Beyond Cash Transfers: How Sri Lanka’s Maternal-health Architecture can guide India

India’s progress on a fragmented foundation

India’s gains sit on a system with three structural weaknesses: The first is a shortage of skilled staff. Rural Community Health Centers report specialist vacancy rates above 70%, including approximately 74% for obstetricians and gynecologists, so many services exist on paper more than in practice.⁸ The second is the nature of the front line. ASHAs are drawn from the villages they serve but paid through performance incentives rather than salaries, for example, about INR 600 for institutional delivery in the poorest states, which leaves the most isolated workers the least paid and the most stretched.⁴ The third problem is fragmentation. Care reaches women through a succession of conditional, eligibility-bound schemes, JSY first and then PMMVY, LaQshya, and PMSMA, which reward facility delivery and tend to leave behind those who cannot comply with the rules.^2,4

Underneath this is the problem of a sequence. JSY drew women into facilities for 12 years before the LaQshya labor-room quality program began in 2017, and the gap shows that contact grew faster than the care behind it ( Figure 1). Antenatal coverage is still low, with only approximately 59% of women completing four or more visits, and the caesarean rate has risen from 8.5% in 2005–06 to 21.5% in 2019–21 and 27.2% in 2023–24, nearly half of which are performed in private hospitals. The result is over-treatment in some places alongside under-care in others.^2,3

Figure 1. India’s institutional-delivery rate and Maternal Mortality Ratio, 2005–2021, against the SDG 3.1 national target (MMR 70) and Sri Lanka’s current MMR (~25).

Access has approached universal while the MMR, though much reduced, remains well above both benchmarks.

Sources: Sample Registration System MMR bulletins (2004–06 to 2018–20); NFHS-3/4/5 (institutional delivery); WHO MMEIG (2025) for Sri Lanka.

How Sri Lanka did it

Sri Lanka’s MMR fell from approximately 1,500 per 100,000 in the early 1940s to about 18 in 2023, a decline of more than 98%, and almost every birth now takes place in a facility with antenatal coverage close to universal.^9,10,11 Three things carry the system. Maternity care has been free at the point of use since the free delivery policy of 1952, provided as a right of citizenship alongside free schooling and free outpatient care.⁹ The Public Health Midwife, a cadre created in 1926 and steadily enlarged, now numbers about 10,000, roughly one for every 2,000 to 3,000 people; each is a salaried professional, trained for a year in hospital and six months in the field, who provides antenatal and postnatal care at doorstep.^9,11 Behind her sits a three-tier referral system that keeps comprehensive emergency obstetric care, including caesarean section, blood, anesthesia, oxytocin, and magnesium sulfate, available at district hospitals as a matter of national policy.^11,12 None of this waited for prosperity: the 1952 commitment was made when income per head in Sri Lanka was lower than that in many low-income countries today.⁹

The decisive design: accountability and universality

The two features set Sri Lanka apart, and both speak to India’s gaps. The first is geographical accountability. Each midwife is answerable to every pregnant woman in a defined area and seeks out those who do not come forward, so the unit of responsibility is the place, not the register ( Table 1). India’s ASHA works mainly with women who have already entered the system, and so tends to miss the very women whose first contact with care would otherwise be an emergency.^9,11 The second factor is universality. Free care for everyone is simpler to run, fairer, and often cheaper than means-tested support because targeting adds administrative cost and tends to exclude the poorest, who are least able to satisfy eligibility rules.⁹ India’s reliance on conditional cash is exactly the target model that Sri Lanka’s experience calls into question.

Architecture, not money, and resilience under stress

Table 2 sets out the comparison. India is, if anything, poorer per head than Sri Lanka and spends a similar modest share of GDP on health, yet Sri Lanka’s MMR is about a quarter of India’s. No difference in income or spending can account for a gap that large. The explanation is the architecture: a free, universal system staffed by professional community midwives, with clear accountability and a referral chain that works, set against India’s volunteer frontline and cash incentives layered onto an uneven base of facilities. The economic collapse of 2022 put this to the test. Sovereign default, a 53% fall in the currency, and the breakdown of medicine imports, which supply about 85% of the country’s drugs, left some 150 essential medicines out of stock and caused oxytocin shortages in facilities.^13,14 The MMR rose, with modelled estimates near 34 per 100,000 in 2021, but the system held and recovered to about 18 by 2023.⁷ The episode cut both ways. It exposed a real weakness in import dependence, and it showed the value of a deep, universal system that could absorb a shock which would have overwhelmed a thinner, incentive-based one.

Why India has not made the choice, and what it would take

Sri Lanka’s model is not secret to Indian policymakers, which raises the real question of why India has not adopted it. A part of the answer is structural. Health is largely a state responsibility in India’s federal system, so reform depends on the center and more than two dozen states moving in step, with no single point of accountability. Part is fiscal. The National Health Policy of 2017 set a target of raising public health spending to 2.5% of GDP by 2025, up from approximately 1.15%, with the union government meeting roughly 40% of it.¹⁵ This target has not been met: government health spending was about 1.9% of GDP in 2023–24, and the center’s own share has drifted down rather than up, leaving most of the recent increase in the state.¹⁶ Part is political. A large private sector now handles a substantial share of births, including most caesareans, and households have built their expectations around it. Therefore, moving to universal public provision means working against established interests rather than on a blank slate. The frontline India built was also designed to be inexpensive: the ASHA is a volunteer in a small honorarium, not a salaried professional, and turning that network into a paid cadre is a fiscal and political commitment that successive governments prefer to defer.

The direction India would need to take, though, is one that it has already chosen on paper. In 2018, the government launched the Midwifery Services Initiative, creating a Nurse Practitioner in Midwifery cadres trained in the International Confederation of Midwives standards to staff midwife-led care units, with a stated goal of about 86,000 professional midwives.¹⁷ Its aims read almost as a description of the Sri Lankan system: respectful, lower-intervention care for low-risk births, fewer unnecessary caregivers, and better outcomes. The gap between vision and the ground is practical rather than conceptual. The rollout has been slow and uneven across states, and the new cadre must find its place against an obstetric establishment accustomed to leading maternity care, the kind of professional boundary dispute that has slowed midwifery reform in other countries, and the initiative has not been tied to the financing or the universal-coverage commitment that would carry it to women who need it most. The price of doing this properly is, at the bottom, that one India has already promised to pay. Closing the gap to the 2.5% target means finding roughly 0.6% of GDP, on the order of ₹1.8 lakh crore (about US$21 billion) a year at current output, and holding it across budget cycles instead of letting the center’s share erode.^15,16 A salaried community-midwife cadre would be a modest part of that set against the far larger sums households already pay out of pocket for care the public system does not reliably provide. On that basis, the sensible sequence is to scale the midwife cadre first where mortality is highest, in the high-burden northern and eastern states; give it geographic accountability and the free, universal backing that makes the Sri Lankan version work, and guarantee the emergency-obstetric referral behind it, rather than waiting for a single nationwide reform.