## Notes from 10 May 2025 [[2025-05-09|← Previous note]] ┃ [[2025-05-11|Next note →]] # Reimagining South Africa’s civil service with variable-geometry approaches __TL;DR__: _Drawing on new [[South African Institute of Race Relations (IRR)|IRR]] and [[Public Affairs Research Institute (PARI)|PARI]] reports diagnosing deep politicization and laying out reform pathways, this text proposes a dual-track, variable-geometry reform for South Africa. Track 1 introduces phased, opposition-informed, merit-based selection of senior civil servants (e.g. Heads of Department), Track 2 establishes a new statutory civil service regime for technical and routine roles, piloted by those newly appointed leaders, and rolled out agency by agency, open only to fresh hires and phased in based on HR capacity._ It's always exciting to see new ideas about civil service reform in South Africa, especially when I have a small grain of sand to contribute to these discussions! This week I came across two new publications that offer substantive perspectives on reforming South Africa's public administration. [The IRR's Blueprint for Growth](https://irr.org.za/reports/the-irrs-blueprint-for-growth/the-irr-blueprint-for-growth-2025/4-the-irrs-blueprint-for-growth-3-in-service-of-the-public-reforming-south-africas-public-administration): Sara Gon and Terence Corrigan's paper "_In service of the public: Reforming South Africa's public administration_" provides a detailed assessment of the public service's current state, diagnosing it as fundamentally unfit for purpose and failing to execute its developmental mandate. Their analysis identifies an inadequate skills base, weak management, endemic corruption, and pervasive political interference as key factors undermining effectiveness. The IRR documents how the system's design grants politicians extraordinarily wide discretion in appointments, with minimal separation between political and administrative elements. Their paper traces a troubling evolution where initial political preferencing has, in many cases, morphed into capture by criminal networks. Their reform proposals emphasize the need for depoliticization and professionalization, though with considerable skepticism about the current government's willingness to implement such changes. [Ryan Brunette's Policy Brief](https://pari.org.za/policy-brief-limiting-political-influence-on-appointments-in-the-south-african-public-administration/): "_Limiting Political Influence on Appointments in the South African Public Administration_" builds on Brunette's [earlier PARI working paper](https://pari.org.za/working-paper-checking-political-discretion-in-recruitment-to-south-africas-public-administration/) to deliver a sophisticated analysis of excessive political discretion in the civil service. Brunette constructs a compelling argument showing how the current statutory framework empowers politicians with nearly unfettered control over administrative appointments, systematically elevating political criteria above competence. His work carefully documents how political factions and parties use these powers for patronage and corruption. Drawing on the [Zondo Commission](https://www.statecapture.org.za/)'s findings, Brunette identifies uncontrolled political discretion in appointments as "the essential mechanism" of state capture. His approach includes more optimism regarding reform possibilities, proposing institutional designs that could gradually establish more professional civil service practices within South Africa's political reality. Both texts cite different works that I have authored or co-authored with [[Ivor Chipkin]], and provide an excellent opportunity to revisit and expand on some of the arguments we made in our policy briefs. Before discussing specific reform strategies, it's crucial to understand four fundamental dimensions that structure any merit system. These dimensions provide a framework for analyzing different approaches: 1. **Selection**: How officials are recruited and appointed 2. **Performance**: How performance is evaluated and managed 3. **Appeal**: What mechanisms exist for appealing personnel decisions 4. **Tenure**: What kinds of protections against dismissal exist (if any) Different countries configure these dimensions differently, with varying degrees of political input and professional/technical control in each dimension. There is no single 'right' approach - each system reflects particular historical, political and administrative traditions. South Africa needs to develop its own approach that addresses all four dimensions in a way that is appropriate to its context. ## The need for differentiation: A principal-agent perspective [I strongly advocate for gradual reforms](https://nsi.org.za/publications/reforming-south-africa-senior-civil-service/). While an [[2025-05-05#^audacious-vision|audacious vision]] for change is indispensable, its successful realization depends on an incremental implementation. This approach allows for building confidence in the reform process, generating new momentum, and incorporating lessons learned to sustain progress. A key element of this gradualism is differentiation. The IRR's point regarding the need for differentiated status and limitations on political appointments is well-taken. This principle of differentiation is also central to Brunette's argument concerning institutional variations, suggesting that reforms can be effectively initiated within specific entities and agencies before wider expansion. From a principal-agent theory perspective, elected politicians (principals) delegate authority to civil servants (agents) to implement policies. The appropriate degree of political control in this relationship is not uniform; it logically varies depending on the nature of the specific role and the particular aspect of the employment relationship under consideration. As a result, the transition from a system characterized by heavy political input across all dimensions of the merit system to one in which political discretion is appropriately constrained and informed cannot be achieved overnight. Rather than attempting to overhaul everything at once, reform should begin by embedding foundational pillars in areas where they are most critical, and then systematically expand. This phased introduction allows for the development of a nuanced framework (akin to a menu) that delineates where each pillar of reform lies and the appropriate level of policy intervention for different contexts. Different countries configure these dimensions differently, with varying degrees of political input and professional/technical control in each dimension. There is no single "right" approach - each system reflects particular historical, political and administrative traditions. South Africa needs to develop its own approach that addresses all four dimensions in a way that is appropriate to its context. The rationale for such tailored approaches is evident in other complex domains. For example, relying on a single public procurement procedure is impractical, as acquiring a massive infrastructure project like a railway, large software requiring multiple user iterations, or standard items like pens and toilet paper each demands distinct procurement methodologies and supplier relationships. The same principle applies to people management within the civil service: the approaches to selection, evaluation, performance management, and tenure protection must necessarily differ for a teacher, a specialized economist, and a department director, reflecting their distinct roles and responsibilities. ## My proposal lies in the following assumptions (in simplifying terms!) In delegation‐theory terms, strategic (leadership) roles carry the highest agency risk: they face the greatest ambiguity of goals and therefore demand the strongest alignment with policy makers. Such roles sit in a high-political-input band, with appointment processes that move from opposition-informed shortlists through executive selection to direct political dismissal. Specialized technical staff occupy a middle band, where expertise is balanced against political considerations via mixed expert panels. Routine roles are low-risk and appropriately managed through closed-concours systems with standardized exams and minimal political interference. Crucially, political input cascades down the hierarchy, creating a new model of oversight: decisions made at the leadership level shape both the scope of expert‐panel review for technical staff and the parameters of closed‐contest exams for routine roles, ensuring alignment with policy objectives at every tier. | Role | Low Political Input | Medium Political Input | High Political Input | |------------------|-------------------------------------|--------------------------------|------------------------------------| | Routine Roles | Closed _concours_ (standardized exams) | Agency-framework assessments | — | | Technical Staff | — | Expert-panel shortlists | — | | Leadership | — | Opposition-informed shortlist → executive selection | Direct political dismissal | My point is that different functions require different degrees of interaction with the political power that develops policy, and different degrees of alignment with political power across these four pillars of merit protection. The separation between policy and administration (politics and implementation) is blurred in some areas more than others. Lets see how other countries deal with this tension ### France and Germany: Strong legal codification Both France and Germany utilize comprehensive statutory codification models that clearly define detailed legal provisions for the key dimensions of their merit systems. This approach, characterized by its "regulatory density" (more rules to regulate more scenarios), effectively reduces discretion and the scope for undue political interference. In both nations, these extensive legal frameworks are interpreted by specific branches of the judiciary or independent administrative courts, ensuring consistent application. Furthermore, this legal structure explicitly distinguishes between political and career positions, providing clear parameters for the management of each category. ### United Kingdom: Regulatory authority vs. Executive appointment power The UK’s [Civil Service Commission](https://civilservicecommission.independent.gov.uk/) oversees the recruitment system for civil servants and serves as an appeals body, functioning purely as an impartial arbiter rather than conducting hires itself. Its independence is bolstered by a statutory requirement that the government consults opposition parties before appointing the First Civil Service Commissioner, embedding cross-party guarantees of impartiality. The [Commissioner for Public Appointments](https://publicappointmentscommissioner.independent.gov.uk/), by contrast, holds an executive mandate to organise and regulate how ministers fill senior roles on national and regional public bodies. While this Commissioner shapes and audits the appointment process to uphold fairness and merit, the ultimate power to appoint remains with ministers. To safeguard impartiality, the Commissioner’s own appointment is subject to rigorous pre-appointment scrutiny by a parliamentary committee with strong opposition representation. ### United States: Separation of recruitment/performance and appeal/tenure functions After Carter's reforms, the US separated the Civil Service Commission's functions between the [Merit Systems Protection Board](https://www.mspb.gov/) (MSPB) for oversight and appeals and the [Office of Personnel Management](https://www.opm.gov/) (OPM) for HR administration. This institutional division creates checks and balances by separating the protection function from day-to-day management functions, a separation that [Ivor and I proposed for South Africa](https://nsi.org.za/publications/the-paradox-of-reform-a-response-to-the-ramaphosa-governments-proposal-for-professionalising-the-public-service/) in our analysis of the professionalization framework in 2022. ## A dual-track reform strategy for South Africa South Africa needs a comprehensive yet incremental reform approach that consists of two complementary tracks that would operate in parallel but with different scopes and implementation timelines. ### Track 1: Leadership selection reform The first track focuses specifically on reforming the selection process for leadership positions (Heads of Department – HoD and equivalents): - Establish independent selection committees that include opposition-nominated experts - Maintain government prerogative to appoint (and dismiss) HoD but only from vetted shortlists - Begin with strategic leadership positions in critical agencies - Gradually expand this model to other agencies and departments over time This approach creates immediate improvement in leadership quality while maintaining the government's legitimate role in selecting leaders aligned with its policy agenda. The focus on leadership positions recognizes their unique role in bridging political and administrative functions. Moreover, these selected leaders must then be empowered with the necessary discretion to effectively cascade merit-based recruitment practices throughout their respective agencies, notably by having meaningful input into the recruitment processes for positions below them. ### Track 2: New civil service regime (for technical and routine functions) In parallel, South Africa should develop a completely new civil service legal regime for non-leadership positions, creating regulatory density to diminish and qualify political discretion (drawing lessons from the [[Public Sector Reform in Peru|Peruvian]] [experience](https://guias.enap.edu.pe/subjects/reforma-antecedentes)): 1. Create a new statutory framework for civil servants with clearly defined rules for selection, performance management, appeals, and tenure  2. Grant access to this new regime ONLY for those selected through newly created merit-based recruitment processes 3. Implement the regime gradually, starting with specific agencies with strong HR capacity 4. Establish a new civil service agency to mediate and oversee the transition of agencies to the new regime 5. Expand the new regime as agencies develop their HR professionalization capacity This gradual approach acknowledges that attempting to reform the entire existing system at once would be politically unfeasible and potentially counterproductive. Instead, it creates parallel systems that can coexist during a transition period, with the new merit-based system gradually expanding its scope. Although Peru’s reforms offer valuable insights, South Africa should critically assess their strengths and shortcomings (especially in implementation speed, institutional capacity and political sustainability) when designing its own approach. ### Supporting institutional reforms To support both reform tracks, South Africa should simultaneously develop: - A non-judicial appeals system with agency-level panels of HR professionals (judges know how to interpret the law, less how to deal with typical HR grievances that require sophisticated expertise) - A central oversight body to establish standards and monitor implementation (a renovated Public Service Commission could play this role)  - Enhanced HR capacity development programs across the public service These institutional supports would provide the technical infrastructure needed for successful implementation of both reform tracks. ## The variable geometry of civil service reform The reforms must recognize what we could call "variable geometry": different functions require different balances of political and professional control across the four pillars of merit protection. Some types of positions will have more or less political input depending on both their function and which aspect of employment we're considering. For leadership positions, some dimensions of the merit protection system should distance political input more than others. [That's why Ivor and I proposed institutionally separating selection from performance management](https://nsi.org.za/publications/the-paradox-of-reform-a-response-to-the-ramaphosa-governments-proposal-for-professionalising-the-public-service/). For South Africa, an incremental reform could start by controlling political input in the selection of leaders while maintaining political control over their dismissal. Could this create high turnover? Yes. But it would be better than the current situation and could be adjusted over time. ## Changing political incentives I genuinely believe South Africa is politically approaching a scenario more conducive to reform - at least there's a chance. The political landscape is becoming more competitive. Of course, as Brunette suggests, the ANC might fear losing its power to appoint people, which gives it electoral advantage, making it less amenable to reform. BUT there's another possibility: the ANC may increasingly fear losing government control to other parties or a future coalition without them. The current widespread politicization model would mean that once the ANC is out of government, it would lose control over all these personnel because the next coalition would have the power to appoint its own people. This could incentivize the ANC to want to "lock in" its people in government today by creating some type of tenure for civil servants that wouldn't allow future rulers to replace them. This could generate some incentive for agreement. As [[Christian Schuster]]’s 2017 research [across 117 countries suggests](https://onlinelibrary.wiley.com/doi/epdf/10.1111/padm.12334), political competition significantly shapes the likelihood of merit-based civil service reform, often more than the presence of legal mandates. His large-scale evidence reinforces my point that any civil service reform in South Africa will only make sense if the opposition begins to be consulted - and today, the "opposition" is inside the so-called Government of National Unity (GNU). Any other approach will ultimately just be the ANC creating committees with its own people to conduct selection and engage in "[merit-façade](https://www.tandfonline.com/doi/full/10.1080/01900692.2021.2009858)". The only way to truly ensure neutrality is to have input from both government and opposition. This will create incentives to form committees with technicians who have input from both parties, allowing these technicians to supervise merit protection in certain circumstances.