Human Development

Does subsidising publicly provided services discipline markets or distort demand?

  • Blog Post Date 16 August, 2024
  • Articles
  • Print Page
Author Image

Utkarsh Kumar

Columbia University

uk2154@columbia.edu

Author Image

Parijat Lal

Columbia University

pl2700@columbia.edu

Prior research has documented the failure of India’s flagship safe motherhood programme in reducing perinatal mortality, despite substantially increasing the share of mothers delivering at public healthcare facilities – presenting a conundrum for policymakers. Examining responses to the programme across various segments of the maternal healthcare market, this article locates the explanation in the interactions between public and private healthcare providers.

Effectively designing large welfare schemes is crucial in the presence of limited government resources. Prior research in economics has demonstrated that ‘equilibrium’ responses, including adjustments by targeted and untargeted stakeholders, can amplify, attenuate, or redistribute gains from welfare schemes (Barahona et al. 2023, Atal et al. 2022, Khanna 2023).

We study India’s Janani Suraksha Yojana (JSY), a nationwide government policy that subsidised maternal healthcare services at India’s public hospitals in a market with ‘vertically differentiated’ public and private suppliers – with the latter providing higher quality of care at a higher price, on average. Theoretically, subsidising publicly provided goods and services can restrict private suppliers’ market power, yielding lower prices (Cunha et al. 2019), higher quality, or both – resulting in improved consumer welfare. On the other hand, such an initiative can also potentially adversely affect the quality of goods and services received by consumers if the subsidised option, which is now more attractive, is of lower quality. Studying this change in demand is especially important in the healthcare context, where variation in quality can translate into differences in health outcomes.

Maternal healthcare in India

Historically, India has underperformed in terms of newborn and maternal mortality. In 2005, the rate of neonatal mortality per 1,000 live births was 38 in India, higher than the global average of 26, as well as compared to the rates for neighbouring countries including Bhutan (27), Nepal (33), and Sri Lanka (7). 1 Around this time, over 70% of Indian mothers gave birth at home, presumably without any medical supervision. In this context, the Indian government launched its nationwide in 2005 to improve health outcomes, particularly for children and mothers.

The goal of JSY, a flagship initiative under NRHM, was to reduce maternal and newborn mortality by offering a one-time cash transfer (ranging from Rs. 600-1,400) to mothers for delivering at a public health facility. Previous evaluations of JSY have documented an alarming puzzle: despite substantially increasing the share of mothers delivering at health facilities, the scheme failed to lower perinatal mortality (Powell-Jackson, Mazumdar and Mills 2015, Andrew and Vera-Hernandez 2022). Given the scale of JSY, its muted impact poses a conundrum for policymakers. To better inform the design of policies moving forward, we investigate responses to JSY across various segments of India’s maternal healthcare market and show that interactions between India’s public and private healthcare providers play an important role in resolving this puzzle.

Data and methodology

Our data come from three rounds of the District Level Household Survey (DLHS), covering the period 2000-2011. This nationally representative dataset was constructed primarily to track the achievements of NRHM. While individual women cannot be tracked over time, it does feature extensive information on the most recent childbirth for the surveyed mothers. This includes details on socioeconomic background, date of delivery, type of delivery (vaginal or cesarean section), type of facility (public, private, or home), out-of-pocket (OOP) costs incurred, and receipt of government assistance, which we use to construct a measure of JSY’s reach. Further, the survey asks women several questions about previous pregnancies, which, along with other variables identified with the help of the medical literature, allow us to estimate a mother’s predetermined risk (and thus the need for medical supervision or care) before the most recent delivery. Our main measure of healthcare quality is perinatal mortality.3

We employ a difference-in-differences4 empirical strategy, motivated by the staggered rollout of JSY across districts. While the policy was announced in 2005, coverage varied across districts over the next few years as necessary personnel and frameworks were put in place. We consider a district ‘treated’ (subject to the intervention) by JSY if: (i) at least 25% of eligible women delivering in public hospitals in a given quarter report receiving financial assistance for the delivery from the

Government, and (ii) this share does not decline over the following year. We use the imputation method from Borusyak et al. (2023) to compare outcomes in treated districts with those in districts not yet treated by JSY. This approach relies on the assumption that treated and yet-to-be-treated districts would have the same trends in outcome variables in the absence of JSY. We provide support for this assumption by checking for differential pre-trends.

The first step in our analysis is to extend findings from earlier research on JSY, which focused either on mothers in rural areas or specific low-performing states, to the entire available sample. This gives a broader picture of the impact of JSY across the entire market. In line with the results from existing studies, despite a 27% increase in institutional births over the two years following a district being treated by JSY, there is no decline in perinatal mortality. The rest of our analysis focuses on the options available to pregnant women and the responses that explain this disappointing outcome.

Public and private providers in India are vertically differentiated

First, we document several facts that highlight the vertical differentiation in the market for maternal healthcare in India. First, controlling for a mother’s wealth and predetermined level of risk, the likelihood of perinatal mortality is smallest at private facilities and highest at home. Second, richer and more educated mothers are more likely to deliver at private facilities and least likely to deliver at home. This is not surprising, as median OOP costs for childbirth at private facilities are up to four times higher than those at public facilities. Third, women delivering at private facilities report receiving higher quantity and quality of antenatal care (as measured by the number and coverage of antenatal checkups). These descriptive facts suggest real and perceived quality differences across public and private facilities.

How does incorporating private maternal healthcare providers and vertical differentiation into the analysis enhance our understanding of JSY’s muted impact?

Subsidising the public option led to worse allocation of medical risk across facilities

To reduce perinatal mortality, the ideal patient-facility match would involve higher-risk mothers receiving care at higher-quality facilities (private hospitals in this context). While JSY resulted in fewer deliveries at home, the financial incentives diverted mothers away from private facilities and into public facilities too. As shown in Figure 1 (left panel), the likelihood of high-risk mothers delivering at private facilities declined by 6.42%. This adverse sorting is mainly driven by high-risk mothers from poorer households, the primary intended targets of JSY, who became 19% less likely to deliver at private hospitals. While sorting out of private hospitals saved them money, it likely increased the risk of adverse health outcomes.

Figure 1. Likelihood of private facility birth for high-risk (left) and low-risk mothers (right)

Moreover, evidence suggests that while high-risk mothers sorted into public hospitals, the quality of these hospitals deteriorated. JSY provided a large stimulus to demand for public healthcare facilities; however relatively little was done to augment public sector capacity (obstetricians and gynaecologists, beds, and nurses). Andrew and Vera-Hernandez (2022) document that limited capacity at public facilities meant that JSY caused increased congestion and declining healthcare quality. Complementing their finding, we show that mothers from higher socioeconomic groups in capacity-constrained districts partly managed to adapt to this worsening public sector quality, by sorting away from public facilities towards more expensive private facilities.7

Private sector response further restricts access to the highest-quality care

Subsidies under JSY could have improved access to higher-quality private facilities if increased competition from the subsidised public option reduced prices for private maternal healthcare. Figure 2 (left panel) shows that JSY decreased OOP costs for deliveries at public facilities by 18%. However, despite this competitive pressure, private facilities responded by raising their prices on average, further restricting access for poorer mothers. This increase is driven by states where mothers from higher socioeconomic backgrounds were ineligible for JSY subsidies, allowing private facilities to counter the effects of lower market share by exploiting the ‘inelastic’ demand of their patients8

Figure 2. Changes in prices at public (left) and private facilities (right)

Private facilities in these states increased prices by 4.5% in response to JSY, without demonstrating any improvements in perinatal mortality, despite now catering to a less risky pool of mothers (a consequence of the result in Figure 1). Instead, the price hikes are explained by increased rates of cesarean-section deliveries at private facilities.9

Concluding thoughts and policy implications

Our findings reinforce the importance of effectively designing large-scale policies with pressing goals. JSY, despite being a costly push to improve health outcomes, did little to reduce perinatal mortality in India. While this puzzle has been documented in previous literature, we demonstrate that interactions between public and private providers contributed to the muted effects of JSY.

Our results suggest two potential avenues of complementary policy interventions: (i) investments in public sector capacity to alleviate ensuing congestion and declines in quality of care, and (ii) improving access to private sector healthcare for India’s poor, potentially via targeted vouchers or comprehensive health insurance schemes. Developments along the latter dimension, such as the Chiranjeevi scheme and PM-JAY10, are promising, although their effectiveness must be rigorously analysed.

Notes:

1. The neonatal mortality rate is the number of children dying before reaching 28 days of age, per 1,000 live births in a given year. Cross-country estimates are developed by the United Nations Inter-agency Group for Child Mortality Estimation and are available here.

2. The World Health Organization (WHO) defines perinatal mortality as the “number of stillbirths and deaths in the first week of life per 1,000 total births”, which is closely linked to the quality of care received during childbirth.

3. Ideally, we would also be able to measure maternal mortality. However, we do not observe the approximately 0.3% of mothers who suffered death during childbirth in the DLHS data.

4. Differences-in-differences (DiD) is a technique used to compare the evolution of outcomes over time in similar groups, where one experienced an event – in this case, early access to JSY – while the other did not.

5. We borrow this definition from Andrew and Vera-Hernandez (2022). Our results are robust to using a series of alternative cutoffs to construct the treatment, including 15%, 20%, and 30% instead of 25%. We also test the robustness of our results by running regressions using the numerical share of eligible women receiving financial assistance as the treatment variable.

6. We split our sample of mothers into four groups based on their risk level (below or above median) and economic background (below or above the poverty line).

7. We modify our empirical strategy to account for changes in patient composition at private facilities because of JSY, controlling for mothers’ socioeconomic status and predetermined risk level.

8. Demand for private facilities is likely to be price inelastic, that is, patients’ demand will remain the same despite changes in the price.

9. The cost of this type of delivery was higher than the average monthly income for Indian households in 2016. Moreover, rates of cesarean sections at private facilities were already very high (over 20%) before JSY, and higher-risk mothers were more likely to shift away from these facilities. This suggests that the increase in cesarean sections is unlikely to be entirely medically necessary or the result of patient demand.

10. The Chiranjeevi scheme was introduced by the Rajasthan government to provide cashless health insurance to all eligible families living in the state. The PM-JAY (Pradhan Mantri Jan Arogya Yojana) provides health insurance cover for secondary and tertiary care hospitalisation to poor and vulnerable families across India.

Further Reading

  • Acemoglu, Daron, “Theory, General Equilibrium, and Political Economy in Development Economics,” Journal of Economic Perspectives, August 2010, 24 (3), 17–32. Available here.
  • Andrew, Alison and Marcos Vera-Hernandez, “Incentivizing Demand for Supply-Constrained Care: Institutional Birth in India,” Review of Economics and Statistics, 2024, 106 (1), 102–118. Available here.
  • Cunha, Jesse M, Giacomo De Giorgi and Seema Jayachandran, “The Price Effects of Cash Versus In Kind Transfers,” The Review of Economic Studies, 2019, 86 (1), 240–281.
  • Atal, Juan Pablo, Jose Ignacio Cuesta, Felipe Gonzalez and Cristobal Otero, “The Economics of the Public Option: Evidence from Local Pharmaceutical Markets,” American Economic Review, March 2024, 114 (3), 615–644.
  • Khanna, Gaurav, “Large Scale Education Reform in General Equilibrium: Regression Discontinuity Evidence from India,” Journal of Political Economy, February 2023, 131 (2), 549–591.
  • Muralidharan, Karthik, Paul Niehaus and Sandip Sukhtankar, “General Equilibrium Effects of (Improving) Public Employment Programs: Experimental Evidence From India,” Econometrica, 2023, 91 (4), 1261–1295.
  • Powell-Jackson, Timothy, Sumit Mazumdar and Anne Mills, “Financial incentives in health: New evidence from India’s Janani Suraksha Yojana,” Journal of Health Economics, September 2015, 43, 154–169.
No comments yet
Join the conversation
Captcha Captcha Reload

Comments will be held for moderation. Your contact information will not be made public.

Related content

Sign up to our newsletter