Content uploaded by Frances Lund
Author content
All content in this area was uploaded by Frances Lund
Content may be subject to copyright.
Is there a rationale for conditional
cash transfers for children in South
Africa?
Frances Lund*, Michael Noble**, Helen Barnes**
and Gemma Wright**
*Associate Professor, School of Development Studies, University of KwaZulu-Natal and
Associate Member of the Department of Social Policy and Social Work at the University of
Oxford, England.
E-mail: lundf@ukzn.ac.za
** Respectively: Professor of Social Policy, Research Fellow and Senior
Research Fellow at the Centre for the Analysis of South African Social Policy, Department of
Social Policy and Social Work at the University of Oxford, England.
Email: Michael.noble@socres.ox.ac.uk
Working Paper No 53
July 2008
ISBN 978-1-86840-664-7
Acknowledgements
Geoff Budlender; Albert Dlwengu; Beth Goldblatt; Annie Leatt; Shirin Motala; Paula
Proudlock
0 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
ABSTRACT
The South African state awards unconditional means-tested cash transfers to
the caregivers of some eight million poor children. Amidst increasing
demands on the state for social assistance, the question has been asked:
should the Child Support Grant (CSG) be made conditional on education or
health related behaviour to enhance its effectiveness? Issues relating to the
popular Latin American conditional cash transfer programmes for children
are summarised. The history, current reach and impact of the CSG are
described, as well as administrative strengths and weaknesses in delivery.
We argue that, to be in line with South Africa’s Constitution, the structural
problems of the supply side of education and health provision should be
addressed, as well as problems with the delivery of the CSG, rather than
considering whether to impose conditionalities that could further exclude
poor children and their caregivers from this modest but vital form of support.
1. INTRODUCTION
Cash transfers are one of a range of measures for addressing poverty. A
regular amount of money is allocated directly to particular groups such as
the elderly, the unemployed, or children. Cash transfers differ from in-kind
interventions, such as the provision of food or food stamps, in that the
former give the recipient choice about how to spend the money.
The nature and extent of cash transfers (CTs) that a country provides can be
viewed as a reflection of its welfare regime. Indeed Esping-Andersen (1990)
based his well known typology of welfare states on the extent to which a
state, through its transfer schemes, de-commodifies labour. A country with
residual means tested provision such as South Africa would typically fall
into his ‘liberal’ classification. Liberal welfare regimes are characteristically
associated with ideologies which see the causes of poverty rooted in
individual pathologies rather than having structural causes. Such liberal
regimes tend to place the responsibility for exiting poverty firmly on the
individual (Handa & Davis 2006; Schubert & Slater 2006). Consequently the
role of the state in poverty alleviation through CTs is seen as undesirable.
Often such transfers, where made, will be buttressed with conditions to
ensure that the poor ‘take responsibility’ for their own redemption from
poverty. To some extent this is a crude over-simplification – there are some
countries which would otherwise be thought of as social democratic regimes
where conditionality is attached to state transfers (Kildal 2001). However,
CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA 1
the defining feature of such social security systems is that their
conditionalities tend to positively support the individual in, for example, a
return to work, and they are characterised by significant state financial
commitment (OECD 2003). In liberal regimes in contrast the focus is often
on creating barriers to inclusion in the social assistance scheme with an
explicit or covert agenda to reduce social security expenditure (Quinn &
Magill 1994).
A simple taxonomy of CTs distinguishes between non means tested and
means tested benefits. Non means tested benefits can be contributory, such
as unemployment insurance, where the worker, employer and the state make
contributions, or non-contributory (categorical), such as Child Benefit in the
UK, payable to all children of specified ages. Means tested benefits are
referred to as social assistance and may be either unconditional or
conditional.
In an unconditional programme, once a person qualifies to enter the scheme,
the amount is an entitlement, for a fixed period. The current South African
CSG is an example of a means tested unconditional social assistance transfer
in respect of children. It was introduced in 1998 as one of a fairly extensive
system of relatively unconditional benefits which go to children, people with
disabilities and elderly people. The system of social assistance has continued
to expand since 1994, even in the context of the overall neo-liberal macro-
economic policy. In early 2006, the three main grants reached well over one
quarter of the total population of 44 million: 6.98 million children were
beneficiaries of the CSG, 2.25 million the Old Age Grant, and 1.3 million
the Disability Grant (Budlender & Woolard 2006).
Conditional cash transfers (CCTs) formally require some type of behavioural
compliance on the part of recipients. CCTs relating to children have a
lengthy pedigree (Farrington & Slater 2006). As early as 1987 some states in
the USA had used the waivers afforded by the Omnibus Budget
Reconciliation Act 1981 to introduce requirements for school attendance as a
condition of receipt of Aid for Families with Dependant Children. These
programmes, collectively referred to as ‘Learnfare’, were typically coercive
rather than supportive and conformed to the liberal ideology referred to
above (Quinn & Magill 1994).
The south has recently seen the rapid spread of CCTs, based primarily on
evidence of their impacts in Latin American countries. CCTs relate to both
2 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
poverty reduction strategies and the Millennium Development Goals and
their main focus is children. They aim to address short-term poverty by
providing poor households with cash income, and in the long-term, through
health-related and education-related behaviour, their goal is to break the
inter-generation transmission of poverty by building children’s ‘human
capital’.
One effect of the South African political transition was to make grants more
accessible, especially to rural people, and this together with HIV/AIDS is
leading to increased demands for social assistance, especially for the
Disability and the Foster Child Grants. In the face of these and other
pressures, the idea of conditionality is gaining favour in South Africa.
This paper focuses on South African cash transfers in respect of children in
poor households. First, the impressive evidence for the reach and impact of
major CCT programmes, particularly in Latin America, is reviewed and
emerging problems are identified (Section 2). The background to the South
African CSG, including conditionalities that were considered in its design, is
discussed (Section 3). The patterns of take-up and reach of this grant are
analysed and evidence from early assessments of its impact is presented
(Section 4). A classification of characteristics, requirements and
conditionalities that have to be fulfilled to access social assistance is
detailed, and using evidence from studies of the CSG it is shown how some
formal conditionalities have been imposed, and how some administrative
requirements have become de facto conditionalities (Section 5). Finally the
appropriateness of conditionality in South Africa is considered in the light of
the available research evidence and given the type of welfare regime
promised for South Africa in both the Constitution and in high level policy
documents. The arguments for attaching conditionalities to nutrition,
attendance at Early Childhood Development (ECD) facilities, and school
performance are considered. It is suggested that these are primarily supply-
side problems which would not be improved by introducing further
conditionalities. Instead, the systemic problems should be tackled in line
with the Administrative Justice Act (AJA) and with Batho Pele (‘People
First’) principles (Section 6).
2. CONDITIONAL CASH TRANSFER PROGRAMMES
Child-oriented CCT programmes are familiar in industrialised countries, and
a range of CCT programmes has recently been introduced in Latin America,
CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA 3
the largest of which are Mexico’s Oportunidades (formerly Progresa), a
health, education and nutrition programme, and Brazil’s Bolsa Escola
(‘school stipend’), which respectively reach four million and five million
households (Rawlings & Rubio 2005). In Brazil, eight cash transfer
programmes have now combined into Bolsa Familia (‘family stipend’) and
together reach eight million or 15.5 per cent of Brazilian households (Soares
et al. 2006). Other countries with CCTs include Bolivia, Colombia,
Honduras, Jamaica and Nicaragua.
The focus of conditionalities is to a greater or lesser degree on poverty
alleviation, civic responsibility, investment in human capital, and
maintaining existing human capital. CCTs require that beneficiaries change
their behaviour in certain ways, and assume that money will enable the
recipient to do so. In other words, they assume that it is a cash constraint that
is keeping the recipient from, for example, going to health services or
school; that health services and schools exist; that cash is the appropriate
incentive to encourage attendance; and that improved health status and
school attendance will impact on school achievement, and thereby improve
life chances.
Different programmes have different characteristics, but typically, monthly
or bi-monthly cash grants are given to the mothers of children of school-
going age or younger in poor households. The transfer is conditioned on
health-related and education-related behaviours. In health, usual
requirements are that all family members must be taken to the health
services a minimum number of times and mothers must attend health
education classes. In education, a child must attend school for a certain
number of days, for example, 75 or 85 per cent of school days in a year.
Some programmes make continuation of benefits conditional on school
performance as well. Some specifically give a larger transfer for girls, to
address the fact that they have significantly lower enrolment and attendance
figures than boys. In addition to the health and education requirements, some
programmes require parents (mothers) to attend community meetings, or do
some hours of ‘community work’.
Programmes vary significantly in terms of the ages of children that they
target. Under fives are often the focus of the health component of the CCTs:
Bolsa Escola includes those from six to 15 years old, in grades 1 to 8;
Progresa focuses on children up to the age of 18, including grades 3 to 9;
and the Programa Nacional de Becas Estudiantiles (‘national programme of
4 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
student scholarships’, commonly known as Becas) in Argentina is used as a
means to encourage transition to and progression through senior school
(Heinrich 2006). Programmes vary also in the extent to which they focus on
child labour.
Eligibility for CCT programmes is determined by a wide range of
assessments of poverty. The most stringent (and costly) is the verification of
household income. The amount of the transfer itself is based on some
criterion connected with the programme purpose. It can be set according to
an assessment of a poverty level, sometimes by determining direct costs of
programme participation, such as transport to get to health services, as well
as by determining the opportunity costs of fulfilling the conditions for
receiving the transfer.
An innovative aspect of many of these programmes is that, while they target
children and families, the money is paid to the mother rather than to the
father or the household head. Some programmes have as an explicit
objective that this should lead to ‘women empowerment’; in others, the
payment to mothers is grounded more in the international evidence which
shows that money will be spent ‘better’, in terms of children’s health and
education, if it goes to women.
CCT programmes, especially Progresa, have been extensively and
rigorously evaluated as part of programme design, with most countries
conducting randomised experimental and control studies (Rawlings & Rubio
2005). The assessments are positive about most aspects of the programmes.
In summary, they show that they can be well targeted, and work effectively
as incentives for investing in human capital (syntheses can be found in
Skoufias 2005; de la Briere & Rawlings 2006; Rawlings 2004; Rawlings &
Rubio 2005).
In Bolsa Escola and Programa de Erradicacao do Trabalho Infantil in
Brazil, Progresa, and the pilot for the Red de Proteccion Social (RPS) in
Nicaragua, the CCTs had a dramatic effect on school enrolment and
attendance, while Becas increased attendance, reduced repetition rates, and
improved school performance (Heinrich 2006).
With regard to child labour, programmes have reduced the probability of
children working, with a stronger effect for boys than for girls. In Nicaragua,
the pilot RPS reduced the probability of children working in high risk
CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA 5
activities. There, although after-school activities were available to all school
children, only the CCT programme children actually spent increased hours
attending, suggesting that it was the incentives that led to the behaviour
change (Rawlings 2004).
In terms of health, again most of the evidence is positive. The Mexican,
Brazilian and Nicaraguan programmes all resulted in an increase in nutrition
monitoring, in immunisation rates, and in growth rates (Behrman &
Hoddinot 2000; Maluccio & Flores 2005; Rawlings & Rubio 2005).
Furthermore Progresa reduced illness in children (Gertler 2004), as well as
sick leave days for adults. The nutritional supplement has been the least
successful of the health components, with families diluting the supplement
to make it stretch further, or sharing it with all household members, thus
weakening the intended focus on the nutritional status of children under five
(Rawlings & Rubio 2005).
Most assessments judge the routing of the CCT via the mother to have
contributed towards women’s sense of self-esteem and self-confidence, as
well as giving them control over (part of) the household purse. Evaluations
of Progresa have also pointed to the way in which the programmes have
raised awareness of the importance of girls’ education (Adato 2000; Adato,
Coady et al. 2000; Adato, de la Briere et al. 2000).
A number of problematic areas would be expected in new programmes that
have grown so rapidly. The first concerns gender relations. A quantitative
and qualitative assessment of the effect of Progresa on women’s status and
intra-household relations found that the programme had indeed led to an
increase in women’s sense of their own empowerment. The programme
placed extra demands on women’s time, but the women found it worthwhile
if the result was better chances of schooling for their children (Adato, de la
Briere et al. 2000); however, they expressed the wish that their husbands
would attend the education classes as well. Molyneux (2006) on the other
hand offers a strong critique of Progresa as a maternalist programme
suggesting that the programmes place additional demands on women’s
scarce time, and they are expected to juggle multiple paid and unpaid work
activities even further.
Second, the long-term effects of the CCT programmes in addressing poverty
are not yet known, as they have been in existence for only a decade. In what
circumstances will the proven increased educational attendance, and
6 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
performance, be translated into increased opportunities for moving out of
poverty? Four factors have been identified as mediating the relationship: the
quality of education, rates of employment, the ability of the labour market to
absorb labour, and rates of return to education (Bourguignon et al. 2002).
With regard to the employment that is available to young people when
leaving school, in most developing countries, informal employment
including self-employment constitutes the larger part of all employment.
Thus critical in the quality of schooling is whether schooling equips children
with the idea that many will have to make their own work, rather than find
work provided by someone else, and whether the curriculum offered will
enhance their chances of doing so successfully. The question that needs to be
answered is what level of schooling, of what sort, is necessary to make a
difference to both finding and making work?
Third, it is not yet clear whether it was the conditionality that made the
difference, or whether just the cash itself would have achieved the same
effects (Gertler 2004). A Brazilian study compared the targeting and poverty
reduction effects of the unconditional benefit for elderly and disabled people
with the conditional Bolsa Escola (Soares et al. 2006). Both were well-
targeted to the poor, and both had marked and similar effects on poverty
reduction. Conditionalities impose costs on programmes, but assessing such
costs is complex.
Finally, most conditional programmes assume that there is a demand-side
problem that can be attended to by providing incentives to individuals.
Increasingly, evaluations of the CCTs suggest this is a limitation, and that
supply-side problems need to be addressed as an integral part of programmes
(Rawlings & Rubio 2005). Nicaragua and Honduras, for example, have
invested part of programme money into improving educational infrastructure
(Caldes et al. 2004). It is suggested that long term benefits are reliant on
better schooling provision and better health services.
Having reviewed the evidence from the conditional transfer programmes in
Latin America, we turn to the unconditional transfer in South Africa.
3. THE SOUTH AFRICAN CHILD SUPPORT GRANT
State social assistance in South Africa started in 1929 with the introduction
of Old Age Pensions for the white and coloured population, and then later
the inclusion of all racial groups. It grew over the course of the last century
CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA 7
to include support for people with disabilities, and for families, women and
children. Cash transfers were means tested for income and assets, and
payable from general revenue. At its inception, and then under the apartheid
regime, there was racial discrimination in every aspect of the system, in
racially separate administrations, level of payments, income testing, and in
timing and method of payment (Lund 1993). In anticipation of the transition
to democracy, and the need to forge uniformity between the different welfare
administrations with respect to welfare entitlements, the mid-to-late 1980s
saw the gradual equalisation of benefits and administrative procedures
across the differently classified racial groups. By September 1993, most of
the inequalities had been removed from the main welfare grants for elderly
and for disabled people (Republic of South Africa 1996).
Severe inequality remained in the State Maintenance Grant (SMG). This
grant had been designed to support poor women and their children up to the
age of 18, where the woman had been widowed or abandoned, or the father
had been committed to a state institution. All racial groups were eligible for
this means tested grant, but in 1993 the SMG went mainly to coloured and
Indian women and children (50 and 40 per thousand children respectively).
Only 13 per thousand white children received it, because of wealth levels,
while very few of the poorest African women could access it: in 1993 it
went to only one in a thousand African children (Republic of South Africa
1996). Eight million African people lived in the Bantustan areas, where the
welfare departments by and large did not administer this grant. Social
workers also appeared to act as gatekeepers who approved of pensions for
elderly and disabled people, but not for younger single women and children.
A committee of enquiry was established in 1996 to investigate all forms of
child and family support, and one major recommendation was to replace the
SMG with a CSG.
The CSG was initially available to children from birth to their seventh
birthday, and then extended in three phases to include children up to their
fourteenth birthday. In making the award to the child’s ‘primary caregiver’,
the CSG broke with former welfare convention which was to channel child-
oriented support only through biological parents or legally defined adoptive
or foster parents. This was an attempt to deal with specific aspects of South
African society which had emerged over the periods of colonialism and then
apartheid: the absence of many parents, especially fathers who went to the
towns, mines or white-owned farms to work; customary marriages and
customary polygamy; fluid and mobile household structures; and high rates
8 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
of caring by grandmothers and aunts. All of these characteristics were
present before the advent of HIV/AIDS; they have become more pronounced
since.
The value of the CSG when introduced in April 1998 was R100 per month
per beneficiary and in July 2006 the grant was valued at R190. As such it
was much smaller than the R820 per month grants for elderly people and
people with disabilities. The committee recommending this cash transfer
initially wanted a universal benefit for all children of a certain age, both as
an appeal to solidarity under the new non-racial regime, and because, given
widespread poverty, means testing is inefficient and costly. A compromise
was reached on what was to have been a simple means test based on the
income of the primary caregiver and her/his spouse or partner, and with
differentials for urban/rural location and for type of dwelling unit. Everyone
who qualifies according to the means test is paid at the same rate.
A number of policy options other than an unconditional cash transfer were
considered. If the SMG was to be phased out, then that R1.3 billion budget
item could be diverted to somewhere else in the welfare budget, rather than
be lost. Alternatives needed to have institutional capacity, deliver on a
relatively large scale, and redress racial and spatial inequities in welfare
provision. Support for Early Childhood Development (which in South Africa
is defined to include children up to their tenth birthday) was one alternative
but the sector was institutionally weak at that time. There was strong support
for child nutrition, but experts argued that, given current government policy
and capacity, a CT was likely to be more effective than in-kind nutrition
supplements or food vouchers for children. General child welfare services in
South Africa had been severely under-funded, and would come under
increasing pressure with the spread of AIDS. However, they were seen as
complementary, rather than an alternative to CTs.
It was decided to concentrate on children in their earliest years. In South
Africa, the vast majority of children are born into the health services, then
have no contact with public institutions until they attend school. The CSG
aimed to address vulnerability during this period, in the form of a CT, to
very young children, related to some measure of the costs of nutrition of
children at this age.
In designing the CSG, a number of measures were considered that would
increase synergies with primary health care, improve on current delivery
CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA 9
systems, and lower costs to the poor, while recognising that taking up
entitlement simply does bear some cost. A basic administrative requirement
was that the primary caregiver applying for the benefit had to have an
identity document, and the child a birth certificate.
Health-related conditionalities were considered. One was for applicants to
have a Road to Health card (which included both growth monitoring and
immunisation status). Another was to move from a complicated and costly
annual review mechanism that was applied to other social grants, to a single
visit to the health services when the child was between 24 and 30 months
old, the period in which child health services are most concerned to check
immunisation and overall health status. Both these ideas were turned down
by the health ministry at the time; the Road to Health card was re-introduced
later as a requirement.
4. THE PERFORMANCE OF THE CHILD SUPPORT GRANT
In order to assess whether conditionalities would be appropriate in the
case of the CSG, it is necessary first to assess the present performance of the
grant both in terms of take-up and reach, and in terms of the extent to which
it alleviates poverty and supports the development of human capital. It is
unwise to draw firm conclusions about the performance of a programme in
existence for only a decade. Nevertheless databases exist which have
allowed reliable early assessments of the CSG even in the absence of
randomised experimental studies.
Eligibility and take-up
Table 1 provides information about the proportion of children under eleven
whose caregivers were eligible to receive the CSG in January 2005. This is
calculated by dividing the number of children under eleven who have an
eligible caregiver by the total number of children under eleven. This can be
seen as a measure of the prevalence of children living in low income
families in an area.
Nationally, in January 2005, 7.39 million children under eleven were eligible
to receive the CSG and there were 10.96 million children under eleven in
total. This means that the caregivers of 67 per cent of children were
potentially eligible to receive the CSG. Given the provincial differentiation
in poverty levels, there is considerable variation in terms of the potential
reach of the CSG at provincial level: 41 per cent of children under eleven in
10 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
the Western Cape have caregivers who are estimated to be eligible for CSG,
while 79 per cent of children under eleven in both Limpopo and the Eastern
Cape have caregivers who are estimated to be eligible for the grant.
Table 1 Proportion of 0-10 year olds whose caregivers are eligible to
receive the CSG (January 2005)
Province Children aged 0-10
whose caregivers
are eligible
Children aged 0-10 Eligibility rate
(%)
Western Cape 386300 950800 40.6
Eastern Cape 1390300 1759500 79.0
Northern Cape 134000 209400 64.0
Free State 470800 659200 71.4
KwaZulu-Natal 1782200 2400600 74.2
North West 629700 892100 70.6
Gauteng 808200 1754600 46.1
Mpumalanga 585200 820900 71.3
Limpopo 1199500 1518400 79.0
South Africa 7386100 10965600 67.4
Source: Noble et al., 2005.
Take-up rates for January 2004 and January 2005 are compared in Table 2.
The take-up rate for South Africa as a whole for the relevant age groups
increased from 63 to 71 per cent between January 2004 and January 2005. In
terms of take-up for under nines only, the increase was greater, changing
from 63 to 73 per cent. This suggests that take-up is lower for the nine and
ten year olds who only became eligible for the CSG in April 2004, and that
this age group brought down the overall take-up rate for January 2005
(Noble et al. 2005). For the relevant age groups, all provinces saw an
increase in take-up rate over the year; the Eastern Cape and North West
Province had the greatest increase at thirteen percentage points, while
KwaZulu-Natal had only a two percentage point increase. The take-up rates
for under nines only also increased between January 2004 and January 2005.
Table 2 Take up of CSG at provincial level
Province January 2004
take up rate 0-8
y
ear olds (%)
January 2005
take up rate 0-8
y
ear olds (%)
January 2005
take up rate 0-10
y
ear olds (%)
Western Cape 78.7 88.5 84.0
Eastern Cape 53.5 68.2 66.3
Northern Cape 58.9 63.4 62.7
CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA 11
Free State 57.5 66.4 64.5
KwaZulu-Natal 61.9 68.3 64.3
North West 59.1 73.7 72.0
Gauteng 75.4 84.4 83.3
Mpumalanga 67.2 74.6 73.7
Limpopo 67.8 74.1 73.9
South Africa 63.5 72.9 70.7
Source: Noble et al., 2005.
Poverty alleviation and human capital development
All studies to date show good targeting for poverty, except that some
fraction of the very poorest are excluded (Budlender & Woolard 2006;
Barrientos and DeJong 2006; Case et al. 2005).
The income thresholds for the means test have stayed the same since
inception (Leatt 2004) meaning that only the extremely poor are now
included, and there has been inconsistency with the way urban and rural
areas are defined for the purposes of the means test, which excludes many
who should be eligible (Budlender et al. 2005; Goldblatt et al. 2006; Hall
2005).
School attendance rates among young children are high across all racial
groups, with attendance rates for six, seven and eight year olds at 83, 97 and
98 per cent respectively (Budlender & Woolard 2006). The Umkhanyakude
study of 10 000 households in KwaZulu-Natal found that receipt of the CSG
led to an eight per cent increase in school enrolment among six year olds,
remarkable because households were also shown to be poor, and enrolment
rates already high (Case et al. 2005). The 2004 General Household Survey
showed a small but positive link between the CSG and school attendance
(Budlender & Woolard 2006).
No existing database contains both CSG and child labour data (Budlender &
Woolard 2006). South Africa does not (yet) have a serious problem of child
labour, and there is no evidence that school attendance is significantly
affected by withdrawal of children in order to work.
5. CONDITIONALITIES OR CONDITIONS OF ENTITLEMENT?
In South Africa, conditionality is presently being used to mean different
things. Some (Goldblatt et al. 2006) refer to the requirement of an ID as a
conditionality, and others (Budlender & Woolard 2006) restrict it to
12 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
requirements such as signing up for development programmes or getting
immunisations. An attempt is made here to distinguish between various
factors, some of which are conditionalities.
With unconditional transfers, the Constitution or other legislation defines a
right which becomes a material entitlement for a person with certain
characteristics who meets certain qualifications. When conditionality is
introduced, the applicant meeting certain requirements to get the benefit
must in addition conduct him or herself in certain ways in order to continue
receiving it. Some things which are not conditionalities, however, act to
exclude some who are eligible, and there are sometimes blurred lines
between conditionalities and other requirements.
It is useful to distinguish between the following.
First, there are the characteristics of the person which are needed to qualify,
such as being of a certain age, or being disabled. These are not
conditionalities.
Second, there are administrative requirements such as possession of a birth
certificate or proof of citizenship. These are not conditionalities, but do act
as barriers to access. The home affairs department which issues documents
has had difficulties with delivery, and the repeat visits and long delays
present significant opportunity costs, in terms of time and money spent on
the process, and in the grant income foregone (Waddell, 2002). Proof of
socio-economic status is needed for the means test. This is not a
conditionality, but is another forceful mechanism of exclusion. In the field
the means test is unevenly applied, which could reflect the fact that it is a
relatively new grant, with officials still learning the procedures, or that
officials decide for themselves to what extent it should be applied.
Other examples of administrative requirements are that applicants have also
been asked for their marital status, marriage certificates and divorce decrees
(Hall 2005) and particular districts or offices have imposed employment-
related requirements, asking applicants for an affidavit saying they were
unemployed (Hall 2005) or requiring the applicant to produce the ‘brown
card’ from the labour department stating the applicant has registered as
unemployed and job-seeking (Children’s Institute 2005; Hall 2005). None of
this is legally necessary, and all serve as barriers to access.
CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA 13
Third, there are requirements in the regulations, called ‘conditions’, such as
‘the child has to be accommodated, fed and clothed’ and stipulating that the
grant has to be spent on the child. These are normative injunctions. They
cannot easily be monitored, and can allow for abuse of powers as they are
not sufficiently rule-based. Some appear to be reasonable: the child must
continue to be in the care of the person receiving the grant; the state must be
allowed reasonable access to the child and the house that the child is residing
in; the grant has to be spent on the child; and the grant is subject to periodic
review. However, a specific requirement, that the child must have
accommodation and be properly fed and clothed, is not reasonable given that
the level of the grant is not linked to any objectively determined estimate of
the costs of accommodation, clothing and food. The level of the CSG was
initially linked only to nutritional cost, and the value of the grant has eroded
over time. Providing shelter, food and clothing for children is a reasonable
thing to expect parents to do, but this grant at its present value cannot enable
it all to happen.
Fourthly, there are requirements created that attempt to regulate conduct,
some of which are in the regulations, for example immunisation, and some
of which are ultra vires, for example providing proof of having sought
private maintenance from the child’s father (Goldblatt 2005; Goldblatt &
Yose 2004a, 2004b; Hall 2005). Such requirements begin to look like
conditionalities, but could be interpreted as once-off conditions of
entitlement.
The possession of the Road to Health Card appeared as a requirement in the
1998 regulations, and then did not reappear in the amended 1999
regulations. The 2002 regulations state that ‘(the caregiver) shall ensure that
the child concerned receives immunisation and other health services where
such services are available, without charge’. In the 2004 regulations (not yet
implemented), the words ‘where such services are available, without charge’
have been removed from this clause.
The onerous requirement of providing proof of having sought private
maintenance did not appear in the 1999 regulations, yet as late as 2005, still
appeared as part of procedures in a number of offices.
Fifth and finally, there is the category of true conditionalities, where the
beneficiary has to keep on doing something, such as attending school and/or
accessing health services. Lack of compliance leads to partial or full loss of
14 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
the benefit. No such conditionalities legally exist in South Africa at the
present time. In the administration of the CSG, some have been imposed that
are outside of law, but formally encoded in official documents, while some
are outside of law and the result of officials’ discretion, for example school
attendance.
School-related conditionalities were not considered in the design of the CSG
not only because it was planned that it should be as unconditional as
possible, but also because it was targeted at pre-school children, or to
overlap only with the very first years of school. Had it been intended for
older children, school-related conditionalities would not have made sense
because of the high rates of primary school enrolment of both boys and girls
in South Africa. The age extension of the CSG has led to some offices now
imposing school-linked requirements which are conditionalities. Sometimes
a letter from the school is required to confirm school enrolment or school
attendance (Children’s Institute 2005; Hall 2005) or the police must certify
school enrolment (Hall 2005).
6. DISCUSSION
The CSG has reached a large number of children in a relatively short period
of time. Analyses show that it has been reaching rural areas, is well targeted
for poverty, and appears to increase the length of time children are in school.
It was designed as an unconditional benefit, with the exception of minimal
qualifications to enable efficient administration, and one or two relatively
minor requirements needed to create synergies with other departments, or
minimise corruption. Over time, a range of additional requirements have
been added, some of which amount to de facto conditionalities that are not
always legal. In the transitional space in South Africa, bureaucrats have
imposed their own rules. There are many examples of officials exercising
discretion in favour of individual clients, or challenging punitive aspects of
the regulations themselves. On the whole, however, administrative discretion
appears to be subverting the aim of the broader social policy by imposing
additional costs on the poor. In addition, spatial settlement patterns imposed
by apartheid and lack of administrative capacity and inefficiency impose
costs on those seeking access.
There is no way of directly comparing the costs of administration of the
unconditional South African CSG and the CCTs in Latin America, but it is
possible to re-visit the problems that have been identified with the CCTs,
CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA 15
and see whether comparable issues might be experienced in the CSG,
whether conditional or not.
Some Latin American schemes have the empowerment of women as an
explicit programme goal, assumed to be achieved by giving women the
income, and through their attendance at group activities. Women have
primary responsibility for signing up, and then for ensuring that children
attend school, and family members go to the health services. In South
Africa, the woman beneficiary is likewise primarily the conduit for the CSG
and she assumes responsibility for the time-consuming application process.
Women are faced with patriarchal forms of authority at every turn. Proof of
residence (a requirement in some areas) has to come from the tribal authority
(Goldblatt & Yose 2004a); affidavits of various kinds have to be secured at
police stations, which are very masculine domains; applying for paternal
maintenance is through a judicial procedure which is known to make women
vulnerable. During screening procedures, monitors observed officials giving
moralising lectures which were ‘patronising and inappropriate’ (Goldblatt &
Yose 2004a).
The Latin American CCT programmes have had positive effects on school
performance. The South African CSG was initially aimed at children up to
the age of seven. The age extension to fourteen years has led to some
informal conditionalities being imposed, even though school attendance
rates are good (though they may become less so with HIV/AIDS). Given the
parlous quality of education for poor South Africans in both urban and rural
areas, it is not necessarily getting children to school that matters in breaking
long term poverty: it is about resources and facilities, or management, or
teaching practice at schools (see Chisholm 2006: 202-203). It is a supply-
side problem, similar to that pointed out in Bourguignon et al. (2002) for
Brazil, and by Britto (2004) for the CCT programmes in general. Poor
teaching and lack of leadership in under-resourced schools are common, and
there are low returns to education for a number of years. Enrolment and
attendance are necessary conditions in trying to escape poverty, but they are
not sufficient.
One argument used for extending the age limit of the CSG was because of
school costs. Regulations do not at present make it a conditionality that the
child has to go to school. If this were ever introduced, it would require a
massive increase in the amount of the grant, to compensate for expenses that
should be paid by the education department, or, at least, should not be
16 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
funded through the welfare department. Expenses paid by households
include school fees (fee exemption for children from poor households does
not always happen in practice), school uniforms, and contributions to school
cleaning equipment and building funds.
Any attempt to create conditions such as family participation in livelihood
activities is undesirable. This harks back to the conditionality initially
imposed on the CSG which was for applicants to sign up for development
projects. This was completely unreasonable given the fact that poverty often
coincides with places where there are no such projects. Expectations that
caregivers should attend health workshops would likewise impose additional
costs on poor women, as suggested by Molyneux (2006).
As noted in Section 2, a core question in the assessment of CCTs was
whether it was the conditionality itself that made for such impressive
positive impacts, or whether the cash alone would have achieved the same
objectives. If conditionalities are to be considered in South Africa, what
could likely candidates be? An official South Africa source mentioned
exploring links with nutrition, ECD, and free primary health care (especially
immunisation) (Plaatjies 2006: 10).
With regard first to nutrition, there is a need for a national programme of
support for nutrition for children in their earliest years. Children of school-
going age have access to the unevenly implemented National School
Nutrition Programme, which reached about half of all primary school
learners over the years 1994/5 to 2003/4 (Kallmann 2005:11). South Africa’s
children need nutritional support most critically between birth and two years
of age, but at this stage, there is no programme with which CSG children
could be linked. Second, linking the CSG to attendance at an ECD facility
will not address the fact that the lack of ECD participation is a supply-side
problem, with only one in six children able to access them as there are too
few facilities or they are too expensive (Department of Education 2001).
Third, in terms of primary health care, there is free access for young
children, though the service is uneven. It would seem reasonable to expect
that in a society at South Africa’s level of development, immunisation
should simply be a free and accessible service for all citizens, without
linkages to other programmes.
This review of the evidence suggests that introducing behavioural
inducements to poor people in South Africa to ensure the best educational
CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA 17
and health outcomes for their children should not be the main focus of
attention for policy makers. Such conditionalities would be inconsistent with
the (essentially) social democratic social policy regime set out in the
Constitution and the Reconstruction and Development Programme White
Paper. In both documents the causes of poverty and unemployment are
unambiguously stated as having a structural rather than personal aetiology –
that is, they form part of the apartheid legacy. Chapter 2 of the Constitution
comprises a Bill of Rights which sets out an unparalleled series of socio-
economic rights in respect of social security, health, education and housing.
These rights speak to a set of inclusive social policies with a significant role
for the state consistent with a social democratic welfare regime rather than
residual minimalist welfare provision which typifies the liberal welfare
regime. Conditional social security, based on assumptions that poor parents
are in some way culpable if their children fail to attend school or attend
clinics is inconsistent with the structural explanations for poverty which are
implicit in the Constitution.
The new South African government has done much to try and address and
redress the racism and repressive aspects of the old apartheid state. It has
introduced the Batho Pele (‘People First’) set of principles into
administration to ensure fairer and more transparent service to the public,
and an Administrative Justice Act (AJA), yet procedures in the
implementation of the CSG contradict both. The AJA requires that ‘organs
of state may not act capriciously and arbitrarily’ and ‘Decisions must be
rationally related to the purpose for which the power was given’ (Currie &
Klaaren 2001: 16-18), but it is clear that there is a great deal of
administrative discretion. The requirement to apply for private maintenance
has no rational relationship whatsoever to the CSG; the widespread habit of
telling applicants to ‘come back within three weeks’ bears no relation to
whether that is an appropriate time to return. Principle 8 of Batho Pele
stipulates that ‘Public services should be provided economically and
efficiently in order to give citizens the best value for money’, but poor
people incur unreasonable costs in seeking and maintaining access to the
system (Republic of South Africa 1997).
It is widely acknowledged that there is a large gap between policy and
implementation in post-apartheid South Africa, and this is no surprise given
the ambitious hopes that there were for fundamental social, economic and
political reforms. The CSG has rolled out very rapidly; by stealth, some of
the policy intent is being subverted by administrative action. Patterns of
18 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
take-up suggest that the means test is an unnecessary and ineffective
measure, and it should be lifted altogether or be replaced by a far simpler
measure, as suggested by Goldblatt et al. (2006). Indeed, a universal
categorical grant for children would sit with greater ease in what on paper
appears to be a social democratic policy agenda.
REFERENCES
Adato M (2000) The impact of PROGRESA on community social
relationships, Final Report, Washington D.C.: International Food
Policy Research Institute.
Adato M, Coady D and Ruel M (2000) An operations evaluation of
PROGRESA from the perspective of beneficiaries, promotoras, school
directors, and health staff, Final Report, Washington, D.C.:
International Food Policy Research Institute.
Adato M, De La Briere B, Mindek D and Quisumbing A (2000) The impact
of Progresa on women’s status and intra-household relations, Final
Report, Washington, D.C.: International Food Policy Research
Institute.
Barrientos A and Dejong J (2006) Reducing child poverty with cash
transfers: a sure thing? Development Policy Review, 24(5): 537-552.
Behrman J and Hoddinott J (2000) An evaluation of the impact of Progresa
on pre-school child height, Discussion Paper Brief, 104 Washington
D.C.: International Food Policy Research Institute.
Bourguignon F, Ferreira F and Leite P (2002) Ex-ante evaluation of
conditional cash transfer programs: the case of Bolsa Escola, Policy
Research Working Paper, 2916, Washington D.C.: World Bank.
Britto T F (2004) Conditional cash transfers: why have they become so
prominent in recent poverty reduction strategies in Latin America,
Working Paper Series, No. 390, The Hague: Institute of Social
Studies.
Budlender B and Woolard I (2006) The impact of the South African Child
Support and old age grants on childrens’ schooling and work.
International Programme on the Elimination of Child Labour, Geneva:
International Labour Office.
Budlender D, Rosa S and Hall K (2005) At all costs? Applying the means
test for the Child Support Grant, Cape Town: Children’s Institute and
Centre for Actuarial Research, University of Cape Town.
Caldes N, Coady D and Maluccio J (2004) The cost of poverty alleviation
transfer programs: a comparative analysis of three programs in Latin
CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA 19
America, Discussion Paper 174, Washington D.C.: International Food
Policy Research Institute.
Case A, Hosegood V and Lund F (2005) The reach and impact of the Child
Support Grant in South Africa: evidence from KwaZulu-Natal,
Development Southern Africa, 22(4): 467-482.
Children’s Institute (2005) Observations: Eastern Cape means test costing
research, March, Cape Town: Children’s Institute, University of Cape
Town.
Chisholm L (2006) The state of South Africa’s schools. In Daniel J, Southall
R and Lutchman J (Eds), 2005, State of the Nation: South Africa
2004-2005, Cape Town: HSRC Press.
Currie I and Klaaren J (2001) The Promotion of Administrative Justice Act
Benchbook, Cape Town and Johannesburg: Siber Ink in association
with the Research Unit for Law and Administration, University of the
Witwatersrand.
De La Briere B and Rawlings L (2006) Examining conditional cash transfer
programs: a role for increased social exclusion? Social Protection
Discussion Paper, No. 0603, Washington D.C.: World Bank.
Department of Education (2001) Nationwide Audit of ECD Provisioning in
South Africa, Pretoria: Department of Education.
Esping-Andersen G (1990) The three worlds of welfare capitalism,
Cambridge: Polity Press.
Farrington J and Slater R (2006) Introduction: cash transfers: panacea for
poverty reduction or money down the drain? Development Policy
Review, 24(5): 449-511.
Gertler P (2004) Do conditional cash transfers improve child health?
Evidence from Progresa’s control randomised experiment, American
Economic Review, 94(2): 336-41.
Goldblatt B (2005) Gender and social assistance in the first decade of
democracy: a case study of South Africa’s Child Support Grant,
Politikon, 32(2): 239- 57.
Goldblatt B and Yose C (2004a) Research findings: gender and the Child
Support Grant – Gauteng, Unpublished paper, Johannesburg: Centre
for Applied Legal Studies, University of the Witwatersrand.
Goldblatt B and Yose C (2004b) Research findings: gender and the Child
Support Grant – North West Province, Unpublished paper.
Johannesburg: Centre for Applied Legal Studies, University of the
Witwatersrand.
Goldblatt G, Rosa S and Hall K (2006) Implementation of the Child Support
Grant: a study of four provinces and recommendations for improved
20 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
service delivery, Johannesburg: Centre for Applied Legal Studies,
University of the Witwatersrand, and Cape Town: Children’s Institute,
University of Cape Town.
Hall K (2005) Costing the means test for the CSG: field report from the
Western Cape, Unpublished report, Cape Town: Children’s Institute,
University of Cape Town.
Handa S and Davis B (2006) The experience of conditional cash transfers in
Latin America and the Caribbean, Development Policy Review, 24(5):
513-536.
Heinrich C (2006) Demand and supply-side determinants of CCT program
effectiveness, Unpublished working paper, University of Madison-
Wisconsin.
Kallmann K (2005) Food for thought: a review of the National School
Nutrition Progamme, in Leatt A and Rosa S (Eds.), Targeting Poverty
Alleviation to make Children’s Rights Real, Cape Town: Children’s
Institute, University of Cape Town. [CD-ROM].
Kildal N (2001) Workfare tendencies in Scandinavian welfare policies,
Geneva: International Labour Office.
Leatt A (2004) Granting assistance: an analysis of the Child Support Grant
and its extension to seven and eight year olds, Children’s Institute
Working Paper, Cape Town: University of Cape Town.
Lund F J (1993) State social benefits in South Africa, International Social
Security Review, 46(3): 5-25.
Maluccio J A and Flores R (2005) Impact evaluation of a conditional cash
transfer program, The Nicaraguan Red de Proteccion Social, Research
Report, 141, Washington D.C.: International Food Policy Research
Institute.
Molyneux M (2006) Mothers at the service of the New Poverty Agenda: the
PROGRESA/ Oportunidades programme in Mexico. In Razavi S and
Hassim S (Eds), Gender and Social Policy in a Global Context:
Uncovering the Gendered Structure of ‘the Social’, Basingstoke:
Palgrave Macmillan.
Noble M, Wright G, Barnes H, Ntshongwana P, Butierrez-Romero R and
Avenell D (2005) The Child Support Grant: a sub-provincial analysis
of eligibility and take up in January 2005, Project on Take Up of
Social Grants, Unpublished paper, Pretoria: National Department of
Social Development, South Africa.
Organisation for Economic Cooperation and Development (OECD), (2003)
OECD Employment Outlook 2003 - Towards More and Better Jobs,
Paris: OECD.
CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA 21
Plaatjies D (2006) Conditional cash transfers in South Africa, Presentation
at Third International Conference on Conditional Cash Transfers,
Istanbul, Turkey, June 26-30. Sourced on 16 August at
http://info.worldbank.org/etools/ICCT06/DOCS/English/Day1/Plaatji
es
Quinn L M and Magill R (1994) Politics versus research in social policy,
Social Service Review, 68(4): 503-520.
Rawlings L (2004) A new approach to social assistance: Latin America’s
experience with conditional cash transfer programs, Social Protection
Discussion Paper, No. 0416, Washington D.C.: World Bank.
Rawlings L and Rubio G M (2005) Evaluating the impact of conditional
cash transfer programs, World Bank Research Observer, 20(1): 129-
55.
Republic of South Africa (1996) Report of the Lund Committee on Child and
Family Support, August, Pretoria: Government Printer.
Republic of South Africa (1997) Government Gazette No 18340, 1 October.
Schubert B and Slater R (2006) Social cash transfers in low-income African
countries: conditional or unconditional? Development Policy Review,
24(5): 571-578.
Skoufias E (2005) PROGRESA and its impacts on the welfare of rural
households in Mexico, Research Report, No. 139, Washington D.C.:
International Food Policy Research Institute.
Soares F V, Soares S, Medeiros M and Osorio R G (2006) Cash transfer
programmes in Brazil: impacts on poverty and inequality,
International Poverty Centre Working Paper, No. 21, Brasilia:
International Poverty Centre, United Nations Development
Programme.
Waddell J (2002) A comparative analysis of social welfare grant
administration in South Africa, Unpublished mimeo, May, Oxford:
Department of Social Policy and Social Work, Oxford University.
22 CONDITIONALITY AND CASH TRANSFERS FOR CHILDREN IN SOUTH AFRICA
