There is a growing unease across government-run health systems about performance. Outcomes remain uneven, variation is wide, and reform efforts often seem to generate more paperwork than improvement.
Clinical decisions are made under uncertainty. Outcomes depend on case mix, timing, infrastructure, and chance as much as effort.
Governments respond in predictable ways: tighter oversight, more audits, sharper targets, stricter enforcement. Yet on the ground, many public health workers experience these measures not as support for better care, but as signals to be cautious, defensive, and risk-averse. When accountability systems are designed primarily around control and punishment, they often produce exactly the opposite of what is intended: not greater responsibility, but careful self-protection.
Accountability in public services has long been discussed through two routes. One emphasises citizen voice and community pressure; the other relies on bureaucratic oversight through rules, inspections, targets, and disciplinary action. Both matter.
Borderline cases are escalated early; interventions are chosen because they are easier to justify in retrospect; flexibility is avoided even when judgement might suggest otherwise.
But in complex services such as healthcare, they also have clear limits. Clinical decisions are made under uncertainty. Outcomes depend on case mix, timing, infrastructure, and chance as much as effort. Yet many accountability systems implicitly assume that outcomes are fully controllable and that deviations signal failure. In such environments, discretion becomes dangerous. Professional judgement gives way to behaviour that is safest to defend rather than best for the patient.
This dynamic is visible in India’s public obstetric services. Intense scrutiny around maternal outcomes and caesarean sections has led many doctors to practise defensively. Borderline cases are escalated early; interventions are chosen because they are easier to justify in retrospect; flexibility is avoided even when judgement might suggest otherwise.
This behaviour is not driven by greed or indifference. It is driven by fear of inquiries, transfers, suspensions, and being left exposed when something goes wrong. In regions with adequate capacity, particularly in southern and western India, this fear has contributed to persistently high rates of C-sections in the public sector.
Performance is monitored, but the emphasis is on continuity, learning, and system-level outcomes rather than individual blame.
Importantly, this is not simply an Indian story. Evidence from several developing countries suggests that systems perform better when frontline providers operate within a clear, credible compact with the state, one that balances expectations with protection, and accountability with autonomy. Thailand’s Universal Coverage Scheme is a case in point. While purchasing and provision were separated, both remained public. Providers were given clear expectations, stable financing through capitation and global budgets, and meaningful managerial autonomy. Accountability was real, but it was exercised through data, dialogue, and periodic review rather than constant punitive oversight. The result was sustained improvement in access, efficiency, and quality in a resource-constrained setting.
A new compact with the public health worker would begin by acknowledging that risk is intrinsic to healthcare and cannot be eliminated by surveillance.
Brazil’s Unified Health System shows a similar pattern. Municipal providers operate with considerable autonomy within long-term financing and planning frameworks. Performance is monitored, but the emphasis is on continuity, learning, and system-level outcomes rather than individual blame.
Across these contexts, improvement did not come from importing rich-country models or tightening bureaucratic control. It came from institutional arrangements in which frontline providers could reasonably expect that good-faith professional judgement would be recognised rather than punished. This is what is missing in our public health systems today. Instead of an explicit compact, we have accumulated layers of control. Expectations are high, but protection is thin. Responsibility is individualised, while risk is pushed downward.
A new compact with the public health worker would begin by acknowledging that risk is intrinsic to healthcare and cannot be eliminated by surveillance. It would shift accountability from single adverse events to patterns of practice and learning over time, provide stability of expectations, and recognise that autonomy without protection produces fear, just as protection without accountability produces complacency.
For non-profit organisations working alongside the state, this matters deeply. Many well-designed initiatives fail not because the ideas are wrong, but because they collide with systems governed by fear. Without a new compact, these frustrations will persist. With one, the same interventions could unlock very different behaviours. That is what a new compact is ultimately about.