The State of ECD in South Africa

The D G Murray Trust, and a number of DGMT funded projects, had the priviledge to be involved in the Towards Carnegie 3 conference held in Cape Town in September 2012. The conference entitled 'Strategies to Overcome Poverty and Inequality' was styled as the Third Carnegie Conference (following on from that organised in 1932 and 1984), and aimed to bring attention to the causes and dimensions of persistent inequality in South Africa, and consider policies and actions aimed at significantly reducing inequality and poverty in both the short- and long-term.

Three papers were presented at the conference that emanated from DGMT: the state of early childhood development in South Africa, the use of mobile information technology to connect young people to opportunity and the role of foundations in transformation.  David Harrison, CEO of the D G Murray Trust, shares his paper the state of early childhood development in South Africa here.

The state of provision of early childhood development services in South Africa

David Harrison, Chief Executive Officer, DG Murray Trust

In an earlier presentation, Chris Desmond made the case for an intensified national focus on early childhood development (Desmond 2012), drawing on the work of the National Diagnostic Review of ECD.   He stressed the importance of integrated service provision and outlined the respective roles of different sectors, including health, education, social development, agriculture and environmental services.

Many of these sectors have well-established service delivery platforms that can be built upon to extend specific aspects of ECD.  For example, a national network of clinics provide preventive and curative child health services and could reach most children in South Africa with an expanded package of services.  But early childhood education is different in that a national infrastructure is not in place.  Provision is piecemeal and idiosyncratic.  There is no ready delivery service platform for going to scale.  For this reason, I would now like to focus on the state of ECD provision through early childhood education facilities and community-based programmes, and make five recommendations for improving it dramatically.

Recommendation 1

The first is that we need a population-based framework for planning ECD that has clear coverage targets for service delivery. The current system of provision is blind to the majority of young children who are outside the system. It only ‘sees’ the children who are in registered ECD facilities.  Despite an increase in the number of subsidies to early childhood development (ECD) centres, still only a third of young children are exposed to formal child care or education outside of the home. Among the poorest 40% of our population, that proportion drops to one fifth [Figure 1].

Figure 1Percentage of children 0 – 4 years attending an ECD facility, 2008


Source:     National Income Dynamics Survey 2008

The problem is that, in underserved communities in particular, children have access by caprice.  Where someone has established a crèche or preschool, children can attend if their parents can afford it or if the facility gets registered and successfully applies for a government subsidy.  In other words, the system of provision is driven by the location of facilities, rather than the other way around.  Public money follows the facility rather than the educational needs of young children. 

The starting point in planning for service delivery should be to assess the numbers and educational needs of all the young children in a defined geographical area; and the aim should be to achieve universal coverage of all eligible children, starting with the poorest municipal wards in each province. In response, programmes should be developed to ensure that the necessary human, physical and financial resources are in place to support and monitor implementation. 

A challenge is that ECD services are largely provided by the non-profit sector, without any State guarantee of funding to achieve specific targets of coverage.  Resource and Training Organisations (RTOs) – there are about 75 of them in South Africa - provide training and in-service support to smaller community-based organisations (CBOs) and to individual crèches that provide ECD services.  They provide a solid grounding for ECD provision in South Africa, but many children fall between the gaps – in communities where there are no RTOS, but even in communities where they are RTOs (Figure 2). Many individual child care facilities have no relationship with an RTO and – the unregistered ones in particular - struggle along without any support or quality assurance.

Figure 2: Current organisation of ECD provision in South Africa


In planning for ECD, we still tend to ask the wrong question: Typically, we ask how many ECD facilities there are in a municipality and how many children are being reached by the subsidy.  We should be asking how many young children there are and how we can provide quality early childhood development to all of them, through different age-appropriate modes of service delivery.

Recommendation 2

This brings me to the second recommendation, namely that we need to define a clear-age differentiated strategy for early childhood care and education.  Currently, we are so stuck in the ECD site-linked subsidy model, that we’ve lost sight of the need for a comprehensive child development strategy that should start during pregnancy and continue through the transition during schooling.  The platform for successful education needs to be built bottom up. To use an analogy, you can’t build a tower from children’s building blocks by starting in the middle.  The building blocks need to be stacked up from the floor. Yet, for most children, we start trying to lay down the building blocks from the middle – when they enter Grade R.   One point of concern in the Report of the National Planning Commission is the recommendation to extend Grade R downwards, to include a pre-Grade R year.  While ultimately it would be good to have every four year old in a pre-Grade R facility, there must be age-appropriate programmes in place for 0-3 year olds as well.     

Not every child between the age of 0 and 4 years should be in a child care facility, because parents and caregivers should be at the centre of their child’s development.  Programmes for 0-2 year olds should focus on the quality of parent-child interaction that promotes cognitive and language development, as well as psychosocial support for parents.  Given the vital role of primary caregivers in this age-group, delivery should be principally through home-visiting programmes that dovetail with other community-level programmes.  Building on this foundation, programmes for 2-3 year olds should focus on socialisation, achieved largely through group-interactions such as community playgroups.  Children aged 3-4 years will benefit from home-visiting and community programmes, but there should be a clear plan to expand access to centre-based programmes.  This could be on a part time basis unless full day child care is required. 

Our primary interest should not be the mode of service delivery, but the quality of the intervention.  In this regard, a minimum package of ECD interventions should be defined for each of the modes of delivery described above.  This package should include specific structured interventions related to: nutrition, parenting and psychosocial support, cognitive and language stimulation, literacy and numeracy development, and access to health and social services.  This package needs to be carefully costed so that the State can make informed decisions about resource allocation. Currently, there is no funding formula for non-centre based provision.  The means-tested per capita subsidy for learners applies only to children in ECD centres. 

Recommendation 3

This brings me to the third recommendation, which is to establish a public funding formula for non-centre-based programmes.  Provision should be made for financing community- and home-based ECD programmes, based on a specified allocation per child per month. To facilitate this, monitoring systems be improved to allow for tracking of individual children, both for financing purposes and to ensure participation.  This will require the use of unique identifiers for every child, namely the identity number, supplemented by a substitute number where the ID number is not available.  Given that ECD provision is poorest in underserved areas – and that it will take years to extend facility-based provision to all children – it is vital that we develop alternative modes of provision that enable us to treble the number of children exposed to formal ECD provision within the next decade.

But we must also address the current obstacles that, in effect, keep the poorest children out of the system. We must break the cycle of exclusion of the poorest children.  This is the fourth recommendation.  Ironically, the centre-based subsidy was intended to expand access for poorer children.  And to some extent it does. There has however been a significant (76%) increase in the number of learner subsidies over the past six years.  This is welcomed.  Of the children reported to be in preschool or crèches, about 40% have access to the means-tested learner subsidy (Figure 3). 

Figure 3. About 40% of children in crèches or preschools receive a learner subsidy


Centre registration is a precondition for registration. At present, the poorest communities get locked into a vicious cycle of exclusion. They don’t have the finances to improve their buildings, so they can’t meet infrastructural standards for registration. Most of the fees paid by parents are then spent on food for the children. That means less money for teachers and fewer teaching materials, and little chance of meeting the quality standards for learning that are required for registration.  It’s a domino-effect that significantly limits a poor child’s prospects of early education.

In terms of the Children’s Act (as effective from April 2011), all ECD facilities and programmes need to be registered in order to operate.  If this provision were implemented, it is likely that about a third of all ECD facilities would have to shut down[1]. Fortunately, there seems to be a pragmatic approach by the Department of Social Development in this regard.

Even if sites are registered, it does not necessarily mean that they will receive the subsidy. Table 1shows that two-fifths of children in registered facilities do not receive a subsidy. The PETS survey in three provinces found that over 40% of ECD facilities had to wait more than 2 years to receive funding after registration (Van den Berg et al 2010). 

Table 1  Status of registration of ECD facilities (as of April 2011)


No. of registered ECD centres

No. of children receiving subsidy

No. of children in registered ECD centres

% of children in registered ECD centres who receive subsidy


18 826

466 217

789 424



2 911

75 880

82 336



2 979

42 969

97 031



3 473

56 082

151 649



3 167

70 815

123 545



2 184

52 813

96 053



1 144

42 444

49 393




30 732

64 161




25 617

30 191



1 417

68 865

95 060



18 826

466 217

789 424


Source: Louise Erasmus, Dept of Social Development (April 2011)

Access for poor children can be improved immediately by making available funding for the approximately 25% of registered facilities that receive no subsidy.   Section 98 of the Children’s Act, which makes provision for conditional registration, should be fully used to further expand access to subsidies for children in quintiles 1 and 2, by conditionally registering facilities that meet basic safety standards, but which require the resources to meet other quality standards.

The requirement that facilities should register as both an ECD programme and facility should be lifted.  There should be a single registration process for all ECD programmes, with specified sections dealing with centre- and non-centre-based programmes. These changes would significantly reduce the access gradient for poorer children.

My final recommendation is that we must ensure that there sufficient capacity at provincial level to provide programme support.  The scale-up of quality early childhood development in South Africa cannot happen without the necessary skilled personnel.  At present, most of the officials involved in providing early childhood development services play an administrative role – managing the site-based linked subsidy scheme.  We need a team at municipal, district and provincial level that can support the development of ECD programmes, including training, the development and distribution of resource materials and quality assurance.

At present, responsibility for ECD is spread thinly across three government departments (DSD, DBE and DoH), and delegated to relatively junior officials in most provinces.  A single agency is required at national level to coordinate implementation of an integrated programme, and similar agencies should be replicated at provincial level.  These agencies should be capacitated to provide the level of support required to implement a province-wide strategy for early childhood development, as contemplated in Section 93 of the Children’s Act.  The focus of the support should be to improve access to quality ECD services.

In conclusion, I want to reiterate that the current organisation of ECD is not geared to scale-up.  If we don’t effect the type of changes described above, we will continue to grumble along with limited coverage – and access skewed towards those who actually need ECD services least, because they come from better-resourced homes.  The knock-on effects will be huge: persistently poor school achievement because children can’t learn properly, with long-wave effects on the national economy.  In a society where families are so fragmented and parents are themselves poorly educated, early childhood development is key to breaking the cycle of underachievement.  We must now seek a quantum leap in the provision of quality services for those children who need it most.  In a society where so many problems are intractable, it’s an obvious opportunity to make new gains in education, economic growth and societal well-being.



Desmond C (2012). Early Childhood Development (ECD): Providing the best first chance for all children. Paper presented at the Carnegie Conference 3-7 September 2012

Van den Berg S, Williams B, Burger C et al (2010), Tracking public expenditure and assessing service quality in Early Childhood Development: Insights from a survey of three provinces, Department of Economics, University of Stellenbosch


This paper draws on the background paper for the National Diagnostic Review of Early Childhood Development, ‘Opportunities for Learning, authored by me (April 2012)

[1] Estimate based on number of children in registered centres compared with number of 0-4 year olds in any preschool or crèche.


 Search for lessons learned:

Leave blank for all. Otherwise, the first selected term will be the default instead of "Any".
Leave blank for all. Otherwise, the first selected term will be the default instead of "Any".