India

India’s health financing and benefit incidence

Lekha Chakraborty

India spent 3.29 per cent of GDP on health, or Rs 6,801 per person, according to the National Health Accounts (NHA) Estimates for India 2023-24, released in November 2026. Out-of-pocket expenditure (OOPE) has fallen sharply as a share of total health spending — from 64.2 per cent in 2013-14 to 42.7 per cent in 2023-24. Yet in per-capita terms it has risen, from Rs 2,767 in 2022-23 to Rs 2,901 in 2023-24. This is the paradox of Indian health policy.

The state is financing a larger share of total spending, and households a smaller proportion. Absolute amounts paid by families continue to climb, however, because the disease burden is becoming more chronic, more expensive and more dependent on outpatient care.

The analytical link between health outcomes and financing is shown in a research paper which I have coauthored with Kausik Bhadra and Rashmi Arora, presented at the LSE–UNU-WIDER meeting at the London School of Economics on 6 November 2026. Our method maps National Sample Survey (NSS) ailment categories to the World Health Organization’s International Classification of Diseases (ICD) and then applies benefit incidence analysis to public health spending.

The evidence base

Two complementary reports now available offer views of India’s health system that have not yet been fully integrated.

The NHA 2023-24, published in November 2026 by the National Health Systems Resource Centre under the Ministry of Health and Family Welfare, measures who pays for health care and where the money goes, using the System of Health Accounts 2011 framework.

The second is NSS Report No. 596: Household Social Consumption: Health (80th Round, January–December 2025), released by the National Statistics Office in April 2026. Covering 139,732 households, it is the eighth comprehensive all India health survey and records morbidity, hospitalisation, outpatient treatment and household medical spending. The NHA captures financing patterns of health, the NSS captures household-level experience of illness and care. Neither, however, reliably links public expenditure to internationally comparable ICD codes of morbidity. That is the missing analytical link the Chakraborty et al. method seeks to research.

The International Classification of Disease (ICD) codes

The approach constructs an ICD code-based, disease-specific map of who benefits from public health spending. Concentration curves and Lorenz curves are used to test whether public spending is pro-poor or pro-rich. If the benefit concentration curve lies above the line of equality, spending is pro-poor; if it lies below the Lorenz curve, it is regressive.

What the ICD mapping with Indian data shows

Public health spending is broadly pro-poor across most ICD categories. This is a significant finding that India’s public health system, whatever its weaknesses, is not being captured by elites. There is no systematic “elite capture” of public health spending.

Significant gender and geographic differentials do appear. In rural areas, women often show higher benefit capture than men across several disease categories. In urban areas the pattern is more mixed, with some categories showing pro-rich utilisation among men. These inferences rest on the penultimate NSS round; the newly published 80th-round data have not yet been fully incorporated.

The mapping also exposes gaps. NSS data do not cover two important WHO ICD categories — developmental anomalies and “symptoms, signs or clinical findings not elsewhere classified.” This is not a minor omission. India cannot fully account for congenital conditions, undiagnosed illness or the early stages of disease presentation.

The NSS 80th Round

Conducted between January and December 2025 and published in April 2026, the survey finds that 13.1 per cent of people reported illness in the preceding 15 days. Women reported higher rates than men (14.4 per cent versus 11.8 per cent). Illness rises sharply with age, 42.5 per cent of men and 45.4 per cent of women aged 60 and above reported an ailment. Urban residents reported slightly higher morbidity (14.9 per cent) than rural residents (12.2 per cent). Hospitalisation (excluding childbirth) over the preceding 365 days stood at about 2.9 per cent nationally. Men had a marginally higher rate than women (3.0 per cent versus 2.8 per cent). The gap was wider in rural areas (2.9 per cent versus 2.6 per cent). Rates peak among the elderly (around 8 per cent for those 60+) and are also elevated among young children (0–4 years).

Disease-specific patterns

Cardiovascular disease is the leading hospitalisation category among people aged 60 and above (18.2 per cent of admissions) and a major contributor to reported illness from age 30 onwards, driven by hypertension and heart disease.

Endocrine and metabolic conditions — chiefly diabetes and thyroid disorders - become significant from age 30 and require lifelong outpatient management.

Infections and respiratory disease remain central in early life, infections account for 43.8 per cent of hospitalisations among children aged 0–4, while respiratory disease accounts for 13.2 per cent.

Psychiatric and neurological conditions peak in young adulthood in reported illness and remain a significant share of adult hospitalisations. Injuries are especially prominent among young adults and a leading cause of hospitalisation in the 15-29 and 30-44 age groups.

State-level variation

Interstate differences in the Proportion of Persons Responded as Ailing (PPRA) are pronounced and deserves a deeper study using these latest data to ICD codes, applying benefit incidence analysis methodology to each state, which is beyond the scope of this article. A summary classification of major states from the recent data shows that Karnataka, Uttarakhand, Jharkhand, Bihar, Haryana, Madhya Pradesh, Uttar Pradesh and Chhattisgarh fall in the 10 per cent range; Odisha, Himachal Pradesh, Telangana, Punjab, Assam, Tamil Nadu, Maharashtra, Rajasthan and Gujarat report 11–20 per cent; West Bengal and Andhra Pradesh lie in the 21–25 per cent band; and Kerala stands out with the highest reported morbidity, exceeding 25 per cent. Kerala’s elevated PPRA is consistent with longer-standing patterns, with higher health awareness, stronger health-seeking behaviour, an older population structure and higher prevalence of chronic conditions. Lower figures in states such as Bihar, Uttar Pradesh, Madhya Pradesh and Jharkhand may reflect differences in underlying disease burden, low reporting behaviour, low access to care or demographic composition rather than strictly lower illness.

Hospitalisation rates vary by age and sex across states, public facilities account for a larger share of care in rural areas, while private providers dominate more in urban settings. Institutional deliveries have become nearly universal (around 96 per cent overall). Health insurance coverage under any scheme has expanded markedly since 2017-18, reaching roughly 45-47 per cent in both rural and urban areas, driven especially by Government-sponsored schemes. Out-of-pocket medical expenditure remains substantial, particularly for private hospitalisation. These differences carry clear implications for resource allocation under the National Health Mission and state programmes, prioritising chronic disease management and elderly care in high PPRA states, while strengthening primary care, diagnostics and reporting systems in lower reporting states. Age and gender gradients, higher burden on females and the elderly, appear broadly consistent across regions.

The financing mismatch

Government health expenditure has tripled in nominal terms, from Rs 1.30 lakh crore in 2013-14 to Rs 4.14 lakh crore in 2023-24. Per capita Government health spending rose from Rs 1,042 to Rs 2,959 over the same period. Yet the share of Government spending in GDP remains low, total health expenditure in 2023-24 was 3.29 per cent of GDP, with Government spending at only 1.43 per cent, still well below the 3-4 per cent threshold widely regarded as necessary for meaningful progress toward universal health coverage.

India has increased public spending, but not enough to offset the growing burden of chronic disease. The result is a system in which households (oope) continue to pay a large and rising absolute amount even as their share of total spending falls. The household budgets go catastrophic, with rising burden of health spending.

The policy imperative

The evidence points to a clear priority. India must raise Government health spending toward at least 3 per cent of GDP while simultaneously improving the targeting and efficiency of that spending. The ICD-based benefit incidence framework in my coauthored paper offers a practical instrument. It allows policymakers to see which diseases consume public resources, which populations benefit, and where gaps remain. It also supports geographical and sex-disaggregated analysis, essential given the gender patterns revealed by the NSS. The 80th-round NSS supplies the morbidity map. The NHA supplies the financing map. The ICD-based benefit incidence method supplies the analytical link between resources and outcomes.

What India now needs is to use that benefit incidence analysis framework systematically to build a health system that is both financially protective and epidemiologically responsive.

The writer is professor at NIPFP, Ministry of Finance, Govt of India and also research affiliate at the Levy Economics Institute of Bard College, New York; Views presented are personal.