Application of the decision space framework revealed system-wide patterns, in which lines of authority are generally well-defined in principle, however with personal networks taking on an important dimension in how stakeholders can act, and particularly when it comes to being accountable. The framework also enabled identification and description of significant informal ingenuity and capacity. The influence of political and personal factors is more clearly identified at higher system levels, across provinces and at district level, whereas at sub-district and facility level, the dominant theme is capacity, which affects all health system components. The emergent decision space is therefore characterised by personal networks and qualities, at higher levels, and on informal coping strategies between facilities and groups of staff at lower levels.
Accountability has some force collectively, in that there is real attention to meeting performance targets as set out in the APP and transparency in reporting. However, on an individual level for managers and staff, accountability is less robust and tends to be punitive and systematically screening out local innovation. For communities, direct action (often negative, in the form of protests, media stories or litigation) is perceived as more effective to enforce accountability than using the formal structures within the health system, and as documented in other sectors (58).
Although many challenges have been highlighted for decision space, which enables health managers and staff to respond flexibly and appropriately to local contexts, assets within the system remain rich. These include many committed staff, expertise, financial resources, well designed national policies, and willingness and ability to find informal coalitions and ways to negotiate constraints (as shown in the example of drug-sharing between clinics to avoid stock outs). Recent achievements in addressing severe acute malnutrition in the province also indicate what can be done by bringing leadership and using local data to identify priority issues and build coalitions to address them (59). This testifies to resilience in the system and indicates how much more would be possible if current limits to decision space were addressed. The enabling of bottom-up, appreciative and reality-based learning and exchange are therefore important routes to further understand and develop decision space.
It is also important to understand some of the contextual factors which influence the health system. Provinces are still struggling with the legacy of apartheid and associated socioeconomic injustices (19). Communities were systematically exploited to provide cheap labour to benefit a white minority, explicitly economically marginalised and deprived of a decent standard of living through denial of access to well-paid work, and displaced, forced to live in 13% of the land, called Bantustans (or euphemistically homelands), most of which was agriculturally unproductive and better suited to cattle and game farming (60). While the regime was dismantled 25 years hence, deepening social and health inequalities along the lines of race gender and economic status are widely documented (61). Unemployment is further key contextual factor. At 35% overall, and higher among youth, it not only adds to challenges for communities in promoting health and accessing health care, but also puts pressure on the public health employment market (62).
Added to this is an organisational culture of low trust, related to perceptions of higher level distortions and state capture, but which creates a highly controlling and disabling environment for staff at mid- and lower levels. Development has been hampered by patrimonialism, deployment of cadres (former anti-apartheid activists) and corruption within the ruling party (63). Beresford describes how ‘gatekeeper politics’ ensure that those in power stay in power, and in rural areas these are predominantly elected ward councillors. Consequently, those most disenfranchised are turning to struggle actions as developed during the fight against Apartheid, such as service delivery strikes.
Reflecting on how our case study complements the existing decision space literature, we make novel contributions to discussions about how sub-optimal resourcing can lead to informal coping strategies (both positive and negative) (4, 64). Our findings are also consistent with studies which show the impact of resource uncertainty on narrowing decision space (65). The difference between responsibility and financial resources to enact those responsibilities is also a common strand in the literature (4), and one which we identify in the current case study. Our case study also adds examples of the importance of infrastructure and information as enabling organisational capacities.
Wider literature (4) suggests that bureaucratic accountability may reduce decision space in some contexts, which we also find in this setting. The perceived lack of decision-space may also have impacted on community engagement, as was documented in India (66), quoted in (4). In the other direction, politicised community participation may have reduced real accountability in our context.
Many studies find unclear authority as a limiting factor for decision space and accountability (4) however in this context, roles and responsibilities were relatively clear and not a major constraint. Unlike in Ghana, where incomplete political and fiscal decentralization ensured that the balance of power in the health system remained at national level (65), in Mpumalanga, there is delegation of authority through the Constitution and resources (through equalisation grants) to provincial level. However, the dynamics we document converge to squeeze decision space at lower system levels. The study is consistent with wider literature on decentralisation, finding that district teams have insufficient management skills or resources to effectively implement the health programmes they oversee (67).
The study draws on mixed sources to examine a transect through the system, but faces limitations of generalisability. However, other studies from South Africa suggest that while Mpumalanga faces some specific contextual challenges, it is broadly similar to other areas. Health district managers in Johannesburg, for example, highlighted poor leadership and planning with an under-resourced centralised approach, as well as poor communication – internally within the service and externally with the community – and poor integration of health strategies and programmes (68). Complexity of tasks, competing demands and lack of support for front line managers and staff within a hierarchical organisational culture is also documented in other parts of the South African health system (55). Limited capacity, inadequate operational resources and irregular monthly supervision visits have been seen to limit stewardship and poor management, with concerns about effectiveness of Ward-based PHC Outreach Teams documented in other provinces ((69–71). Further, Coovadia et al. highlight widespread challenges of gaps in competence and lack of effective leadership, stewardship and personal accountability, rooted in historical legacies of colonialism and apartheid. A recent study also found growing corruption in the health sector in South Africa, with the largest number of reports focussing on the provincial level (72).
The recent South African Quality of Care Commission’s first recommendation focuses on ethical and effective leadership, while its second is on strengthening community structures for engagement and accountability (73). Research conducted in Mpumalanga, Gauteng and Western Cape also highlighted the lack of accountability to patients, and tendency to focus on upward accountability instead, which supports abusive relationships with patients but is also driven in part by lack of support for staff (74). While many quality improvement initiatives are focused at facility level, these will struggle to be effective and be sustained without more meso and macro level changes (75).
Some express concern that the compliance culture in the public sector as a whole – deepened by recent anti-corruption measures - is crowding out innovation, developmental, non-hierarchical and cross-departmental approaches and responsiveness to users:
"Protests happen every day, but officials worry more about what the Auditor-General will say, or whether politicians will throw them under the bus, than what the people think of them"
Debate concerns whether new political windows of opportunity are opening in South Africa currently, which can combat corruption, challenge the culture of cover up and open up decision spaces for committed and skilled managers at all levels of the health (and wider public) system.
While these factors are amenable to change, it is important to acknowledge the deep structural influences from social, historical and health systems contexts and organisational cultures. Our analysis above indicates that it will take time and commitment to motivate staff and provide meaningful, distributed leadership. Support for informal leadership development strategies may also be an important element in building capacity at lower system levels to expand and use decision space (76), along with a greater focus on system ‘software’, such as building trusting relationships, improved communication and dialogue skills. Better communication and direct contact between government officials and citizens are also key (77). A review of the VAPAR process to date suggests that willingness and commitment for cooperative reflection and action exists within the local health system, evidenced in both sustained engagement and participation in the process and formalised partnerships for health systems (42). This cooperative action, including across sectors, has been in evidence in the recent effort to tackle COVID in the province.