty. Report of the
Director-General for the International Labour Conference, Geneva: ILO.

Islam, I. (2003). Labour Market Policies, Economic Growth and Poverty Reduction:
Lessons and Non-lessons from the Comparative Experiences of East, South East and
South Asia, Issues in Employment and Poverty, Discussion Paper 8, Recovery and
Reconstruction Department, International Labour Office, Geneva, April.

Islam, I. (2002). Poverty, Employment and Wages: An Indonesian Perspective, In

Labour Market Policies and Poverty Reduction Strategies in Recovery from the Asian
Crisis. Report of a Seminar held in Jakarta, 29 April- 1 May. ILO, Bangkok.

Islam, R. (2004). The Nexus of Economic Growth, Employment and Poverty Reduction:
An Empirical Analysis, International Labour Office, January.

Jahed, M. (2007). Pro-poor Trade and Industry Policies: Considerations for South
Africa, Oxfam South Africa, Johannesburg.

Kakwani, N. (1993). On a Class of Poverty Measures, Econometrica, Vol.48, No.2,


Kakwani , N. (2000). On Measuring Growth and Inequality Components of Poverty

with application to Thailand , Journal of Quantitative Economics.

Kakwani, N. and E. Pernia, (2000). What is Pro-poor Growth, Asian Development

Review, Vol. 16, No.1, pp. 1-22.

- 51 -
Kakwani, N. and H. Son, (2002). Pro Poor Growth and Poverty Reduction: The Asian
Experience, The Poverty Center, Office of Executive Secretary, ESCAP, Bangkok.

Kakwani, N., S. Khandker and H. Son, (2003). Pro Poor Growth Strategies in Africa:
Poverty Equivalent Growth Rates with applications to Korea and Thailand, Economic
Commission for Africa, Economic Policy Research Center, Expert Group Meeting, June.

Klasen, S. (2003). In Search of the Holy Grail: How to Achieve Pro-Poor Growth,
Ibero-America Institute for Economic Research, George-August-Universitt Gttingen,
Discussion Paper No. 96, October.

Kuznets, S. (1995). Economic Growth and Income Inequality, American Economic

Review, Vol. 45, pp. 1-28.

Lal, D. and H. Myint, (1996). The Political Economy of Poverty, Equity and Growth: A
Comparative Study, Oxford: Clarendon Press.

Lipton, M. (1996). Rural Reforms and Rural Livelihoods: The Contexts of International
Experience In M. Lipton, M. de Klerk and M. Lipton (eds). Land, Labour and
Livelihoods in Rural South Africa, Durban: Indicator Press.

Lbker, M., G. Smith, and J. Weeks (2000). Growth and the Poor: A Comment on
Dollar and Kraay, Center for Development Policy and Research, University of London,
School of Oriental and African Studies, October.

Meth, C. (2007). What is pro-poor growth? What are some of the things that hinder its
achievement in South Africa? Oxfam South Africa, Johannesburg.

Orshansky, M. (1965) "Counting the Poor: Another Look at the Poverty Profile", Social
Security Bulletin, Vol. 28, No. 1, January 1965, pp. 3-29 reprinted in Social Security
Bulletin, Vol. 51, No. 10, October 1988, pp. 25-51.

Osmani, S. R. (2002). Exploring the Employment Nexus: Topics in Employment and

Poverty, A Report prepared for the Task Force on the joint ILO-UNDP Programme on
Employment and Poverty.

Osmani, S. R. (2004). The Employment Nexus between Growth and Poverty: An Asian
Perspective. A Report prepared for the Swedish International Development Agency
(SIDA), Stockholm and the United Nations Development Programme (UNDP), New
York. (mimeo).

Oxfam. (2000). Growth with Equity is good for the Poor, (A Response to Dollar and
Kraay), Oxford, June.

- 52 -
Ravallion, M. (2001). Growth, Inequality and Poverty: Looking Beyond Averages,
World Development, Vol. 29, No.11, pp. 1803-1815.

Roy, R. and A. Heuty (2005). Investing in Development: The Millennium Development

Goals, Aid and the Sustainable Capital Accumulation, Columbia School of International
and Public Affairs Review, Spring, Vol. 58, No. 2.

Seers, D. (1977). The New Meaning of Development, International Development


Seers, D. (1979). The Birth, Life and Death of Development Economics, in

Development and Change.

Statistics South Africa, (1995). October Household Survey. Statistics South Africa:

Stewart, F. (1985). Development and Modernization Basic Needs in Developing

Countries. John Hopkins University Press, Baltimore.

Streeten, P., S. Burki, M. Ul Haq, N. Hicks and F. Stewart, (1981). First Things First:
Meeting basic human needs in developing countries, Oxford University Press.

Sutcliffe, B. (1995). Development after Ecology In V. Bhaskar and A. Glyn (eds). The
North the South and the Environment: Ecological Constraints and the Global
Environment. London: Earthscan.

United Nations Conference on Trade and Development (2002). World Investment Report.
New York and Geneva.

United Nations Copenhagen Declaration and Programme of Action, (1995). The World
Summit for Social Development 6-12 March.

UNDP (1994). 1994 Human Development Report New Dimensions of human security,
New York: Oxford University Press.

UNDP (1995). 1995 Human Development Report Gender and human development,
New York: Oxford University Press.

UNDP (1996). 1996 Human Development Report Economic growth and human
development, New York: Oxford University Press.

UNDP (1997). 1997 Human Development Report Resisting new forces of Poverty in a
changing World, New York: Oxford University Press.

UNDP (1998). 1998 Human Development Report Consumption for human

development, New York: Oxford University Press.

- 53 -
UNDP (1999). 1999 Human Development Report Globalization with a human face,
New York: Oxford University Press.

UNDP (2000). 2000 Human Development Report South Africa: Transformation for
Human Development. Pretoria.

UNDP (2003). 2003 Human Development Report South Africa: The Challenges of
Sustainable Development: Unlocking Peoples Creativity. Oxford University Press, Cape

UNDP (2005). 2005 Human Development Report International cooperation at a

crossroads: Aid, trade and security in a unequal world. New York: Oxford University

UNDP. World Population Prospects. Department of Economical and Social Affairs

(UNDESA). United Nations Population Division.

UN Millennium Project (2005). Investing in Development: A Practical Plan to Achieve

the Millennium Development Goals, London: Earthscan.

United Nations Volunteers (2001). Volunteering and the United Nations System. Bonn:
UN Volunteers.

Wade, R. (2004). Is Globalization Reducing Poverty and Inequality? World

Development, Vol. 32, No.4, pp.567-589.

Wade, R. (2001). Winners and Losers. The Economist 26 April.

Wallerstein, I. (1979). The Capitalist World Economy. Cambridge University Press.

Weeks, J. and R. Roy (2005). Making Fiscal Policy Work for the Poor, UNDP,

World Conference on Environment and Development (1987). Our Common Future.

Oxford: Oxford University Press.

World Summit for Social Development (1995). The Copenhagen Declaration and
programme of Action. New York: United Nations.

- 54 -
Table 1. Policy Scenarios and Evolution of Unemployment Rate (2007-2015)

Scenarios Indicators 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Employed 12,451,098 12,571,663 12,695,190 12,809,410 12,932,317 13,054,484 13,178,972 13,305,849 13,432,494 13,558,539
A1 Unemployed 4,323,877 4,416,666 4,492,499 4,566,205 4,617,680 4,658,115 4,702,016 4,749,426 4,803,087 4,863,493
Unemployment Rate 25.8% 26.0% 26.1% 26.3% 26.3% 26.3% 26.3% 26.3% 26.3% 26.4%
Employed 12,451,098 12,690,932 12,933,680 13,182,953 13,435,450 13,694,584 13,957,394 14,227,745 14,500,710 14,788,651
A2 Unemployed 4,323,877 4,297,397 4,254,009 4,192,662 4,114,547 4,018,015 3,923,594 3,827,531 3,734,871 3,633,382
Unemployment Rate 25.8% 25.3% 24.8% 24.1% 23.4% 22.7% 21.9% 21.2% 20.5% 19.7%
Employed 12,451,098 12,813,377 13,186,389 13,571,378 13,967,476 14,373,092 14,795,490 15,573,140 15,742,055 16,133,264
A3 Unemployed 4,323,877 4,174,952 4,001,300 3,804,237 3,582,521 3,339,507 3,085,497 2,804,717 2,493,526 2,288,768
Unemployment Rate 25.8% 24.6% 23.3% 21.9% 20.4% 18.9% 17.3% 15.5% 13.7% 12.4%
Source: ADRS model of South African Tax and Transfer Simulation Model (SATTSIM-Plus)

- 55 -
Table 2. Evolution of Poverty and Inequality Indicators (2007-2015)

Scenarios Indicators 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Poverty Rate 44 39.7 39.2 39.0 38.7 38.4 38.0 37.7 37.3 37.0
Gini Coefficient 0.658 0.660 0.659 0.657 0.656 0.655 0.654 0.653 0.652 0.651
Poverty Rate 44 39.4 38.6 38.1 37.1 36.5 35.6 34.9 34.4 33.9
Gini Coefficient 0.658 0.658 0.655 0.651 0.647 0.643 0.639 0.634 0.632 0.630
Poverty Rate 44 39.1 37.6 36.7 35.7 34.9 34.1 33.2 32.7 31.8
Gini Coefficient 0.658 0.656 0.650 0.643 0.637 0.633 0.628 0.614 0.612 0.606
Poverty Rate 44 39.1 37.6 36.7 35.6 34.7 33.8 32.5 31.8 31.2
Gini Coefficient 0.658 0.656 0.650 0.643 0.636 0.632 0.629 0.611 0.608 0.605
Poverty Rate 44 39.1 37.5 36.7 35.5 34.4 33.2 32.2 31.1 30.4
Gini Coefficient 0.658 0.656 0.650 0.643 0.636 0.631 0.628 0.622 0.604 0.600
Poverty Rate 44 38.4 36.9 35.9 34.7 33.3 32.1 31.1 30.0 29.3
Gini Coefficient 0.658 0.652 0.646 0.639 0.632 0.627 0.623 0.618 0.600 0.596
Poverty Rate 44 35.4 34.2 33.5 32.4 31.2 30.3 29.3 28.2 27.7
Gini Coefficient 0.658 0.631 0.626 0.621 0.616 0.612 0.609 0.604 0.588 0.584
Poverty Rate 44 35.2 33.7 32.5 31.1 30.0 28.9 27.9 26.7 25.4
Gini Coefficient 0.658 0.630 0.624 0.617 0.611 0.606 0.602 0.596 0.579 0.575
Poverty Rate 44 34.9 32.7 31.4 29.9 28.5 27.0 25.5 23.7 22.6
Gini Coefficient 0.658 0.630 0.621 0.613 0.606 0.600 0.594 0.588 0.570 0.564
Poverty Rate 44 32.0 30.5 29.2 27.7 25.3 24.6 23.5 22.3 21.2
Gini Coefficient 0.658 0.618 0.610 0.604 0.597 0.592 0.588 0.582 0.565 0.560
Source: ADRS model of South African Tax and Transfer Simulation Model (SATTSIM-Plus)

- 56 -
Table 3. Total Elasticity of Poverty and Policy Scenarios

Group D
Group A Scenarios Group B Scenarios Group C Scenarios
Calculation of Elasticities Scenario
A1 A2 A3 B1 B2 C1 C2 C3 C4 D1
yv Poverty elasticity of income -0.273 -0.332 -0.343 -0.363 -0.380 -0.407 -0.437 -0.483 -0.540 -0.593
Gv Poverty elasticity of inequality 14.273 5.395 3.527 3.641 3.530 3.557 3.310 3.333 3.418 3.498
y Inequality elasticity of income -0.019 -0.062 -0.097 -0.100 -0.108 -0.114 -0.132 -0.145 -0.158 -0.169
E Income elasticity of employment 6.526 3.676 2.739 2.688 2.563 2.582 2.670 2.743 2.828 2.743
Employment Elasticity of Growth 0.161 0.341 0.536 0.541 0.577 0.577 0.577 0.577 0.577 0.577
Sy Income elasticity of social provisioning 0.870 1.034 1.214 1.203 1.223 1.058 0.659 0.510 0.408 0.388
Social provisioning elasticity of growth 0.491 0.491 0.491 0.491 0.491 0.572 0.950 1.262 1.624 1.658
Inequality elasticity of growth -0.020 -0.077 -0.143 -0.145 -0.159 -0.171 -0.203 -0.230 -0.258 -0.268
Total Elasticity of Poverty -1.095 -1.586 -1.923 -2.013 -2.145 -2.313 -2.566 -2.917 -3.361 -3.578

Calculation of Compound Elasticities

Employment induced Pov elast of growth (poverty
elasticity of employment)
-0.574 -0.832 -1.009 -1.056 -1.125 -1.213 -1.346 -1.531 -1.763 -1.877

SP induced poverty elast of growth (poverty

elasticity of social provisioning)
-0.233 -0.338 -0.410 -0.429 -0.457 -0.493 -0.547 -0.621 -0.716 -0.762

Inequality elasticity of growth -0.020 -0.077 -0.143 -0.145 -0.159 -0.171 -0.203 -0.230 -0.258 -0.268
Poverty elasticity of inequality 14.273 5.395 3.527 3.641 3.530 3.557 3.310 3.333 3.418 3.498

inequality induced elasticity of growth -0.287 -0.416 -0.504 -0.528 -0.563 -0.607 -0.673 -0.765 -0.882 -0.939
Total Elasticity of Poverty -1.095 -1.586 -1.923 -2.013 -2.145 -2.313 -2.566 -2.917 -3.361 -3.578
Source: South African Tax and Transfer Simulation Model (SATTSIM-Plus), ADRS

- 57 -
Table 4. Pro-Poor Growth Indicators

Group A Scenarios Group B Scenarios Group C Scenarios Group D Scenario

A1 A2 A3 B1 B2 C1 C2 C3 C4 D1

Growth Elasticity of Poverty -0.404 -0.585 -0.709 -0.743 -0.791 -0.853 -0.947 -1.076 -1.240 -2.238
Inequality Elasticity of Poverty -0.691 -1.001 -1.214 -1.271 -1.354 -1.460 -1.620 -1.841 -2.121 -3.731

Total Elasticity of Poverty -1.095 -1.586 -1.923 -2.013 -2.145 -2.313 -2.566 -2.917 -3.361 -5.969
Notes: Growth elasticity of poverty (Kakwani 1993) is the percentage change in poverty due to a 1 percent increase in economic growth, provided that the growth process does not
change inequality (i.e. the benefits of growth are distributed equally among everyone in the country). Inequality elasticity of poverty is a measure of how much the total poverty index
changes as a result of changes in inequality that accompany the growth process. The growth is pro-poor (pro-rich) if the change in inequality that accompanies growth reduces
(increases) the total poverty. Therefore, the growth is pro-poor (pro-rich) if the total elasticity of poverty is greater (less) than the growth elasticity of poverty (Kakwani et al 2003).

Source: South African Tax and Transfer Simulation Model (SATTSIM-Plus), ADRS

- 58 -
Table 5. Distinguishing Features of Economic Scenarios

Group D
Group A Scenarios Group B Scenarios Group C Scenarios
Features Scenario
A1 A2 A3 B1 B2 C1 C2 C3 C4 D1
1 Average annual real growth rate of 5 percent X X X X X X X X X X
2 Average annual inflation rate of 5 percent X X X X X X X X X X
Low employment intensity of Growth
3 (employment elasticity of growth equals 0.20) X

Medium employment elasticity of growth

4 (employment elasticity of growth equals 0.40) X

High employment elasticity of growth

5 (employment elasticity of growth equals 0.60) X X X X X X X X

6 Pro-poor allocation of new employment X X X X X X X

7 Pro-poor increase in average salaries X X X X X X
Extension of child support grant to poor children
8 X X X X X
up to 17 years old
9 Adult grant for the unemployed X X X X
10 Annual real increase in grant amounts X X X
Increase in government in-kind social transfers
11 to poor households X X

- 59 -
What is pro-poor growth?
What are some of the things
that hinder its achievement in
South Africa?


Charles Meth

Honorary Research Fellow School of Development Studies University

of Kwa Zulu Natal and Research Associate South African Labour
and Development Research Unit, University of Cape Town

Prepared for

Oxfam GB South Africa

24 May 2007

- 60 -

Occasions when research findings do not answer the question they set out to confront,
are probably more numerous than we know. One way of dealing with this problem is
to re-title the findings to reflect more accurately, the question that ends up being
addressed. In this spirit, the original title of this piece of commissioned research,
which was The limits to redistribution (inequality reduction) through pro-poor
investment, has been treated in the same way as the working titles so common in
the film industry.

The reason for this deviationism is that it transpired, once research commenced, that
references to pro-poor investment in the literature are exceeding rare. Not only that, it
did not take long to discover that there is, if not ignorance, then disagreement about is
meant by the term pro-poor growth that such investment is presumably intended to
stimulate. To deal with the notion of pro-poor investment, it was thus necessary to
grapple first with that of pro-poor growth. It requires but the briefest of encounters
with the economic growth literature to discover that the connections between growth
and investment are still but poorly understood. Hopes of dealing with pro-poor
investment in a satisfactory manner were, therefore, pushed into second place, and the
goal of understanding pro-poor growth was allowed to dictate the research agenda.

After a brief review of the poverty problem in South Africa, the paper dives into the
international literature on pro-poor growth. The disagreement about what pro-poor
means may be seen in the way it is understood by those who deploy it. One group
(clustered around the World Bank), defines it as any growth that raises the incomes of
the poor. The other, loosely grouped around the UNDP, accepts that limited
definition, but argues that growth must be progressive in the sense of reducing

Increasing concern with the impact of inequality on poverty reduction strategies is

reflected in a torrent of literature from the World Bank. This culminates in a paper by
Bourguignon (2004) on the poverty-inequality-growth triangle, which speaks to the
difficulties of redistributing assets, the favoured (conservative) form of
redistribution. It ends with Bourguignon acknowledging the efficacy of social grants
as a means of reducing poverty and inequality, admitting that they could be viewed as
contributing to human capital formation (and hence, qualify as pro-poor

The spotlight is then turned on South Africa, ostensibly to examine the scope for
redistribution, but actually to consider the extent to which the country has followed
the package of policies generally regarded as favourable to pro-poor growth (in the
sense in which the UNDP defines it). Obstacles to reductions in inequality (through
pro-poor growth) are sought and found in the fields of education, healthcare,
extension of the social grant system, black economic empowerment, job creation, and
the reduction of regional disparities, where, it is argued, government is more or less
paralysed by commitments to developmental state objectives which it lacks the
capacity to discharge.

- 61 -
Preface 1
Poverty in South Africa is so obvious and so extreme that there is scarcely any need to
justify the assertion that it should be eradicated. Inequalities, of income, of wealth, of
opportunity, of life chances, are also extreme (the magnitude of the much-vaunted
social wage notwithstanding)gross deprivation is everywhere to be found, often
cheek-by-jowl with extremes of opulence. At least as far as income inequalities are
concerned, there is a widespread perception that these are unacceptably high (SASAS,
2003). Yet the argument that more income should be redistributed than is currently
the case does not enjoy much support among the well-off. Self-serving assertions
about the deleterious effects on the poor of handouts, which trip easily off the
tongue whenever the topic is raised, drown the more persuasive arguments in favour
of social justice.

Redistribution is an emotive and controversial topicthis paper tries to take some of

the heat out of the discussion by looking in some depth at what the great and the good
in the research community have had to say about the relationship between poverty,
inequality and growth. The paper also attempts to look at the constraints, of structure
and agency that inhibit (or facilitate) the implementation of anti-poverty (and anti-
inequality) policies. By these means, it is hoped to take some of the sting out of what
must seem to many, a set of unpalatable conclusions.

Thanks are due to Asghar Adelzadeh and Mohamed Motala for the critical but constructive comments
they offered. Needless to say, responsibility for the piece as it now stands, is mine alone.

- 62 -
Table of contents

Abstract ....................................................................................................................61
Structure of the remainder of the paper ...............................................................68
What is pro-poor growth? ........................................................................................69
A brief history of the notion of pro-poor growth.................................................73
The glacier melts: Inequality is bad for the poor .................................................77
What is to be redistributed? .....................................................................................82
The scope for reducing inequality in South Africa..................................................85
Redistributive growth: Moments of inertia.........................................................90
Human capital formation: The domino transition .........................................93
Dis-ease in the health sector?...........................................................................96
Income inequality: Type As vs. Type Bs, Round 2 .........................................99
Social grants: More than a glimmer of light ..................................................100
Black economic empowerment......................................................................101
Employment creation .....................................................................................103
Regional disparities........................................................................................106
Developmental state delusions.......................................................................112
What to do until the state gets developmental? .............................................114
Conclusion: When all else fails, lower your standards ..........................................116

Index of tables

Table 1 Monthly per capita income in South Africa and its nine provinces among the
African population group, September 2004.................................................................66
Table 2 Poverty impacts and the non-correlation of inequality and growth...............77
Table 3 Policies to promote pro-poor growth: Research findings, consensus policies,
and remaining debates..................................................................................................86
Table 4 Process issues in promoting pro-poor growth: Research findings, consensus
policies, and remaining debates ...................................................................................89
Table 5 Reading and mathematics skills in 15 African countries...............................94
Table 6 Nos. of people reporting problems at clinics during June 2005 ....................98

Index of figures

Figure 1 How growth affects the poor ........................................................................70

Figure 2 Map of Eastern Cape Index of Multiple Deprivation .................................108
Figure 3 Map of Limpopo Index of Multiple Deprivation .......................................109
Figure 4 Map of Western Cape Index of Multiple Deprivation................................110

- 63 -

Substantial redistributions of income (from poor to rich, or vice-versa, or to or from

the middle classes), and the circumstances in which they occur, are rare enough to
excite comment whenever they do. Invariably, such circumstances involve shifts in
the balance of power between different groups or classes, 2 it being generally the case
that redistributions of income are zero-sum games. As will be seen below, the
proposition that a reduction in inequality is desirable in countries like South Africa
where both poverty and inequality are extreme, enjoys fairly broad acceptance. That,
however, is not sufficient to ensure that policies for the reduction of both will
necessarily be pursued. Recognition of deprivation among some significant
proportion of a population may be a necessary condition for the alleviation (or better
still, the eradication) of poverty through redistributive measuresmoral suasion,
based upon that recognition, is unlikely to be sufficient to bring about the desired
change. Within any economy, the limits to redistribution (or inequality reduction) are
set by a variety of structures, institutions and agents. When the research commenced,
the hope was entertained that it would be possible to understand the significance of
these constraints for attempts to bring about redistribution by pro-poor investment.

What, exactly, constitutes pro-poor investment is not a settled matter. So rare are
the uses of this term (which presumably owes its existence to the emergence, in recent
times, of the debate about pro-poor growth), and so complex is the connection
between investment (in general) and growth (in general), that the task of saying
anything substantial about it appears to be impossible. This means that one of the
tasks of the paper, that of discovering what has been found to be, or what has been
proposed as, pro-poor investment, has not been (cannot be?) discharged very well.

And thereby hangs a tale. There may be researchers who can write the introduction to
a paper before the conclusion is wrapped up, I am not one of them. Writing the
introduction after the research has been completed means that one has before one,
whatever has turned up in the attempt to address the research questions under
consideration. The pessimistic findings of the present paper suggest that the way to
start the paper is to make it clear from the very outset, that the paper makes no
meaningful contribution to the meagre stock of knowledge about pro-poor
investments. Instead, therefore, of commencing by discoursing on the joys that only
pro-poor investment can bring, we begin by looking at some of the dimensions of
South Africas poverty problem.

The South African government, possibly influenced by a set of estimates that register
a substantial decline in poverty between 2000 and 2004 (van der Berg et al, 2005), has
taken of late to boasting that not only will South Africa virtually eliminate poverty
by 2014 , but also that it will radically reduce inequality (AsgiSA, 2006, p.17). If
poverty is conceived of as a ravenous monster, then the success of anti-poverty policy

The Conservative (Thatcherite) revolution saw the Gini coefficient rise from about 0.28 in 1985 to
roughly 0.37 in 1990. Source, (National Statistics
online). Downloaded: 12th January 2007. The share of the top decile of income earners in the USA
went from about 32 per cent in 1979, to about 44 per cent in 2000 (Piketty and Saez, 2006, p.11). They
outline competing explanations of the phenomenon, one of which posits an increased ability of
executives to set their own pay and extract rents (p.7).

- 64 -
may be judged by the extent to which choices made among the menu of policy options
succeed in sating its hunger. That appetite, it will be observed, does not disappear
after a single mealthe hunger must be faced every day. Getting things right means
that the monster gets smaller, not that it goes awaypoverty persists even in the
wealthiest countries. A hint of the size of the monster is given by the figures in Table
1. In the table, the distribution of income among the African population group is
given in the form of a set of estimates of the numbers and percentages of individuals
in various income categories. Provinces in the table are ranked by severity of poverty,
the Eastern Cape being worst-off, and Gauteng best-off. At the bottom end of the
distribution, two of the commonly used poverty lines in South Africa are used to
describe the extent of the poverty problem. The first, used by van der Berg et al
(2005) is set at R250 per capita per month in 2000 prices. Inflated using the CPI to
2004 prices, this amounts to about R309 per month, or about R10 per person per day.
The other line, R400 per month in 2004 prices, is the inflated version of the R322 per
capita per month in 2000 prices, applied in Hoogeveen and zler (2004). The income
measure used in the table includes all earned income reported in the LFS, all migrant
remittances in cash, and all social grants. To correct for under-reporting an error of
75 per cent is assumed in the reported values of earned income and remittances (i.e.,
reported incomes and remittances are multiplied by 1.75). This brings survey income
totals roughly into line with national accounts income estimates (Meth, 2006a).

As may be seen, a little less than half of the African population is below the
exceedingly modest line of R309 per month, while more than 55 per cent is below
what will probably be the official bottom bound of the poverty critical range. 3 With
income poverty rates approaching 70 per cent, Limpopo and the Eastern Cape, South
Africas poorest provinces, are home to about 40 per cent of the countrys poor.
KwaZulu-Natal, a distant second to Gauteng in employment and remuneration terms,
has the largest absolute number of poor, many of them remote from the industrialised
metropolitan areas. Between them, these three provinces contain more than 60 per
cent of South Africas poor Africans. The per capita incomes of a further 1.1 million
people of the Coloured population group (among a population of about 4.12 million),
over 900 000 of them in the three Cape provinces, fall below the R400 line. 4

Of the 18 million or so people below the R309 poverty line in 2004, 14 million lived
in workerless households (most containing working age people, but in which nobody
had employment). These zero-income (from employment, that is) households
survived on a mix of social grants and/or remittances. Among them were about 1.8
million people in households receiving no incomes at all in the survey reference
period, subsisting, we know not how. 5 The remaining four million people below the
poverty line were located in households containing about 800 000 workers. Although
the bulk of poverty is caused by unemployment, the problem of the working poor still
looms fairly large.

The R400 per month (R322 in 2000 prices) line originates in the work of Babita et al (2003). The use
of a wide poverty critical range is recommended, inter alia, by Woolard and Leibbrandt (2001, p.49).
Doing so makes it possible to establish whether [our] poverty rankings are robust.
Source of estimate: own calculations on September 2004 LFS data set. See Meth, 2006a.
People cannot survive on nothing. The likely explanation for these true zero income households is
the shortness of the reference period during which information on earnings is collected. See Meth,

- 65 -
Table 1 Monthly per capita income in South Africa and its nine provinces among the African population group, September 2004
Eastern KwaZulu- Mpuma- North West Northern Western
Cape Limpopo Natal langa Province Free State Cape Cape Gauteng Total
<R100 17.5 16.9 17.3 11.5 16.6 14.4 15.6 17.6 15.6 16.2
<R309 42.5 42.5 34.9 35.3 30.0 30.1 22.8 14.9 14.7 32.0
Sub-total <R309 (%) 60.0 59.4 52.2 46.9 46.5 44.5 38.3 32.5 30.4 48.2
Sub-total (people) 3 689 000 3 268 000 4 221 000 1 382 000 1 589 000 1 094 000 133 000 373 000 2 081 000 17 803 000
Sub-total (No. <= 15 years) 1 698 000 1 513 000 1 847 000 627 000 613 000 427 000 57 000 127 000 694 000 7 588 000

R310-400 7.9 8.4 6.9 9.5 8.3 8.4 9.2 3.9 3.8 7.1
Sub-total <R400 (%) 67.8 67.8 59.1 56.4 54.9 52.9 47.6 36.4 34.2 55.3
Sub-total (people) 4 173 000 3 730 000 4 775 000 1 663 000 1 874 000 1 301 000 165 000 418 000 2 343 000 20 416 000
Sub-total (No. <= 15 years) 1 899 000 1 692 000 2 079 000 734 000 708 000 506 000 67 000 138 000 785 000 8 590 000

R400-1199 21.6 21.0 24.9 28.1 27.6 29.5 32.3 38.9 29.8 26.0

R1200-2499 5.8 5.5 9.4 9.3 9.3 8.9 13.0 17.8 20.2 10.5

R2500-4999 3.4 3.7 4.3 4.4 4.7 6.3 4.6 4.2 9.2 5.1
R5000-9999 1.2 1.7 1.9 1.4 2.9 2.0 3.7 2.9 5.2 2.5
R10 000+ 0.4 0.3 0.5 0.6 0.6 0.5 1.2 0.2 1.7 0.7
Total no of people 6 152 000 5 499 000 8 082 000 2 949 000 3 414 000 2 458 000 347 000 1 148 000 6 853 000 36 901 000
Total no. <= 15 years 2 545 000 2 295 000 3 135 000 1 133 000 1 163 000 838 000 113 000 322 000 1 988 000 13 529 000
Remuneration/capita/annum 5018 2179 8194 8920 6783 6496 9300 13 124 28 618 11 211
Relative per capita
household expenditure
Africans, 2003 42 43 56 62 66 61 71 65 88 62
Source: Income distribution - own calculations from September 2004 Labour Force Survey data set. See Meth (2006a).
Per capita remuneration per annum is obtained by dividing the value of remuneration of employees based on regional services levies received for the year 2004 by the
mid-year population estimate for that year. Remuneration figures are from P9149, 30 June 2006, Table 1, and population figures from P0302, 27 July 2004, Table 20.
Relative per capita expenditure figures are from Department of Education (2006a, Table 10, p.15). Mean expenditure for the country as a whole is 100.

- 66 -
A gloomy paper by members of the Harvard panel convened to study constraints and
opportunities confronting the states AsgiSA (Accelerated and Shared Growth
Initiative for South Africa) concludes that the labor market is very near the steady
state so it is unlikely that the unemployment rate will fall without intervention or an
external shock. (Banerjee et al, 2006, p.46). Urging caution, these authors conclude
that the problem is amenable to policy intervention. Other than to refer in passing to
the old standbys, educational reform, training and subsidies, they steer clear of any
prescription, arguing that their analysis is simply insufficient to distinguish options.
In place of the big push, they propose well-designed, carefully monitored small scale
interventions (p.60). The unemployed, it would seem, are going to have to be patient
for a while longer. 6

Such reductions in the number of poor as have taken place between 2000 and 2004, 7
did so primarily because of the rapid expansion in the number of child support grants
paid. Since these grants are relatively small (but welcome nonetheless among the
poor), those lifted above the poverty line are likely to have been the less poor. Even
so, the benefits have been widely distributed. When that grant programme begins to
approach full take-up, poverty reduction via the social grant route will cease. It is a
simple matter to demonstrate that there are too few children, relatively speaking, to
raise household incomes in the poorest segment of the population by much. The
absence of any significant social protection for those in the age group 15-60 (females)
and 15-65 (males), and the unwillingness of the state to consider social grants for the
unemployed, 8 thus leaves the majority of the poor completely exposed.

The paragraph in which the boasts (repeated above) about South Africas coming
onslaught on poverty and inequality are located, reads thus:

We believe that we have built the basis for a national effort to achieve faster and
shared economic growth. With this programme [AsgiSA] we can achieve our
social objectives and we can more than meet the Millennium Development Goals.
Our second decade of freedom will be the decade in which we radically reduce
inequality and virtually eliminate poverty. We know now that we can do it,
working together around an initiative which has the support of the nation. 9

Three aspects of this statement are noteworthy: the first is the assertion that poverty
will be virtually eliminated by 2014-2015. The second concerns the form of growth,
My analysis of changes in employment patterns in the period 2001-2004 suggests overwhelmingly
that the jobs that were created were primarily at the top end of the labour market (Meth, 2006a, Table
11, p.46), the same sort of skill bias reported by other commentators.
The extent of this reduction is the subject of disagreement between myself and van der Berg et al,
(2005). According to van der Berg et al (2005, p.17), between 1993 and 2000 the poverty headcount
ratio rose (as did the poverty headcount itself, from 16.2 to 18.5 million), the poverty gap ratio rose,
and so did the poverty gap squared (the poverty severity ratio). Between 2000 and 2004, they have
poverty fallingthe headcount goes from 18.5 to 15.4 million, all of the ratios improve as well. My
estimates have poverty falling between 2001 and 2004 by between 1.2 and 1.5 million (Meth, 2006a,
Table 8, p.37)
In line with conservative ideology that insists that unrequited transfers are bad, the state offers the
Extended Public Works Programme (EPWP) instead of grants. Although some of the projects are
worthy, the state simply does not have the capacity to administer a programme of the magnitude
required to address the unemployment problem.
Downloaded from, 19th January 2007.

- 67 -
faster and shared (the latter, as we see below is a synonym for pro-poor); the third
is the commitment to a radical reduction in inequality. South Africas record in
addressing poverty since the advent of democracy is spotty. It is more or less
common cause that (income) poverty rose slightly between 1993/95-2000. What
happened thereafter is contested. While the so-called social wage has undoubtedly
improved the lot of the poor (Bhorat et al, 2006), it is easy (and tempting on
governments part) to exaggerate these improvements (Meth, 2006b). The harsh truth
is that income poverty is still a severe problem in South Africa, one on which vast
spending on education, health, housing, free electricity and water, and a host of
other programmes has yet to make a serious dent.

If all the virtually eliminate poverty boast rests on is an extrapolation of the van der
Berg et al results (and there is little else in the AsgiSA which could justify such an
extraordinary claim), then to describe it as unrealistic, is to be charitable. Similarly
with the radical reductions in inequalitythe AsgiSA document gives little sense of
how this is to be achieved. Nor does it suggest what inequality is to be reduced:
inequality of income? inequality of wealth? inequality of opportunity? In this regard,
it is on all fours with other, similar pronouncements about the promotion of pro-poor
growth. In short, the punchline of the analysis of the concept of pro-poor growth that
follows is that although the concept is capable of being defined in technical terms (the
definitions are disputed), it is best understood as a political aspiration or goal.

Not only must those who would understand the potential of pro-poor growth (and
investment) to address the problem of poverty and inequality be familiar with the
extent and nature of the problem, they must also be able to decipher the term pro-
poor, to find out what really is intended by its use. Having sketched the barest of
outlines of the problem of poverty in South Africa (nothing has been said about
inequality), it is time now to look at the debate (technical and political) about the
adjective, pro-poor.

Structure of the remainder of the paper

We have had our first taste, in the form of a brief reference to the response of the
Harvard panel members to South Africas unemployment problem, of the kind of
policy advice (stolid?) that emerges in the face of seemingly overwhelming odds.
Advice available elsewhere on how to deal with poverty does not differ all that much.
Lest this be thought to smack of arrogance, let it be recorded that it is said in full
awareness of the fact that the social ills of unemployment, poverty and gross
inequality have long defied the very best brains in the business and, in many places,
look set to continue doing so.

Although there have been many development success stories, and millions have been
raised out of poverty, there are also well-documented instances of long-run growth
coinciding with rising poverty (Cord et al, 2003). The clamour for pro-poor may be
seen as arising out of the concession (grudging?) by World Bank economists that
poverty will not readily be eradicated by growth in those countries where inequality is
severe, or rather that it will take an unpalatably long time for it to do so. How then to
tackle the subject? One starting point, for want of any more obvious entry port, is
with the debate over pro-poor growth. Having surveyed the attempts to define pro-

- 68 -
poor growth concept, and having documented the confidence in (any) growth as a
poverty alleviating mechanism giving way to a much more nuanced understanding of
the obstacle that inequality constitutes for poverty reduction, the paper dives into the
poverty-inequality-growth triangle (Bourguignon, 2004). Once inside this part of
the literature, it is possible to approach the question of what is regarded as being
capable of being redistributed in a theoretically rigorous manner.

The laundry list which this excursion generates is supplemented by another, more
comprehensive list reproduced from Klasen (2003), the latter being the fruit of an
extensive examination of policies that foster pro-poor growth (being both pro-poor
and pro-growth). South African policy is squashed through the grid he offers, and a
few critical areas where implementation falls far short of aspiration are selected for
critical review. These include schooling (human capital formation); income
inequality; black economic empowerment; employment creation, and regional
disparities. 10

What is pro-poor growth?

It is by no means unheard of for controversy to rage around concepts in the social

sciences. The concept of pro-poor 11 is a case in point. Given that the purpose of
this section of the paper is to define pro-poor growth, the history of its emergence that
follows may seem to be unduly long. There is, however, virtue in digging deeply into
its background, for only in that way can we understand why confusion about precisely
what pro-poor growth is supposed to means lingers on. To pre-empt the findings of
this excursion, it is concluded (as noted above) that the most useful way to view the
notion of pro-poor growth is as a political aspiration or goal. Like many political
goals, pro-poor growth means different things to different people. In the absence of a
consensus view of what pro-poor is, policymakers whose job it is to design and
implement pro-poor policy, are hampered in their efforts to do so by the lack of clarity
of what is to be achieved by the policy. Behind the confusion, it will be argued, lies a
strong disagreement between different constituencies about how much (and what kind
of) redistribution is desirable.

In the international literature, two views on what constitutes pro-poor growth may be
discernedreconciling took them a surprisingly long time. Figure 1 below,
reproduced from a DfID (Department for International Development of the UK
government) briefing paper (DfID, 2004, p.1), may be used to illustrate the dispute
and its (uneasy) resolution.

The laundry list approach to policy formation is not without its dangers. There is some discussion
of these in the concluding section of the paper.
Concept is possibly too grand a label to attach to the notion. Most commentators tiptoe around it,
only one of the pieces at which I have looked has the courage to describe the term as the politically
correct adjective it so obviously is (Bhalla, 2004, p.1). This notwithstanding, Bhallas paper, like
those of so many other researchers, is about the search for an adequate definition of pro-poor growth.
In passing, it is worth noting that most of the papers consulted in constructing this overview restrict
themselves to defining growth in terms of simple measures like per capita GDP. In general, the poor
emerge as people only when we turn to questions of the policies required to achieve pro-poor growth.
Questions about the nature of growth itself emerged in only one of the pieces I consulted. May (2004,
pp.146ff), following Douthwaite (1997) uses the notions of good and bad growth (as developed
among green economists) to address the question of when growth may be deemed to be pro-poor.

- 69 -
On the vertical axis is measured the rate at which the incomes (or consumption) of the
poor grows. The horizontal axis records growth rates of income (consumption) for
the population as a whole. The faster growth measured along the vertical axis takes
place, the faster poor people are raised out of poverty. Crudely speaking, if the rate of
growth of the incomes of the poor is slower than that of the population in general,
including the non-poor, income inequality must increase. That would have occurred
in Brazil, Bangladesh, India and Chile for the time period under consideration (mid-
1980s to late nineties). In Ghana, although growth was sluggish, the poor fared better
(in income growth terms) than the population overall (growth rates for the poor
exceeded those in Brazil in Bangladesh). 12 If the data are to be trusted, income
inequality in Ghana will have fallen.

Figure 1 How growth affects the poor

How growth affects the poor

(% per anum)

Growth rate 4
for the poor

3 India

-4 -3 -2 -1 0 1 2 3 4 5 6 7
Overall growth rate



Source: DfID, 2004, p.1

The position adopted by one group in the argument (frequently advanced by World
Bank economists) is that any growth on the vertical axis is pro-poor. This is also the
position taken by DfID. It is explained as follows: DfID distinguishes absolute pro-
poor growth from relative pro-poor growth. The former considers only the incomes
of the poor, the growth of which, as DfID observes, is closely related to the rate at
which absolute poverty is reduced. Relative pro-poor growth takes into account
income growth among the non-poor as well. If, as in Ghanas case, growth on the
vertical axis exceeds that on the horizontal, then both absolute and relative pro-poor
growth can be said to have taken place. DfID goes on to say that in the light of their

In Zambia, incomes of the poor fall at a slower rate than incomes of the population in general. The
only consolation to be had from this sad performance is that inequality did not worsen (if the data are to
be believed).

- 70 -
commitment to the Millennium Development Goal of halving poverty by 2015, it is
the absolute definition of pro-poor growth that is preferable (DfID, 2004, p.2). 13

The disdain of the UNDP for this position is thinly disguised. Those of progressive
disposition will be pleased by the UNDPs ditching of technicist terms like absolute
and relative pro-poor growth, in favour of the adjective progressive to describe
relative pro-poor growth. 14 The Human Development Report 2005 introduced their
discussion of progressive growth in the following terms:

Like motherhood and apple pie, everybody is in favour of pro-poor growth.

The concept, like its increasingly popular and more recent variant shared
growth, captures the idea that the quality of growth, as well as the quantity,
matters for poverty reduction. But the concept means very different things to
different people. (HDR 2005, p.65)

Arguing that the use of progressive to describe the form of pro-poor growth the
UNDP prefers to foster, is no mere semantic difference, the HDRs compilers claim
that focusing:

on the relative position of poor people highlights the potential for small
distributional shifts to produce major gains for poverty reduction. (HDR 2005,

In the UNDP approach, redistribution takes its place firmly beside growth as the
object of policy, and that, on two grounds. The first of these is social justice. Along
the 45 line in Figure 1, growth is described as distribution neutral. Hard-line growth
fundamentalists argue that increasing inequality (further movement into the quadrant
containing Brazil, Bangladesh, India and Chile in Figure 1) is not a problem as long as
the poor gain (move up the vertical axis) in the process. 15 To show how difficult such
thinking is to square with notions of social justice (itself a hotly disputed concept), the
HDRs authors use the example of Brazil to show how little distribution neutral growth
(growth in which relative shares do not change) benefits the poor. For every $1

Difficult though it may be for many countries to achieve, the first Millennium Development Goal
Eradicate extreme poverty and hunger is modest. It translates into: Reduce by half the proportion of
people living on less than a dollar a day. and Reduce by half the proportion of people who suffer from
hunger (see The adoption of such a goal does not mean that
headcount itself must fall as wellit is quite possible for population growth to cause headcounts to rise
while headcount ratios fall. Whether or not one could then say that poverty had gone down is one the
hard questions posed by Kanbur (2004, pp.6-7).
As far as the Eradicate extreme poverty and hunger Millennium Development Goal is concerned, the
indicators for poverty halving target are (i) the proportion of the population below $1 (PPP) per day (or
below a national poverty line), (ii) the poverty gap ratio (incidence x depth of poverty), and (iii) the
share of the poorest quintile in consumption. It looks as though only the first of these (the least
onerous?) is relevant for meeting the target. The desired changes in (ii) and (iii) are not specified.
Those of conservative bent, presumably will not.
An example of this type of thinking is to found in a compendious comparative analysis of 21 studies,
sponsored by the World Bank, undertaken in developing countries (Lal and Myint, 1996). The authors
conclude their first chapter (Poverty, Equity and Growth) thus:
The distributivist objections (based on the Kuznets curve) to the classical liberal presumption that
growth would alleviate poverty thus seem to have little empirical support. On the existing evidence,
mass poverty can be alleviated by rapid, efficient (labour-intensive) growth, and we need not worry
about the distributional consequences. (Lal and Myint, 1996, p.44)

- 71 -
increment to total income, the top 20 per cent would receive 85 cents, as opposed to
the three cents received by the bottom 20 per cent. If society attaches more weight to
the wellbeing of the poor, such an outcome is inconsistent with basic principles of
fairness and social justice (HDR 2005, p.65). Since not everyone agrees with that
vision of social justice (nor indeed, believes that what constitutes social justice can be
agreed upon), we note that there are reasons other than the merely(?) ethical or
normative for supporting policies to reduce inequality. These are considered further

A second and related advantage of progressive growth addressed by the HDRs

authors (and numerous others) is that if poverty reduction [or better still, poverty
eradication] is a goal of policy, then the reductions in inequality that go hand-in-hand
with a greater (proportional) capture of the benefits of growth by the poor will (ceteris
paribus) raise the rate of poverty reduction.

Increasing [poor peoples] share of additional growth, they argue, can

accelerate the rate at which rising prosperity reduces poverty, while at the same
time raising the overall growth rate. (HDR 2005, p.65)

The big debate, of course, and one that is central to the question of the limits of
redistribution, is that of whether the other things being equal condition holds,
another topic to be addressed further below.

While the distinctions above between absolute and relative pro-poor growth (or
between growth and progressive growth) both seem unexceptionable, Klasen (2003),
in search, as the title of his paper indicates, of the holy grail, proposes a fourfold set
of criteria by which to define pro-poor growth. The definition must address the

1. The definition should differentiate between pro-poor and other kinds of

2. It must at least imply that the poor have benefited proportionately more from
growth (i.e., it must be progressive in the sense described above)
3. It must be sensitive to the distribution of income among the poor, giving
greatest weight to the poorest
4. Finally, the measure must not focus exclusively on the poor, but must permit
as well, assessment of the growth performance overall (Klasen, 2003, pp.3-4)

To meet all these requirements, he argues in favour of the use of the population and
poverty weighted measures proposed by Ahluwahlia and Chenery in 1974 (Klasen,
2003, p.4). The former is the average of the percentile growth rates. Since overall
growth rates in unequal economies are more or less determined by growth rates at the
top end of the distribution, a population-weighted measure would give greater weight
to growth rates among those not fortunate enough to be located there. 16 The poverty-

The following simple example illustrates the proposition. Suppose that the distribution of income in
an economy among the five quintiles in period t 1 is 30; 70; 100; 250 and 550. By period t, these
have grown to 30.5; 71.2; 101.5; 255 and 561 respectively. The overall growth rate in the economy is
1.92 per cent, whereas the population-weighted growth rate is 1.78 per cent. The overall rate is pulled
down by growth of 1.67 per cent in the bottom quintile; growth of 1.71 per cent in the second from
bottom, and of 1.50 per cent in the middle quintile.

- 72 -
weighted measure is more controversial, requiring as it does greater weight on the
income growth among the poor and lesser weight on that of the well-off. As Klasen
points out, such a process is akin to the Atkinson inequality measures (p.4). He refers
to empirical work of his own that applies these measures (p.4n) but does not suggest
ways of resolving the disagreements over how relative weights should be
determined. 17

Summing up, we may see that stipulations (1) and (2) above land squarely in the
debate reviewed above, with (2) expressing the UNDP position. The approach
implied by the DfID approach (see Figure 1) would not meet condition (3) because
that simply lumps together the poor. The failure of that approach to take into
account differential conditions among the poor means that the overall assessment
required in (4), if produced without population-weighting the growth estimate, may
overstate progress. Unlike DfID (and the World Bank), Klasen is not about to be
satisfied by a formulation which says that any increase in the incomes of the poor is
evidence of pro-poor growth. Inequality has moved to centre stage.

For Klasen and for the UNDP (as exemplified, for example, by Kakwani et al, 2004
or the 2005 Human Development Report) the route by which inequality came to
occupy this position is direct. The World Bank has dragged itself to a similar
position, but by a somewhat different path. Before we examine that story, it will be
useful to take a brief at the way the concept of pro-poor growth entered the
development discourse, noting once again the use above by the AsgiSAs scriptwriters
of the fashionable concept of shared growth, a lapse into jargon that is entirely in
keeping with the boast promising a radical reduction in inequality.

A brief history of the notion of pro-poor growth

Although the evils of inequality have been debated with the passion and vigour they
merit for centuries, one of the pieces responsible for the re-introduction of the
question of redistribution into polite (neo-classical) circles was probably Ravi
Kanburs (1987) paper. In it he writes:

Having discussed the measurement of poverty, we now turn to the question of

alleviation. To some, however, this is a nonquestion. Or, at least, it is not a
question separate from the question of achieving the fastest possible rate of growth.
It is suggested that trickle down will solve the problem in due course and that
redistributive measures that dampen growth will hurt the poor more than they
benefit them in the long run.

Taking a hypothetical set of poverty and growth figures that do not seem to be far
off the mark for many developing countries he estimates crossover periods, noting

There is an echo in this exercise of the debate over whether a consumer price index should be
democratic or pluotcratic, i.e., weighted by population or by share of consumption. In practice,
CPIs are plutocratic. This means that if costs of the consumption basket of the poor increased at a
greater than average rate, the index would have difficulty in detecting the change.
These difficulties are dwelt upon at some length by Cord et al, 2003, p.4.

- 73 -
it will take more than twenty years for the average poor person to be lifted out
of poverty. (his emphasis)

Playing around with the figures, he manages to get the crossover period down to
fifteen years, observing after doing so that:

It is in this context that the urgency of poverty alleviation measures has to be

seen. Waiting for three or four five-year plan periods for poverty alleviation
(which does not take into account the poorest of the poor) may be too long given
the objectives of some governments. Explicit redistributive strategies may well be
introduced in response to slowness in trickle down it is simply a matter of
political arithmetic. (Kanbur, 1987, pp.70ff)

At some point after the Kanbur article referred to above appeared, the precise date is
not important, the term pro-poor slipped into the development-speak lexicon. 18 It
emerged in an atmosphere of spirited contention between what, on the one hand,
could be described as growth fundamentalists, and on the other, a group that places
greater emphasis on equity, and on the problem of addressing the welfare of the poor
until such time as they can be rescued by growth. An uneasy truce seems to have
settled over the debate now, with the publication by the UNDPs International Poverty
Centre of a paper by Kakwani et al (2004) on the nature of pro-poor growth. One of
Kakwanis co-authors was Shahid Khandker, of the World Bank in Washington, a fact
which in and of itself, does not sound particularly interesting, until consideration is
given to the history of a few short years earlier, when the World Bank and the UNDP
were, if not at daggers drawn over what pro-poor meant, were certainly mutually
critical of their respective standpoints.

Energy was infused into the debate by the publication of the Dollar and Kraay (2000)
paper on the impact of economic growth on the welfare of the poor. The central
message of the paper, published by the World Bank, is captured in its titleGrowth
is good for the poor. While not saying that growth was everything, as far as poverty
alleviation was concerned, the ideological slant of the Dollar and Kraay work was the
primacy of growth as a means of addressing poverty. The Dollar and Kraay paper
was widely seen among progressives as an attempt to drag the World Bank off a path
down which its former head had attempted to send it(articulated in his 1997 annual
address at the IMF-World Bank meeting entitled The challenge of inclusion.
(Wolfensohn, 1997)one which entailed a clear commitment to growth with
equity (Oxfam, 2000, p.5). If the Dollar and Kraay paper had been a just a research
exercise, Oxfam observes, it could have been dismissed as weak and largely
irrelevant. Sending the message that it did (leave it to the markets and let the
benefits of growth trickle-down), and bearing as it did, the imprimatur of the
World Bank, (Oxfam, 2000) 19 the Dollar and Kraay piece aroused a storm of protest
(Dagdeviren et al, 2000; Lbker et al, 2000; Oxfam, 2000; Wade, 2001).
The earliest reference I can find, in a far from comprehensive search, is a paper by Ravallion and
Datt (1999). No doubt there are other, earlier uses of it.
Strictly speaking, the Dollar and Kraay paper does not bear the imprimatur, as Oxfam claimed. As a
matter of course, the Bank uses the standard disclaimer that the views expressed in any of the papers it
publishes are those of the authors and not necessarily those of the Bank. How believable such attempts
at distancing an institution from controversial views is not obvious.

- 74 -
Misrepresentation of the views of opponents is the stock-in-trade of the politician and
the fundamentalist alike (and, if we are honest, of all of us, on occasion). One of the
criticisms of the Dollar and Kraay effort is that they are guilty of that offence. The
Oxfam (2000) piece, for example, states that:

Much of the Dollar - Kraay report is spent caricaturing the arguments of others in
order to create convenient straw men. The view (held by almost nobody but
presented as widespread) that growth is bad for the poor is subjected to a withering
critique. Oxfam (2000)

Dagdeviren et al are more temperate, but equally dismissive. The observe that:

In a recent paper, Dollar and Kraay (2000) reach the conclusion, based on cross-
country regressions, that the typical outcome of the growth process in developing
countries is to leave the income share of the lowest quintile unchanged; ie.,
distribution-neutral growth. For some reason that is not clear, the authors
characterise this with the phrase, growth is good for the poor (italics in the
original). This is a rather strange statement, for it challenges the imagination to
produce any growth pattern that would provoke the converse phrase, growth is
bad for the poor; ie., a growth pattern in which the poor become worse off.
Strictly speaking, if the elasticity of the income share of the poor with respect to
growth is positive, growth is good for the poor. Why an elasticity of unity
should be the borderline between growth being good or bad for the poor is not
clear; indeed, it would seem arbitrary. The policy issue is not whether growth is
or it is not good for the poor (it is except in extraordinary circumstances), but what
policy measures can make it better for the poor. (2000, pp.15-16)

According to Lbker et al (2000), who reject the Dollar and Kraay findings thus:

(i) the empirical work is based on theoretically unsound equations; (ii) the
data are seriously flawed; and (iii) the policy variables are not defined
appropriately and are tested in an inconsistent manner. These problems imply that
the policy conclusions of the authors are unsafe.

the paper:

might have gone largely unnoticed had it not apparently been associated with the
well-publicised controversy over the World Development Report 2000. (2000,
p.1) 20

Recounted in Wade (2001), the controversy was an intellectual (ideological)

struggle within the Bank that, amongst other things, changed the shape of the January
2000 red-cover draft report. The red-cover draft, although beginning with an
acknowledgement of growth as the engine of poverty reduction, laid a stress on
empowerment and security that highlighted these two over the growth-oriented
section on opportunity which proved to be highly controversial. In essence, the
According to the Left Business Observer (November 2000), when questioned at a seminar, Dollar
and Kraay retreated on many of their claims. Kraay, was particularly forthright and honest about the
paper, noting at one point that they could just have well have titled it Growth Is Good for the Rich .

- 75 -
argument was over the question of whether effective safety nets needed to be created
before free-market reforms were pushed through. Opponents of adherents to the err
on the side of safety school (if one may call them that) were reformers who insisted
that safety nets had to be built simultaneously with market reforms, not made a
precondition for them. (Wade, 2001, p.132) It was this group that carried the day. 21

The struggle to push equity considerations well into second place behind growth came
at a time when work conducted by researchers at the Bank was beginning to disclose
that the hoped-for impact of Bank lending on poverty was not being achieved.
William Easterly (another critical employee who was encouraged to leave) 22 reported

the poor benefit less from expansions during a structural adjustment program
than in expansions without an adjustment program, while they are at the same time
less hurt by contractions. [in other words, the elasticity of poverty falls].

[I]t is disappointing that the poor do not share fully in growth in those cases where
there are recoveries that accompany adjustment lending. Since the Bank and the
Fund ultimately wish to restore growth in the economies to which they make
adjustment loans, it is worrisome that positive growth has less of a poverty-
reducing impact with high Bank-Fund involvement. (2001, pp.8-9)

Part of the explanation of why poverty does not fall with recovery, for at least some of
the countries that Easterly examined, lies in the fact that, as Ravallion has pointed out:

Inequality matters to the pace of poverty reduction that is achieved at any

given rate of growth. (2001, p.17)

In simple technical terms, the finding of the Easterly (2001) piece referred to above, is
that the growth elasticity of poverty in those countries he examined fell (or was low)
in countries with high Bank and Fund involvement. His analysis takes into account
the levels of inequality in the countries concerned. If an optimist is one who can find
consolation no matter what the circumstance, then Easterly must surely qualify.
Discussing the relationship among poverty, inequality and growth, he observes that:

Ten percentage points higher Gini will lower the growth elasticity of poverty by
0.6 percentage points. A not-often-noticed implication of this result is that the
poor will be hurt less by output contraction in a highly unequal economy than in a
relatively equal one, simply because the poor have a low share of output to begin
with. 23 (2001, p.8)

Both Joseph Stiglitz, chief economist at the Bank, and Ravi Kanbur, whom Stiglitz brought in to take
charge of drafting the red cover version of the report, left the Bank (Wade, 2001, pp.129 and 134).
In the Preface to the paperback edition of his Elusive Quest for Growth, Easterly says that when
readers asked if his statement in the original prologue to the effect that my employer the World
Bank encourages gadflies like me to exercise intellectual freedom, was really accurate, his reply
was Well, almost. It should be modified to the World Bank encourages gadflies like me to find
another job. (2002)
It is odd that he does not note that the capacity of the very poor (some of them hidden away in the
statistics) to absorb even a tiny negative shock may be so limited that any contraction of income could

- 76 -
Easterlys foray into the field marks the beginning of a steady stream of research,
none of which is so un-collegial as to dismiss the Dollar and Kraay (2000) results
outright, but all of which shows that averages conceal at least as much as they reveal.
Our first encounter with this body of work is with a paper by Ravallion (2001).

The glacier melts: Inequality is bad for the poor

Ravallions contribution to the debate has been substantial. The first piece of work of
his at which we look, which appeared not long after the Dollar and Kraay piece,
shows how it is possible for two apparently diametrically opposed views on the
impact of globalisation and growth on poverty, both to contain a morsel of truth. One
is the growth is good for the poor line treasured by conservatives (as though anyone
disagreed). The other (from Oxfam) is that widening disparities are putting a brake on
poverty reduction (Ravallion, 2001, p.2). 24 Table 2 below, reproduced from Table 1
in his paper, tells an interesting story about growth experiences in different
developing countries. There is, as Ravallion reminds us, a mechanical relationship
between poverty, inequality and growth. Given this, if inequality is rising, it will
accelerate the rate at which poverty increases when average household income is
falling. If average household income is rising and inequality is rising, the
combination of forces will slow down the rate of poverty reduction. Effects of the
opposite type occur when inequality is fallingif income is falling, the poverty
impact is cushioned. If incomes are rising, poverty reduction is given a boost.

Table 2 Poverty impacts and the non-correlation of inequality and growth

What is happening to average household
income between the surveys?
Falling Rising
Rising (16% of spells) (30% of spells)
What is happening to Poverty is rising at Poverty is falling at
inequality between the a median rate of a median rate of
surveys? 14.3% per year 1.3% per year
Falling (26% of spells) (27% of spells)
Poverty is rising at Poverty is falling at
a median rate of a median rate of
1.7% per year 9.6% per year
Note: Based on 117 spells between two household surveys covering 47 developing
countries in the 1980s and 90s. Poverty is measured by the percentage of the
population living below $1/day at 1993 purchasing power parity. Inequality is
measured by the Gini index.
Source: Ravallion, 2001, Table 1, p.18

have catastrophic consequences (a reality of which someone of his vast experience cannot be unaware).
Ravallion (2005a, pp.11-12) makes a similar observation with respect to low inequality.
An article waxing lyrical on the Dollar and Kraay findings that Ravallion cites appeared in The
Economist, May 27, 2000, p.94. The response in the form of a letter from Oxfam to the editor was in
The Economist, June 20, 2000, p.6. The publications eagerness to embrace the Dollar and Kraay
position confirms (if confirmation were necessary) its conservative ideological stance.

- 77 -
Ravallion warns against simply assuming that falling inequality guarantees large falls
in poverty. Whether or not such falls are achieved, depends critically on exactly how
the reduction in inequality is achieved. As far as the magnitude of the falls in
inequality associated with falling poverty are concerned, Ravallion observed that:

The changes in a summary measure such as the Gini index often seem small in
magnitude. However, this can be deceptive, since even small changes in overall
distribution can matter greatly to how much the poor share in growth. (2001,

Initial conditions are another important determinant of the rate of poverty reduction.
Drawing once again on the mechanical nature of the relationship between inequality,
poverty and growth, Ravallion points out that when inequality is low, the share of
additional income resulting from growth that goes to the poor will tend to be greater
than it would be when inequality is high. If the concern is with poverty reduction,
mere growth rates in an economy do not convey much information, for as Ravallion

an important determinant of the rate of poverty reduction is the distribution-

corrected rate of growth in average income, given by a measure of initial inequality
(100 minus the measure of inequality) times the rate of growth. Indeed, the
distribution-corrected growth rate knocks out the simple rate of growth when both
are used in a regression for the rate of poverty reduction between surveys across
countries and time It is not the rate of growth that matters, but the distribution-
corrected rate of growth. (2001, pp.18-19)

Next on the stage are Cord and his co-authors (2003), whose paper begins with a
critique of the concept of pro-poor which reads thus:

Development practitioners and policy makers are fond of the term pro-poor
growth because, rather like Humpty Dumpty in Alices Wonderland, to each of
them it means exactly what [they] want it to mean. (Cord et al, 2003, p.1)

Staying pragmatically in the furrow later to be ploughed by Ravallion in his (2004)

response to the UNDPs International Poverty Centre One Pager on pro-poor
growth (Zepeda, 2004), 25 Cord et al conclude that further debate over the meaning of
pro-poor would probably be counter-productive to poverty reduction goals. On
these grounds they work with the weaker formulation of pro-poor growth: growth
that benefits the poor (p.4). Doing so does not prevent them, however, from
puncturing any residual complacency among the growth fundamentalists about the
relationship between growth and poverty reduction. Like Ravallion, they challenge
the notion that growth is enough, and produce as well, a two-by-two typology to show
the relationship between inequality and growth. The four quadrants of their Table 3
(2003, p.19) are labelled respectively, starting in the bottom right-hand (the south-east
quadrant) and proceeding clockwise, Pro-poor biased growth (positive growth,
inequality falls); Pro-poor recession (negative growth, inequality falls), Anti-poor

In a two-page response to the UNDP, Ravallion (nd), says that: The real issue is not whether
growth is pro-poor but how pro-poor it is. (His emphasis)

- 78 -
recession (negative growth, inequality rises); and finally in north-east quadrant,
where one finds two outcomes Broadly shared growth and Not pro-poor by any
definition (positive growth, inequality rises). All told, in the north-east quadrant
there are 15 countries in the broadly shared growth category, and nine in the not
pro-poor by any definition. If nothing else, a finding of this sort should serve to
staunch the flow of fundamentalist platitudes about growth. In the south-east
quadrant (pro-poor biased growth) there were 17 countries. Cord et al report that
despite the presence, among the shared growth group, of countries where income
growth was so low as to have had little impact on poverty:

the median growth rate of the incomes of the poor in countries in the shared
growth category (NE quadrant) exceeds that for countries in this pro-poor biased
growth category. (2003, p.13)

Inevitably, this poses the question of whether or not there is a growth-inequality trade-
off? 26 The question cannot be ducked, but before facing it, we look at a more
emphatic intervention on the inequality question by Ravallion (2005a), one whose
title was used as the heading for this sub-section of the paper, Inequality is Bad for
the Poor.

Summing up the argument thus far, it may be said that the UNDP and those of like
mind opted for what Cord et al describe as Pro-poor biased growth from first
principles. That does not mean denying that income increases for the poor of the type
garnered under a positive growth, rising inequality scenario are worth striving for.
The Bank, by contrast, starting from the weaker definition of pro-poor growth, is
driven to a confrontation with the problem of inequality by the growing evidence of
its ill-effects. Ravallions 2005a effort, one of the latest in a series of papers on the
matter, is concerned with the important question of the:

number of ways in which the extent of inequality in a society, and how it evolves
over time, influences the extent of poverty today and the prospects for rapid poverty
reduction in the future. (2005a, p.2)

Prominent among the lessons learned appears to be the likelihood that distribution
neutral growth is likely to leave many of the poor behind, the antecedent conditions
effectively acting as a barrier in high inequality countries. As yet but imperfectly
understood, the specific factors that constrain some poor people from participating in
the benefits of a growing economy require the attention of researchers so that
policies for rapid poverty reduction that promote economic growth can be devised
(2005a, pp.28-29).

By 2004, redistribution was so firmly on the agenda that Franois Bourguignon,

Senior Vice President and Chief Economist of the World Bank, could be found
declaring that to formulate the question of whether the main focus of development
strategies should be placed on growth, or poverty, and/or on inequality was to pose a
false dilemma. For him, the answer to the development conundrum takes the form
of two statements:

The ubiquitous Ravallion has not shied away from this question eitherwe will look briefly below at
a recent paper of his (2005b) on the topic.

- 79 -
First, the rapid elimination of absolute poverty, under all forms, is a meaningful
goal for development. Second, to achieve the goal of rapidly reducing absolute
poverty requires strong, country-specific combinations of growth and distribution
policies. (Bourguignon, 2004, p.1) 27

Exploration of the poverty-growth-inequality triangle by Bourguignon (2004) 28

leads him to point out that change in poverty (as noted earlier) is a function of growth
in mean incomes, changes in the distribution of income and the initial distribution (he
expresses them in the form of an identity (2004, p.5). Digging into the relationships
among them, he shows that though both growth and inequality changes are important
for changes in poverty, their relative impacts depend on their initial levels. A series of
simulation exercises, followed by an empirical examination yields the following

That inequality matters as much as growth for poverty reduction, with the
caveat that inequality reduction cannot be expected to proceed indefinitely (for
political economy reasons), so that in the long run, growth is main factor for
poverty reduction

That country specificity matters a great deal it is likely that changing the
distribution is probably more important for middle-income and inegalitarian
countries, while growth is probably more important, in relative terms, for low-
income and egalitarian countries. 29 (Bourguignon, 2004, p.12)

South Africa, it need hardly be said, is a middle-income, inegalitarian country. To

maximise the impact of poverty reduction policies, policies to reduce inequality are,
therefore, probably necessary. Recall, at this point, that during the discussion of the
results in Figure 1, where the notion of social justice was raised, it was noted that
reasons for redistribution other than mere norms would be adduced. Here is the
reasonif a country like South Africa has as one its values, the reduction of poverty,
then it is likely that the achievement of that goal would be hastened by policies that
seek to reduce inequality as well.

Note that it is only likely, not inevitable. Two indeterminacies stand in the way of
any inevitability: the effects of growth on distribution, and the effects of inequality on
growth. Bourguignons analysis of the first of these two relationships concludes that
after the early certainties of the Kuznets inverted U-curve (inequality increases over
time then begins to decrease) although there are relationships, there is too much
country specificity in the way growth affects distribution for any generalization to be
possible. (2004, p.13).

In other words, there is now a convergence around the position articulated by Cord et al (2003, p.1)
to the effect that:
we conclude that economic growth and growth-oriented policies, while necessary for sustained
poverty reduction, do not guarantee that it will occur at the country level. For pro-poor growth to
take place policies must be both pro-growth and pro-poor.
The Bourguignon paper in question (2004) does not contain the usual disclaimer saying that it
represents the writers, and not (necessarily) the Banks position.
A large simulation exercise by Dagdeviren et al (2000, pp.37ff) arrives at a similar conclusion.

- 80 -
As far as the effects of inequality on growth are concerned, Bourguignon reports that
the dominant view today is that inequality is not a final outcome of growth but plays
a central role in determining the rate and pattern of growth. (2004, p.14). He reports
on the tests of a number of hypotheses that might explain why progressive
redistribution may be growth enhancing , and also when it might not be. These
include the effects of credit market imperfections (redistribution can increase
efficiency investment and growth); on the relationship between democracy;
redistribution and growth (a democracy may achieve redistribution at the cost of
capital accumulation); and on redistribution and social conflict (too much inequality
can lead to violent redistributionindividual or organised). Aggregate cross-country
analysis yields somewhat inconclusive results. Of the two ways he suggests for
getting around this problem, one is probably too data-hungry to be practicable. The
other entails checking whether the micro-economic mechanisms behind the
hypothesised explanations are verified, and then hazarding a rough estimate of the
likely impact on growth. Once again, generalisations are not in order.

Whether or not there is a tradeoff between poverty and inequality is a question that
has attracted considerable attention. 30 After examining the experience of 70
developing countries in the 1990s, Ravallion (2005b) concludes that the results differ
depending on whether the concern is with relative inequality and absolute poverty (no
systematic relationship), or with absolute inequality, where the intention is to reduce
to lower the absolute gap between he rich and the poor. In such cases, he observes,
there will tend to be a trade-off; those harbouring aspirations of lowering absolute
inequality must in general be willing to see lower absolute levels of living for poor
people (pp.1 and 15).

On balance, the literature favours a reduction of inequality, if the circumstances are

right. The question is: redistribution of what? Klasen (2003) notes that:

[a]part from supporting a traditional emphasis on furthering economic growth,

[the findings he reports on] suggest that pro-poor growth policy should have a
significant emphasis on reducing inequalities, particularly asset and gender
inequalities, in order to benefit from the triple pay-off to redistribution [reducing
poverty immediately, increasing growth and enhancing the poverty impact of such

He continues, however, with the observation that:

While this has been recognized in theory [in particular, by the World Bank]
there is no clear equality-enhancing policy agenda spelt out to achieve the
necessary inequality reduction, particularly for high inequality countries where
such an agenda would be most urgent in order to further pro poor growth.
(Klasen, 2003, p.8)

On the question of trade-offs, Lopez (2004), asks whether a pro-growth strategy [is] always the best
pro-poor strategy? He finds that all of the pro-growth policies he considers lead to poverty reduction
in the long run. Some of them, however, lead to higher inequality. Higher poverty levels in the short
run are also a possibility. This leads him to suggest that if a pro-growth policy forms the core of a
poverty reduction strategy, complementary measures to counter short run increases in poverty should
be implemented.

- 81 -
Equality-enhancing policies are what is soughtwhere are they to be found?

What is to be redistributed?

For proponents of the argument that under certain circumstances, the easiest and best
way to lower inequality is to transfer income from those that have to those that do not,
an initial encounter with the empirical evidence in World Bank research papers comes
as something of a dampener. Calls for the redistribution of income, if they do not
elicit the response populist, then at least call forth a barrage of cautions from Bank
economists and those influenced by them. 31 Caution is not misplaced, but it can be
taken to such lengths that opportunities for highly beneficial transfers are overlooked.
This, it will be argued, is what is happening in South Africa. Empirical foundations
of an argument against the redistribution of income as a means of addressing
inequality are present in the observation by Ravallion that:

The above results are unrevealing about what specific aspects of inequality
matter. The theoretical arguments based on credit-market failures point to the
importance of asset inequality, not income inequality per se, and there is evidence
of strongly adverse effects of asset inequality in growth (2001, p.23)

His primer on pro-poor growth, commenting on the need for more research into the
relevant sources of inequality, mentions access to both private (human and
physical) assets and public goods, to infrastructure and social services, but not of
income (Ravallion, 2004, p.18). A comprehensive list of the policy interventions,
most involving redistribution of one sort or another is given in Ravallion 2005a,
pp.28-30. The list begins with a warning that if poverty reduction is the over-riding
goal, then no fiddling with policies that reduce inequality at the expense of poverty
reduction should be tolerated (p.28). His suggestions (presumably not intended to be
exhaustive) can be boiled down to include:

community-driven development projects whose benefits are not captured by

local elites
security of access to land in rural economies, through tenancy reform and
titling programs
sound public investments in rural infrastructure,
better policies for delivering quality health and education services to poor
people (more responsive to the needs of the poor)
better instruments for credit and insurance, both for smoothing consumption
and underpinning otherwise risky growth-promoting strategies

Bill Easterly, former Bank economist, and author of the magisterial The Elusive Quest for Growth,
displays the scars of an economist burned early on his career by an idealist support for an income
distribution programme gone horribly wrong. His account of the Mexican experience in the early
1970s we were venturing from an area where we still understood littlethe determinants of
growthinto one where we knew almost nothinghow to redistribute income toward the poor without
harming growth. (Since then, the cycle swung back to growth, but now we once again are shifting
toward redistribution, still lacking much knowledge about how to achieve it). (2002, p.224). Apart
from the odd scattered reference to populist governments seeking to redistribute to their followers
(p.267) [on which we will have more to say below], there is very little more in the book about
redistribution of income. The topic does not merit an entry in the books index.

- 82 -
policies that allow key product and factor markets (for land, labor and credit)
to work better from the point of view of poor people
removing biases against the poor in taxation, spending and regulatory
(including migration) policies

The many poor-area programs and attempts to set up growth poles have not
answered all of the questions of how do address marked regional disparities in living
standards. 32 Are infrastructure investments in poor areas (often with poor natural
resources) effective in reducing poverty? Ravallion asks. Does it make more sense to
move jobs to people, or people to jobs? Is there a trade-off between achieving greater
regional equity such as by focusing on areas with high poverty rates but low
poverty densities and poverty reduction in the aggregate? (Ravallion, 2005, p.29)
Questions such as these are of particular relevance in South Africa.

An absence of overlap between policy prescriptions from economists of more or less

the same stripe would be cause for concern. It is reassuring, therefore, to note the
similarities between the suggestions above from Ravallion (2005a) and those from
Bourguignon (2004) below. In discussing the sort of redistributive steps that could be
taken to counter the adverse effects of initial inequalities 33 , Bourguignon, however,
goes well beyond Ravallion, plunging straight into the controversy over whether it is
wealth or income that is to be redistributed. The route into the argument is via a
rehearsal of nave views which suggest that a redistribution of wealth (through
various pro-poor policies) would set in train a virtuous circle of sustainable pro-poor
growth (2004, pp.19ff).

Reproducing the standard argument that since the distortionary effects of income
transfers that are not lump-sum can reduce growth, redistribution should be of wealth
rather than of income if it is to be growth and efficiency enhancing (p.20),
Bourguignon, having set up the position, meticulously takes it apart, by working
through the practical difficulties of redistributing wealth. In the case of land
redistribution, for example, he shows that the subsidised re-allocations of land to the
poor invariably involve an income transfer (via taxes levied to finance subsidies).
Pointing to the paradoxical inclusion of non-lump sum income transfers in so-called
wealth transfers, he notes that the positive wealth (redistribution) effect may outweigh
the negative income effect (p.21).

That discussion provides a springboard for a leap into an engagement with the
conventional wisdom on pure income transfers, namely, that they are bad (p.21). The
received wisdom emphasises the adverse effects on both recipients and those whose
taxes fund such transfers (these arguments form the backbone of the South African
governments objections to the extension of the social grant system to cover those
presently without any social protection). Noting that this is now being challenged, he

Klasen warns that: cross-country analyses suggest that [labour productivity improvements]
improvements in agriculture may have a much smaller or even negligible effect in high inequality
countries, further emphasizing the need to tackle distributional issues. (2003, p.10)
Crime, with its massive social and private costs, is high on the list of social ills whose impact on the
poor is much greater than it is on the well-to-do (Bourguignon, 2004, p.17). A redistribution of
resources towards the combating of crime (obviously highly topical in South Africa) could benefit the
poor disproportionately. The well-off would gain as well, probably more by the reduction of the
private costs of purchasing security, an option that is not available to poor.

- 83 -
makes a claim which is sweet music to those advocating the extension of social grant
systems as a means of addressing poverty. His argument runs thus:

To the extent that beneficiaries may improve their standard of living, income
transfers may contribute to the accumulation of human capital among them. Their
nutrition could improve, for example. Under these conditions, apparently pure
income transfers in effect lead to some particular wealth accumulation among the

Another channel through which income transfers may affect the assets owned by
poor people is through insurance. Indeed, many economists now consider that in
the presence of a high and possibly increasing macro-economic volatility, targeted
transfers can be useful instruments for social protection. They may also
contribute to pro-poor growth (i.e. growth that reduces poverty) by avoiding dis-
savings, for instance by taking children out of school or by helping credit-
constrained poor people to be productive workers or take up productive
opportunities for self-employment. (Bourguignon, 2004, p.21)

In support of his claims about the efficacy of income transfers as a vehicle for
promoting pro-poor growth (and inequality reduction) he cites the experience of
smart transfers like Mexicos Progresa/Oportunidades and Brazils Bolsa
Escola/Bolsa Familia (p.22). 34 By contributing to directly to human capital
formation, income transfers become respectable, so that:

possible adverse consequences of growth on the distribution of income may be

corrected by redistribution at low cost, and possibly even at a negative cost.

So much for income transfers. Bourguignons treatment of asset transfers is sensitive

to the political economy of such endeavours. In place of simplistic claims about how,
for example, increased levels of education benefit the poor, so that redistribution of
this asset in their favour would be pro-poor, Bourguignon looks to the possibility of
successful transfers to the poor by considering levels of political participation
(including the determinants of the capacity to mobilise) necessary to secure transfers
away from the elite (2004, p.23). The general lesson drawn from this is that asset
transfers in general involve an economic payoff to the incumbent elite (gains from
economic growth spurred by the increased productivity) accompanied by a potential
loss of political power as the newly-empowered acquire the economic capacity to
assert political claims (which the very poor generally are unable to do). With the
introduction of class differences (Bourguignon prefers to refer to social stratification)
and institutions (p.23), 35 the analysis moves well beyond the often trite
generalisations characteristic of much of the literature (pro-poor propaganda as
motherhood of which everyone approves).

De Oliveiras (2006) analysis of the Bolsa Famlia is much less sanguine. While recognizing its
impact on poverty, he sees it as a vote-buying device in a country where progressive politics has all but
disappeared, and one moreover, that traps the poor in their poverty.
In a remark which puts one in mind of the cultivation of a black middle class in South Africa,
Bourguignon comments on the strategic promotion of such a group by the dominant elite (2004, p.23).

- 84 -
Having started out with a laundry list of redistributables, we end up with the
suggestion that:

income transfers are desirable because they can stimulate human and possibly
physical capital formation, and would hence be growth enhancing, and
to gauge the feasibility of asset transfers, a political economy approach is

The scope for reducing inequality in South Africa

Of the many ways to approach this question, the method selected begins with another
laundry list, this time, one compiled by Klasen (2003). Klasen gathered together the
research findings, agreed policy implications, and areas of remaining debate of a wide
range of policies to promote pro-poor growth, and presented these in a table. In
another table, he did a similar thing with the process issues in promoting pro-poor
growth (Klasens, 2003, Tables 1 and 2, pp.27-29). These are reproduced below as
Tables 3 and 4, but with a few differences: one of them is that instead of the areas of
debate column being populated by findings from the international arena, the debates
considered are specifically South African. As will be seen, this means that in many
cases, little or no debate is required. Another is that a few more categories, two of
them not usually conceived of as policy issues, have been added to the table. One of
these concerns the ideological climate(s) fostered by the dominant elites (the ANC in
politics, and the capitalist class in the economy). The second is that of the quality of
analysis and the statistical data on which it is based. The third is the area of
employment policy. Working our way through the items in the table isolates a set of
topics to be dealt with in greater detail. (Some of) these will be tackled as they arise.

Before commencing, a few observations about Klasens list are in order. In the first
place, it is the most complete catalogue I have stumbled across in my meander
through the literature on pro-poor growth, of policies to foster (or hinder, if not
implemented) such growth. In the second, it is necessary to bear in mind, as one
scans the list that Klasens definition of pro-poor requires that such growth be
progressive in the sense in which the UNDP uses the term, i.e., that inequality
fallsnone of this wishy-washy stuff in which any growth that sees the incomes of
the poor rising is dubbed pro-poor. In the third, while he apparently does not
disapprove of the extension of social grant systems, even though he is wary of the
poverty trap created by means-tested benefits (2003, p.13), he does not favour the use
of static measures (like a minimum wage) to alter the functional distribution of
income unless such measures are also compatible with (or better) foster growth
(2003, p.11).

- 85 -
Table 3 Policies to promote pro-poor growth: Research findings, consensus policies, and remaining debates
Policy issue Research Finding Agreed policy implication Areas of debate
Macroeconomic Macroeconomic stability critical necessary Monetary and exchange rate policy should aim Some disagreement about severity of policy of
Stability (see also (though not sufficient) condition for pro poor for low inflation and competitive exchange rates; inflation targeting (3-6 per cent)
individual areas of growth; poor hurt particularly by high inflation fiscal policy should aim for low budget deficits
macro policy) and high macro volatility
Monetary and Overvalued exchange rates and high black A competitive and possibly undervalued Occasional disagreements between proponents
Exchange Rate market premia hurt economic growth and tend to exchange rate a critical ingredient to ensure of different values of the Rand
Policy be anti-poor macro stability; government intervention
necessary to manage capital inflows
Fiscal Stance Large budget deficits hurt growth and are Governments should aim for moderate budget Tax cuts for the well-off have gone too far?
unsustainable. Rapid expenditure cuts can often deficits through broadening of the tax base and, Complaints from progressive camp about
undermine delivery and quality of critical if necessary, a refocusing of expenditures (esp. developing country running a surplus calls for
services (e.g. health and education) and hurt the cuts in subsidies to state owned enterprises and deficit spending to combat poverty and
poor. unproductive sectors). During crises not feasible unemployment
or desirable to cut expenditures fast.
Privatization Loss-making state-owned enterprises undermine Reform of loss making state-owned enterprises Enthusiasm for further privatisations muted.
fiscal stability, with negative implications for the and parastatals critical. Privatization processes Reforms in hand to make major parastatals (e.g.,
poor. Some privatizations have been captured must be transparent and competitive. rail services) more efficient
by local elites and have not led to better services
for poor.
Financial Sector Severe financial repression hurts savings and Capital account and financial sector reform Moves to control worst excesses of kerbside
promotes capital flight. Poorly sequenced should be phased slowly, be implemented only if lenders. Private sector expansion of banking
financial sector reforms can be counter- macro stability has been achieved, and be services to poor
productive and destabilizing accompanied by tight regulation, competition
policies, and policies to improve access of the
Trade Policy Anti-export bias hurts growth and the poor; Focus on removal of anti-export bias competitive Extensive liberalisation undertaken, but more is
import liberalization can be anti-poor and not exchange rate, duty draw-back schemes, etc; required, according to Edwards and Lawrence
sufficient to generate supply response. provision of infrastructure to assist exports, esp. (2006). Opposed by trade-union movement
Diversification essential for long-term growth for export diversification

- 86 -
Policy issue Research Finding Agreed policy implication Areas of debate
Agriculture Raising agricultural productivity critical for pro Renewed emphasis on agricultural research and Major problems with productivity of
poor growth. Removal of price distortions extension, rural infrastructure, and competitive redistributed farm land. Extension services,
necessary but not sufficient in the presence of marketing and input supplies. Open access to infrastructure and inputs all compromised
other market failures. Protection and subsidies in OECD markets and removal of OECD subsidies
North hurt poor in South critical
Industrial Policy Removal of distortions necessary but not Focus on providing infrastructure and services to Industrial policy about to be launched by
sufficient for vibrant industrial sector, esp. small industrial sector Department of Trade and Industry. Limits of
and medium enterprises manufacturing sector as engine of growth
Human Capital Lack of human capital of the poor hurts growth Increased investment in education and health, Health and education expenditure already high
and poverty reduction. Education and health particularly basic education and primary health in absolute and relative terms. Problems are of
services have suffered greatly under economic care; greater focus on quality; reallocation of quality of service
crises and SAPs. Credit constraints and high public spending to the poor, lowering costs of
costs for health significant deterrent for the poor primary health care and education through
greater subsidies and use subsidized community
Asset Inequality Asset inequality (in particular land) reduces On land inequality: Removal of subsidies to Progress in land redistribution painfully slow.
economic growth and poverty impact of growth large landowners land taxes to increase land for Primary focus of official policy on asset
sale; land redistribution necessary. Other asset inequality is on controversial policy of broad-
inequalities: microcredit and subsidies for based black economic empowerment
infrastructure extensions for the poor (e.g.
electricity hook-ups)
Income Inequality High income inequality associated with higher Safety nets, social funds, and some targeted cash Substantial social protection network, but with
poverty and lower poverty impact of growth; and in-kind transfers to the poor huge hole through which those of working age
high initial income inequality may reduce fall. Relative silence on question of how to
subsequent growth tackle extreme inequality

- 87 -
Policy issue Research Finding Agreed policy implication Areas of debate
Gender Inequality Gender inequality reduced growth and makes More supply of education for girls plus targeted Nominal commitment to empowerment of
growth less pro poor subsidies to boost enrollments; removal of women masks serious structural
restrictions on female control of other assets; disempowerment
political empowerment of women
Regional Regional inequality can sharply reduce impact of Targeting of state transfer programs and safety Capacity of regional jurisdictions to absorb
Inequality growth on poverty; possibility of regional nets on regions with high poverty concentration; development funds limits local development
poverty traps Focus on improving infrastructure; Regional initiatives
inequality to be considered in programs of
decentralization and fiscal equalization
Population Policy High fertility among the poor a constraint to pro Emphasis on female education and employment Tolerably well-developed. Disagreement over
poor growth. Inequality reduction often a result as well as access to reproductive health services impact of AIDS on population growth
of fertility decline among the poor
Security Physical and social security essential for pro Safety nets and greater physical security Threats to physical and mental wellbeing of
poor growth essential measures promote pro poor growth crime

Ideology Conservative ideology fosters anti-welfare Competition between Treasury (promotion of Extent to which current anti-poverty policy can
stance. Commitment to notion of developmental vigorous (Victorian) values of self-reliance) and succeed. Disagreement over future shape of
state Social Development over social justice welfare policy. Delusion over capacity to
deliver development pushes achievable
policies aside
Research Poverty and inequality data not good enough to Disagreement over impact of policy leads to Appropriate forms of measuring instrument for
support robust findings on either. confusion about policy implications monitoring and evaluating policy not settled
Unemployment data questioned
Employment Predicated on private sector employment Dismantling of barriers to entry, formal training Efficacy of EPWP as residual employer.
policy creation, skills upgrading, small-business structures, learnerships. Effectiveness of training institutions (SETAs).
support Employment subsidies currently under
consideration. Two-tier labour markets to
address unemployment, especially among

- 88 -
Table 4 Process issues in promoting pro-poor growth: Research findings, consensus policies, and remaining debates
Policy issue Research Finding Agreed policy implication Areas of debate
Governance Poor governance, corruption, political instability Reducing incentives and possibilities for Mixture of official denial and grudging
and civil strife a major deterrent to investment, corruption by simplifying rules and regulations acknowledgement of significance of corruption,
growth, and poverty reduction. Poor suffer more that invite rent-seeking behavior; merit-based governance weaknesses, crime (social violence)
under poor governance pay and recruitment; increase public stifles debate
accountability through greater transparency, cf. also the debate on the developmental state
better institutional oversight of governments
(parliaments, independent boards) and
decentralization. Donor support for conflict
prevention resolution, and post-conflict
reconstruction critical
Private Sector Indigenous private sector critical for State assistance with capacity-building, finance Powerful private sector capable of stifling
employment growth and dynamic economy (esp. microfinance), dialogue between state and initiatives which it sees as against its interests
domestic private sector
Political Economy Domestic political economy crucial for success. Dialogue to replace donor conditionality. Internal variant in terms of which powerful
of Reform Pro-poor coalitions necessary to implement Empowerment of poor and local analytical and social partners (business, organised labour,
package research capacity critical for implementation state) dominate policy formation. Exemplified
by weakness of Development Chamber in
Donor Policies Donors can assist with pro poor growth when aid Aid should be focused on poorest countries that Not applicable
and advice is focused on poorest countries and promote pro poor growth, should flow through
those with highest poverty impact of policies budget and be accounted for using national
processes, and observe country leadership

- 89 -
And finally, it is important to bear in mind the comments Klasen makes about
selection among the policies in his list:

It would be ideal, he says, for governments to be able to focus on policies that

have the largest marginal effect on Pro Poor Growth. Some policies have a large
effect on growth and may not be particularly pro-poor; others do not have such a
large effect on growth but are very pro-poor. The best policies (win-win) are
those that have a large effect on both. But not enough is knows (sic) which
policies fall into what category. In addition, at times it is critical to implement a
package of policies as individual policies alone may not be enough. Here more
work is clearly needed as the current wish-list for pro poor growth is very long and
some kind of sequencing, prioritizing, and packaging is critical. (Klasen, 2003,

Redistributive growth: Moments of inertia

Filling in the cells in the last column of Tables 3 and 4 allows us to reflect both on
those policies that are not obviously pro-poor (if anything, they are pro-growth), as
well as those that, if successfully implemented, would be pro-poor (in the
redistributive sense). The former group includes the so-called macro fundamentals
which government has striven so hard to get right. Having more or less done that,
the reward has been the longest unbroken period of steady if unspectacular growth
since World War II (Frankel et al, 2006). 36 There are quibbles over the severity of
anti-inflation policy, major complaints about the (primary) surplus, the argument
being advanced that a larger deficit 37 could be used to finance anti-poverty policy,
especially social grants.

Trade policy is an occasional battleground between losers (unionised workers) and

policymakers (backed by academics) who claim that there have been significant
benefits from increasing openness to trade, and that there needs to be more of it
(Edwards and Lawrence, 2006). The links from here to redistribution in favour of the
poor are fairly tenuous. Increasing employment of unskilled and semi-skilled workers

The dangers of fetishising macro-fundamentals were addressed in one of the first of the Harvard
project papers (Hausmann, 2006). In response to the claim by Summers that: the rate at which
countries grow is substantially determined by three things: their ability to integrate with the global
economy through trade and investment; their capacity to maintain sustainable government finances,
and sound money; and their ability to put in place an institutional environment in which contracts
can be enforced and property rights can be established. I would challenge anyone to identify a country
that has done all three things and has not grown at a substantial rate. Hausmann said that he could
think of many, naming South Africa, Mexico, El Salvador and Morocco (2006, p.1, emphasis in
original). There are also dangers in the premature celebration of the growth acceleration that South
Africa has enjoyed in recent years. Hausmann also points out that not all growth accelerations are
sustained. About 40 percent revert back to either slow or negative growth after the 8-year period.
(2006, p.7)
In the 2006 Budget Speech, the Minister of Finance pointed out that the actual overall deficit in
2005/2006 was 0.5 per cent of GDP (15th February 2006, p.8), rather than the projected deficit 3.1 per
cent. In the 2007 Budget Speech, the Minister made it known that he was budgeting for a fiscal surplus
of 0.6 per cent of GDP in the year 2007/2008 (21st February 2007, p.11). A moderate deficit was
forecast for the end of the projection period.

- 90 -
would be one possible channelthis appears not to have happened. 38 Cheaper
imports would benefit consumers in general. The extent to which the poor benefit
would depend on the import intensity of their consumption. Odds on the probability
that this is not a particularly important source of pro-poor growth are quite high.

Industrial policy (which is closely connected to trade policy) has been on something
of a rollercoaster. The two papers suggesting ways in which the manufacturing sector
could be re-invigorated produced by the Harvard team (Rodrik, 2006; and Aghion et
al, 2006), seem to contradict each other. The latter argues that competition policy
aimed at reducing excess mark-ups (and, presumably, profits?) is necessary, while the
former claims that manufacturing has shrunk, relatively, because it has suffered a
crisis of profitability. In discussing the relevance of the composition of growth for
poverty alleviation, Loayza and Raddatz (2006) make the point that the largest
contributions are to be expected from the labour-intensive sectors; construction,
manufacturing and agriculture. Employment in the latter two, unfortunately, appears
to be static or falling (one cannot be certain with agriculture). Attempts at stimulating
certain industries, call-centres, for example, have been successful, but the numbers
employed, relative to the number wanting work, are trivial.

As far as agriculture is concerned, South Africa differs from most developing

countries. There are similarities in some of the ex-colonies, but the structures in the
sector bequeathed by apartheid capital accumulation pose a clutch of problems, the
solution of which will require a much greater effort than appears to have been made
thus far. These problems exist on both the supply and the demand side for all
participants in the agricultural and related sectors.

Assisting the poor to gain ownership of land would be an excellent pro-poor

investment. Not all of those who want to gain access to land wish to do so because of
its usefulness as an income-generating asset. For many, its symbolic value is a major
attraction. A huge majority (81 per cent) of Africans polled in the 2003 South African
Social Attitude Survey (SASAS) thought that government should redistribute land to
black South Africansonly 17 per cent of whites shared this view (Roberts, 2006,

By contrast, a survey conducted recently in the 21 nodes that together make up the
Integrated Sustainable Rural Development Programme (ISRDP) and Urban Renewal
Programme (URP), 13 of which fall under the ISRDP and eight of which fall under
the URP, found that as far as peoples economic expectations were concerned, 60 per
cent of those in the ISRDP nodes, and 73 per cent in the URP wanted a job, as
opposed to 22 per cent in the rural nodes, and ten per cent in the urban, who wanted
land for farming (Everatt et al, 2006, p.10).

On the question of who it would like to see returning to the land, it has become
increasingly clear that government favours the emergence of a kulak class, rather than
the mass of impoverished [b]ricoleurs people under multiple-use value-oriented
livelihood regimes seeking subsistence, with limited connections to the formal

Berry (2006, Section 7.1) expresses doubts as to the ability of export expansion to provide an easy
answer to its employment challenge. In his view, there is also little chance that the benefits of an
export boom will trickle down through the population.

- 91 -
economy, they are are hardly seen in government circles and not recognised in
national land reform policy (Gran, 2006, p.20). He argues that:

if land is reorganised from within existing rural communities and

infrastructure is improved so that livelihoods gradually improve for millions of
rural people that would itself be a major improvement, beneficial perhaps for
urban and modern developments as well. (2006, p.2)

Instead, it appears that if all of the necessary infrastructure, physical and institutional
required to make it possible for people with limited knowledge of the complexities of
commercial farming to farm successfully, is to be provided, government would prefer
to direct that support to emerging and commercial farmers and not to ordinary
livelihood-seeking people in general (Gran, 2006, p.2, citing Cousins, 2003).
Commenting on the programme of land reform, Cousins argues that it is in deep
trouble. With a target of 30 per cent of land in black hands by 2015, just four per
cent of white-owned land had been redistributed by 2006. 39

For a variety of reasons, mega-projects to tackle poverty and deprivation seem

unlikely to succeed in non-urban 40 areas of South Africa in the near future. Writing
about conditions in the ISRDP (and URP) nodes where employment is mainly
unskilled, and unemployment is very high, Everatt et al observe that:

Ironically, in such an economic context, the project-based approach to

developmental social welfare offered by DSD may be appropriate. Where large-
scale economic interventions have little prospect of success, there is still space for
small-scale local projects to help small groups of people, so long as expectations
of scale, impact and sustainability are very firmly rooted in local realities. Project-
based development will never transform these huge and poor nodes into
economically thriving and socially cohesive areas on their own, but they have an
important developmental role to play at the micro level. (2006, p10)

With very little stretching, this approach will probably, in the absence of a change in
the direction of agricultural policy, have to serve the poor rural populations who do
not stand to benefit from these policies. Redistribution of land will take place, but it
does not look, at this point, as though the effects of this redistribution will be spread
very wide. One of the more promising avenues for pro-poor investment, is, if not
closed, is certainly choked with the tangled weeds of conflicting interests. A detailed
investigation of projects underway (of which there must be many hundreds, if not

These findings were reported in an article by Chris McGreal in The Guardian Unlimited, Tuesday
January 30, 2007 headed White farmers, black land hunger: sparks fly in the tinder box
The use of an awkward phrase like non-urban is necessitated by the difficulties of defining rural
adequately, and the attendant political problem of ensuring that rural areas are not neglected in the
allocation of resources and provision of services. Prior to 2004, the term non-urban instead of rural
was used in official statistics in South Africa. It appears, for example, in the 1995 October Household
Survey (P0317, 27 November 1996). Its last use in surveys appears to have been in about 1993 (see for
example, the September 2003 Labour Force Survey (LFS) (P0210, 25 March 2004). Surveys up until
that date used 1996 population census demarcations. Both non-urban nor rural disappear from the
September 2004 LFS (P0210, 31 March 2005), the reason being that a switchover to the 2001 census
geographical structure of six Metropolitan (Category A) areas and 47 District Councils (comprising
231 Category B municipalities and 25 district management areas), had taken place. The issue is so
complex that it merited a long reportsee Statistics South Africa, 2003, p.2.

- 92 -
thousands), with the intention of discovering ways in which those responsible can be
assisted, is probably the most useful contribution that can be made at this point.

Human capital formation: The domino transition

Nowhere is the damage done by apartheid more apparent than in the school system. It
was hardly necessary, therefore, to suggest to South Africas first democratic
government that investment in the schooling system should be pro-poorthat precept
has guided the reconstruction and transformation of the educational sector from the
very beginning. Reading the reports and documents of the national Department of
Education on progress made to date is heartwarming, reading critical assessments by
outsiders somewhat less so. In 2006 there were 12.3 million kids at school, 340 000
of them (2.8 per cent of the total) in private or independent schools, the remainder in
the public school system. 41 More than 6.6 million learners are at school in the three
provinces with the highest poverty rates among Africans (Eastern Cape, Limpopo and
KwaZulu-Natal). Progression rates beyond grade 10 fall off rapidlyin 2006, 8.9 per
cent of learners were in grade 10, and 4.6 per cent in grade 12 (in the independent or
private schools, the percentage in grade 12 was 8.4). 42 The Gross Enrolment Rate
(GER) for the country as a whole was 98 per cent. Four provinces, Free State,
Northern Cape, North West and Western Cape (surprisingly?) had GERs in the low
90s, the rest were 99 per cent or more. The Gender Parity Index (GPI, the ratio of
girls to boys) is 1.01 for the country as a whole, close to unity or close to it in six
provinces (none lower than 0.98), and above it in three (1.05 in the Western Cape)
(Department of Education, 2006b). Girls handsomely outperform boys at school; in
the 2004 Senior Certificate (Grade 12) examination, 47.8 per cent of the boys (40 689
of them) achieved the university entrance endorsement. Corresponding figures for the
girls were 52.2 per cent and 44 428 (Department of Education, 2005, p.8).

Although the glossy publications by the Department of Education do not attach

numbers to the problems facing the education system with the same ruthlessness as
the academics who study the system and its workings, references to the problems
(challenges) facing the system can be found in the official documents. The
Education for all status report, for example, lists 11 major challenges, ranging from
the influence of poverty and unemployment on dropout rates and achievement, to the
draining away of teaching skills (Department of Education, 2005, p.11). Digging
through the 2005/2006 annual report one finds slow progress with the achievement of
strategic objectives like that of [expanding] the reach of education programmes to
disabled out of school children and youth. (One of whose performance measures is
implementation strategy to identify out of school vulnerable children in the feeding
areas of the designated schools) (Department of Education, 2006c, p.19). Open a
research publication like the Departments monitoring and evaluation report on the

The Department of Educations Education for all status report (2005, p.9) interprets this low
percentage as evidence of confidence in the quality of the public education system. Relative
learner/educator ratios (pupil/teacher ratios, we used to call it) in the private and public systems are
17.5 and 32.6 (Department of Education, 2006a, p.1). While these ratios do not guarantee success, it
would take a high level of innocence not to suspect that most of the countrys future elite will come, as
they did in the past, from the private schools. The likely reason why most people with high aspirations
for their children do not send them to private schools is that the fees are astronomically high.
The proportion of learners in grade 12 in the Eastern Cape was only 3.5 per cent of the total.

- 93 -
impact and outcomes of the educational system, almost at random, and one finds
evidence of the seriousness of the problem, and the slow (or lack of) progress. 43

It is when one turns to the critics outside that holes yet to be plugged become more
visible. Critiques range from the exposure of the rankest corruption and graft, 44 to the
measurement of performance against peers, and against national objectives. One of
the most trenchant critical analyses is that by van der Berg (2005). Socio-economic
status (SES) is the major determinant of school performance in South Africa, i.e.,
inequality matters. It matters so much, that he concludes that: the great majority
of our schools are dysfunctional (2005, p.67)

To place the matter in perspective, the South African performance needs to be viewed
in the context of that of other Africa countries, many of which are much poorer.
Table 5 shows the comparative results of grade 6 children in 15 countries taken from
the SACMEQ II (Southern African Consortium for Monitoring Educational Quality)
survey conducted in 2000. Things may have improved since then, but the dismal
results below show just how far there is to go. The results are (broadly) for the top
and bottom quintiles of the population as measured by socio-economic status.

Table 5 Reading and mathematics skills in 15 African countries

Country Low SES High SES Mean Country Low SES High SES Mean
Seychelles 561.8 594.4 582.0 Mauritius 550.0 607.7 584.6
Kenya 525.3 577.5 546.5 Kenya 546.9 587.1 563.3
Tanzania 528.8 575.2 545.9 Seychelles 532.4 567.8 554.3
Mauritius 508.3 555.1 536.4 Mozambique 527.5 532.6 530.0
Swaziland 519.1 541.0 529.6 Tanzania 509.0 545.5 522.4
Botswana 502.5 543.6 521.1 Swaziland 511.3 522.2 516.5
Mozambique 510.5 523.0 516.7 Botswana 498.9 529.8 512.9
South Africa 440.2 543.6 493.3 Uganda 496.3 519.2 506.3
Uganda 472.3 495.5 482.4 South Africa 446.8 524.3 486.3
Zanzibar 468.1 492.2 478.2 Zanzibar 474.0 483.9 478.1
Lesotho 449.2 454.5 451.2 Lesotho 448.6 444.9 447.2
Namibia 421.5 486.1 448.8 Zambia 425.5 444.8 435.2
Zambia 423.6 456.5 440.1 Malawi 428.2 442.2 432.9
Malawi 422.9 440.7 428.9 Namibia 408.7 461.3 430.9
Average 500.0 Average 500.0
Source: van der Berg (2005, Table 3.1.1, p.64)

Several features of these results jump out at us; the first being the huge difference in
performance between the bottom SES and the top, in both tests, as compared with
most of the other countries (Namibia and Kenya have fairly substantial differentials
but theirs are dwarfed by South Africas). A second feature of note is that even those

Figure 20 on p.35 of the report (2006a), for example shows falling proportions of age-appropriate
school attendance in grades 6 and 9 between 1995 and 2003. Figure 22 on p.38 shows falling
attendance among the disabled in the same period.
noseweek Issue 88 (February 2007, pp.8-11) for example, in its section on Trickle-up economics,
carries a detailed report on an abuse of the school feeding system. The report Let them eat kak,
carries the strap A bunch of well-connected ANC fatcats are using an Eastern Cape school feeding
scheme as a cash cow at the expense of starving kids. The Department of Education has overall
responsibility for the school feeding scheme.

- 94 -
in South Africas top SES do not fare all that well. In the reading test, they were less
than 20 points ahead of Kenyas low SES group. In mathematics, they would have
ranked at position number 5 in the Low SES column, behind Mozambique. 45 Because
of South Africas high enrolment rates, these results are little deceptivemany poor
children the poorer countries do not make it to school at all. Nonetheless, they point
to the massive and continuing damage that extreme inequality does to school
performance. The bad news, as van der Berg points out, is that:

Even middle-income South Africans, 46 with all the advantages that means, get
dragged down by their schools, rather than being provided with opportunities by
them. Truly transformational education should mean that schools are able to turn
around the effects of the lower SES and counteract some of the effects of poverty
on education. We are far from this state. (2005, p.67)

Soutien (2005, p.61) uses the term domino transformation to describe the movement
from lesser-privileged to more-privileged schools by kids formerly excluded by
apartheid. Dictated largely by class position (SES), this movement entrenches the
poverty of the poor; the better-off parents leave the communities, the schools they
leave behind become dumping grounds, plagued by low morale and criminal

No-one with any knowledge of conditions in the South African education system is
unaware of the need for policy to be pro-poor. Each of them recognises that asset
transfers to the poor are necessary. Other than to say that the quality of education,
and especially primary education, must improve, what form these asset transfers
should take, however, is far from obvious. Tax revenues with which to fund
improvements are not in short supply. What is sorely needed is better quality
instruction (both of instructors and the institutions in which they instruct), towards
which steps are already being taken (Flscher, 2006). As van der Berg (2005, p.70)
points out, however, better data with which to identify good and poor performing
schools is necessary before targeted interventions are possible. The other side of the
story, the absorptive capacity of poor children, is more opaque. 47 On the one hand,
we have evidence that poor children in the rest of Africa can perform reasonably well.
On the other, there may be something about the social setting in which poor South
African children are stuck that prevents improvements in the quality of instruction
from being translated into improvements in performance. The survey conducted
recently in the 21 ISRDP and URP nodes found very high levels of anomie (People
like me cannot influence developments in my community) and alienation (No-one
cares about people like me). Levels of anomie peak at 46 and 53 per cent respectively
in Inanda (an urban node) and Mitchells Plain (another urban node). Levels of
alienation reach 41 and 46 per cent in Bohlabela and Sekhukhune (Everatt et al, 2006,

A table in Taylor (2006, Table 3.2.1, p.65) describes performance in mathematics at the high school
level with chilling precision. Two-thirds of all passes in the subject at the higher grade in the final
school year come from just seven per cent of the schools (i.e., from 414 schools, 34 of them African
schools, 380 formerly privileged schools). These are the top performers among 6000 odd high schools.
He says that The reading performance data for South Africa show that there is little improvement in
test scores until one reaches the top fifth of SES values (2005, p.65)
The Department of Education survey referred to above (2006a) found that in 2003, about 3.4 million
of the children who attended school regularly experienced hunger. Eight hundred thousand of them
were in the Eastern Cape and almost nine hundred thousand in Kwazulu-Natal. (2006a, Figure 28,

- 95 -
paragraphs 160-167). Attitudes like these, if communicated to the children in the
household (how would it be possible to avoid doing so?), are unlikely to foster

Even the briefest acquaintance with the critical literature on education leaves one
feeling overwhelmed by pessimism. A recent article by Taylor (2006), discoursing on
the qualities with which a good schooling should endow learners (why are they not
pupils any more?), a strong work ethic, the ability to perform under pressure, a
sense of initiative and responsibility, had this to say:

Without these habits of mind and knowledge skills, school leavers do not have
the wherewithal to make a contribution to society; consequently, they are easy
prey to a life of unemployment, crime or corruption. South Africas great tragedy
at the present time is not poverty, high levels of crime or rampant inefficiency in
the civil service. Profoundly problematic as these issues certainly are, they are
manifestations of a deeper problem underlying our society the inability of most
schools to provide young people with the attitudes and intellectual skills required
to build a modern state. (Taylor, 2006, p.65)

Taylor is not all gloomhe has some success stories to relate, but the brute fact of 80
per cent of schools providing poor education is hard to escape. On that unhappy note
we take leave of the school system.

Dis-ease in the health sector?

There are problems in health as well. We limit ourselves here to a few scattered
observations, when what is required is a full-scale examination. Three of the major
problems in the South African health system are (a) the grossly inequitable
distribution of resources between private health care for the minority that can afford
them, and the majority that has to make use of the public health system, (b) the
difficulties of transforming the health system so that primary health care (PHC)
becomes its foundation while limiting the collateral damage suffered by the public
tertiary system in the process, and (c) the challenges to the system resulting from the
AIDS pandemic. The literature on the last of these three is huge; we shall not attempt
to engage with it other than to note that the epidemic diverts resources (material,
financial and human) away from other pressing medical needs.

As far as the first of the problem areas is concerned, Blecher and Harrison (2006) cite
the Draft Health Charter released by the DoH in 2005 to the effect that the:

most significant challenge facing the South African health system is to address
the inefficient and inequitable distribution of resources between the public and
private health sectors relative to the population served by each. (p.32)

Giving the dimensions of this inequity, they state that of the roughly R135 billion
flowing through health sector financing intermediaries (about eight per cent of GDP
in 2006), about R59 billion (44 per cent) goes through the public sector, while the rest
(R75 billion) goes through the private sector. Private health care is available to 15 per
cent of the population, mainly through medical scheme coverage (2006, p.32).
Disparities in per capita expenditure between the two parts of the system are

- 96 -
immenseBenatar estimated the averages at $(US)850 in the private sector vs. $150
in the public sector (2004, p.82). While not all of the difference between the two
figures is accounted for by quality of service, much of it is. Illustrating what this
means in practical terms he points out that:

nationally, there are 8.7 cardiac surgeons per 1 million population who
perform 774 operations per 1 million people annually in the private sector; in the
public sector, there are 0.6 cardiac surgeons per 1 million people who perform 69
operations per 1 million population. (p.82)

While there is general agreement that the huge expansion in primary health care
(PHC) facilities and treatment that the past few years have witnessed 48 was absolutely
necessary, the costs, in terms of lost capacity in the major teaching hospitals, where
[h]ospital funding growth has been unimpressive and where the implementation of
the proposals in the 2004 Modernisation of Tertiary Services report are lagging
behind (Blecher and Harrison, 2006, p.47), have been severe. One anticipated
consequence of the expansion of PHC should be an increase in referrals to the major
hospitals of acute cases, as the means of detecting them in outlying areas improves.
Starving the major public sector hospitals of resources has (predictably) reduced their
capacity to deal with this problem, as the following figures demonstrate:

At Groote Schuur Hospital [in Cape Town], cardiac surgical operations in adults
have been reduced from 700 per year to fewer than 250 per year. In the
orthopedics department, budgetary reductions have resulted in the limitation of
joint replacements to 60 procedures per year in 2003, as compared with 350 in
1993. In the ophthalmology department, there has been a 60 percent reduction in
faculty and a 50 percent reduction in beds over the past decade. In the general
surgery department, the waiting time for surgery for breast cancer is now 8 weeks
(as compared with a wait of 2 weeks 10 years ago). (Benatar, 2004, p.85)

Unlike school, where attendance is compulsory, people who are unwell are not
obliged to seek medical help. During the reference month for the 2005 General
Household Survey (June 2005), 5.9 million people who reported being ill or injured.
They may be divided into those who sought attention (4.8 million people) and the
remainder, who did not. The 1.08 million people who did not receive medical
attention, may be sub-divided into those who said that it was unnecessary to attend a
medical facility (681 000 people), and those who needed medical attention but were
unable to obtain it. Among this group of 392 000 people, who, it may be assumed,
deemed that they did want or need medical help, 83 000 did not receive it because the
nearest medical facility was too far away), and 208 000 because it was too
expensive. 49 In other words, if the reference month was a typical month, then it is
possible that during the year, about three million people would have been denied
medical attention because they were too poor.

The quality of help they receive in public sector clinics varies enormously. Although
the bulk of users report being satisfied, a large number of people experience problems.

Blecher and Harrison report that PHC visits increased from 81 million 101 million per annum over
the period from 2000/01 to 2005/06 (2006, p.45)
It is not clear why so many people report that care was too expensive. As Blecher and Harrison point
out, primary health care (at least) is free at point of service (2006, p.32).

- 97 -
An indication of the type and extent of problems may be obtained from the GHSs.
Table 6 lists the numbers of individuals who reported experiencing one of four
problems with the quality of service provided at public sector clinics during the
reference month for the 2005 GHS. They are divided along income lines into two
groups: individuals with per capita expenditure of less than R412 per month in 2005
prices (R322 in 2000 prices, the possible lower bound of South Africas poverty
critical range); and those with per capita incomes below R824 per month, a figure that
is not much higher than the upper bound of R760 per month.

Table 6 Nos. of people reporting problems at clinics during June 2005

No. below No. below
R412/month R824/month Total
Facility not clean 24 400 39 000 104 400
Long wait 150 900 298 200 872 200
Drugs needed not available 60 100 121 000 329 800
Staff rude, uncaring, patient turned away 45 200 80 400 221 900
Total 280 600 538 600 1 528 300
Source: Estimated from 2005 GHS data set.

The number of people inconvenienced during the course of the year depends on two
things; whether or not the figures are typical of any month in the year, and the
frequency of visits of individuals who are inconvenienced. If the figures for the
month of June 2005 are typical for those of the year, then there would have been
about 10 million episodes where people were made to wait too long for medical
attention, during the year. Likewise, there could have been about four million
episodes in which people did not get the drugs they needed, and nearly three million
episodes where people will have experienced rudeness or uncaring attitudes from
staff, or having been turned away from the facility. Roughly one-third of all of these
problems would have been experienced by poor people.

This account squares with the report by Blecher and Harrison that in PHC facilities,
quality of care is sometimes sub-optimal, public facilities often have long waiting
times and primary care facilities tend to have few doctors (2006, p.32). Comparisons
between service provided in the private sector and public sector are stark: as opposed
to the 46 per cent who complained about too long a wait at public sector clinics, about
ten per cent of those who saw private doctors made a similar complain. In clinics, the
necessary drugs were not available to about 18 per cent of patients; three per cent of
those consulting a private doctor made the same complaint. Less than two per cent of
those attended to by private doctors complained that staff were rude and uncaring, as
opposed to the almost 12 per cent who made this complaint about service at public
sector clinics.

The GHS does not ask people what service they would choose if income were not a
constraint. It looks as though about 130 000 people with per capita incomes less than
R412 per month chose to make the sacrifice involved in consulting a private doctor.
This is close to half of the number (303 000) in similar circumstances who chose to
use a public sector clinic. Voting with ones feet when one is so poor is possibly the

- 98 -
strongest criticism of the system that one can make. Despite the obvious progress in
making health care available to more South Africans, especially among the poor, there
apparently remains considerable scope for reduction of inequalities between the
quality of publicly- and privately-provided health services.

Income inequality: Type As vs. Type Bs, Round 2

Referring back to Table 3 to get our bearings, we may be seen to have collapsed
issues of asset inequality, human capital and gender inequality 50 into the foregoing
discussions. After asset inequality in the table comes income inequality, a subject that
should not need to be discussed herewhole forests have needed to be felled to allow
details of the conditions in one of the most unequal countries in the world to be
disseminated. Even so, much more remains to be said. Debates over economic policy
to deal with poverty and the maldistribution of income lead easily to the adoption of
polar positions. So, as an entre to the subject, let us look at a paper that tries to take
some of the heat out of the argument between the contending parties. Kanbur (2001)
identifies, at one extreme, what are perceived as conservatives in the treasury at the
other, a gaggle of activists, variously portrayed (not by Kanbur!) as impractical
dreamers or profligates who agitate for greater social spending (dismissed as macro-
populism). Arguing that despite vociferous disagreement, there has been, if not a
rapprochement, then at least a narrowing of the field of that disagreement to three
major areas, of which one, time horizon (how long will it take to reduce inequality
to acceptable levels), is of particular relevance in the South African case. In
essence, Kanburs Type Bs (those promoting greater social spending, especially on
welfare), possibly closer to the poor, demand short-term measures to alleviate
poverty. The Treasury, by contrast (his Type As) takes a medium- to long-term view
of poverty reduction. Applying this framework to the South African conditions, it
may be argued that not enough is being done to address poverty and inequality. To
proceed thus is not to ignore the many short-term steps that government has taken it
is rather to claim that not only is it not certain that the longer term steps will work, it
is also the case that even if they do, they will take too long to do so. In addition to
that, the short term measures taken are inadequate. As Kanbur says, paraphrasing
Keynes, in the short run, (too?) many of the poor will be dead.

Unfortunately, in South Africa, there is as yet, little sign of rapprochement. If we ask

the question of government; what pro-poor investments may be made to reduce
inequality? one possible answer is that the social wage has already done so, and to a
significant extent. Academic findings of increasing poverty and inequality that
dominated the debate until the early years of the new millennium have been
challenged by government on the grounds that such claims ignore the massive
expenditure on in-kind transfers (free or subsidised goods and services): electricity;

Not for one second is it imagined that the fleeting reference above to the performance of girls in
school being superior to that of boys, and the fact that Gender Parity has been achieved, or even
surpassed in schools, can be taken as an adequate treatment of gender inequality. Despite the
commitment to gender equality in South Africa, there is still a long way to go before women achieve
real equality with men. We will content ourselves here with just one illustration, drawn again from
Everatt et al (2006). The issues, treated in paragraphs 212-219, are those of consensual decision
making and gender-based violence. One statistic will suffice: 21 per cent of men in the 21
development nodes believe it is appropriate to hit or beat their partners if the partner ignores the
children. We have a long way to go.

- 99 -
water; transport; housing; education; school feeding schemes; sanitation, and health
care. There is merit in governments challenge. Unfortunately, in its haste to
confront its critics, government published, in the Ten Year Review (PCAS, 2003), a
set of wholly implausible estimates of the impact of social spending (and taxation).
Based on research commissioned for the Review, research that examined the fiscal
incidence of social spending, government claims that social spending in 1997 reduced
the Gini coefficient from a level of 0.68 (excluding social transfers) to 0.44. The Gini
coefficient for the year 2000 (excluding social transfers) was estimated at 0.59 (how
the pre-transfer fall came about is not made clear). At any event, taking social
transfers into account allegedly causes the figure of 0.59 to fall to 0.35 (PCAS, 2003,
pp.90ff). Not only are these estimates methodologically unsound, it is by no means
clear what the re-estimated Gini coefficients mean. 51 Trying to engage government
on the issue of income inequality is surprisingly difficult. Part of the reason for this
might be an unwillingness to believe that spending so much money 52 has not had the
desired effect. Yet as we have seen in the analysis of the state of education, 80 per
cent of the school-going population is not reaping the benefits that the expenditure of
large sums should provide. Transformation of the health care system still has a long
way to go.

Social grants: More than a glimmer of light

What is working well, by common consent, is the social grant system. 53 Virtually
every study that looks at poverty concludes that social grants reduce it. Yet
government has set its face resolutely against further expansion of the system (it is
intent on reducing the numbers of claimants) on the grounds that it creates and
reinforces dependency. 54 Social security spending in 2006, not all of which is on
grants, at about R80 billion odd, amounted to a little less than five per cent of GDP in
that year. The budget for fiscal year 2006/2007 allocates R57.7 billion for
expenditure on social grants (National Treasury, 2007, Table 6.6, p.106). In the years
2003-2004 and 2004-2005 increases in final consumption expenditure amounted to
respectively R42 and R47 billion (in constant 2000 prices). Inflating those increases
to the prices of 2006 converts them into increases of R58 and R66 billion. 55 Most of
this will be consumed by the well-off. If economic growth continues in the way it has
Despite its rejection by economists of every stripe, the claim has been repeated in a work that bears
the title A Nation In The Making: A Discussion Document On Macro-Social Trends. Produced by the
Policy Co-ordination and Advisory Services in the Presidency (PCAS, 2006), the document has, as one
of its objectives, the shifting of the focus on macroeconomic performance to one that gives social
indicators greater prominence. The inequality reduction claim is worded thus: The Gini coefficient,
another widely used measure of inequality, was 0,59 in 2000 when social transfers were excluded. If
these were included, it was 0,35. (PCAS, 2006, p.14)
Of allocated expenditure of R460 billion for the tax year 2006/2007, R262 billion goes to Social
services. Education gets R92 billion, health R54 billion, social security and welfare R81 billion, and
community development, R25 billion (National Treasury, 2006a, p.105). A little arithmetic suggests
that expenditure on social services amounts to about 15.6 per cent of GDP; education, 5.5; health, 3.3;
social security, 4.8; community development, 1.5 per cent.
Even government acknowledges the efficacy of social grants as opposed to its alternatives. After
boasting about the 2182 community assets left behind after R6.5 billion expenditure on
infrastructure, Government acknowledges that: public works [the other major anti-poverty
programme] are not as efficient as income grants in alleviating income poverty (PCAS, 2003, p.19).
Repeated claims that teenage girls are falling pregnant (and abandoning their babies) in order to
qualify for the child support grant were laid to rest by a piece of research commissioned by the
Department of Social Development (Steele, 2006).
A CPI of about 140 is assumed. The CPI for 2006 was 138.8 and the CPIX 141.1. (P0141.1)

- 100 -
over the past few years, in a short while, the annual increment to final private
consumption will be as almost large as the social welfare budget, and larger than total
the amount spent on grants in 2006/2007. Despite this flood of extra income, not
much reduction of income inequality can be expected from current social policy. If it
happens at all, it will have to come from employment growth, an unlikely prospect.

Until such time as serious engagement around a number of issues can take place,
amongst them, those of the quality of services, and those of the need to expand the
social grant system, no sensible discussions with government about income inequality
in South Africa can be expected to take place. Until clear indications of how income
inequality is to be reduced are published, vague statements like those in the AsgiSA
document about radical reductions in inequality should be treated as just so much
hot airpolitical posturing intended to take the steam out mass dissatisfaction.
Redistribution to the poor will almost certainly not come from so-called broad-based
black economic empowerment, the next matter to engage our attention.

Black economic empowerment

In the bad old days, the apartheid regime tried hard, but without much success, to
create a black middle class, to act as a bulwark against rising popular discontent.
Some of the fruits of this enterprise, in the form of corrupt and incompetent former
bantustan officials, are still with us. For the rest, the enterprise is mainly a bad
memory. The new attempt to create a black middle class and a black bourgeoisie, if it
succeeds, will provide protection (at bargain basement prices) against anti-capitalist
sentiment, precisely because it will be black and capitalist (MacDonald, 2006, p.154).
The impact on poverty and the mal-distribution of income are (and are likely to
continue to be) minimal (p.158). Expectations that this fabricated class will behave:
differently and better than the white bourgeoisie because its interests coincide
with the immediate interests of [the] majority , and that it will:

promote job creation, the fostering of skills development, the empowerment

of women, the strengthening of the popular organs of civil society, and active
involvement in the fight to end poverty (MacDonald, 2006, p.156, citing Pallo
Jordan) 56

are touching, but completely out of touch with the reality of capitalist accumulation. 57
To damp down on popular resentment to what is perceived by many as black
economic enrichment, it is important, however, for the state to keep the propaganda
machine steaming ahead.

Tackling this issue in his recent book on power in the South African economy,
Calland quoted two extracts from a South African Communist Party document which
dismissed the emerging black capitalist as excessively compradorist and parasitic
The reference is to a paper written by Jordan for the ANCs 50th National Conference. See
MacDonald, 2006, p.215.
Swilling et al (2005), who should know better, engage in similar delusions, when they speculate on
the possibility that black economic empowerment (BEE) and broad-based black economic
empowerment (BBBEE or BBEE) will manage to break white control of investment decision-
making quickly enough to ensure that private sector investment levels climb back up over the 15% of
GDP mark (2005, p.74). No reasons are offered as to why black capitalist behaviour should differ so
dramatically from white.

- 101 -
(2006, p.264). A comprador bourgeoisie, in one definition, cannot function on its
own but relies on imperialist capitalists, sacrificing national interests to do so. The
curious interlocking elite interests that have formed since 1994 rely on special
share deals, affirmative action, BEE quotas, fronting, privatization and trading on its
one real piece of capital (access to state power) to establish itself. As Calland
points out, before 1994, South Africa had a well-developed capitalist class. Self-
preservation requires that it cuts this emerging fraction a slice of the action in order
to remain in favour with the new political reality. (p.264)

If the new elite were disposed to behave in the idealistic fashion suggested by Pallo
Jordan, actively promoting income redistribution, the policy of black economic
empowerment would indeed be worthy of celebration. All the signs point in the
opposite direction. To illustrate the corrosive effects of the policy on development,
two very different stories will be told. The first is from a communist, the other, a
former ANC cadre. First, the communistJeremy Cronin. The long quote that
follows tells a tale too seldom heard:

the policy focus on achieving market-driven growth, has managed to

reproduce a very problematic form of individualism. We used to talk about black
people in general, and Africans in particular, being nationally oppressed. I do not
remember anyone ever using the words historically disadvantaged individual
during the struggle. But that is now the discourse and epitomises the paradigm
within which we are operating. To be oppressed is to be systemically oppressed
by a system an injury to one is an injury to all and the solution to this
historical oppression has to be systemic transformation. But when you talk about
disadvantage (and an individuals disadvantage), you imply that the system is
okay, but you are just lacking some advantages within it. And when you call the
disadvantage historic, then you are even saying the systemic problems are now
just history. Again, we get back into that redistribution model of change, instead
of addressing the transformation of the fundamental realities of our society.

I welcome what President Mbeki said in his critique of the prevailing

consumerism that is so devouring both our society and our ANC-led movement.
To take action against this we need to look at ourselves in the ANC, and we need
to look at our policies. I offer by way of example some of the recent statements
made by the inner circle of the Mbeki government. One very prominent member
said that blacks should go out and become filthy rich. Another one said, I didnt
struggle to be poor. Another said, Why should black capitalists behave any
differently from white capitalists?

What has been encouraged is a sense of self-righteousness: I am historically

disadvantaged, and I am self-righteously and individually entitled to a slice of the
action. And when you fall out of the inner circle of promotion, or favouritism, or
whatever, then the explanation has got to be a conspiracy and you
demagogically call on millions of poor South Africans who are unemployed, who
are disillusioned, who are marginalised, to follow you, to love you, and to see in
you a mirror image of themselves as the disempowered, the marginalised, the
neglected, and so forth.

- 102 -
Both approaches (market-driven growth and demagogic mobilisation around
grievance) elide the systemic issues. As a result, there is an absence of a concrete
programme of transformation and the lack of perspective on what the systemic
issues are. This forces us in the direction of a narrow politics of politicians. It is
the politics of the palace, of succession debates and factionalism, and not the
politics of active citizens within a democratic society. (Cronin, 2006, p.9)

Cargill (2005) takes a very different line, but the end-result is the same, a capitalist
class with little appetite for entrepreneurship. If there was to be any hope that a black
capitalist class would meet at least the employment creating objective foreseen for it
by dewy-eyed analysts, then that hope is bound to be dashed unless the structures
within which accumulation among the black population is being fostered, are
changed. Modelled on the Malaysian approach, BEEs elimination of the principle of
risk-taking among those it seeks to empower is likely to have the same effectthe
creation of a class of Ali Babas who simply trade in state licenses and contracts
empowerment speculators. (Cargill, 2005, p.25, citing the former Prime Minister of
Malaysia) Arguing that there are compelling reasons for getting BEE ownership
policy right, she concludes that:

While BEE ownership answers a strong political need right now, the numbers
suggest that if commercial value cannot be injected into the process, the net results
could be extremely negative for the economy in the medium to longer term
growth and poverty alleviation could be put at risk. (2005, p.27)

The imperative to get BEE right is stronghowever, even if government succeeds in

devising the appropriate policies to do so, no expectation that much by way of
redistribution to the poor will occur should be entertained. The qualifier broad-
based will stick to the policy of black empowerment, but the number that will benefit
is relatively small. 58 There is likely to be as little enthusiasm among the nouveau
riche for inequality-reducing redistribution (i.e., significant redistribution) as there is
among the established (mainly white) bourgeoisie.

That leaves only employment (actually, unemployment), at which we have glanced in

passing. Let us go back to it.

Employment creation

There are about three million households, as noted earlier, that contain people of
working age, none of whom is employed, most of whom say they want jobs. That is
the refrain one hears from every survey. In high unemployment areas, this is the first
and most obvious demandrecall the figures of 60 per cent of those in the ISRDP
nodes, and 73 per cent in the URP wanting a job (cited above from the Everatt et al
piece). Their chances of finding one are slendernot for nothing is economics called
the dismal science. A simulation exercise conducted recently by Altman (2006),
argues that from an employment creation perspective, five broad sources of jobs in the
economy may be distinguished. They are:

Every now and then, a deal that touches the lives of more than just a few of the usual suspects
makes news. An early example is recounted in Ensor (2004). The deal in question seems to have
extended to 2 192 employees, each of whom had a stake in R5 million holding, i.e., about R2280 each.

- 103 -
poverty alleviating activities, ranging from the EPWP to survivalist activities,
the public service
resource-based sectors
follower industries, such as retail or construction (relatively low paying)
dynamic goods and services industries that take advantage of global markets
(2006, p.4)

It is not the intention to engage critically with this taxonomy; suffice it to say that
with assumptions of varying degrees of optimism, she manages to halve the rate of
unemployment by 2014 59 (that would probably leave us with about two million
officially unemployed, and four million or more unemployed according to the
expanded definition, a problem that is still considered to be unacceptably high by
global standards (Altman, 2006, p.14). The plausibility of the assumptions used is
up for debatesome fairly spectacular (sustained) sectoral growth rates60 are
required, if hopes are not to come to rest on the EPWP, a burden for which it is not
well suited (Altman, 2006, p.13).

Building a simulation model to take some of the guesswork out of the question of
whether or not the unemployment halving goal will be met is not difficult; populating
it with reliable assumptions about the behaviour of some of the variables, e.g., the rate
of growth of participation rates, both official and expanded is not so easy. For what it
is worth (and that may not be a great deal) a model I built (Meth, 2006b) suggested
that about 3.6 million jobs needed to be created between 2004-2014 (2006, p.379), a
target that current rates of job creation suggest is not impossible. That outcome,
however, depends on the assumption that while the working age population grows at
one per cent per annum, the participation rate (extrapolated from past behaviour)
declines at 0.1 per cent per annum. If that prediction is not met, and the official
participation rate rises at a modest 0.2 per cent per annum, i.e., potential workers
respond to the year-upon-year sustained growth that AsgiSA promises, in such a
restrained manner that the participation rate only goes from 54.6 per cent in 2003 to
56.6 per cent in 2014, the number of jobs required would rise to 4.6 million. More
robust responses to perceived increases in job opportunities, in which the discouraged
change labour market status by actively seeking work, could push the official
participation still higher, causing the chances of meeting the halving goal to recede.

As far as interpretation of the official statistics on job creation is concerned,

government, unsurprisingly, has repeatedly attempted to place them in the best
possible light, or as the jargon would have it, to put upon them, a favourable spin
(Meth, 2007). In the 2007 State of Nation Address, for example, the President
claimed that:

Over the past three years, the economy has created some one-and-half million
jobs. It is encouraging that in the year March 2005 to March 2006 alone, 300 000
of the jobs created were in the formal sector outside of agriculture, representing a
growth rate of about 4%.

Using a wide array of policies, Pollin et al (2006, p.xv) manage to get the (official) rate of
unemployment down to 15.4 per cent by 2014.
For sources of growth, AsgiSA relies to a significant degree on infrastructure investment. The
likelihood that this will stimulate private investment (as hoped) is investigated by Frankel et al (2006,
p.60). Their conclusions are not encouraging.

- 104 -
The official figures (which the President has questioned in the past) estimate that
March 2003 total employment was 11.3 million. By March 2006 this had risen to
12.45 million, an average increase of 382 000 per annum. 61 It is only in the last year
(March 2005-March 2006) that the employment has grown by more than half-a-
million (544 000). If one excludes jobs in informal agriculture on the grounds that
accurate employment estimates in this sector are notoriously difficult to make, then
the total number of jobs created over the year 2005-2006 falls to 354 000. Over the
period 2003-2006, it would drop to 887 000, to give an average rate of job creation of
296 000 per annum.

The stability of the official unemployment rate in the region of 26 per cent
(movements of one percentage point either side of this figure are not significant) from
September 2004 onwards, points to the depressing possibility that the conclusion
drawn by the Presidents Harvard team of economists, namely that the existing rate of
unemployment is an equilibrium rate brought about by structural changes in the
economy. One can draw from this the inference that this rate is not likely to change
much unless countervailing structural changes can be made. Of these, there is some,
but not much evidence.

Because of the implications it has for poverty and inequality, the question of who
would get such jobs as are created, has to be faced. 62 This is so regardless of whether
or not the halving target is met. 63 It seems reasonable to suggest that those who are
best placed, either by qualification, by social location (ability to network), or
geographical location (proximity to employment possibilities) will be first in the
queue for jobs. People with such characteristics are unlikely to be at the very bottom
of the income distribution. 64 They may also be those who, harbouring unrealistic
expectations (high reservation wages) resulting from a belief in the propaganda about
the value of education, 65 are most easily capable of adjusting those (downwards) to
market realities (Banerjee et al, 2006, p.53).

Pro-poor employment growth would want to see those obtaining jobs whom the
analysis seems to declare least likely to find gainful employment. As a barrier to
successful redistribution of income, this is most formidable. It is true that in the long
run, economic growth is the only sustainable route out of poverty. If the pattern of

If we take the period September 2002-September 2005, then the numbers grow from 11.3 million, to
12.3 million. For this set of figures, if we exclude informal agricultural employment, the annual
increase in employment over the three years is about 400 000. Treating the agricultural employment
figures would possibly give a more balanced picture of employment creation. Note that the sources of
all of the employment figures cited here are the relevant Labour Force Surveys.
In order to understand the impact on poverty of job creation, the formidable job of modeling who
gets jobs must be undertaken. For an example of how this may be tackled, see Filho and Horridge
(2004, ppp.6ff). The model they use is more complex (by several orders of magnitude) than the simple
device I built for looking at the halving question. Even so, they must still make heroic assumptions.
The goal of halving unemployment by 2014 is at once both ambitious and modest. It is ambitious in
the sense of being difficult to achieve, and modest in the sense that even if it could be achieved, it
would still leave far too many people wanting jobs.
If Banerjee et al (2006) are to be believed, those most likely to find paid employment will have a
matric (grade 12) plus some post-matric qualification (p.26) and will not face excessive job-search
costs (p.47). This rules out areas remote from the major metropoles.
The huge increase in the number of school-leavers with a grade 12 certificate will have pushed down
the rate of return to such investment in human capital (Banerjee et al, 2006, p.53).

- 105 -
growth is constrained in such a way that it bypasses most of the poor, a growth pattern
with rising inequality, and possibly rising poverty as well, may result. Certainly, the
precedents for such an outcome existwe reported on them in our literature survey

Regional disparities

The story of South Africas division into white areas and bantustans which were to
be granted independence is so well known that it should not have to be recalled here.
It is useful though, when thinking about the investment required to bring development
to the poor, to have before a picture of the regions in which that development is to
take place. To present this, we reproduce maps of three provinces (as Figures 2, 3 and
4 below) depicting electoral wards in the province ranked from least to most deprived.
Two of the provinces, according to the figures in Table 1, are among those that
contain most of the poor people in South Africa (Eastern Cape and Limpopo), while
the third is the second most prosperous, the Western Cape. The maps are part of a
project for estimating Provincial Indices of Multiple Deprivation (PIMD).

The results of work on a joint project headed by the Centre for the Analysis of South
African Social Policy (CASASP) in the University of Oxford (Noble et al, 2006a and
2006b), the (ordinal) indices used to derive the rankings are based on Census 2001
data. 66 They measure the combined effects of deprivation across five domains:
Income and material deprivation; Employment; Health; Education and Living
environment. The work, a valuable addition to the poverty literature in South Africa,
reviews earlier attempts at the measurement of poverty at small area level (2006a,
pp.9-11), highlighting the pioneering nature of some of the work, but also drawing
attention to their weaknesses, of which the lack of adequate conceptual frameworks is
quite prominent. Noble et al offer four reasons for constructing small area
deprivation indices:

The non-randomness of geographical patterns of disadvantage (or advantage)

The effectiveness (if deprivation is adequately measured) with which the most
deprived areas may be targeted,
The likelihood that high concentrations of poor will place heavy demands on
the capacity and resources of local providers, and
The ability of multiple deprivation indices to detect the varying mix of
problems experienced in different areas. (2006a, p.8)

The conceptual model underpinning the indices is described thus:

[It] is based on the idea of distinct domains of deprivation which can be

recognised and measured separately. These are experienced by individuals living
in an area. People may be counted as deprived in one or more domains, depending
on the types of deprivation that they experience. The overall index of multiple
deprivation us conceptualised as a weighted area level aggregation of these
specific domains of deprivation. (2006a, p.13) 67

The discussion that follows, lifted almost in its entirety from Noble et al, is intended only to serve as
an introduction to the undertaking.
Note the difference between this weighting process, in which very poor performance in one domain
can be offset by better performance in another, with the insistence by Bourguignon and Chakravarty

- 106 -
An important feature of the individual domain indices is that each reflects a
particular aspect of deprivation. In their words:

the Employment Deprivation Domain captures exclusion from the world of

work and conditions of work not the low income that might flow from it. The
Income Deprivation Domain can be used separately from a PIMD to examine low
income alone. The Education Deprivation Domain represents educational
disadvantage and does not include non education indicators which may contribute
to education deprivation such as the lack of electric lighting to undertake
homework. Such an indicator would be captured in the Living Environment
Deprivation Domain. This approach avoids the need to make any judgements
about the complex links between different types of deprivation (for example the
links between poor health and unemployment), and enables clear decisions to be
made about the contribution that each domain should make to the overall PIMD.
(2006a, p.13-14)

Domain indices are combined into multiple deprivation indices by first being
transformed to an exponential distribution, with each transformed domain range lying
between zero and 100. This has the effect of stretch[ing] out the most deprived 25%
of wards in each province. Transforming each domain so that it has:

a common distribution, the same range and identical maximum/minimum

value [means] that when domains are combined into a single index of multiple
deprivation the equal weighting is explicit; there is no implicit weighting as a
result of the underlying distributions of the data (2006a, p.29)

Noble et al have developed this work in a number of ways, all of them of interest to
the policymaker intent on directing resources to the areas in which they are most
urgently needed (the national Department of Education has seized upon the indicators
with great gusto). The first of these is a set of national indicators of multiple
deprivation. Currently under wraps because of their extreme sensitivity (since they
rank areas from most to least deprived across the country as a whole, they would
immediately upon release, be seized upon by provincial politicians, fighting for extra
allocations for their provinces). The other development addresses a weakness in the
PIMDs, that is easily visible by inspection, namely, that (many of) the areas
(jurisdictions) for which the results are presented are too large to be used for any
practical policy-making exercise. Data are available to produce indicators at almost
any geographical levelreplicating the techniques used in the UK to look at
genuinely small areas, Noble et al have done the preparatory work for a similar
exercise in South Africa, in areas containing about three thousand people. It wants
only the co-operation of the relevant authorities for this work to go ahead.

What is remarkable about these maps is the way in which they illustrate what we all
know to have been one of the major features of apartheid accumulation, namely, that
it made possible the development of the white areas by putting in place the
infrastructure necessary for such development. The infrastructural investment most

(2003, pp.27-28) that: a multidimensional approach to poverty defines poverty as a shortfall from a
threshold on each dimension of an individuals well being.

- 107 -
plainly visible in the maps is the set of major roads. The way they skirt, or cut
through areas of high deprivation at their thinnest points is truly remarkable.

Figure 2 Map of Eastern Cape Index of Multiple Deprivation

Note how the major road, the N2, links the former white towns in what used to be
the Transkei Bantustan. Although they contain substantial pockets of poverty, now

- 108 -
clustered around former white towns, the parts of the Eastern Cape outside of the
Transkei are relatively prosperous. One surprise is the former Ciskei Bantustan,
centred around Bisho. Ones expectation is of much greater deprivation than the map
suggests. Once more, smaller areas are probably required to see it clearly.

Figure 3 Map of Limpopo Index of Multiple Deprivation

- 109 -
The map of Limpopo tells a similar storythe major arterial road, the N1, links
former white towns such as Messina and Pietersburg (now renamed Musina and
Polokwane). The light coloured portions of the map are mainly prosperous white-
owned farmlands (many of them the subject of land claims) while the dark areas are
former Bantustans like Venda.

Figure 4 Map of Western Cape Index of Multiple Deprivation

- 110 -
In the Western Cape, the huge dark area centred around Beaufort West is the Central
Karoo Integrated Sustainable Rural Development Programme (ISRDP) node, with its
adult population at census time (2001) of only 33 000. 68 The dark areas around the
N2 motorway in the inset, are informal settlements that have sprung up in recent
years, and older apartheid creations like Mitchells Plain and Khayelitsha, both among
the nodes earmarked for special attention as part of the Urban Renewal Programme

In our trawl through the literature, we pointed to some of the questions that have been
raised in connection with regional under-development, citing a paper by Ravallion
(2005a). Among the questions raised were these: Does it make more sense to move
jobs to people, or people to jobs? Is there a trade-off between achieving greater
regional equity such as by focusing on areas with high poverty rates but low
poverty densities and poverty reduction in the aggregate?

Probably the most coherent attempt to begin answering these questions is to be found
in the Integrated Sustainable Rural Development Programme (ISRDP) and Urban
Renewal Programme (URP) covering the 21 nodes containing South Africas poorest
people. We have already encountered this enterprise above. The aim, as Everatt et al
explain, is to:

transform their respective nodes into economically vibrant and socially

cohesive areas initially through anchor projects to kick-start the programmes, and
then through better co-ordination between departments geared to providing an
integrated suite of services to all citizens, especially those living in poverty. The
point of both programmes is the more efficient and effective use of existing
government resources, rather than operating as standard, stand-alone programmes
with a dedicated budget. (2006, paragraph 2).

Budget allocations for the ISRDPs and URPs are channeled through the Department
of Provincial and Local Government. As may be seen by consulting the expenditure
allocations (National Treasury, 2006b, pp.59ff) the budgets for the programme are
tiny. This is in line with the intention of using the vote to make, as Everatt et al point
out above, more efficient and effective use of existing government resources.

Whether or not the state has the capacity to achieve the kind of integration necessary
to foster development in these benighted areas is moot. Certain it is that poverty in
them has declined, and for two reasons. The first of them is the increased provision of
basic services; housing, electricity, water and sanitation. This is recorded in some
detail in Everatt et al (2006). The second is the flood of social grants (mainly the
child support grants) into these areas. Paragraphs 33, 34 and 35 in Everatt et al
(2006) show graphically just how dependent people in the development nodes,
especially the rural nodes are on the grant system. Even in the urban nodes, with their
greater access to employment, grants are only just beaten into second place as source
of income by earnings from employment.

The Everatt et al study interviewed 400 households in each node. With an adult population of
33 000, Central Karoo is dwarfed by O R Tambo with its population of 757 000, where 400 households
were also interviewed (2000, paragraph 8 and Table 1).

- 111 -
Now here is the crunchas noted before, governments ideology of development is
one which eschews, as far as possible, reliance on social grants. The South African
state has adopted, hook, line and sinker, 69 the notion of the developmental state,70
which emerged out of the debate about the reasons for the success of the so-called
East Asian tigers. Effectively, the adoption of this ideology blocks the single most
important channel for redistribution, one even approved, as we have noted above, by
the Chief Economist of the World Bank, Franois Bourguignon (2004), social grants.

Developmental state delusions

In recent times, two substantial pieces on the topic of South Africa as a developmental
state have made their appearance. The older of them is by Swilling et al (2005). The
more recent paper is by Southall (2006). Swilling et al are concerned with the
proclaimed ambition to be a developmental state because of the significance of such a
claim for understanding governments approaches to poverty alleviation in particular,
and development policy more broadly. Arguing that the South African state has
followed neither the classic neo-liberal state model, 71 nor the developmental state
model (p.16), they try to make sense of the messiness of policy by focusing on the
search (by government) for the elusive developmental state during the first
democratic decade, and into the second, where the search intensifies (2005, p.6).
Although parts of their analysis seem facile, they are certainly no apologists for the
state. In any case, one need not swallow their whole analytical apparatus. To the
uninitiated, the developmental state may sound cuddly. For those without
knowledge of the forebears of this state form Swilling et al spell out the
characteristics of the classic developmental state. It has the following elements:

a determined developmental elite committed to the modernisation project;

relative autonomy from major capitalist economic interests who are always
keen to capture the state;
a powerful, competent and insulated economic bureaucracy that enjoys the
highest possible political support but operates without too much political
a weak and subordinated civil society which means there are no rival centres
of alternative policy formation;
the effective management of non-state economic interests via formal
structured compacts, incentives and penalties; and
accessible and usable institutions of repression, legitimacy and
performance 72

In an era in South Africa where the discourse of policy formation was generously
sprinkled with calls for transparency among stakeholders in finding the way

Following the publication of the White Paper on welfare in 1995, the former Department of Welfare
and Population Development changed its name to the Department of Social Development. Its
aspiration, despite the fact that the bulk of the cash it spends is disbursed in the form of social grants, is
to engage in welfare work in a developmental way.
In rejecting pleas for a basic income grant to address destitution, government has invoked this
As Seekings (2002) has pointed out: the democratic state inherited a highly redistributive budget
from the late apartheid state in 1994 (p.5). To describe all the steps taken since then to address the
conditions of the poor (regardless of their success) as neo-liberal, as many critics do, is not useful.
The authors are citing Leftwich, 1995, pp.400-406. See Swilling et al (2005, pp.8-9).

- 112 -
forward, open acknowledgement of a desire to see the third of these preconditions
met would not have been politically advisable. The preferred kind of developmental
state (at least in public discourse) would have been what Swilling et al term the
pragmatic developmental state. Such a model would:

look for a leading developmental role for the state (with a focus on human
capacity, knowledge and skills), but via creative and complex partnerships with
non-state actors (especially the ethical and responsible wings of the corporate
sector), and with civil society formations rooted in poor and working class
communities in particular. (2005, p.15)

This vision may (or may not) have been in the minds of those thinking about
development policy in the first development decade, but, argue Swilling et al:

there is a lot of evidence that key thinkers in the South African state who are
leading the post-2004 developmental state perspective are using a classic
developmental state model. (2005, p.15)

Whatever the case, the fledgling state could not match word and deed. As Swilling et
al observe:

Despite aspirations to be a developmental state, the necessary capacity for

purposive, decisive and autonomous policy coordination was not successful prior
to 2004, despite various attempts. (2005, p.16).

The reality was one of a desperate scurrying to fill posts, write reports and compile
budgets, forcing many:

progressives in government to make a trade-off between no progress in

achieving broadly defined transformation goals, and achievable more narrowly
defined short term strategic advances. (2005, pp.16-17)

Their conclusion on the upshot of this muddling through is worth citing at length:

Policy choices during the first decade of democracy in South Africa, they
argue, were based on what was invariably an imperfect and incomplete reading of
a complex reality (including the balance of class forces), after which choices were
made under extreme time pressures and developed their own relatively
autonomous momentum as they got sold to many diverse constituencies to mean
different things hence the (suspect and dangerous) assumption that the more
bland a policy document was, the wider the support base would be. What often
started off as a half-baked idea at a pressurised meeting must suddenly be
packaged by desperate (and quite often ignorant) political leaders to win
stakeholder support by government officials and advisors who were often plagued
by doubt and / or shallow mandates. But often it was too late to reverse the
process and all that could be done was to pretend that this was all part of a well-
conceived plan. (Swilling et al, 2005, p.18)

It may well be that in the second decade of freedom that the capacity constraints
(Joffe, 2005, p.29) that so hamstrung government in its earlier endeavours, will be

- 113 -
eased. Hamstrung though those efforts may have been, they did succeed in putting a
lot of infrastructure on the ground, albeit often of dubious quality. The portents are,
however, not good.

From a completely different starting point, Southall (2006) arrives at conclusions not
all that different from those reached by Swilling et al. The arguments of the three
different schools of thought that Southall considers are too complex to do justice to
here, so we must be content with a summary of his concluding thoughts. He starts
these by noting that:

while the aspiration for South Africa to become a developmental state is

understandable and in many ways admirable, it must nonetheless be regarded with
caution. Most certainly he argues we need to go well beyond the employment
of the term as a feel-good label if the concept of developmental state is to retain
any heuristic value. (2006, p.xli)

From this flow a number of conclusions, the first of which is that conditions in South
Africa are so different from those facing ruling elites in the countries that inspired the
developmental state literature; Japan, South Korea and Taiwan after World War II,
that no easy analogies can be drawn. A second conclusion is that the strengthened
public sector mooted as the development leader by AsgiSA:

does not add up to the planning rationality which saw developmental states
politically establishing economic goals, devising a far-reaching industrial policy
and directing the investment behaviour of their largest corporations.

To the extent that South Africa may be regarded as a developmental state, its focus is
almost entirely on internal strategies and arrangements: its external links being
hugely problematic (pp.xli-xlii). Next, although there are important lessons to be
learned from the developmental state experience elsewhere, about which analysts
(critics) of the South African governments differ, they appear to agree that the
capacity of the South African state is seriously limited, and that there is a tension
between its simultaneous pursuit of demographic representivity and efficiency
objectives (p.xlii). Finally, Southall raises the taboo question of whether
development may require constraints upon democracy (pp.xxxviii and xlii-xliii).
Although it is unlikely that hard-won political freedoms will easily be surrendered,
posing the question has the virtue of opening up a debate about what compromises
and coalitions may be necessary to embark upon such a project.

What to do until the state gets developmental?

Be all that as it may, resolution of the questions hanging over the debate on how to
eradicate poverty, is becoming a matter of increasing urgency. Poverty has
undoubtedly declined, yet as Louw (2006) has pointed out, resentment has risen. She
suggests three reasons why an improvement in living conditions observed at the
individual level (in poverty analysis) has not translated into greater political and social
stability. They are:

the inequality divide has sharpened

grants are highly imperfect substitute for jobs, and as such, are resented

- 114 -
poverty perceptions are relative rather than absolute, so that poor individuals
feel worse off in relation to middle class (2006, p.86)

It is not clear that the available evidence provides unambiguous support for these
propositions, but for what it is worth, we look briefly at them. Views about inequality
in South Africa are contradictory. There is widespread aversion to it, but there is no
agreement about the means proposed for reducing it. Social actors split sharply along
racial lines when it comes to proposed solutions. Nor is there agreement about the
relative importance of inequality, as opposed to other social problems. Roberts
(2006) suggests that the concern about inequality among South Africas elite may be
traced to self-interest. 73 Their concern, he argues:

about the income disparity in the country might stem from the perception that
inequality breeds crime and poses a threat to property rights. Social consciousness
among elites could also explain their concern about the economic divide, but
research suggests that such social solidarity is poorly developed among most elites
in South Africa a situation complicated by the geographical, social, educational,
economic, and psychological distances between elites and the impoverished

Resistance to redistributive policies may arise from the fact that, unlike the poor,
many elites have not benefited directly from such interventions. Poor delivery,
corruption, nepotism and possibly a fear of a Zimbabwean-style redistribution of
assets may also have tainted the redistribution process. (Roberts, 2005; Roberts,
2006, p.117)

Turning once again to Everatt et al (2006, paragraph 168) we note that when it comes
to perceptions of the greatest threats to unity in South Africa, political and racial
issues ranked as most important, overshadowing all others. Scores for political were
41 and 33 in ISRDPs and URPs respectively, and for racial, 25 and 43. Presumably
the high score on racial in URPs reflects closer proximity to other, more privileged
population groups. Inequality registered only 11 points in ISRDPs and 9 points in
URPs, a somewhat surprising result, and one that at least raises questions about the
significance of inequality.

Louw has argued, as noted above, that grants give rise to resentment, allegedly
because they are imperfect substitutes for employment. Before a claim such as this
could be accepted, a lot more evidence in support of the proposition would have to be
forthcoming. It could well be that what is being picked up here is the resentment
expressed by people who feel stigmatised by their dependence on grants, or by people
whom the means testing procedures have excluded, on grounds that those so treated
regard as arbitrary. In short, it may the nature of the grants themselves, rather than the
fact that of dependence on grants which is the major irritant. It is well-known that
means-testing gives rise to ill-feeling, let alone the equally well-known welfare traps
which they create when the entry into paid employment places receipt of the grant at
risk. The literature on social grants is huge, with much of it being of a highly

Whatever the case, there is a surprising amount of agreement with the proposition that government
should provide work (Roberts, 2006, p.120)

- 115 -
conservative nature. The literature has much less to say about unconditional grants,
such as the proposed basic income grant.

Finally, as Rose (2006) has pointed out, the debate about relative deprivation,
reference groups and social comparisons is complex, with many unanswered
questions. Information on what the reference group for the poor in South Africa may
be (who they resent most, if they do, the middle class or the elite) is a little scarce.
The most probable split is along racial linesthere certainly does seem to be a
division of this sort when it comes to the measurement of satisfaction domains, with
Africans being least satisfied regardless of income position (Roberts, 2006, p.111). It
is possible that the South African Social Attitude Surveys could answer questions of
this sortclearly, if resentment continues to rise, then the need to answer such
questions becomes more acute. For the moment, it does not seem particularly helpful
to suggest that the middle class is the primary target of the resentments of the
poorthat surely cannot be where the major obstacles to redistribution are to be
found. The real story must be much more complex.

Conclusion: When all else fails, lower your standards

If one were to search for a single word to sum up the sense of most of what has gone
before, the word pessimism would be as good a contender as any. There are bright
spotstumbling out of the hurly-burly of the debate over what pro-poor growth
means has come an unexpected insistence by World Bank researchers (and very
senior ones at that) on the barrier that inequality constitutes for efforts to reduce
poverty. So, while it may remain true (as nobody has denied) that sustained growth is
required for poverty eradication, it has also been recognised that countries with high
inequality have to run merely to stand still. It has been acknowledged that social
grants are not mere dependence-creating machinesthat not only do they provide
sustenance when there is little else on which to live; they also foster human capital
accumulation (i.e., that grants constitute a wealth and not merely an income transfer).
The social grant story in South Africa is a huge success. 74 Although they have but
little effect on income poverty and on inequality, the assets and services provided by
government (sometimes lumped together under the heading of the social wage) have
made a substantial difference to the lives of millions. Apart from these successes, the
corners into which we have pried face some serious problems.

Avenues for further pro-poor redistribution, while not non-existent, are certainly no
freeways down which policymakers can coast to a state of bliss. Following
Adelzadeh 75 we can attempt to inject some rigour into a discussion of the limits to
redistribution by making the following distinctions: first, we may distinguish between
the limits related to the unwillingness of policymakers to adopt a given policy (e.g.,
a basic income grant, and limits related to the underlying workings of the South
African economy (e.g., the capital intensive pattern of investment). Doing so can help

Albeit with qualifications that arise because of perverse incentives designed into the system, e.g., the
possibility of obtaining foster care grants instead of child support grants, or the temptation to try for
disability grants because of the almost total absence of any form of social protection for the working-
age unemployed.
pers. comm. March 2007. The passage that follows is taken verbatim from Adelzadehs comments
on the first draft of the paper.

- 116 -
treat the limits differently in terms of expectations for change. It is also possible to
re-define aspects of the different aspects of limits to redistribution through pro-poor
growth in terms of constraints related to policy formulation and implementation
capacity, current structure of production and production relations (the accumulation
mode); international trade and finance. With that template to guide us, we can work
through the list of problem areas tackled in the section of the paper headed The scope
for reducing inequality in South Africa.

Modesty is called for in doing solittle of what has been said on each of the limit
areas that have been addressed, can lay any claim to originality. As a basis for
advocacy, the research work carried out above cannot be regarded as anything more
than a first step. It has wandered into fields inhabited by scholars with years of
experience, and by policymakers familiar with the problems and their potential
solutions. 76 Finding ways to offer advice (as opposed to issuing statements laced with
catchy sound-bites) is not going to be easy. The notion that an individual researcher
can range over so many fields and uncover much that is new is probably mistaken. It
is likely that the best use to which a report such as the present one can be put is as a
background document to a workshop in which a group of researchers and activists
brainstorm the problems discussed in the report. Using the framework suggested
above by Adelzadeh, it is possible that new ways of looking at the problems raised,
may emerge.

In education, changing from a situation where 80 per cent (or at any event, some large
proportion) of South African schools are said to be dysfunctional, to one where the
education system contributes to pro-poor growth requires the overcoming of many
constraints, not all of which lie within the capacity of those in the education sector to
influence. In the first place, although sound heretical to say it, it is not obvious that if
the quality of schooling received by the majority of kids were improved beyond all
expectation, that they would be able to obtain employment upon leaving school. In
considering the prospects for redistribution through human capital formation, it is
necessary to take into account the underlying workings of the economy, not too
mention the international context within which that economy functions.

There are two sets of actors in the classroom, pupils and teachers, they used to be
called, now they are learners and educators. Whatever name one gives them,
changing for the better what happens in the classroom is going to require substantial
social engineering. It is rumoured, for example, that one of the major stumbling
blocks to teacher improvement (if necessary, by getting rid of incompetent teachers)
is union obstructiveness. This may simply be management propaganda, but if it is
not wholly untrue, then improving the quality of teachers, if necessary, by dismissing
rotten eggs, should have a high priority assigned to it. As for the pupils, some of them
come from such poor households that the school feeding schemes constitute a major
attraction. Here are two examples:

The Department of Education, for example, has produced a substantial report (DoE, 2006a) using all
of the relevant surveys and censuses (LFSs, OHSs and the 2003 GHS) to look at the impact and
outcomes of education on South Africas population. While one might want to criticize the report for
lacking the critical edge that independent researchers bring to the subject, apart fro celebrating the
many achievements of the department, it also identifies a number of serious problems.

- 117 -
The families who send their children to the centre they struggle. You will find
they dont get grant and no one is working in the house. You will see the people
are suffering from hunger. You can even see the child enjoys to be at school more
than home because at least they do get something to eat. How can you ask them
for school fees? Other people are living in the worst situation. Sometimes the
children tell you how many days they sleep without getting food. You feel sad
after hearing these stories. (Caregiver, North West)

But you find when the school is near the closing date there are those who cry,
who do not want to go home because they only get meals at school. We serve two
meals here at school. Some have only these meals because of circumstances at
home, difficulties, so they do not want to go home. (Teacher, Eastern Cape)
(Cited in Philpott, 2004, p.96)

The poverty which occasions these pitiful accounts has not escaped anyones
attention. The Department of Education knows much (but not all) about what ails the
systemthe 2005 Country Status Report (referred to above) reports on progress and
offers a long list of the challenges still faced. There is not much, however, that the
department can do about poverty. It may be that there is scope for an analysis of the
education system that locates the problems identified in the Status Report within the
framework outlined above; for example, by relating it to the underlying workings of
the economy, (and thinking heretical thoughts). If the intention is to engage in
advocacy in the field of education then it is clear that such a course of action must be
backed by further researchonly by asking fresh questions (has the question about
the potential contribution of educational transformation to redistribution been posed
and answered?) can any advance be made. As the noseweek article (cited above)
suggests, other profitable lines of inquiry may exist in those areas where public and
private meet, especially in the field of procurement, where opportunities for robbing
both government and the poor appear to be plentiful. The constraint to be tackled in
that case would be that of implementation and governance.

Healthcare would also seem to offer some space for the development of a critical
perspective. One field of particular interest is that of the impact of the development
of the primary health care (PHC) system on tertiary health care. While it is accepted
that PHC is of paramount importance, the mode of implementation of the policies to
develop that part of the healthcare system appears to have inflicted severe damage on
tertiary institutions. The impact of this on the poor is likely to have been severe.
Implementation of the 2004 Modernisation of Tertiary Services report referred to by
Blecher and Harrison (2006) is noted as lagging behind. Although constructed in
consultation with professionals, some of the recommendations may be not be capable
of being implemented if medical departments lose the critical mass necessary to
maintain the requisite levels of scholarship. In the short- to medium-term, the wealthy
can avoid the problems this may pose for specialist servicesthe poor, who need
these services as much (or even more) than the wealthy, cannot. In the long-run,
medical services as a whole may suffer. The literature in this area seems a little thin,
as does that on the training of nurses and midwives. Occasional references to the
crisis in nursing may be found, but an admittedly brief search did not turn up much.
Here, one would be working in the fields of policy design and implementation.

- 118 -
Governments obdurate resistance to proposals to extend the social grant system, is as
far I can tell, based on very slender evidence. Treasury asserts, apparently on the
basis of a model that nobody outside has seen, that extensions to the system beyond
those already proposed, are unaffordable, and there is an end of the matter. In this we
have an example of limits to redistribution related to the unwillingness of
policymakers to adopt a given policy. Several steps need to be taken to address this
problem. In the first instance, it is necessary to recognise that it may be rational to
reject demands for extending the grant system. That, however, needs to be
established using the best evidence available. Politicians prejudices are not evidence
of anything other than the fact they are prejudiced. As in the other limits addressed
in this paper, there are numerous researchers working in the field. There are still
gaps, though, some of them quite large. Chief among these is the absence of the
multiplicity of (independently constructed) models necessary for arguing about the
likely outcomes of welfare policies of various kinds. Building such things is a big
undertaking, well beyond the capacity of individual researchersinstitutional support
is necessary for such a task. It may, however, be possible to find was of contributing
to a project whose time must come soon.

It is nothing short of amazing that in a country which boasts so often about

democracy, that government has been able to get away with rejecting the demands for
a basic income grant advocated by the Taylor Committee (DoSD, 2002), offering
nothing in its place but a series of empty statements about the building of a
comprehensive social security system on the one hand, and a couple of hundred
thousand temporary job opportunities annually through the Extended Public Works
Programme (EPWP), on the other. With the ANCs biennial conference looming, this
looks set to change. At the time of writing, a flurry of activities in two departments,
the National Treasury (NT) and the Department of Social Development (DoSD),
centred around the design of comprehensive social security policies, is taking place.
Presumably this activity is not unconnected with anticipated questions at the
conference (from those of populist bent?) about the success or otherwise, of
governments anti-poverty programmes. It is not obvious what sort of demands
ought, at this point to be made. Some well-placed questions in Parliament about why
it has taken so long to address the issue of a truly comprehensive social protection
system might not go amiss.

A gut response to the issue of Black Economic Empowerment is that the less said
about BEE (some folk call it Black Economic Enrichment, and not without good
cause), the better. As pointed out above, however, this is something (especially the
part about making it genuinely broad-based) that has to be got right. This is easier
said than done. BEE policy, if Cargill (2005) is to be believed, creates structures that
undermine incentives to engage in entrepreneurial activity (a difficult enough thing to
foster anyway). Critical scholarship in this field is not lackingwhat is in relatively
short supply is criticism from more middle-class black people.77 Exploration of the
limits to entry of a would-be black capitalist class, given the existing structures of
the economy, and the harsh international context within which it operates, if it has not
already been done, would probably be a worthwhile research project. Looking more
deeply into the drive to create (rapidly) a black capitalist class, the deployment of a

The Communist Party, as we saw above, has articulated strong criticisms of the BEE and BBBEE
endeavours. It is not known how much effect, if any this has had on the ANC.

- 119 -
public choice theoretic analytical framework (Cullis and Jones, 1992) could enhance
the good work done by Cargill (2005) on understanding the vicious circle to which
this imperative has given rise. It looks as though there is involuntary collusion of the
state with voracious would-be rich (responding, as economic theory would predict, to
a set of incentives which encourage such behaviour) uniting, it could be argued, to
make policy that is both flawed and incapable of being implemented. What one
would do with such insights, if evidence to support the hypothesis offered above were
forthcoming, is, of course, another matter altogether. In its first Policy Brief (October
2006), the Senior Policy Seminar of the African Economic Research Consortium
(AERC) argued that:

good governance is the most important factor in poverty alleviation efforts in


Issues of governance are raised at every turn by BEE and its relative, Broad-based
Black Economic Empowerment (BBBEE). So too are questions of the relationship of
the initiative to pro-poor growth.

Employment creation, as a potential field for activist intervention, offers little scope.
Everybody agrees that jobs are requiredmany researchers have devoted energy to
the question of where such jobs might be created. Despite the concentration of high-
powered brains on the job, the Harvard team, one of whose tasks it is to discover to
the constraints and opportunities confronting the states AsgiSA has not, so far,
offered any earth-shattering advice on the matter. There is a whiff of disagreement, as
we noted, among the analysts that does not bode well for policy formulation. The
unemployment analysis of which the paper by Banerjee et al (2006) is the first fruit,
has not yet reached the point where firm policy recommendations can be offered. The
only suggestion to emerge from the admittedly cursory analysis in the present paper of
the employment creation problem, is that government be more careful in its use of
employment statistics, hardly the sort of advice to which politicians are likely to pay
any heed.

As far as regional inequities are concerned, there is room to engage critically with the
ISRDPs and URPs, a task that has been given a massive boost by the recent report
compiled by Everatt et al (2006). In a document that contains many surprises (and
confirms many suspicions) one of the more surprising findings was that when asked
what the main problems were facing communities in some of the poorest areas of the
country, crime was first in both the rural and urban settings, with poverty a distant
third or fourth. In the URPs, 43 per cent of respondents placed crime in first position,
with HIV/AIDS in second place (19 per cent of respondents), housing in third (eight
per cent), with poverty and child abuse joint fourth (seven per cent). Corresponding
figures for the ISRDPs were crime: 31 per cent, HIV/AIDS: 15 per cent). Poverty
was third at 18 per cent of respondents (Everatt et al, 2006, paragraph 246, Figure 49).
The proportion of people who had been victims of crime was simply astonishinga
little less than one in five (18 per cent) in URPs and one in ten (nine per cent) in
ISRDPs (paragraphs 241 and 242). Corresponding to this is a massive sense of
insecurity50 per cent of those living in Khayelitsha felt very unsafe (paragraph
240). In this, as in so many other aspect of illfare, the burden is unequally distributed.
The wealthy have the resources to purchase security (or to live in secure areas), most
of the poor do not. How much crime inhibits people in the pursuit of income-

- 120 -
generating activities is not easy to determine, but the effect is unlikely to be small.
An area of regional disparity that seems worth engaging with is thus that of crime and
personal security. One of Klasens (2003) research findings is that Physical and
social security [are] essential for pro poor growth (see Table 3 above), a conclusion
with which it is impossible to disagree. South Africas crime problem is typical of
those experienced by countries undergoing transitions, both the policies designed to
deal with it, and the means of implementing those policies are problematic. The
debate in South Africa has, however, taken a very unfortunate turn, at least in part
because the most vociferous critics of governments attempts to deal with crime are
those with the highest incomes. If the debate could somehow be shifted to focus more
on the everyday sufferings of the poor, in a way that improves their lot, it would
probably be the single most significant pro-poor investment that could be made.

Finally, we come to the question of governments illusions (delusions) about its

capacity to function as a developmental state. Two pieces of work, from very
different quarters and not obviously linked to the developmental state debate, will be
used to back up a request for greater realism on the part of government when it comes
to an evaluation of what can, and what cannot be achieved. The papers in question
are by Hausmann (2006) and Wright (2006). From the piece by Wright (a challenging
paper which all those interested in creating alternatives to the virulent capitalism that
dominates our lives should read) I wish to refer to his discussion of an emancipatory
social science (2006, pp.94ff). There are three basic tasks such a science faces in
fulfilling its mission:

first, to elaborate a systematic diagnosis and critique of the world as it exists;

second, to envision viable alternatives; and third, to understand the obstacles,
possibilities and dilemmas of transformation.

Hardly an original insight, one would have thought: surely this is what social policy is
all about? Sometimes, however, it pays to go back to first principles. This would
certainly appear to be the case as far as the question of the developmental state is
concerned. Although the abstract idea of such a state may be noble, springing as it
does out of a view of social justice with which all progressives can identify, it is in
the attempts to operationalise the concept that we begin to see how far government is
from having taken even the first of the steps outlined by Wright. Although there are
areas of agreement, there is no systematic diagnosis and critique that finds broad
acceptance (opposing views within the National Treasury and the Department of
Social Development about what the shape of South Africas social security system
should be are evidence of this). The Southall (2006) piece on the developmental state
makes it quite clear that there are conflicting views on what the thing is supposed to
be. This means that viable alternatives cannot be envisioned. Although a lot of work
has been done that attempts to understand the obstacles, possibilities and dilemmas of
transformation, in the absence of a shared vision (compare, for example, the views of
the Communist Party with those of the ANC in government, on a range of issues),
such understandings must remain elusive. Until these shortcomings are resolved, the
danger to which Southall points of the concept being used as a feel-good label is
surpassed only by the danger of its being used as a cloak behind which to hide an
unwillingness to act at all.

- 121 -
A completely different animal, the paper by Hausmann (2006) provides a timely
warning against the dangers of slipping into the belief that not only must the
developmental state achieve the (still vague) outcomes it promises, but that in
addition, it must also have done everything else in Table 3 above. It is worth citing at
length from the paper, which was presented to an audience made up of high-ranking
officials from the G-20 countries in Pretoria. Among them were South Africas
Minister of Finance, and the Governor of the Reserve Bank. Taking issue with the
kind of advice offered by international financial institutions, he said that:

The typical message that countries have been getting is clear: do everything.
This involves coming up with a long laundry list of necessary reforms, like the one
implied by the three strategies mentioned above. There is typically little sense of
priorities or optimal sequencing and the implied message from Larry Summers is
that you cannot overdose on reforms while the message from Anne Krueger is not
to expect too much until all reforms are in place.

There is an implied assumption in this agenda. It is that any reform is good at any
time; the more areas reformed, the better and the deeper the reform in any area, the

The only problem with this approach is that economic theory provides no reason
to think any of the above is true. And the reason is the so-called theorem of the
second-best, which says that if you have an economy with several distortions,
eliminating one of them is not necessarily welfare-enhancing. If you eliminate
one distortion and you still have the rest of the distortions in place, you can make
things better or worse. (2006, p.7)

Having worked his way through what works (in theoretical and empirical terms),
Hausmanns advice is:

if countries face low growth, focus on the binding constraints. That would be
the paradigm that emerges here. This is in essence an optimistic message. It says
your task may be easier than you think. (2006, p.8)

So far, the Harvard team gets an A for effort, and probably a less-elevated score for
identifying the binding constraints (on the would-be developmental state). It may
well that the constraints preventing the achievement of the growth necessary to
eradicate poverty in one of the most unequal countries in the world are so numerous
and so severe, that we cannot reasonably expect them to be removed. If that turns out
to be the case, we are back with Wrights questionwhat are the viable alternatives?
If growth cannot eradicate poverty fast enough, the question of what can alleviate it in
the meanwhile (more social grants?) needs to be faced with greater openness of mind.

We end with a warning from a paper on Brazil by de Oliveira (2006). Brazil and
South Africa share a number of unfortunate characteristics, notably that of extreme
inequality. Look at the way de Oliveira described Brazil in an earlier paper: he said it
was a:

social formation emerging from the mutations of an industrialized, semi-

peripheral economy, under the pressures of capitalist globalization, privatization

- 122 -
and the Third (molecular-digital) Industrial Revolution, as a duck-billed platypus:
part mammal, part bird. The creature combines external dependency with
casualized labour, truncated accumulation with an unremittingly inegalitarian
social order; despite thirty years of democratization, its level of consciousness
remains doubtful. Among its most conspicuous features is a new social class,
defined by its access to and control over public funds. (2006, p.10)

This is not South Africa, and some of the differences are very important. So too,
however, are the similarities. The de Oliveira paper ends, pessimistically, with a
direct reference to South Africa, which, while overstating the case for an alleged
surrender to neoliberalism, nonetheless seems uncomfortably close to the truth in its
claim that:

South Africa was probably the first announcement [of a new paradigm] in which
the oppressed appear to assume the moral leadership of society, while capitalist
relations become ever more brazen. (2006, p.21)

The de Oliveira paper is a strong argument in favour of a national-developmentalist

project, and far from lauding the Bolsa Famlia, as is so often done, points instead to
the effect it has of depoliticizing poverty, turning inequality into an administrative
problem (2006, p.22).

If the South African government lacks the capacity to act as a developmental state,
and if AsgiSA, which as government is keen to remind us:

is not a government programme. It is a national initiative supported by key

groups in the economy business, labour, state-owned enterprises (SOEs),
government economic agencies, entrepreneurs and all spheres of government. 78
(The Presidency, 2007, p.4)

does not deliver on the social goals, will the grant system depoliticize poverty? If
only government were less strident about its aim to be a developmental state, and
instead tailored its activities to match its competencies; if only government were less
vehemently opposed to doing that which will certainly alleviate poverty, namely to
extending the social grant system in a democratic way, Brazilianization might be less
of a possibility.

Where are the representatives of the poor in all of this?

- 123 -

African Economic Research Consortium (AERC). Policy Brief, No. 1, October


Aghion, Philippe; Braun, Matias, and Fedderke, Johannes. Competition and

Productivity Growth in South Africa, Center for International Development at
Harvard University (CID) Working Paper No. 132, August 2006.

Alderman, Harold; Babita, Miriam; Lanjouw, Peter; Makhata, Nthabiseng;

Mohamed, Amin; zler, Berk, and Qaba, Olivia. Combining census and poverty
data to construct a poverty map of South Africa, in Hirschowitz (ed), 2000, pp.5-52.

Altman, Miriam. Elements of employment scenario building: do sources of job

creation matter?, unpublished paper prepared for Employment growth &
development initiative, HSRC, Pretoria, November 2006.

Ahluwhalia, Montek and Chenery, Hollis. Redistribution with Growth, Oxford:

Oxford University Press, 1974. Cited in Klasen (2003).

Babita, Miriam, zler, Berk; Shabalala, Nozipho and Thema, Harry. Changes
in Poverty and Inequality in South Africa: 1995-2000, Unpublished manuscript,
Statistics South Africa, 2003. Cited in Hoogeveen and zler, 2004.

Banerjee, Abhijit ; Galiani, Sebastian, Levinsohn, Jim and Woolard, Ingrid.

Why Has Unemployment Risen in the New South Africa? , Center for International
Development at Harvard University (CID) Working Paper No. 134, October 2006.

Benatar, Solomon R. Health Care Reform and the Crisis of HIV and AIDS in
South Africa New England Journal of Medicine, 351(1), July 2004, pp.81-92.

Berry, Albert. Employment and income distribution experiences of minerals

exporters and of countries achieving growth acceleration, unpublished paper
prepared for Employment growth & development initiative, HSRC, Pretoria, August

Bhalla Surjit S. Pro-Poor Growth: Measurement and Results, Oxus Research &
Investments, New Delhi, March 15, 2004.

Bhorat, Haroon, Naidoo, Pranushka and van der Westhuizen, Carlene. Shifts in
Non-Income Welfare in South Africa: 1993-2004, DPRU Working Paper 06/108,
Development Policy Research Unit, University of Cape Town, May 2006.

Blecher, Mark and Harrison, Stephen. Health Care Financing in South African
Health Review 2006, Durban: Health Systems Trust, October 2006, pp.31-64.

Bhuwanee, Karishmah. Pro Poor Policy: A Selected Annotated Bibliography,

Overseas Development Institute, July 2006.

- 124 -
Bourguignon, Franois and Chrakravarty, Satya A. The measurement of
multidimensional poverty, Journal of Economic Inequality, 1, March 2003, pp.25-49.

Bourguignon, Franois. The Poverty-Growth-Inequality Triangle, paper presented

at the Indian Council for Research on International Economic Relations, New Delhi,
February 2004.

Brutigam, Deborah A. The Peoples Budget? Politics, Power, Popular

Participation and Pro-Poor Economic Policy, International Development Program,
School of International Service, American University, Washington, DC, March 29th,

Brown, Susan (ed). Taking Power in the Economy. Economic Transformation Audit
2004, Cape Town: Institute for Justice and Reconciliation, 2004.

Brown, Susan (ed). Conflict and Governance: Economic Transformation Audit

2005, Cape Town: Institute for Justice and Reconciliation, 2005.

Brown, Susan (ed). Money and Morality: 2006 Transformation Audit, Cape Town:
Institute for Justice and Reconciliation, 2006.

Buhlungu, Sakhela; Daniel, John; Southall, Roger and Lutchman, Jessica (eds).
State of the Nation: South Africa 2005-2006, Cape Town: HSRC Press, 2006.
Referred to as Buhlungu et al (eds), 2006.

Calland, Richard. Anatomy of South Africa: Who Holds the Power?, Cape Town:
Zebra Press, 2006.

Cargill, Jenny. Black corporate ownership: Complex codes can impede change, in
Brown (ed), 2005, pp.21-27.

Cord, Louise; Lopez, J. Humberto and Page, John. When I Use A Word
ProPoor Growth and Poverty Reduction, Poverty Reduction Group, The World
Bank, August 2003 (Paper No. 32886).

Cramer, Christopher. Inequality, Development and Economic Correctness,

Department of Development Studies, School of Oriental and African Studies,
University of London.

Cronin, Jeremy. Liberate South Africa from anti-public discourse, SA

Reconciliation Barometer, Vol. 4, Issue 3, November 2006, pp. 8-9.

Cousins, B. The Zimbabwe crisis in its wider context: The politics of land,
democracy and development in southern Africa. Unpublished paper, Programme for
Land and Agrarian Studies, University of the Western Cape, July 2003. Cited in
Gran, 2006.

Cullis, John and Jones, Philip. Public Finance and Public Choice: Analytical
Perspectives, London: McGraw-Hill Book Company, 1992.

- 125 -
Dagdeviren, Hulya; van der Hoeven, Rolph, and Weeks, John. Redistribution
Matters: Growth for Poverty Reduction, Working Paper Series No. 99, Department
of Economics, School of Oriental and African Studies, University of London, October

de Oliveira, Francisco. Lula in the Labyrinth, New Left Review, 42,

November/December 2006, pp.5-22.

Department of Education. Monitoring and Evaluation Report on the Impact and

Outcomes of the Education System on South Africas Population: evidence from
household surveys, Pretoria, September 2006a.

Department of Education. School realities 2006, Pretoria, October 2006b.

Department of Education. Annual Report 2005-2006, Pretoria, August 2006c.

Department of Education. Education For All 2005 Country Status Report:

SOUTH AFRICA, Pretoria, April 2005.

Department for International Development (DfID). What is pro-poor growth and

why do we need to know?, ProPoor Briefing Note 1, February 2004.

Department for International Development (DfID). How to accelerate pro-poor

growth: a basic framework for policy analysis?, ProPoor Briefing Note 2, September

Department of Social Development. Transforming the Present Protecting the

Future: Report of the Committee of Inquiry into a Comprehensive System of Social
Security for South Africa, (Consolidated Report), Pretoria, 2002 (Referred to as the
Taylor Report, or DoSD, 2002)

Dollar, David and Kraay, Aart. Growth is Good for the Poor, World Bank Policy
Research Working Paper 2587, April 2001 [2000].

Douthwaite, R. Good growth and bad growth, downloaded from,
version dated May 1997 with subsequent minor revisions. Cited in May, 2004.

Easterly, William. The Effect of International Monetary Fund and World Bank
Programs on Poverty, World Bank Policy Research Working Paper 2517, 2001.

Easterly, William. The Elusive Quest for Growth: Economists Adventures and
Misadventures in the Tropics, Cambridge, Massachusetts: The MIT Press, 2002.

Edwards, Lawrence and Lawrence, Robert. South African Trade Policy Matters:
Trade Performance & Trade Policy, Center for International Development at Harvard
University (CID) Working Paper No. 135, October 2006.

Ensor, Linda. Creative Empowerment: Genuinely Broad-Based, For A Change, in

Brown (ed), 2004, pp.23-25.

- 126 -
Everatt, David; Smith, Matthew J, and Solanki, Geetesh. Baseline survey of the
21 ISRDP and URP nodes: Topline report and data tables, Pretoria: Department of
Social Development, November 2006. (Downloaded from, 9th February 2007)

Filho, Joaquim Bento de Sousa Ferreira and Horridge, J Mark. Economic

Integration, Poverty and Regional Inequality in Brazil, Centre of Policy Studies and
the Impact Project, Monash University, General Working Paper No. G-149, July

Flscher, Alta. Education: Expenditure, quality, in Brown (ed), 2006, pp.61-64.

Frankel, Jeffrey; Smit, Ben and Sturzenegger, Federico. South Africa:

Macroeconomic Challenges after a Decade of Success, Center for International
Development at Harvard University (CID) Working Paper No. 133, September 2006.

Gran, Thorvald. Land politics, trust relations in government and land reform in
South Africa: Experiences from the Western and Northern Cape provinces,
Programme for Land and Agrarian Studies (PLAAS), University of the Western Cape,
Occasional Paper No. 29, April 2006.

Hardstaff, Peter. From GATS to New WTO Investment Rules Undermining Pro-
Poor Investment Regulation, Paper by Head of Policy, World Development
Movement (WDM) to one-day Symposium on Investment, Sustainable Development
and the WTO, Washington, May 22 2003.

Hausmann, Ricardo. Economic Growth: Shared Beliefs, Shared

Disappointments?, Center for International Development at Harvard University
(CID) Working Paper No. 125, June 2006

Hirschowitz, Ros (ed). Measuring poverty in South Africa, Pretoria: Statistics South
Africa, 2000.

Joffe, Hilary. Divisions about the developmental state, in Brown (ed), 2005,

Kakwani, Nanak; Khandker, Shahid and Son, Hyun H. Pro-Poor Growth:

Concepts and Measurement with Country Case Studies, International Poverty Centre,
Working Paper No. 1, UNDP, Brazil, August 2004.

Kakwani, Nanak and Son, Hyun H. On Assessing the Pro-Poorness of

Government Programmes: International Comparisons, International Poverty Centre,
Working Paper No. 6, UNDP, Brazil, May 2005.

Kakwani, Nanak; Neri, Marcelo and Son, Hyun H. Linkages Between Pro-Poor
Growth, Social Programmes and Labour Market: The Recent Brazilian Experience,
International Poverty Centre, Working Paper No. 26, UNDP, Brazil, August 2006.

- 127 -
Kanbur, Ravi. Economic Policy, Distribution and Poverty: The Nature of the
Disagreements, Cornell University,
draft dated January 2001.

Klasen, Stephan. In Search of the Holy Grail: How to Achieve Pro-Poor Growth,
Ibero-America Institute for Economic Research, Georg-August-Universitt Gttingen,
Discussion Paper No. 96, October 2003.

Lal, Deepak and Myint, H. The Political Economy of Poverty, Equity and Growth:
A Comparative Study, Oxford: Clarendon Press, 1996.

Left Business Observer. World Bank news 95, November 2000.

Leite, Phillippe G; McKinley, Terry and Osorio, Rafael Guerreiro. The Post-
Apartheid Evolution of Earnings Inequality in South Africa, 1995-2004,
International Poverty Centre, Working Paper No. 32, UNDP, Brazil, October 2006.

Lopez, J Humberto. Pro-growth, pro-poor: Is there a tradeoff?, World Bank

Policy Research Working Paper 3378, August 2004

Louw, Megan. Poverty Declines, Resentment Rises, in Brown (ed), 2006, pp.84-

Lbker, Malte; Smith, Graham, and Weeks, John. Growth and the Poor: A
Comment on Dollar and Kraay, unpublished paper, Centre for Development Policy
and Research & The Department of Economics, School of Oriental & African Studies,
University of London, October 2000.

Loayza, Norman and Raddatz, Claudio. The Composition of Growth Matters for
Poverty Alleviation, World Bank Policy Research Working Paper 4077, December

MacDonald, Michael. Why Race Matters in South Africa, Scottsville,

Pietermaritzburg: University of Natal Press, 2006.

May, J. Constructing the social policy agenda: conceptual debates around poverty
and inequality, In Padayachee, V. (ed), 2006, pp.143-159.

Medeiros, Marcelo. Poverty, Inequality and Redistribution: A methodology to

Define the Rich, International Poverty Centre, Working Paper No. 18, UNDP, Brazil,
May 2006.

Meth, Charles. Income poverty in 2004: A second encounter with the recent van
der Berg et al figures, Working Paper No. 47, School of Development Studies,
University of KwaZulu-Natal, September 2006a.

Meth, Charles. Half measures revisited: The ANCs unemployment and poverty
reduction goals, forthcoming in Poverty and Policy in Post-Apartheid South Africa,
edited by Haroon Bhorat and Ravi Kanbur, Cape Town: HSRC, 2006b.

- 128 -
National Treasury. Budget Review 2006, Pretoria, February 2006a.

National Treasury. Estimates of National Expenditure 2006, Pretoria, February


National Treasury. Budget Review 2007, RP 02/2007, Pretoria, February 2007.

Noble, Michael; Babita, Miriam; Barnes, Helen; Dibben, Chris; Magasela,

wiseman; Noble, Stefan; Ntshongwana, Phakama; Phillips, Heston; Rama,
Sharmla; Roberts, Benjamin; Wright, Gemma and Zungu, Sibongile. The
Provincial Indices of Multiple Deprivation for South Africa 2001, University of
Oxford, UK, March 2006a. (Referred to as Noble et al, 2006a).

Noble, Michael; Babita, Miriam; Barnes, Helen; Dibben, Chris; Magasela,

wiseman; Noble, Stefan; Ntshongwana, Phakama; Phillips, Heston; Rama,
Sharmla; Roberts, Benjamin; Wright, Gemma and Zungu, Sibongile. The
Provincial Indices of Multiple Deprivation for South Africa 2001: Technical Report,
University of Oxford, UK, March 2006b. (Referred to as Noble et al, 2006b).

Oxfam. Growth with equity is good for the poor, [A response to Dollar and Kraay,
2000], Oxford, June 2000.

Padayachee, V (ed). The Development decade: Economic and social change in

South Africa 1994-2004, Cape Town: HSRC Press, 2006.

Philpott, Sue. Budgetting for children with disabilities in South Africa, Cape Town:
Idasa, 2004.

Piketty, Thomas and Saez, Emmanuel. The Evolution of Top Incomes: A

Historical and International Perspective, NBER Working Paper No. 11955,
Cambridge, Massachusetts: National Bureau for Economic Research, January 2006.

Pillay, Udesh; Roberts, Benjamin, and Rule, Stephen. South African Social
Attitudes: Changing Times, Diverse Voices, Cape Town: HSRC Press, 2006.
Referred to as SASA, 2006.

Pollin, Robert; Epstein, Gerald; Heintz, James, and Ndikumana, Lonce. An

Employment-Targeted Program for South Africa, New York and Brasilia: UNDP,

Ravallion, Martin and Datt, Gaurav. When Is Growth Pro-Poor? Evidence from
the Diverse Experiences of Indiass States, Policy Research Working Paper 2263,
World Bank, December 1999.

Ravallion, Martin. Growth, Inequality and Poverty: Looking Beyond Averages,

World Bank Policy Research Working Paper, 2558, February 2001.

Ravallion, Martin and Chen, Shaohua. Measuring pro-poor growth, World Bank
Development Research Group Working Paper No. 2666, August 2001.

- 129 -
Ravallion, Martin. Pro-Poor Growth: A Primer, Development Research Group,
World Bank, 2004.

Ravallion, Martin. Looking beyond Averages in the Trade and Poverty Debate,
World Bank Policy Research Working Paper 3461, November 2004

Ravallion, Martin. Comment on Pro-Poor Growth: What is It? One Pager No.1,
International Poverty Center, UNDP, September 2004. No date. Referred to
Ravallion, nd.

Ravallion, Martin. Inequality is Bad for the Poor, World Bank Policy Research
Working Paper 3677, August 2005a

Ravallion, Martin. A Poverty-Inequality Trade-off?, World Bank Policy Research

Working Paper 3579, April 2005b

Rodrik, Dani. Understanding South Africas Economic Puzzles, NBER Working

Paper Series, Working Paper 12565, October 2006.

Roberts, Benjamin. The happy transition? Attitudes to poverty and inequality after
a decade of democracy, in Pillay et al (eds), 2006, pp.101-130.

Roberts, Ben. Confronting the Divide: Attitudes to inequality, Downloaded from
ntent 1 February 2007. Referred to as Roberts, 2005.

Rose, David. Social Comparisons and Social Order: Issues Relating to a Possible
Re-study of W.G. Runcimans Relative Deprivation and Social Justice, Institute for
Social and Economic Research (ISER), University of Essex, Working Paper, 2006-48.

Seekings, Jeremy. The Broader Importance of Welfare Reform in South Africa,

Social Dynamics, Winter 2002, 28(2), pp.1-38.

Son, Hyun H and Kakwani, Nanak. Global Estimates of Pro-Poor Growth,

International Poverty Centre, Working Paper No. 31, UNDP, Brazil, October 2006.

Son, Hyun H and Kakwani, Nanak. Global Estimates of Pro-Poor Growth,

International Poverty Centre, Working Paper No. 31, UNDP, Brazil, October 2006.

Soutien, Crain. Education: Wrestling with legacy, in Brown (ed), 2005, pp.59-61.

Southall, Roger. Introduction: can South Africa be a developmental state?, in

Buhlungu et al (eds), 2006, pp.xvii-xlv.

Statistics South Africa. Census 2001: Investigation into appropriate definitions of

urban and rural areas for South Africa: Discussion document. Pretoria: Statistics
South Africa, [Report No. 03-02-20 (2001)], 2003.

Steele, Mark. Report on Incentive Structures of Social Assistance Grants in South

Africa, presented by Kesho Consulting and Business Solutions (Pty) Ltd, contracted

- 130 -
by GeoSpace International, and commissioned by The Department of Social
Development, July 2006.

Swilling, Mark; van Breda, John; van Zyl, Albert and Khan, Firoz. Economic
Policy-Making in a Developmental State: Review of the South african Governments
Poverty and Development Approaches, 1994-2004, Research Report 3, Economic
Policy and Poverty Alleviation Report Series, Centre for Civil Society, no date, but
apparently published in 2005. Referred to as Swilling et al, 2005.

Taylor, Nick. School Reform and Skills Development, in Brown (ed), 2006,

The Presidency. Towards a Ten Year Review: Synthesis Report on Implementation of

Government Programmes, Policy Co-ordination and Advisory Services, (PCAS),
referred to as (PCAS, 2003).

The Presidency. A Nation in the Making: A Discussion Document on Macro-Social

Trends in South Africa, Policy Co-ordination and Advisory Services, (PCAS),
referred to as (PCAS, 2006).

The Presidency. Annual Report 2006. Accelerated and Shared Growth Initiative for
South Africa (ASGISA), no date, but probably 2007. Referred to as (The Presidency,

van der Berg, Servaas. The schooling solution: Primary school performance is the
key, in Brown (ed), 2005, pp.62-70.

van der Berg, Servaas, Burger, Ronelle; Burger, Rulof; Louw, Megan and Yu,
Derek. Trends in poverty and inequality since the political transition, Stellenbosch
Economic Working Papers, No. 1 / 2005. Referred to as van der Berg et al, 2005.

Wade, Robert. Showdown at the World Bank, New Left Review, 7, January/February
2001, pp.124-137.

Weeks, John and Rathin, Roy. Making fiscal policy work for the Poor, paper
presented at the Fifth Annual Conference of the Global Development Network, New
Delhi, January 2004.

Watt, Patrick. Social Investment and Economic Growth: A Strategy to Eradicate

Poverty, Oxfam Insight Series, 1999.

Wolfensohn, James D. The Challenge of Inclusion, Address to the Boards of

Governors of the World Bank Group, at the Joint Annual Discussion with the
International Monetary Fund, September 1997, Hong Kong.

World Bank. Reducing Poverty, Sustaining Growth: Scaling up poverty

reduction, Case study summaries from A Global Learning Process and Conference in
Shanghai, May 2004.

- 131 -
World Bank. World Development Report 2006: Equity and Development,
Washington: The World Bank, 2005.

World Bank. Pro-Poor Growth in the 1990s: Lessons and Insights from 14
Countries, (jointly with Agence Franaise de Dveloppement, Bundesministerium fr
Wirtschaftliche Zusammenarbeit und Entwicklung, U.K. Department for International
Development), June 2005.

Wright, Erik Olin. Compass Points: Towards a Socialist Alternative, New Left
Review 41, Sept/Oct 2006, pp.93-124.

Zepeda, Eduardo. Pro-poor growth: what is it?, International Poverty Centre, One
Pager, Number 1, UNDP, Brazil, September 2004.

- 132 -
Pro-poor Trade and Industry
Policies: Considerations for
South Africa

Mohammed Jahed
Associate Professor Graduate School Of Public and Development
Management, University of The Witwatersrand

Prepared for

Oxfam GB South Africa

24 May 2007

- 134 -
Table of contents

Title Page

1. Introduction 2
2. Research on South Africas Trade and Industrial Policies 2
3. Background 4
3.1 Definition of Concepts 6
3.1.1 Pro Poor and Growth 6
3.1.2 Trade and Industrial Policies 13
4. Elements of Pro Poor Trade and Industrial Policies 14
5. Experience of Developing Countries with Pro Poor Policies 16
6. Pertinent Issues for South Africa 20
Conclusion 22

1. Introduction
Developing and developed economies seem to gravitate towards both ends of the
spectrum on this issue. This implies that developing countries need to determine their
own trade and industrial policies in their specific country context that socially and
economically benefit their poor constituents.

South Africa defines itself as a developmental state (Edigheji, 2005; Buhlungu et al

2006) and emerging from an oppressive regime of apartheid, its government actively
seeks to half its poverty levels in line with Millennium Development Goals and integrate
itself into the global market with the aid of trade and industrial policies that benefit the
country individually and its trading partners.

This paper provides an overview of the development of pro-poor trade and industrial
policies and raises pertinent issues for South Africa as a developing country. The paper is
broadly divided into five sections. Section 2 broadly contextualises the study by broadly
reviewing research on South Africas trade and industrial policies. Section 3 examines
the key concepts of pro poor; trade; and industrial policies that are used in the paper;
Section 4 provides the elements of a pro poor trade and industrial polices. Section 5
outlines the experiences of developing countries with trade and industrial policies and the
resultant effect on the poor, otherwise termed as pro poor. Section 6 provides a
conclusion by raising pertinent issues for South Africa to consider in developing pro poor
trade and industrial policies.

2. Research on South Africas trade and industrial policies

Recent studies performed on South Africas trade and industrial policies reveal that these
policies have had an adverse effect on the countrys poor. Edwards & Stern (2006) reveal
that the poor have not benefited in terms of employment. This is partly because poor
households are largely disconnected from the formal wage economy and labour income
in the traded sectors. In addition, the study notes that economic and export growth has
been insufficient to draw new entrants into the labour market, which points to the
conclusion that trade growth has not been able to reduce unemployment and poverty
especially amongst the unskilled and rural poor. Hoogeveen & Ozler (2004) draw similar
conclusions and reiterate that South Africa has failed to generate pro-poor growth.

Bhorat (1999) finds that increased trade during the 1990s only benefited skilled labour,
with lower-skilled employment declining. Edwards (2002) establishes that the effect of
increased trade was to raise the skill-intensity of production. Edwards (2003) concludes
that trade-induced technological change explains some of the shift towards skill-
intensive production and falling unskilled labour employment. Trade liberalizations bias
towards higher-skilled labour may be due to the rising capital-intensity of production
that took place during the 1990s. Edwards (2003) finds that firms affected by trade

liberalization invested more heavily in capital equipment. This corroborates observed
labour trends, since increased investment has been found to be associated with a rising
skill intensity of employment (Fedderke et al., 2003).

Thurlow (2006) determines the coexistence of substantial trade liberalization and rising
poverty and inequality. This raises concern that trade policies may have worked against
the countrys development objectives and questions the need for further liberalization.
While the evidence suggests that higher-skilled workers have benefited more than lower
skilled workers, the high level of unemployment in South Africa makes it difficult to
infer the effects of trade on the distribution of household incomes and poverty. This study
examines the relationship between trade liberalization and pro-poor growth in South

The findings suggest that liberalization has not contributed to the rise in poverty during
the 1990s. Rather it has prevented many households from falling into poverty. This is
because increased openness has fostered higher levels of investment and employment by
reducing the cost of capital goods and facilitating productivity-enhancing technological
change. However, liberalization has not benefited all workers and households equally.
High levels of unemployment and inadequate human capital meant that poor households
have been disconnected from most of the benefits of liberalization. Furthermore, rising
import competition has contributed to the fall in manufacturing employment during the
1990s. While this has been more than offset by improved employment opportunities in
the non-manufacturing sectors, the associated short-term adjustment costs may have
increased the vulnerability of the poor and undermined their ability to participate in
subsequent trade-induced growth (Thurlow, 2006).

This study has shown that, while trade liberalization has helped establish a faster-
growing economy that creates employment, it has not encouraged a redistribution of
incomes in favour of the poor. Therefore, while there may not be a trade-off between
trade liberalization and pro-poor growth, the country should not rely on further
liberalization to reduce poverty or address inequalities between different population
groups. In this regard, the government should engage more heavily in targeted pro-poor
strategies, such as public works programs and social assistance. These interventions are
more likely to reach poor and vulnerable households.

Furthermore, liberalization may cause some people to become structurally unemployed,

especially older and lower-skilled workers. Therefore, the countrys development
strategy should address the adjustment costs associated with trade reforms by increasing
its emphasis on social protection and job retraining. Without these explicitly pro-poor
interventions, the governments export-led growth strategy will be insufficient to achieve
the countrys development objectives (Thurlow, 2006).

Research on South Africas textile industry reveals a negative effect has been
experienced in recent years by textile industries policies in South Africa (see for example,
Wood, 2001). South Africa is in the process of developing its industrial policy and its
Department of Trade of Industry anticipates releasing the draft document for public

comment later in 2007. Notwithstanding it will be important to note whether the areas
that Chang (1998) suggests are taken into consideration. These areas are: the accurate
determination of targets utilising limited financial and administrative resources;
formulating a clear vision for South Africas future economy; considering the effect of
demand side policies; substantially increasing and improving on its infrastructure;
establishment of a permanent coherent policymaking institution; and finally interactions
with larger firms in the economy during the formulation of its industrial policy (Chang,
1998: 70)

In conclusion, the empirical evidence suggests that trade reforms over the last decade
have been significant and have contributed positively to economic growth. However,
import competition and technological change may have undermined employment,
especially amongst lower-skilled workers. Poverty and inequality have also risen
dramatically. In addition, it is evident that the presence of an industrial policy that
promotes industrialisation which are labour-intensive will increase employment and
produce positive impacts for pro poor growth.

3. Background

Pasha (2002) recognizes that the previous development paradigm which stressed the
pursuit of growth failed to improve the lives of the poor in developing countries.
Presently, the need to place poverty reduction at the crux of the process of development
requires a national strategy that intends for human development to be sustainable,
equitable and empower the majority of the population. The Millennium Declaration
represents a commendable global commitment achieve this through the Millennium
Development Goals (MDGs).

Trade and industrial policies could serve as catalysts for poverty reduction for developing
countries. The statements below for example, signify the importance of these policies for
poor countries.

International trade can play a major role in the promotion of economic development
and the alleviation of poverty. We recognize the need for all our peoples to benefit from
the increased opportunities and welfare gains that the multilateral trading system
generates...We recognize the particular vulnerability of the least-developed countries and
the special structural difficulties they face in the global economy. We are committed to
addressing the marginalization of least-developed countries in international trade and to
improving their effective participation in the multilateral trading system.
The Doha WTO Ministerial Declaration (2001), par. 2, 3.

World trade has the potential to act as a powerful motor for the reduction of poverty,
as well as for economic growth, but that potential is being lost. The problem is not that
international trade is inherently opposed to the needs and interests of the poor, but that
the rules that govern it are rigged in favour of the rich...
...where good policies enable poor countries and poor people to participate in markets on

equitable terms, trade can act as a force for poverty reduction.
Oxfam, Rigged Rules and Double Standards (2002), pp. 5, 47.

Even if trade rules are radically reformed and a pro-development round is achieved
. current behind the borders problems mean that poor countries will continue to
lose out on the potential benefits of global trade (Oxfam, 2005).

Trade could play a significant role in reducing poverty but not while international
trade rules are so biased against the poor. For years poor countries have been
pressured to open their markets to competition.. The historical evidence shows that
no country has developed through indiscriminate liberalisation. Flexibility to use a
range of policy instruments has been the key to growth with equity (An Agenda for
Trade Justice, Dchas and ICTU, 2003).

Sechler, & Guinan (2005) estimate that the annual trade increases which are as high as
$680 billion, one third of this goes to developing countries and the significant rest to
developed countries. It is important to note that
too often, Northern governments have failed to practice what they preach on
trade issues. They argue for free trade, yet erect barriers against imports from
developing countries. They argue for rules, yet abuse anti-dumping or health and safety
legislation to unfairly restrict imports. They support development, but then deter
countries from processing their own products by tariff escalation. A new and pro-poor
trade regime requires a change of mentality among the rich and powerful (Townsend,

The United Nations MDGs, also recognise the importance of trade. The eighth MDG
motivates for a global partnership for development, with a target to create an open,
rule-based, predictable, non-discriminatory trading and financial system and in
particular address the needs of Least Developed Countries (LDCs). Trade is also
important for achieving the first MDG, eradicating poverty and reducing extreme hunger,
which would see both those living in poverty (defined as less than $1 a day) and those
suffering from hunger reduced by half.

The recognition that that trade is a crucial element in the fight against global poverty is
significant. The need to establish a conducive trade environment for poor countries
resulted in the Doha Development Round in 2001. Developing countries were
demanding policy space within the trading system to address the developmental needs
of poor countries.

The most important agreements reached in Doha that embody its `development' content
include those relating to:
(i) agriculture;
(ii) trade-related intellectual property rights (TRIPS) and public health;
(iii) special and differential treatment (S&D); and
(iv) technical assistance for capacity strengthening in the least developed

However, the collapse of the WTO Ministerial meeting in Cancun, Mexico, in September
2003 was a setback for the vision espoused by both the Doha Declaration and the MDGs.
The UN World Summit in September 2006 reviewed progress set for achieving the
MDGs. It concluded that progress was slow on trade and on the other goals, while
highlighting the importance of the development dimension of the Doha Round
(Townsend, 2006). This collapse is a stalemate that arises from the issue of erosion of
policy space as earlier mentioned. It has been argued that industrial policies are critical
for developing countries to stimulate sustained growth as was the experience of
developed countries in the earlier periods of the industrialisation (Akyz, 2005; Chang
2005a; Rodrik, 2006).

Furthermore Chang (2005b) argues that industrial policies are critical instruments in this
globalisation era to stimulate growth. This contention has been categorically illustrated
during the import substitution industrialisation period (1960-1980) as opposed to the era
of neoliberal reform (1980-2000).

Therefore, the bias in global negotiations reflected above indicate a need for developing
countries to advance pro poor trade and industrial policies that will stimulate internal
production that benefits local consumers as well as secure global markets for local
industries and more importantly have the discretion to do so through their individual and
collective policy space.

In South Africa although the industrial policy has not been fully developed (as will be
discussed later), South Africa has been engaged in bilateral, regional, and global trade
negotiations described above and seeks to promote agreements that will stimulate growth
in its economy. It has adopted policies that reflect this and more importantly that will
benefit second economy of the country. The Accelerated and Shared Growth Initiative
for South Africa (AsgiSA) for example, provides the framework within which an
industrial policy that is pro poor may be developed (South Africa, 2006: Internet Source).

3.1 Definition of Concepts

The concepts pro poor and trade and industrial policies are clarified below to
accentuate the importance of pro poor trade and industrial policies for developing

3.1.1 Pro poor and Growth

Pro-poor growth has been broadly defined as growth that leads to significant reductions
in poverty (Lopez, 2005). It is noted that there are two broad definitions of pro-poor
growth and the main distinction between them is whether for a given growth period they
focus on inequality outcomes (Kakwani & Pernia, 2000) or on poverty outcomes
(Ravallion & Chen, 2003).

Kraay (2004) has identified three potential sources of pro-poor growth. These are: (i) a
high growth rate; (ii) a high sensitivity of poverty to growth; and (iii) a poverty reducing
pattern of growth. Against this background, Ravallion (2004) concludes that "growth will
be quite a blunt instrument against poverty unless that growth comes with falling
inequality". Implying that changes in poverty are mainly driven by changes in inequality.

In addition, two different definitions of pro-poor growth in the literature are observed.
The first definition emphasises a situation where any distributional shifts observed with
economic growth favours the poor. The definition implies that if poverty falls more than
it would have if all incomes had grown at the same rate (McCullock & Baulch, 2000;
Kakwani & Pernia, 2000). This definition focuses on changes in inequality during the
growth process. Pro-poor growth in this definition requires that the incomes of the poor
grow at a higher rate than those of the non-poor. A concern with this definition is that
distributional changes can be pro-poor with no gain to poor people and even falling
living standards for the poor. On the other hand, a pro-rich distributional shift during a
period of overall economic expansion may result in major absolute gains to the poor
(Ravallion, 2004).

The second definition focuses on what happens to poverty. The economic growth process
is said to be pro-poor only if poor people benefit in absolute terms, as reflected in an
appropriate measure of poverty (Ravallion & Chen, 2003). The extent to which growth is
pro-poor by this definition depends mainly on the rate of change in poverty. Both
definitions then beg the question: What measure of poverty should be used?

These two definitions of pro-poor growth can be further decomposed to relative

definition and absolute definition. In the first instance, growth is deemed to be pro-poor if
the average income of the poor increases more than the average income of the non-poor.
This implies that growth is accompanied by a change in income distribution in favour of
the poor; i.e. inequality decreases. In this definition growth is strictly pro-poor. In the
second case, growth is determined to be pro-poor if the poor benefit in absolute terms.
This definition considers only the income of the poor - inequality as such is not

It is observed that with any of these definitions, the major consequence is that policies
must lead to active policies in favour of the poor. Pro-poor growth is therefore opposed to
the old trickle-down approach. The arguments observed further emphasise that while
growth is an important factor for reducing poverty, it is not sufficient for sustained
poverty reduction. This raises the question regarding how growth can be made more
effective in reducing poverty. Three major factors have been identified:

Initially, more unequal countries need to grow faster than countries enjoying a more
equal income distribution. This implies that countries with high inequality need high
economic growth rates to achieve a significant reduction in poverty. The initial level of
income distribution, as well as changes in income distribution can therefore become an
important impediment for the poor to benefit from economic growth.

Secondly, the geographical as well as the sectoral pattern of growth where the high
concentration of the poor in specific regions and economic sectors, the geographic and
sectoral patterns of economic growth in a country are important factors that determine the
impact of growth on poverty. The extent to which economic growth affects the rural
sector is often very important.

Finally, high public expenditures for human capital such as health and education are
required. In overall growth theory, public investment in education and health is necessary
to foster growth through the labour force. This is even more necessary for pro-poor
growth where the poor are those most negatively affected from weak access to education
and health. Therefore, pro-poor requires high economic growth; success in reducing the
initial inequalities; and finally a more pro-poor pattern of growth.

Pro-poor macroeconomic policies have also been defined as policies that promote
economic inclusion, empowerment and social investment. Sobhan (2002) describes a pro-
poor macroeconomic framework as one that serves to transform the poor from being the
incidental beneficiaries of economic growth to becoming one of its prime movers. An
interesting consideration in the definition debate notes that changes in poverty can be
attributed in a part due to economic growth and in part due to changes in the income
distribution that is a change in inequality (SDC, 2004).

Apart from the direct effects on poverty, economic growth and income inequality may
also have indirect effects where economic growth can reduce inequality and thus poverty,
for instance, having the poorest segments of the population earning proportionately more
than the other segments; or where lower inequality can increase economic growth and
thus reduce poverty, for example through better investment opportunities for the poor and
better incentives to work.

In contextualizing the debate, the growth-inequality link literature has focused on

whether countries will have to face trade-offs between reducing inequality and improving
growth performance or whether there is a virtuous circle where growth leads to lower
inequality and lower inequality in turn leading to faster economic growth. The theoretical
literature review has observed different explanations for the potential link between
growth and inequality. It is noted that on the one hand the literature emphasises a growth-
to-inequality type of causality .On the other hand, more emphasis is on an inequality-to-
growth direction of causality.

The growth-to-inequality causality, is rooted in the Kuznets hypothesis. The hypothesis

explains that the distribution of income would fall during the initial stages of
development as an economy transforms from rural to urban and from agricultural to
industrial. As a result, inequality will decrease as the labour force in the industrial sector
grows and that of the agricultural sector falls. More recently, however, a number of
economic models have argued that technological progress may lead to higher inequality
when it affects the productivity of different types of labour. With regard to the impact of
inequality on growth, the theoretical literature is divided between those who argue that

inequality is detrimental for growth and those who argue that inequality is conducive to
higher growth.

It is further noted, there are also models which argue that inequality growth-enhancing. In
these scenarios, Kaldor's hypothesis that the marginal propensity to save of rich people is
higher than that of poor applies. Then if the investment rate is positively related to the
saving rate, and growth is positively related to investment, more unequal economies can
be expected to grow faster. Bourguignon (1981) builds a more elaborate model and
shows that with a convex saving function, aggregate output depends on the initial
distribution and is higher the more unequal society is. The previous discussion suggests a
clear division of opinion in the theoretical literature. In considering the empirical
literature on the growth to inequality relationship, the results are unanimous. The results
in Deininger & Squire (1996), Ravallion & Chen (1997), Easterly (1999) and more
recently Dollar & Kraay (2002) all suggest that growth does not have an impact on

With regard to the inequality to growth link the empirical literature is less unanimous,
reflecting that the same division that the theoretical models suggest. The results suggest
that inequality in income is negatively associated with subsequent growth (Alesina &
Rodrik (1994) and Perotti (1996). Barro (2000) finds no relationship between inequality
and growth.

On the impact of redistribution on growth it is noted that the work by Easterly & Rebelo
(1993) as well as by Perotti (1996) applies. Using several measures of redistribution
(marginal tax rates, average tax rates, social spending) Easterly & Rebelo (1993) find that
redistribution is likely to have a positive impact on growth. Similarly, Perotti (1996) tests
whether income inequality has an impact on the marginal tax rate, and whether the latter
affects growth. His results suggest that while inequality may play no role in setting the
marginal tax rate, higher marginal tax rates will have a positive impact on growth. These
results would suggest a less than conclusive picture regarding the impact that
income inequality has on growth.

The discussion above notes that there is some consensus on the lack of causality from
growth to income distribution in one or other direction. However, on the potential
causality from inequality to growth, the literature is more divided, with some studies
concluding that inequality leads to faster growth, and others suggesting that inequality is
likely to lower growth. Lundberg & Squire (2003), note that in practice, most pro-growth
policies have an impact on inequality and in some cases even conflict with the growth

Therefore, Pro-Poor Growth means a virtuous cycle between growth, inequality and
poverty reduction. Three major assertions are observed from the above namely (SDC,

Growth and income distribution are important for poverty reduction.

High economic growth can offset the increased inequality impact by still reducing
poverty as has been the case for several Asian countries.
Low economic growth, coupled with increasing inequality will result in increased
poverty as evidenced in a majority of African and Latin American countries.

A strong relationship between growth and poverty reduction holds through the creation of
employment in the process of trade and industrial expansion along with labour market
conditions, public sector institutions and fiscal regime. Economic growth would help
rural poverty reduction only if there is promotion of local resource-intensive or labour-
intensive industries which use simple and cheap technology (UNESCAP, 2001). Trade
mainstreaming should emphasises the goal of policy coherence by incorporating trade
policy in a countrys overall development framework so that it complements a countrys
other economic and social priorities. It is important therefore that trade mainstreaming
not only identify opportunities for trade liberalization and export promotion, but also
identify opportunities to reduce poverty through trade policy.

In addition, developing economies face infrastructure and other supply side bottlenecks
and constraints that make it difficult for them to compete effectively and take advantage
of existing market access and other trade opportunities, regardless of whether negotiation
outcomes are to their advantage.

With respect to supply-side constraints, UNDP is assisting developing countries to

address three key capacity development needs: (i) the capacity to compete internationally
by overcoming institutional, human, and other supply side bottlenecks; (ii) the capacity to
negotiate, interpret and implement trade agreements (multilateral, regional and bilateral)
which prioritize poverty and human development concerns; and (iii) the capacity to
incorporate pro-poor trade policy in poverty reduction strategies.

Interestingly, new perspectives are being considered. These perspectives consider

perspective the necessity to promote domestic commerce and domestic market
engagement. The new perspectives emphasise that domestic commerce deepen
development linkages than does export-led growth. Two major reasons for the
shallowness of export-led development are that (1) it often relies heavily on imported
inputs, and (2) no domestic distribution, marketing, retailing, or servicing are developed,
because the market is offshore.

The argument is that trade is good for growth, but to promote development and poverty
reduction, trade must also be accompanied by complementary policies that ensure (1)
investment in health and education (human capital) and (2) investment in transportation
and communications infrastructure to enable distribution and marketing. This is a
broader, more attractive policy agenda.

A successful poverty reduction strategy will need to focus on achieving a high and
sustainable rate of economic growth, with such growth possessing two key
characteristics: high rate of employment generation and rapid agricultural growth. These

appear to be the key conclusions drawn from the Asian experience vis--vis poverty
reduction during the last three decades whereby for economic growth to be favourable to
the poor, then it should have a pattern that directs resources to the sectors in which the
poor work (agriculture), areas in which they live (relatively backward regions), factors of
production which they possess (unskilled labour) and outputs which they consume (such
as food). (Pasha & Palanivel, 2004).

It is noted that the macro- policy environment will have to promote a higher labour-
intensive growth pattern. This will include the development of a system of incentives and
institutions that discourages high capital intensity of investment (e.g. maintenance of an
overvalued exchange rate that permits the import of cheap machinery, the promotion of
large mechanized corporate farms instead of small holders, making intensive use of
family labour).

The UNDP pro- poor economic strategies are based on two principles (Fihlo, 2005). In
the first case, ensures that pro-poor growth benefits the poor more than the rich. To
deliver benefits to the poor, growth must be conditional on the reduction of absolute and
relative poverty and the improvement of the living standards of the majority of the
population. The choice of a pro-poor growth strategy does not depend on the existence of
a positive relationship between equity and growth. The case for pro-poor policies is based
on the intrinsic value of economic equity, its importance for democracy and human
rights, and the recognition of the imperative to achieve pro-poor outcomes as fast as

In the second instance, macroeconomic stability is regarded as a constraint to growth,

distribution and the structural transformation of the economy. The achievement of pro-
poor goals is dependant on addressing several economic constraints, for example, balance
of payments equilibrium, real exchange rate stability and the minimisation of economic
volatility, inflation and the domestic public debt.

The strategy for equity-based growth requires that growth will improve the absolute
condition of the poor and will be equitable enough to improve their relative position.
Preferably achieving equity at the start of the growth process will be attained (such as
through land reform or universal basic education) or by decreasing high inequality over
time (such as through pushing up wages by generating widespread employment among
low-skilled workers).

The UNDP strategy notes that Equity-based growth can be achieved through a variety
of strategies, depending on each countrys initial conditions. It is argued that if growth is
to reduce poverty, it should have a pattern that directs resources disproportionately to the
sectors in which the poor work (such as small- scale agriculture), the areas in which they
live (such as underdeveloped regions) or the factors of production that they possess (such
as unskilled labour or land) (McKinley, 2003).

As discussed in the previous section of this paper, the relationship between trade
liberalization and poverty remains in many respects unclear. However, one of the issues

where consensus exists is the recognition that many trade policies and practices in both
developed and developing countries have detrimental effects on the potential
development gains. These policies could be grouped into the following categories:

High protection in labour-intensive agriculture and industrial goods trade

Anti-dumping tariffs, stringent regulations and standards and other non-tariff
measures potentially hindering the poverty reduction
Limited access to services markets, especially for services in which developing
countries have comparative advantage (including, so called, Mode 4 services).
Complex intellectual property rights protection regime with no recognition of
protection and promotion of traditional knowledge (TK) embodied in products,
services or processes of the LDCs and developing countries
(Ravallion, 2004)

Trade liberalization is a key component of conventional economic policy aimed at growth

and development. Trade liberalization is regarded to have strong positive effects on
resource allocation, economic growth and poverty reduction (Dollar & Kraay, 2000). The
link between trade and growth has been established through empirical works. But the
empirical evidence of the trade-poverty linkage is new area of study and this is
complicated by the debate over the definition of poor and trade openness. One
conclusion to emerge from an evaluation of this literature is that the effect of
liberalization on poverty is likely to be complex. It is likely to depend on a host of factors
including the profile of the poor in a country, the structure of production and
consumption, and the effects of other reforms and so on. It is difficult, therefore, to
conclude that trade liberalization and openness unambiguously results in a reduction of
poverty. Nevertheless, there is consensus that the essential precondition for sustained
poverty alleviation is rapid economic growth and trade is an engine of economic growth.

The economic theory suggests that trade liberalization results in productivity gains
through increased competition, efficiency, innovation and acquisition of new technology.
Trade reform is supposed to expand the set of economic opportunities by enlarging
market size and increasing the effects of knowledge spillovers. All these effects together
are expected to induce growth of output and consequent poverty alleviation. Thus in the
medium term, trade reform has an important role in poverty alleviation through its effects
on the rate and sectoral pattern of growth.

As such, external sector liberalization has had a significant reduction in poverty in many
East and South-East Asian countries. South Asian experience further confirms the global
empirical finding that relationship between trade openness poverty emanating through
growth performance (World Bank, 2004).

Trade openness that has shown in the literature to have an impact on economic growth is
not observed to directly influence the income of the poor in some studies (Ghura et al,
2002; Stryker & Pandolfi, 1997). It means that the poor are not able to benefit from
greater trade, because their lack of access to capital and physical isolation prevent them
from taking advantage of all the opportunities created.

A problem with the new trade and development framework is that it remains anchored in
the export-led growth paradigm that has dominated development economics for the last
two decades. This paradigm is problematic (Blecker & Razmi, 2005). At bottom, it is
doubtful that all developing countries can grow via export-led manufacturing. The global
market is limited, and an excessive focus on exports stands to generate degraded
competition. Another problem with the trade-as-development model is its lack of
attention to each countrys readiness to engage in international trade.

3.1.2 Trade and Industrial Policies

Trade policy generically refers to the complete framework of laws, regulations,

international agreements and negotiating stances adopted by Government to achieve
legally binding overseas market access to overseas markets for domestic firms. This
definition reflects the prominence of trade on a countrys GDP, and balance of payments,
which indicates that trade does have positive and negative implications, the following
literature confirms this.

Barro (2000) finds evidence indicating that openness to trade leads to more inequality
which would be more pronounced in poor countries. Dollar & Kraay (2003) determine
that more trade and a better rule of law would lead to higher equality, whereas higher
inflation, higher government consumption, and additional financial development would
lead to higher inequality. Lopez (2005) concludes that financial development, trade
openness, and decreases in the size of government would be associated with increases in

On the other hand, Lundberg et al (2003) establish that higher education, lower inflation,
and land distribution would lead to lower inequality and also lead to faster growth. In the
same case, trade openness (as measured by the Sachs-Warner index) and more civil
liberties would likely pose a potential conflict between the goals of faster growth and
more equitable distribution.

Ishihara (No date) provides two definitions of industrial policy. The first is a broad
definition that encompasses policies that promote and protect individual industries and
those that preserve and strengthen the basis of the market economy. These policies
consist of:
policies that regulate infrastructure improvement for industries,
industrial structure policies that pertain to resource allocation between industries,
that is, industrial promotion policies and industrial adjustment policies,
industrial organization policies that effect market structure and conduct of
individual industries.

In addition, the narrow definition that distinguishes between industrial and competition
policies. Industrial policies in this case are those that affect economic welfare of country
by allocating resources between industries / sectors or the industrial organization of
specific industries/ sectors, these policies include:

the policies that affect industrial structure of a country for instance, the protection
and nurture policies to strategic industries, and adjustment policies for waning
policies that correct market failure associated with technological development and
imperfect information,
policies that seek to raise economic welfare through administrative intervention in
the industrial organization of individual industries.

For purposes of this paper, the narrow definition of industrial policies will prevail. In
view of the above discourse on trade and industrial policies, the following elements
provided are considered necessary in the discussion of pro poor trade and industrial

4. Elements of Pro-poor trade and industrial policies

Evidence provided above point to the need to design and implement pro poor trade and
industrial policies that include

Macroeconomic policies: choose policies which affect the poor the most (for
example, controlled inflation, selective openness to trade)
Sectoral policies: promote growth where the poor are active (agricultural reform,
industrial policy, human capital development via education, health, promoting
Restructuring policies: promote growth in areas where the poor live
(corresponding rural and urban development)
Redistributive policies: choose policies which enable poor people to gain access
to income and employment (such as redistribution of land to the tiller, land titling
as possibility to gain access to credit, smart income transfers). These can be
illustrated as below:

Policies utilized : Positive Impact of policies
Macroeconomic SECTORS
poor work

Policies Pro Poor AREAS
Trade and poor live
Redistributive Policies
poor possess


poor consume

Although it is not possible to recommend a single policy framework that is ideally suited
to promoting trade, recent capacity-building efforts point to several features or
arrangements that have tended to promote success. Developing countries should seek to
construct policy frameworks with the following elements (OECD, 2001):

A coherent strategy that is closely integrated with a countrys overall

development strategy.
Effective mechanisms for consultation among key sets of stakeholders:
government, the enterprise sector, and civil society.
Effective mechanisms for intra-governmental policy co-ordination.
A strategy for the enhanced collection, dissemination, and analysis of related
Policy networks, supported by indigenous research institutions.
Networks of support institutions.
Private sector linkages.
A commitment by all key trade stakeholders to outward-oriented regional and
global strategies.

In turn, a sound policy framework and effective policy process will:

Enable developing countries to address a wide range of challenges and
opportunities including those that cannot be anticipated over an extended
Enable developing countries to sustain and upgrade capacities after donors have
departed, permitting dependence on external assistance gradually to decline. The
participatory character of effective policymaking will help strengthen local
capacities, as stakeholders learn by doing and learn from each other.
Facilitate local ownership of development efforts. The involvement of a wide
range of actors in the policy process will facilitate genuine local ownership of
development and ensure that strategies and policies are demand-driven.
Reduce the risk that developing country policies will be influenced by the policy
priorities of donors. A decrease in external assistance and an increasing reliance
on local capacities will minimise the chance that a developing countrys policies
will be skewed by the commercial or trade policies of donors.

The rhetoric on pro poor policies in evident in recent development literature, we briefly
examine the issues raised briefly from literature.

5. Experiences of developing countries with pro poor policies

The following is an overview of literature on pro poor policies. On the one hand,
debates around trade policies have resulted in the relationship between trade
liberalization and poverty reduction and inequality has been one of the most prominent
issues of the current debate on pro-poor growth. This debate is reflected in a study
published by the WTO Secretariat in 2000 on Trade, Income Disparity and Poverty.
While there is generally a consensus that expansion of exports can contribute to faster
economic growth, there is less clarity on the direct impact of exports on poverty, once
there is control for the overall rate of growth.

Studies such as Agenor (2002), Ghura et al (2002), and Epaulard (2003) find that, once
the effect of income has been taken into account, trade openness has no significant direct
influence on poverty incidence or on the income of the poor or on the elasticity of the
poverty with respect to growth. Overall, it appears that export growth does not have a
significant direct impact on poverty. Its effect has to be seen primarily via its bearing on
the overall rate of economic growth. Therefore, exports cannot be said to play a
significant role in influencing the extent to which the process of growth is pro-poor or
not. Clearly, for a given growth rate, the implications for the level of poverty are likely to
be more favourable if there is simultaneously a reduction in inequality such that the
income of poor rises disproportionately in relation to the increase in average income in
the economy.

In the rural areas, where there is a lack of infrastructure, credit, marketing channels and
public services, poverty reduction strategies should be geared to address these
constraints. Another consideration of the problem is the unequal distribution of benefits

across countries. There are some countries that are concerned about completely opening
up their economies because of the potentially negative effects on their industrial and
agricultural sectors, especially if industrial-country markets for exports of agricultural
commodities and labour-intensive manufactures from developing countries remain
protected. The Pro-Poor growth strategies that countries tend to advocate for is boosting
domestic savings and investment and using public investment as a stimulus for private
investment. This implies a more activist policy role for the state and a larger revenue
base, with which it can finance capital expenditures and direct them to poverty-reduction

If trade liberalization is not complemented with other more pro-active measures

(especially poverty- focused interventions), such as the building of rural infrastructure,
financing of agricultural development or the provision of adequate credit to small and
medium enterprises, trade liberalization can exacerbate inequality and bypass the poor,
especially the rural poor. To be most effective, liberalization of trade should be designed
carefully and go hand- in-hand with a pro-active industrial strategy. Favouring the
adoption of trade policies with pro-active industrial policies emphasises protection of key
domestic sectors and focusing development on sectors, such as agriculture, where poor
workers are concentrated and are likely to suffer from liberalization. To increase the pro-
poor impact of growth, this alternative policy stance also places priority on some sectoral
measures, such as employment generation and agricultural and rural development, as
critical elements of macroeconomic policies. For job creation, the emphasis is on not only
fostering a more employment- intensive pattern of growth but also taking explicit public
measures to boost the productivity of poor workers. For agricultural and rural
development, the emphasis is on deploying public investment for critical public goods
(such as rural roads and irrigation), improving agricultural terms of trade and providing
equitable and sustainable access to productive resources, such as through secure use-
rights to land.

Knowing how the poor obtain their income is important in assessing the impact of
economic policies. In addition, the sources of poor household income depend on
variables such as region, location, educational level, and ethnicity. A pro-poor structure
would have low tariffs on goods consumed by poor households, and high tariffs on the
goods sold by the poor (Duygan & Bump, 2006).

An open trade regime can be very positive for growth with huge market potential as well
as cheaper inputs. However, it is important that for purposes of distributional effects,
some form of protectionism might be necessary. With regard to Industrial Policy, giving
preference to industries where poor people work and labour- intensive instead of capital
intensive modes of production are utilized is important. Developing countries need to
formulate pro poor trade and industrial policies, in order to achieve this, they will be
required to understand the impact of trade reforms on the poor and to develop strategies
to support pro-poor growth (Siphana, 2003).

Countries which moved more rapidly towards more outward looking trade policies found
it easier to sustain high rates of productivity growth. For example, Korea in the 1960s,
dismantled its import restrictions and reoriented its industrial policy towards export
industries. Chile in the 1980s eliminated all quantitative restrictions and trade controls
and reduced its tariffs substantially whereas India in the 1990s began to dismantle its
import restrictions during the 1980s, and continued with a far-reaching reform
programme in 1991, including a reduction in tariffs from 85% to 25%. As a result, these
countries experienced a rapid acceleration in growth, following the adoption of reforms
including substantial reductions in their barriers to foreign trade.

It is of significance to note that some countries have managed to reduce poverty even in
periods of relatively slow growth by ensuring that whatever gains accrue from this
growth, accrue relatively more to the poorer segments of the population (McKinley,
2003). However delays in accessing cheaper imports will have a negative effect on
consumers, who may also be poor; and the risk of efficiency losses offsetting
distributional gains is high, particularly if the product is not effectively targeted on the
poorest. In Indonesia for example some studies estimate that reducing agricultural tariffs
will lead to poverty increases, but others show how the poor and the very poor have
conflicting interests in rice liberalisation. Either way, higher protection rates should be
viewed as a transitional measure, and should be coupled with measures to reduce
dependency. Otherwise there is a real risk of isolating the poorest in low productivity
sectors, with little prospect of growth.

In the countries covered by the Asia-Pacific regional programme, the impact of trade
liberalization on poverty has been mixed. The studies recommend that liberalization
should be approached cautiously, with some protection of domestic industry being
combined with export promotion. The studies further argue that trade policies be linked
with a pro-active industrial strategy in order to maximize the benefits for development
(Pasha & Palanivel, 2004). As the benefits of trade liberalization have been unequally
distributed. Therefore, there is a call for pro-poor interventions that can compensate, at
least, for the adverse effects of liberalization. These could run the gamut from the less
controversial, such as improved social security (as advocated by the China case study), to
the conventional responses, such as the provision of public goods and agricultural
development, to the most controversial, industrial policies, as advocated by the Indonesia
case study.

For the African context, regional trade agreements have several advantages. First,
countries in a specific region tend to be at similar stages of development and have
common trade constraints. Second, the value created in production and trade is captured
within the region rather than being siphoned off to the North. Third, such trade can
economize on the use of expensive foreign-exchange reserves.

An alternative approach is value-chain analysis. Value-chain analysis follows a product

from the ground to the final consumer and identifies who captures how much value along
the entire chain. This incorporates a stages-of-production, value-added focus. An

alternative focus involves a bargaining-power perspective and emphasizes value-capture.
This pattern of capture has changed significantly owing to globalization and product
branding, which have altered bargaining power along the stages of production. The twin
lenses of domestic commerce promotion and value-capture suggest a new set of policy
questions (OECD, 2001).

The experience in sub-Saharan Africa (SSA) shows that even under a significantly more
open trade regime, poor countries find it difficult to capture the benefits from trade fully
unless they address major supply-side constraints. Overall development policy must
highlight some specific constraints particularly affecting trade capacity such as:
high transaction costs, including transport costs;
low human capital;
monopolistic distribution structures;
barriers to flexible entry and exit for firms; and
limited access to financial services;

Many countries in SSA have faced common difficulties in diversifying their economies
and export structure while managing a long-term decline in terms of trade. The UNCTAD
(2003) Report on Africas economic performance ascribes the loss of market share to
both the inability of SSA countries to maintain their market share in traditional
commodity exports, and their failure to diversify into more dynamic agricultural exports
like horticulture or manufactured exports. Furthermore, growth in other commodities
such as cotton and sugar has been heavily constrained by subsidies in OECD countries.
(Hoekman, et al, 2001) proposes that knowledge of how the poor obtain and how the
poor spend their income is important in designing pro-poor policies The sale of unskilled
labour tends to be the most important source of income for the poor, complemented by
the value of own consumptionthat is, the imputed value of what the poor consume
from their own production. In general, the impact on the sources of income of the poor
will be a more important determinant of the effect of liberalization than the effect on the
prices of the things that they consume. The reason is that trade reform will affect many
relative prices, some of which will move in offsetting directions. As the poor generally
have limited assets, the most important of which is low-skilled labour, the impact on
wages and employment therefore dominates. Although the poor are very diverse, they
frequently work in the rural sector and in the urban informal sector. Thus, policies that
affect agricultural and labour markets are important complements to trade reform for the

On the other hand, industrial policies continue to be more contentious because of their
protectionist nature. Asian countries exemplify the more recent effort to simulate
economic growth through industrialisation. Adopting protectionist industrial policies
during the oil boom period in Indonesia for example, resulted in rapid industrial growth
(Wie, 2006: 342). Another prime example is the policy engagements between Home
Based Workers and policymakers in South Asia in the recent Policy Conference on Home
Based Workers (UNIFEM, 2007: Internet Source).

Hence, policy discretion in the formulation of industrial policies plays a critical role in
shaping production which in turns increases a country growth. Evidence suggests here
that protectionism was important in facilitating growth for developed economies during
their periods of industrialisation see Akyz (2006:12). The experiences of industrial and
successful developing countries also provide two critical lessons. Firstly, that economic
integration with the world economy is most often an outcome of growth and
development, not a prerequisite. Secondly, institutional innovations require domestic
policy space and flexibility in order to be successful development strategies and
outcomes. This is a contradiction in itself when global debates indicates the suggestion
that developing economies should abandon their policy discretion and open their fragile
economies to the global market which has resulted in the impasse at the Doha talks. A
tentative suggestion provided by Akyz (2006:45) is that for individual country industrial
development, policy discretion is necessary.

In addition, Ohno (2003) argues that pro poor growth should not be the sole
developmental instrument as this provides too narrow a development goal. The radical
option provided has often been criticised by developed economies as undemocratic is
known as authoritarian developmentalism which has enabled the Eastern Asian
economies to break from the vicious cycle of poverty and realise growth. The following
are necessary conditions for this radical regime to be effective: a strong and economically
literate leader; a capable team to support the leader; top-down coherent decision-making;
an ideology that affirms material progress as the supreme national goal; and finally
popular support for this regime. Ohno (2003) reiterates that this is a temporary dynamic
regime of convenience that is necessary to propel the country from poverty to

6. Pertinent Issues for South Africa

Therefore, although there is deep resonance amongst developing countries to design and
develop pro poor trade and industrial policies that reduce the incidence of poverty
amongst their poor constituents. It is acknowledged that this may be a tedious process
because of the numerous constraints poor countries face in accessing and utilising a voice
in global markets that is economically beneficial to them.

Capacity constraints within the poorest countries are also a major constraint affecting
not only their ability to take advantage of new export markets, but also their ability to
realise the benefits of reductions in their own tariff barriers. Poor infrastructure, low
human capital and constraints on the development of the private sector all hamper the
economys capacity to respond to the opportunities which greater openness offers. If
trade is to be an effective driver of growth, these constraints need to be addressed
alongside trade reform, as part of a comprehensive growth-promoting and poverty-
reducing development strategy.

The above discussion raises some pertinent issues for South Africa and its trade and
industrial policies (forthcoming) in the effort to bring out pro poor growth of what the
South African government refers to elimination of the second economy.

These issues may be considered from a two-pronged approach that incorporates an

internal viewpoint and a global stance. There are several considerations that South Africa
should take into account so as to ensure that pro poor growth through its trade and
industrial policies. In the first instance, the importance of policy discretion allows the
government to develop policies that benefit its poor. This will require the consideration
and possible abandonment of multi-lateral rules that constrain its policy formulation.
Secondly, continuous engagements with the vulnerable stakeholders in the country
especially during the development of its industrial policy will assist the government in
identifying and inducing pro poor growth. Thirdly, Carim (2005:10) advises in
negotiating trade agreements, South Africa to align its industrial capabilities that would
induce growth and reinforce development; work to ensure that competitive imports do not
completely erode local industries that the vulnerable workforce are situated as has been
the recent case of closure of textile industries. The next consideration involves the South
Africas continuous engagement at the trade talks in the regional and global arena as an
developing country to adopt agreements that benefit the poor in areas that they live and
work as illustrated above.

Finally although authoritarian developmentalism has been considered too undemocratic

and approach, South Africa may borrow the strategic industrial targeting through
protectionism to build the technology of its infantile industries to make them globally
competitive. It is also anticipated that the international donor community can help
provide significant additional development assistance to poor countries needing to
undertake investment in human and physical capital to complement greater trade
openness and lead to poverty reduction.


Agnor, P-R. (2002). Business Cycles, Economic Crises, and the Poor
Testing for Asymmetric Effects. The Journal of Policy Reform .Vol 5, (3):145 160.

Alesina, A. & Rodrik, D. (1994). Distributive Politics and Economic Growth,

Quarterly Journal of Economics 108: 465-90.

Akyz, Y., (2005). The WTO Negotiations on Industrialisation Tariffs: What is at Stake
for Developing Countries. Third World Network. Unpublished.

Balat, J. & Porto, G (2005). Globalization and Complementary Policies. Poverty

Impacts in Rural Zambia, in Harrison, A. (ed). Globalization and Poverty, National
Bureau of Economic Research. Boston: Massachusetts.

Barro, R. (2000). Inequality and Growth in a Panel of Countries, Journal of Economic

Growth 5: 5-32.

Bhorat H (1999). The October Household Survey, Unemployment and the Informal
Sector: A Note. South African Journal of Economics. Vol 67 (2): 321-326.

Blecker R, & Razmi A.(2005). Developing Country Exports of Manufactures: Moving Up

the Ladder to Escape the Fallacy of Composition?, Working Papers 2005-02, University
of Massachusetts Amherst, Department of Economics, revised May 2006.

Bourguignon, F. (1981). Parento Superiority of Unegalitarian Equilibria in Stiglitz

Model of Wealth Distribution with Convex Saving Function in Econometrica, vol 49(6)
p 1469-1475.

Buhlungu S., Daniel J., Southall R, & Lutcham (eds). (2006). State of the Nation 2005-
2006. Pretoria: HSRC Press

Carim, X. (2005). Trade and Competition Policy and the Relationship to Industrial
Development. in Trade and Industry Monitor. Vol 35: 10-14.

Cernat L, (2004). The role of competition in the promotion of competitiveness and

development: Experiences from a sample of developing and least developed countries.
Paper prepared for the 3 International CRC Conference, Pro-poor Regulation and
Competition: Issues, Policies and Practices Cape Town, South Africa 7-9 September

Chang, H. (1998). Evaluating the Current Industrial Policy of South Africa. in

Transformer vol 36: 51-72.

Chang, H. (2005a). Policy Space in Historical Perspective- with Special Reference to

Trade and Industrial Policies. Paper presented at Queen Elizabeth house 50th

Anniversary Conference The Development Threats and Promises, Oxford University, 4-
5 July 2005.

Chang, H. (2005b). Is a strong Industrial Policy possible in the Current Global

Environment? in Trade and Industry Monitor. Vol 34:2-4.

Deininger K & Squire L (1996). A New Data Set Measuring Income Inequality. The
World Bank Economic Review Vol 10(3): 565-591.

Dollar D & Kraay A (2002). Growth is Good for the Poor. The Journal of Economic
Growth. Vol 7(3): 195-225.

Dollar D & Kraay A. (2003). Institutions, Trade, and Growth: Revisiting the Evidence.
World Bank.

Dollar D & Kraay, A. (2004). Trade, Growth, and Poverty. The Economic Journal
Vol 114 (493) p 22.

Duygan B, & Bump J.B. (2006) Can Trade Help Poor People? The Role of Trade ,
Trade Policy, and Market Access in Tanzania. Paper prepared for Tanzanias Diagnostic
Trade Integration Study by World Bank. 21 April 2006.

Edigheji, O. (2005). A Democratic Developmental State in Africa? A concept paper.

Research Report 105, May 2005. Johannesburg: Centre for Policy Studies. Available at

Edwards L. & Stern M., (2006). Trade and Poverty Research Project in South Africa:
Lessons and Policy Recommendations, Trade and Poverty Project, SALDRU: University
of Cape Town.

Easterly, W., & Rebelo, S. T. (1993). Fiscal Policy and Economic Growth: An Empirical
Investigation, Journal of Monetary Economics 32(3), 417458.

Easterly, W. (1999). Life During Growth, Journal of Economic Growth 4, 239276.

Epaulard, A. (2003). Macroeconomic Performance and Poverty Reduction IMF Working

Paper No 02/118.

Fihlo, A.S. (2005).(Breaking the Hegemony of Neoliberalism: Alternative Economic

Policies for Brazil Paper presented at International Conference on Acts of Resistance
From the South Against Globalisation, 5-7 September 2005, Ankara, Turkey.

Fedderke, J. & Shin, Y. (2003). Trade, Technology and Wage Inequality in the South
African Manufacturing Sectors, ESE Discussion Papers 106, Edinburgh School of
Economics, University of Edinburgh, August, 2003.

Frankel, J.A. & Romer, D. (1999). "Does Trade Cause Growth?," American
Economic Review, Vol. 89 (3): 379-399

Ghura, D. Leite, C.A, & Tsangarides, C.G. (2002). Is Growth Enough? Macroeconomic
Policy and Poverty Reduction. IMF Working Paper No 03/72.

Goldberg, P.K & Pavcnik, N. (2004), Trade, Inequality, and Poverty: What Do We
Know? Evidence from Recent Trade Liberalization Episodes in Developing Countries.
Paper presented at the 2004 Brookings Trade Forum on Globalization, Poverty &
Inequality: What Do We Know? Where Are We Going? held in Washington, DC, May
13-14, 2004.

Hoekman, B. Michalopoulos, C., Schiff, M. & Tarr, D. (2001) Trade Policy Reform and
Poverty Alleviation. World Bank, December 2001.

Hoogeveen, J., & Ozler, B. (2004). Not separate, not equal: Poverty and inequality in
postapartheid South Africa. Unpublished Draft.

Ishihara, T.( No Date) Industrial Policy and Competition Policy. Accessed 7 December
2006. (Available at ).

Kakwani, N., & E. M. Pernia, 2000. What is Pro-Poor Growth? Asian Development
Review 18(1):1-16.

Kraay, A. (2004). When Is Growth Pro-Poor? Cross-Country Evidence. IMF Working

Paper 04/47.

Lopez, J.H. (2005). Pro Poor Growth: A Review of what we know (and of what we dont).
Background Paper for Pro Poor Growth program by the World Bank PREM Poverty

Lundberg, M. & Squire, L. (2003). The simultaneous evolution of growth and

inequality. In The Economic Journal, vol 113 (487): 326.

Maier, R. (2005). Trade Policy and Pro-Poor Growth. Econ WPA, International Trade
Series No. 0504007.

McCullock, N. & Baulch, B., 2000. Simulating the Impact of Policy Upon Chronic and
Transitory Poverty in Rural Pakistan, The Journal of Development Studies, Vol. 36, pp.

McCulloch, N. & Ota, M. (2002). Export Horticulture and Poverty in Kenya. IDS
Working Paper 174. Institute of Development Studies, Brighton.

McKinley, T. (2003). The Macroeconomics of Poverty Reduction, Initial Findings of the
UNDP Asia-Pacific Regional Programme. Discussion Paper Bureau for Development
Policy, New York, August 2003.

OECD (2001). The DAC Guidelines: Strengthening Trade Capacity for Development.
Paris: OECD.

Ohno, K. (2003). The Role of Government in Promoting Industrialisation under

Globalisation: The East Asian experience. Paper presented at workshop on Capacity
Enhancement of Myanmar Civil Service, Yangon, 28 November 2003.

Oxfam, 2005, Making Trade Work for Development: What the UK should do. Oxfam
Briefing Paper No. 73, April 2005.

Pasha, H.A. (2002). Pro Poor Policies. Paper presented at the 4th Global Forum on
Citizens, Businesses and Governments: Dialogue and Partnerships for the Promotion of
Democracy and Development. 12-13 December 2003, Marrakech, Morocco.

Pasha, H.A. & Palanivel, T. (2004). Pro Poor Growth and Policies: The Asian
Experience. UNDP: Asia-Pacific Regional Programme on the Macroeconomics of
Poverty Reduction.

Perotti, R. (1996). Growth, Income Distribution and Democracy: What the Data Say,
Journal of Economic Growth 1, 149187.

Ravallion, M. & Chen, S. (1997). What Can New Survey Data Tell Us about Recent
Changes in Distribution and Poverty? World Bank Economic Review, 11(2): 357-82.

Ravallion, M. & Chen, S. (2003). Measuring pro-poor growth. Economics Letters 78

(2003) 9399.

Ravallion, M. (2004). Development Research Group, World Bank 1818 H Street NW,
Washington DC World Bank Policy Research Working Paper 3242, March 2004.

Rodrik, D., Subramanian, A. & Trebbi, F. (2002) Institutions Rule: The Primacy of
Institutions over Integration and Geography in Economic Development. IMF Working
Paper No. 02/189 Available at SSRN:

Rodrik, D. (2006).Industrial Development: Stylized Facts and Policies. Draft chapter,

August 2006 for UN-DESA Industrial Development for the 21st Century. Unpublished.

Sechler, S & Guinan, J. (2005). Trade-plus: Development and the Politics of the Doha
Round, World Bank 2005 Annual Report on World Hunger, Strengthening Rural
Communities. Washington: World Bank Institute.

Siphana, S. (2003). Mainstreaming Trade for Poverty Alleviation: A Cambodian
Experience, paper presented at the EUs Conference on the Future of Textiles and
Clothing after 2005, held in Brussels, 5-6 May, 2003.

Sobhan, R. (2002) Pro-Poor Fiscal and Monetary Policies: Toward Correcting

Structural Injustice in South Asia. Dhaka: Centre for Policy Dialogue.

South Africa, (2006). Accelerated Shared Growth Initiative for South Africa. available at

(SDC) Swiss Agency for Development and Cooperation (2004). Pro-Poor Growth with
Focus on Income Poverty. Synthesis of Agora Presentation, June 2004, Bern.

Stryker, D. & Pandolfi, S. (1997), Impact of Outward-Looking Market-Oriented Policy

Reform on Economic Growth and Poverty, Harvard Institute for International
Development, CAER II Discussion Paper (April).

Thurlow, I. (2006). Has Trade Liberalisation in South Africa Affected Men and Women
Differently? Washington: International Food Policy Research Institute, DSGD Discussion
Paper No. 36.

Townsend, I. (2006). The WTO Doha Development Round: Where next for world Trade.
Parliamentary Research Paper 06/43, House of Commons Library 21 September 2006.
(Available at

UNCTAD (2003) Report on Economic Performance in Africa: Trade Performance and

Commodity Dependence.

UNCTAD, 2003. Trade in Services and Development Implications. 16 December 2003,


Wie, T.K. (2006). Policies affecting Indonesias Industrial Technological Development

in ASEAN Economic Bulletin. Vol 23(3):341-359.

Winters, L.A., McCulloch, N., & McKay A. (2004). Trade Liberalization and Poverty:
The Evidence So Far. Journal of Economic Literature. vol 42 (1):72-115.
Wood, G. (2001). "South African Trade Unions in a Time of Adjustment," Labour/Le
Travail 2001 vol 47 (6). Available online

UNIFEM, (2007) UNIFEM Currents: March 2007. Available at