Institutions and Policy Change: The Development of the Child Support Grant in South Africa

Abstract

Despite the attempt to transform the system of state child support in South Africa shortly after the transition to democracy, the initial changes resulting in the implementation of the Child Support Grant were only partial in nature. This paper explains why institutional stickiness in the shape of failed reform efforts occurred in certain areas, while radical change took place in others. This effort involves the sequential integration of insights from the historical and rational choice variants of neoinstitutionalism. The resulting analysis accounts for the formation of distinct reform preferences, the strategic interactions which shaped eventual outcomes, as well as the ultimate incompleteness of policy transformation in a process of institutional layering. The paper therefore represents a practical attempt at overcoming the divisions between these neoinstitutionalist approaches, while also producing an analysis of a policy arena that is undertheorised.

Notes

1. Exchange rates were calculated throughout using data from the Penn World Table (Heston, Summers, and Aten 2012).

2. While the programme has subsequently been expanded and adapted in a piecemeal fashion to cover more beneficiaries, the focus here is strictly on the specific configuration when it was originally introduced in 1998.

3. Change agents and veto players clearly did not constitute homogenous groups of purely like-minded actors. The diverse backgrounds of change agents naturally meant that there was a diversity of viewpoints within the ranks of the Lund Committee and its successors. It is clear that there were disagreements about various issues, many of which were intensely debated during the sessions of the Itala Think Tank (Lund 2008). However, despite the plurality of opinions, the fact is that the Committee produced one final report containing a single set of policy recommendations. While it is thus acknowledged that there were disagreements among change agents, the focus of the present paper is to understand why only a limited number of the comprehensive policy recommendations contained in the Committee's official final report were ultimately adopted. The same caveat applies in the case of veto players: there was clearly a great deal of diversity within this group, with some ‘progressive’ actors firmly wedded to the ideal of transformation, while others were more resistant. In spite of this, the fact is once again that a single final set of decisions were taken in the name of the group. Intra-group dynamics clearly plays an important role in determining the policy positions of actor sets. However, the present analysis is focused on the interactions among the outcomes of intra-group dynamics. This means that the focus is largely limited to the between-group dynamics and interactions which ultimately determined where the proposed reforms were to be implemented.

4. It is important to note that this statement only refers to the preferences held by change agents—what they wanted to have happen. The implementation of their preferences was a different matter, and this was eventually to be constrained by the competing considerations of state actors, but the impact was limited when it came to change agents formulating their desire for what they wanted the system to look like.

5. All figures related to the SMG are based on the regulations dating from 1 July 1996.

6. Apartheid classified the population into different racial groupings and the use of these terms is relevant to the subject of this paper. The terms ‘black’, ‘coloured’, ‘Indian’ and ‘white’ are therefore used in the context of racially discriminatory welfare provision.

7. There are various technical problems related to the process of capturing data related to this subject in surveys, including reasons for non-disclosure.

8. It is important to keep in mind that, from a historical and international perspective, there was nothing particularly exceptional about the notion that the reform of the social security system would take many years. The constitutional provision of progressive realisation does not take away from the fact that South Africa has one of the most ideational constitutions in the world, precisely because of the explicit inclusion of social and economic rights. There is thus nothing normatively ‘wrong’ with the idea of progressive realisation. However, it is important to recognise that within the context of this paper, the concept undoubtedly played an important role in legitimating concerns about ‘financial sustainability’ and shaping the institutional context.

9. The fact that this was the only true consensual preference in no way diminishes its significance. On the contrary, the existence of such powerful policy consensus across actor groups serves to highlight just how important it was to reform these discriminatory aspects. Normatively, deracialisation and equalisation thus represented tremendous progress, and this fact is not to be minimised.

10. The discussion in Section ‘The preferences of change agents’ of this paper reveals that the SMG was a supremely targeted programme. It not only applied an income-based means test but also incorporated geographic targeting and crude targeting through racial classification.

11. See Lund (2008, 84–90) for a discussion on the Committee's analysis of the policy alternatives.

12. It should be noted that the pre-existing system of delivering the Old Age Pension did in fact penetrate rural areas quite effectively (Lund 2012).

13. Breaking the SMG's link between private and public child maintenance was part of the Lund Committee's overall terms of reference.

14. While the initial CSG regulations did suggest that caregivers be in possession of a Road to Health Card, this was done ‘even though the Minister of Health had distanced her department from it’ (Lund 2008, 73). The important point here is that change agents failed in their attempt to ‘[create] links with the health services in … [this area]’ (Lund 2008, 69).

15. It is worth noting that the partial nature of the ultimate policy reforms does not diminish the role played by change agents throughout the process. They clearly had significant power in setting the agenda for reform due to the extensive knowledge and expertise embodied by the Lund Committee. The government also acknowledged its trust in the expertise of change agents by appointing the Lund Committee in the first place. Assessing the impact that change agents had on the reform process therefore has to take into account this important agenda-setting power. However, the influence that change agents had beyond agenda-setting was ultimately severely constrained by their sheer lack of institutional power and policy-making authority.