In the News and on the Record
Early February is an exciting time to be an observer of India’s economy. Not only can one who falls into this eclectic group read the now unmissable Economic Survey, but we are also met by the vastly scrutinised federal budget. Refreshingly in 2018, a larger part of the discussion that follows has revolved around healthcare. Perhaps the world is truly flat, like Thomas Friedman likes to say, and even the perpetual discussion about healthcare policy in the United States has spilled into the Indian zeitgeist. Alternatively, health care as a topic of public interest could finally be making its way up the ranks of the highly concentrated set of leitmotifs in the Indian political discussion.
The stimulus behind this discussion is the National Health Protection Scheme (NHPS), an ambitious health care policy that aims to provide 100 million families with secondary and tertiary care of up to Rupees 500,000 annually. An approximate 110 billion Rupees ($1.7 billion) would be required to cover insurance premiums, 60 percent of which will be covered through federal funds, leaving the rest to state funding.
Within its broad framework, the NHPS aims to serve ‘deprived households’ in both rural and urban settings. While the government may rely on a more specific definition of the term ‘deprived’, the most appropriate qualification parameter for the term is one developed by the McKinsey Global Institute, in their report ‘India’s path from poverty to empowerment’. MGI has created the Empowerment Line, a measure of the level of consumption required to meet eight basic needs – food, energy, housing, drinking water, sanitation, healthcare, education and social security. The 2012 report found that 56 percent of the Indian population (or 680 million people) fell below this empowerment line. While a comparatively small (and yet larger than the entire population of Brazil) 270 million people fell below the official poverty line. The NHPS thus at least aims to reach a large portion of the people that fall under this empowerment line.
It should be noted at the onset that to discuss the implications of such a policy before any major details are announced would be irresponsible. It is important, however, to analyse and understand the effects of providing large scale affordable health insurance. This analysis does not necessarily cover NHPS directly. Instead, it considers some of the lenses through which the policy should be analyzed as more details emerge about its mechanics and implementation.
A vast majority of the Indian population are not covered under any scheme of health insurance. In fact, as high as 86 percent of the rural population and 82 percent of the urban population are uninsured. A majority of those with some coverage are under a government funded insurance scheme. Only 0.9 percent of rural households and 5.9 percent of urban households have either an employer or a self-sponsored health insurance. This distribution is better illustrated in Figure 1.
Given that such a large majority of the Indian population is currently underserved in their health insurance, we should explore from where their health expenditure is funded. A majority of the population’s healthcare funding needs are met through their household income and savings. These numbers are 67.8 percent for the rural population and 74.9 percent for the urban. Between 20 and 25 percent of this financing originates in debt, and less than 1 percent from sale of physical assets. Figure 2 illustrates these numbers.
Note: The data also mentions ‘other’ sources of funding. But this is not included in the above infographic. The numbers for ‘other’ are small and unspecified.
There are massive differences even within rural and urban areas, especially in terms of wealth. While 3.5 percent of the urban population use private insurance, that number is 12 percent for the highest standard of living quintile. Those in the highest standard of living spend a lot more on healthcare, predictably as a result of relying on private facilities. Those with the lowest standard of living are far more likely to be uninsured and not seek out healthcare because of financial constraints; the proportion of uninsured is 90 percent for this category.
These data also tell us that the lack of health insurance led to high out-of-pocket expenses, which in many cases can be financially ruinous. As stated above, the share of borrowing to meet medical costs in rural areas is higher than that in urban areas. These expenses can be acute during times of hospital admittance, with total expenses averaging as high as Rupees 26,000 in urban areas and 17,000 in rural areas. Under serious circumstances, expenses in urban areas can even be as high as Rupees 45,000. The problem of financial security is further exacerbated by the fact that 57.4 percent of the rural population and 68.3 percent of the urban population do not seek medical advice because of financial constraints. This apparent lack of financial security can certainly be addressed through policies like the NHPS.
Learning from the past
We must consider both financial and health-related aspects of the provision of healthcare insurance. I believe the answer can be found in two of the world’s largest healthcare experiments. The first at the RAND Corporation, and the second in the US state of Oregon. Both these experiments have been subject to a large amount of research and analysis since they were conducted. At their core, both of these experiments attempted to answer two questions: Does healthcare usage increase with more generous insurance coverage? And, does the availability of health insurance lead to better health outcomes?
It should be noted at the onset that these experiments were conducted in the United States, and the policy framework surrounding them is different from that in India. The results will, however, help answer these important questions about the effects of health insurance, and will help us better understand the potential effects of a universal health insurance policy in India.
In the RAND experiment, families were assigned to plans with one of six coinsurance rates. The assignment of families to these plans ensured that the coverage rates ranged widely in its generosity. According to basic demand theory, the demand for a good should increase as that good gets cheaper. The RAND experiment vindicated this theory – participants who were assigned to more generous healthcare plans ended up using more health care. The decision to admit a patient is usually made by the doctor, and is made when the condition of the patient requires immediate attention. Inpatient care thus showed a lower sensitivity of these admissions to price. It held that when the cost of health care was less of a matter of concern, more of these services were consumed. However, the differences in health outcomes were largely insignificant among different coinsurance groups. Outcomes for blood pressure, cholesterol levels, and overall health were similar among groups and were not dependent on the generosity of coverage.
The US public insurance plan for those who cannot afford insurance, Medicaid, is a hot topic of discussion. The random assignment of Medicaid coverage is one way to answer the two questions posed earlier in this essay. The Oregon experiment aimed to do just that – by providing Medicaid to low-income individuals via a lottery. It was found that assignment into Medicaid increased the probability of hospital admissions by 30 percent, outpatient visits by 35 percent, taking of prescription drugs by 15 percent, and emergency room visits by 40 percent. At the same time, the probability of having unpaid medical expenses decreased by 25 percent, and the probability of having out-of-pocket medical expenses decreased by 20 percent. Increases in self-reported physical and mental health were observed. Similar to the RAND experiment, there is no clear evidence supporting the improvement of clinically measured physical health (cholesterol levels, blood sugar, etc.).
Both the RAND and Oregon experiments remain the most influential studies about health insurance. The finding that increased coverage leads to a sharp increase in the usage of health care services, but doesn’t necessarily improve physical health has had profound effects on public policy. This influence on public and political narrative can be summarised thusly: If insurance covers fewer costs, then associated services will only be used when necessary, making people more efficient about their health care usage. But this argument assumes that people consume health care just like they consume other services – which is plainly wrong.
Insurance markets work because, as John Nyman suggests, participants value the additional monetary resources that they gain when they use health services over the monetary losses incurred when they do not. Presuming similarity of the effects of insurance across markets is usually tied to the idea of moral hazard – a situation where a party takes on more risk because it knows it is insured against losses.
I’d like to borrow an analogy from a New Yorker article about insurance and moral hazard. Take the financial services industry. Both the Savings and Loans crisis during the 1980s and the Mortgage Backed Securities Crisis of 2008 were based upon institutions who made risky bets after insuring themselves against losses. The logic in play here is that the availability of insurance allowed these institutions to take risks under the assumption that they would be covered if things went wrong. Let’s extend this logic to automobiles. Would an individual who has car insurance drive rashly and risk damage to their car just because they have insurance? It is certainly true that people who incur minor costs of repair will have a financial safety net through car insurance, but it is hard to believe that a driver would willingly risk severe damage just because insurance is available. Similarly, it is unlikely that people insured for chemotherapy would smoke more just because they know the procedure is covered. Moral hazard is an immensely important concept, but its effects vary across markets.
Increased insurance coverage does result in increased usage of health care services, but this usage is not always unnecessary. The Oregon experiment found, for example, that participants who received Medicaid showed increased early diagnosis and compliance with recommended preventative care. Increased early diagnosis of diseases like diabetes was thus observed. It also finds that insured patients fell less of a financial shock as a result of being covered under an insurance scheme.
All of this is to say that the NHPS should be judged based on what health insurance can accomplish. The provision of health insurance did not lead to better health outcomes in some key categories in both the studies considered in this essay, and the NHPS should not be expected to do anything different. NHPS should, however, provide a measure of financial stability to the currently disempowered segment of the Indian population. In India, where some of the highest medical expenses are for cancer and cardio-vascular diseases, early diagnosis and financial stability can be hugely beneficial. Of course, the population that feeds into the RAND and Oregon studies is different from the Indian population. The results will not be exactly replicable, but the core theory will probably stand.
India faces a number of important supply side issues in the provision of health care. The most significant categorisation of these issues is the divide between public and private healthcare services. Poorer households depend more on public sector hospitalisation and treatment, both in urban as well as rural areas. As one may expect, a growing share of the population is now seeking private treatment because better funded private hospitals and clinics tend to provide better services. In both rural and urban areas, the share of patients opting for private treatment was 56 percent in 1995. As of 2015, these numbers have increased to 58 and 68 percent respectively. With a population that is largely uninsured, and where the largest source of insurance is government funding, the increasing costs of shifting to private care can create situations of financial instability. Private care on average costs 4 times as much as public care. The difference between costs of private and public hospitals are the largest for treatments that are the most necessary in India – general injuries, and obstetric & neonatal care.
NHPS has received its criticism. Many believe that it is just a rebranding of the government’s previous RSBY policy – which, if true, isn’t encouraging. Further, it isn’t clear if NHPS will be based on insurance, or on remittances. Remittance based policies have not had a successful history in India, especially in rural areas. Just 12 out of 1000 rural households received partial or full reimbursements of their medical expenses. This number is 62 in urban areas. NHPS must integrate primary care into its paradigm in order to create a comprehensive healthcare system in India. Additionally it must include treatments in private hospitals, and outpatient care – a large portion of spending for chronic conditions. Non-hospital treatment are more expensive in private facilities as well. Bridging the public-private care gap in India is a challenge, and it may take time for quality to reach the same levels. Until then, insuring citizens for treatment in private care facilities could bring about the previously discussed financial and health benefits – as the RAND and Oregon studies inform us. All of this would mean India rightly increasing its share of GDP spent on healthcare – which has anemically hovered around 1 percent for the last two decades.
A program for universal healthcare insurance like NHPS is a step in the right direction for India. It has the potential to tackle the problem of financial insecurity in healthcare spending. It may also increase the usage of healthcare facilities, potentially leading to better diagnoses of diseases, and broader access to good care for people across the wealth spectrum. NHPS will fall short, however, if better infrastructure for primary care is not funded and if private and outpatient care are not included in the mix. Additionally the NHPS should take the form of an insurance program that forms the backbone for a comprehensive healthcare framework. Before making any final judgements about the policy, we must first learn about its inner workings – about which I look forward to hearing more.
 Primary healthcare refers to the first stage of contact between individuals and the health care system. This may include health education, treatment of common diseases, immunisation, and family planning (among other services). Secondary healthcare refers to patients from the primary stage being referred to specialists for treatment. This may include district hospitals or community health centres. Tertiary healthcare refers to specialised care in facilities like Intensive Care Units (Source: Indian Economic Service). While the NHPS will focus largely on secondary and tertiary care, the government plans to create 150,000 health and wellness centers – which will presumably engage in primary care.
 Since the announcement of NHPS, states like Kerala and Tamil Nadu have indicated their intention to opt in and take whatever funding they can get. Others have asked the administration for more details, while some (like West Bengal) have declared their intention to sway free of the policy’s reach. Health care policies among Indian states is fragmented, with many not having any public health insurance about which to speak. It is far too early to discuss the integration of central and state health care systems.
 All these statistics are collected from ‘Key Indicators of Social Consumption in India – Health, Ministry of Statistics and Program Implementation’. http://www.mospi.gov.in/sites/default/files/publication_reports/nss_71st_ki_health_30june15.pdf
 These are non-governmental employers.
 Standard of living is found as a proxy of per capita consumer expenditure.
 Health in India, National Sample Survey Office.
 Amy Finkelstein has authored and co-authored some absolutely fantastic papers about health insurance in the United States. Anyone interested in exploring these issues further should add her work to their reading lists. Here’s a link to her MIT webpage.
 The RAND Health Insurance Experiment, Three Decades Later by Aviva Aron-Dine, Liran Einav, and Amy Finkelstein (2013)
 The Oregon Health Insurance Experiment: Evidence from the First Year by Amy Finkelstein, Sarah Taubman, Bill Wright, Mira Bernstein, Jonathan Gruber, Joseph P. Newhouse, Heidi Allen, Katherine Baicker, and The Oregon Health Study Group NBER Working Paper No. 17190 (July 2011).
 All of these percentages are in relation to the control group.
 ‘Key Indicators of Social Consumption in India – Health, Ministry of Statistics and Program Implementation’ Statement 3.7a
 ‘Key Indicators of Social Consumption in India – Health, Ministry of Statistics and Program Implementation’ Statement 3.7b
Mihir Baxi is an economic analyst and researcher. His work focuses on international economic affairs.
Tweet at Mihir: @baximihir95