Targeting children’s social grants in South Africa: Does the ‘means’ justify the end? ____________________________

1
Targeting childrens social grants in South
Africa Does the means justify the
end?____________________________

• Anti-poverty programmes and childrens
  development
• A Mexican South African dialogue
• SARPN Workshop 26 May 2005
• By S Rosa, A Leatt K Hall
• Childrens Institute
• University of Cape Town

2
Introduction

• Under international human rights law and
  Constitution of the Republic of South African Act
  108 of 1996 (Section 27, 28 29) children have
  various socio-economic rights, including the
  right to water, social security, basic nutrition,
  shelter, basic health care services, social
  services, and education.
• Under the SA Constitution, the state is obliged
  to realise the rights of everyone, including
  children. The state has therefore instituted
  several poverty alleviation programmes.
• The State has designed a variety of targeting
  mechanisms to ensure that these programmes reach
  the identified population groups. For eg. social
  grant programmes are primarily targeted via
  age-based and income-based (means-tested)
  mechanisms.
• This presentation seeks to critically evaluate
  the targeting mechanism for the pre-eminent
  social assistance programme aimed at children,
  the Child Support Grant (CSG) known as the
  means test. I will also make recommendations for
  improvement to the targeting mechanism.
• In light of this and the conceptualisation and
  implementation of other poverty alleviation
  programmes, I will then briefly comment on
  whether CCTs are an appropriate targeting
  mechanism for the CSG in the South African
  context.

3
Background Means to Live Project

• Presentation based on the Means to Live Project
  of the Childrens Institute - focuses on the
  targeting of various poverty alleviation
  programmes for the realisation of childrens
  socio-economic rights.
• The Means to Live Project will soon release
  discussion papers on various sectors and will
  then conduct primary research on the intersection
  of the various targeting mechanisms and the
  consequences for people living in poverty.
• The Project is looking at the targeting
  mechanisms for the following poverty alleviation
  programmes and how they intersect
• National School Nutrition Programme
• School fee exemption policy
• National housing subsidy scheme
• Free basic water policy
• Child Support Grant
• Free health care services for children and
  pregnant women
• As part of the project, a costing of the means
  test has also been conducted based on primary
  research in the Western Cape and Eastern Cape, in
  March 2005, and calculations by Debbie Budlender
  at the Centre for Actuarial Research at UCT.
  (Paper forthcoming 2005).

4
Ensuring social security meets its targets

• Targeting can be described as a way of
  identifying beneficiaries for a benefit or good.
• It can be universal by, for eg. universal free
  primary education or universal free basic health
  care for children under 6 or
• Narrower targeting would seek to identify
  specific individuals, households, communities or
  entities to which scarce resources or public
  goods can be transferred. These narrower
  targeting objectives require more specific
  mechanisms to identify their intended
  beneficiaries.

5
Types of targeting mechanisms

• David Coady of the International Food Policy
  Research Institute and Margaret Grosh of the
  World Bank recently completed a 122 programme and
  48 country study of different means testing
  mechanisms and their implications. (Coady et al
  2003) They outline three broad categories of
  targeting mechanisms
• Individual or household assessment - used to
  assess the income or means at the disposal of the
  individual or the household to support
  themselves. Verifiable, simplified/
  non-verifiable or proxy means tests.
• Categorical targeting - uses some basic
  characteristic of the person or household as a
  basis for inclusion in a beneficiary group, eg.
  age categories.
• Self-Selection targeting - Particular
  interventions where although eligibility is
  universal, the mechanism for accessing the goods
  or benefits is intentionally designed to
  discourage the non-poor from accessing it.

6
Child Support Grant (CSG)

• The CSG was implemented in 1998 to help to
  alleviate the poverty experienced by many
  children in South Africa.
• The CSG is the states primary programme to
  realise the right to social assistance and other
  related socio-economic rights for poor children.
• It currently targets about 5.5 million children
  between the ages of 0 and 14.
• The CSG is a flat-rate cash transfer targeted at
  poor children between the ages of 0 and 14 years,
  via the primary care-giver of the child (R180
  per month from 1 April 2005) .
• The grant is paid to those who qualify in terms
  of a means test of the PCG and her spouse.
  (Deductions are made for UIF, pensions, tax,
  medical aid and other grants, namely the CSG and
  OAP.)
• Means test R800 per month if the child and
  primary caregiver live in an urban area in a
  formal dwelling or R1 100 per month if the child
  and his or her PCG live in a rural area or in an
  informal dwelling.
• There are no conditions for health or education
  attached to the grant as per the Social
  Assistance Act 59 of 1992.

7
Targeting the CSG

• The CSG is thus targeted via two types of
  targeting mechanisms
• Individual assessment assessment of the means
  or the income of the PCG and spouse and the use
  of proxy indicators of poverty, namely housing
  type and area
• Categorical targeting delineation of age-groups
  who may be eligible for the grant.

8
Conditional Cash transfers

• Rationale
• Conditional Cash Transfer (CCT) programmes, are
  designed to provide cash, benefits or services to
  the poor, conditional upon investments in human
  capital such as sending children to school or
  bringing them to health centres on a regular
  basis. (Das et al, 2004, World Bank Rawlings
  Rubio 2003, World Bank)
• It is viewed that by investing in the human
  capital of children, the intergenerational cycle
  of poverty will be broken.
• CCT programmes essentially have one of the
  following objectives (Das et al, 2004, World
  Bank)
• To reduce future poverty - investing in human
  capital by increasing the health and educational
  attainment of children
• Targeting of resources and pro-poor
  redistribution so when governments are unable
  to directly discern individual characteristics,
  CCTs induce self-selection so that the targeted
  group participates in the program and others opt
  out.

9
Conditional Cash transfers

• Overview
• Have been established in recent years,
  particularly in Latin America and the Caribbean,
  eg. Mexico, Brazil, Colombia, Honduras, Jamaica,
  Nicaragua. (Many funded by the WB)
• Programmes have been evaluated to measure their
  effectiveness and broader impact on households
  behaviour.
• Evaluations have shown positive results and
  programmes have been extended.

10
Conditional Cash transfers

• Education and health components
• Mostly targeted to primary school children, but
  in countries with higher educational attainment
  such as Mexico, seek to benefit secondary
  school-age children
• In Mexico and Honduras, education grant covers
  both direct costs (school fees, school supplies,
  transport etc.) indirect opportunity costs of
  sending children to school rather than work.
• Health and nutrition grants targeted to younger
  children, pregnant and lactating women. Aimed at
  food consumption, health care and nutrition
  education. Cash transfer dependent on attending
  health centres regularly.

11
Conditional Cash transfers

• 2. Supply side support
• In some countries, also provide support to
  strengthen supply of health and education
  services.
• Grants also directly to schools and health
  centres.

12
Conditional Cash transfers

• 3. Poverty targeting
• Targeting the poor is a critical feature of CCTs.
  
• Most rely on both geographic and household level
  targeting. (Mostly proxy means-tests of poverty
  at household level)
• CCT programs playing an increasing role in
  countries poverty reduction strategies through
  major income transfer.

13
Conditional Cash transfers

• 4. Evaluation results
• Available for Mexico, Brazil and Nicaragua.
• Show that CCTs can provide effective incentives
  for investing in poors human capital.
• Positive effects on enrollment rates for boys and
  girls.
• Impact on attendance rates are mixed!
• Effective in reducing child labour. (Mexico)
• Child health and nutrition improved. Increase in
  nutrition monitoring and immunisation rates.
• Improved consumption levels on food. (Mexico)
• CCT investments delivered in cost-effective way.
  (Mexico)

14
Conditional Cash transfers

• 4. Evaluation results cont
• Gap need for promotion of income-generating
  activities.
• Gap need to assess effectiveness of CCTs as
  permanent institution for addressing chronic
  poverty.
• Caution not to assume that positive evaluation
  results in a handful of countries can be
  replicated in other areas, especially areas
  facing supply constraints in health and education
  OR where capacity to administer a CCT program
  would be limited. (Rawlings Rubio 2003, World
  Bank)
• Caution CCTs not necessarily the BEST approach
  to achieving a particular outcome. Need for
  comparison to identify the most effective and
  efficient approach.

15
Issues with targeting of the CSG

• Eligibility
• Age categories
• Problems with the means test
• Difficulties with proof of income lack of
  verification
• Income thresholds have not changed
• Household size not taken into account
• Administrative costs of Department of Social
  development
• Burden of time and costs imposed by means test on
  poor applicants

16
Eligibility and targeting of CSG

• Using a poverty line of R430 a month, 74.9 of
  children aged 0-17 in South Africa are poor
  (Woolard 2003) more than 13 million children.
• Using a R215 per month poverty line, Woolard
  found that 54.34 of children across South Africa
  are ultra-poor 9.7 million children. (Coetzee
  Streak 2004)

17
Eligibility and targeting of CSG

• Figures for how many children would be eligible
  under the inflated thresholds were calculated by
  Budlender, based on the GHS 2003.
• Eligibility was calculated for each age category
  in terms of the current means test thresholds of
  R800 and R1 110, as well as in terms of the
  inflation-adjusted cutoffs of R1 120 and R1 541.
• In terms of age, the first set covers children
  under 9 years (the situation for most of 2003),
  the second set children under 11 years (the
  situation for most of 2004), the third set
  children under 14 years (the situation from April
  2005), and the fourth set children up to the age
  of 18.
• The change resulting from adjusting the cutoffs
  for inflation is about four percentage points.
  Overall, about 66 of children in the chosen age
  group are eligible under the current thresholds
  and 70 under the inflation-adjusted threshold.
  (Budlender forthcoming 2005)
• So we see that eligibility under current and
  inflation-adjusted MT based on GHS are not far
  off poverty figures.

18
Age categories

• The gap is that the CSG is available only to
  children between the ages of 0 and 14.
• If the Department of Social Development manages
  to reach all the poor children under the age of
  14, there will still be
• three million 15-18 year olds living in poverty
  without any assistance, of which 2.5 million are
  living in dire poverty, while
• 60 of the adult population living in poverty
  continue to receive no assistance.

19
Problems with the means test

• 1. Problems with proof of income/ lack of
  verification
• People are unable to prove their lack of income
  or small informal earnings.
• Proof of income may be necessary in order to
  avoid fraudulent claims being made and to ensure
  that the support reaches the intended
  beneficiaries.
• BUT requirements of proof placed on CSG
  applicants are ineffective because large numbers
  of people are working in the informal sector
  many kinds of independent and irregular
  contracts, eg. seasonal workers on farms in the
  WC, and many of the poor in South Africa are
  unbanked.
• There is no standard verification of the
  documentation presented in the application
  process, as it would simply be too complicated
  and too costly to administer.
• This is then a waste of time and administrative
  costs in checking and double-checking forms, when
  the checking is potentially ineffective.
• It is also argued that this is a waste of time
  and administrative costs because most people who
  would take the trouble to apply for the CSG would
  need it it is assumed that the wealthy would
  thus not be accessing the grant for these reasons.

20
Problems with the means test

• 2. Income thresholds have not changed since 1998
• The means test threshold levels of R800 and R1
  100 per month have not increased with inflation
  since they were instituted in 1998.
• The value of the grant has kept pace with
  inflation in recent years.
• Without raising the threshold poverty levels in
  line with inflation, the threshold becomes lower
  and lower in real terms.
• Means that many children who would have been
  eligible for the means test under a means test
  adjusted for inflation, are unable to access the
  CSG although they are as poor as other children
  who did access the CSG in 1998.
• Budlender calculated the impact of not adjusting
  the means test thresholds. She shows that to keep
  pace with inflation, the thresholds would need to
  be set at R1 123 and R1 544 respectively in 2004.
  Instead, in 2004 the value was equivalent to
  buying power of R570 and R784 in 1998.
  (Budlender forthcoming 2005)

21
Problems with the means test

• 3. Household size
• A further flaw in the means-testing mechanism is
  that it does not take into account the number of
  people living in the household and the number of
  children that have to be cared for within the
  financial means of the PCG.
• In the interviews conducted for the costing, the
  size of the PCGs families and the number of
  children in their care varied considerably, yet
  the same income threshold applied to all of them.
  
• The inequity created by the means test in this
  regard undermines and contradicts the gains made
  with the open care-giver concept in trying to
  get away from discrimination against certain
  family structures.

22
Problems with the means test

• 4. Administrative costs of Department of Social
  development
• CI commissioned Debbie Budlender, at the Centre
  for Actuarial Research (CARe), to undertake a
  costing of the means test for the CSG in December
  2004. Results not finalised yet, but here are
  preliminary results.
• In order to do the above costing, the CI
  conducted primary research to gather the
  following information
• How many staff hours and costs are spent by DSD
  and SAPS on administering the means test portion
  of the CSG application?
• What are beneficiaries required to do, how long
  does it take them and how much money do they
  spend or sacrifice by having to collect all the
  information together for the means test portion
  of the application of the CSG?

23
Time estimates for implementation of the means
test for DSD

• Usual time Means test only
• W Cape E Cape W Cape E Cape
• Screening officer 10.2 6.7
• 1st attesting officer 19.2 11.7
• 2nd attesting officer 13.3 8.8
• Assessment clerk 10.2 6.2
• Data capturer 10.0 7
• Verifying officer 8.5 4.8
• Screening officer 3.0 1.5
• Attesting officer 5.8 2
• Verifying officer 4.7 1.4
• Data capturer 3.7 2
• Approval officer 3.6 2.1
• STANDARD TOTAL 71.4 20.9 45.2 9.0

24
Costs to applicants

• 5. Burden of time and costs imposed by means test
  on applicants
• We estimate the total time and money costs to
  applicants of complying with the means test part
  of an application for the CSG, requires close on
  six hours of time and costs about R25 per
  applicant.
• Time costs were calculated by adding together the
  SAPS and non-SAPS time costs. Total money costs
  were calculated by adding together the SAPS and
  non-SAPS money costs and adding the estimate for
  lost earnings of applicants.

25
CCTs for South Africa?

• Rationale
• As investment in human capital?
• Education outcomes
• South Africa has an extremely high enrolment
  figure, despite the existence of school fees.
  (Fleisch Woolman 2003)
• Estimates of the level of enrollment in South
  African schools vary between 93 97. (Fiske
  Ladd)
• BUT need a distinction between enrolment and
  attendance. Many children are temporarily or
  permanently barred from school because of
  inability to pay school fees and fees for
  uniforms and transport, despite being enrolled
  (Wilson 2003).
• Problem is lack of implementation of fee
  exemption policy in South Africa together with
  insufficient funding for schools who then are
  forced to charge learners school fees. CCTs would
  not solve this unless linked to funding/subsidies
  for schools for taking in poor children
  (including secondary level) and where children
  who received the CSG would be exempt from school
  fees automatically.

26
CCTs for South Africa?

• Health outcomes
• Adequate immunisation is an indicator of
  individual and population-based health as it
  plays a key role in reducing the chances of a
  child contracting a disease or reducing the
  severity of an infection and may boost a childs
  immunise system to counteract other infectious
  diseases. (WHO UNICEF 2004)
• The bulk of immunisation occurs within the
  childs first year of life, ending in South
  Africa at 18 months with a booster at 5 years.
• South Africa children lt5 immunisation national
  coverage 71 (2004) compared to 63 in 1998.
• Department of Healths target is 90.

27
CCTs for South Africa?

• 2. As a targeting mechanism?
• Woolard shows that grants are exceptionally
  well-targeted. She shows that the poorest 20 of
  households receive the largest amount from state
  grants, equivalent to two-thirds of their income
  (Woolard 2003). And grants are well-targeted at
  children via mostly female PCGs.
• Administration capacity at municipal and
  provincial level will be difficult need for
  links between municipalities and schools and
  health clinics.
• Need for effective administration with
  cross-communicating databases and coordination of
  government departments, which at present do not
  work well together in planning and implementation
  of poverty alleviation programmes.
• Extra barriers created through need to prove
  school attendance and health care facility
  attendance may serve to exclude more people in
  need.
• Banking sector need to come on board to provide
  bank accounts and relief from bank charges.
• Difficult while we still have user fees for
  health services for children over the age of 6
  and a school fee exemption policy that is barely
  implemented due to low funding support to
  schools. Money will be handed from one government
  department to another government department in
  payment of fees.
• May also be very difficult for the most
  vulnerable children to access, as they may not
  have resources, even with a CSG, to pay for
  transport to school, or may have to look after
  sick and dying family members etc.

28
Alternatives

• It is argued therefore, that CCTs are not
  appropriate for SA at this point in time, but
  perhaps in future when departments responsible
  for various poverty alleviation programmes are
  working in a more integrated manner, and poverty
  alleviation programmes are all in fact being
  implemented effectively.
• In order to deal with current problems of
  targeting the CSG, the following solutions are
  recommended.
• Firstly, the threshold for both tiers of the
  means test for the CSG should be adjusted in line
  with inflation.
• Secondly, since the means test used for the CSG
  is not verifiable, and therefore creates
  unnecessary costs for the government and
  applicants, a universal targeting mechanism of
  self-selection should be implemented and the
  means test done away with.

29
Alternatives

• 3. Finally, the most sensible alternative would
  be to make the CSG universal by giving it to all
  children under the age of 18 - as a first phase
  of a Basic Income Grant - and reclaim the money
  from the better-off households through the tax
  system. (Financing Report on BIG)
• This would be a means test in reverse and would
  hopefully be more carefully constructed on the
  basis of absolute and not relative poverty
  levels.
• The administrative burden to households would be
  eradicated, and would leave the cost of setting
  up a system whereby everyone is able to get
  access the money via a bank account or pay point.
  
• This kind of financing model is also the fairest
  and most redistributive as the richer part of the
  population takes a certain responsibility for the
  poorest.