Reaching People through Micro Insurance: The Case of India & Nepal

Can we assume that people understand insurance principles?

Von David M. Dror

When approaching people at the “base if the pyramid” (BOP) in Low-Income Countries (LIC) with the idea of insurance, the first challenge is to establish a basis for communication. The question that arises is whether they understand insurance principles. We examined this issue by analyzing replies to a series of questions in household surveys we conducted in several locations in India and in Nepal.

Involvement of the stakeholders (Copyright by MIA)

The following examples serve to illustrate attitudes and conceptions we observed among the target population regarding health insurance. These examples are based on HHS conducted in 2009 in Dhading and Banke Districts, Nepal, in Khorda, Kalahandi and malkangiri districts in Orissa (India) and 2007 in Bihar, Maharashtra, Karnataka and Tamil Nadu. (Dror / Radermacher, 2010)

About 70% of the respondents understood that “With insurance you have to pay money upfront, but you do not know whether you will get something out of it”. More than 70% thought that: “insurance is like savings; you will get all your money back”. And more than 50% agreed with the statement “If I do not claim I will get my premium back”.

These replies indicate that respondents were aware about the terms applying to endowment life insurance, but not to the terms of health insurance. Surprisingly, replies did not differ according to the level of education of the household head (most often the respondent) in both places.

On the other hand, when asked about health insurance, respondents exhibited different attitudes altogether. For instance, 70% agreed with the statement: “It is all-right that my neighbor was sick and got money from health insurance, and I was not sick and did not get money” (both districts, Nepal, 2009). However, more than 50% in Dhading but only about 20% in Banke agreed with the statement “I do not want health insurance because it is possible that I pay a premium but get no money because my healthcare costs are not covered”. The different replies in the two cohorts suggest that whereas the Banke group understood and accepted that health insurance cannot cover all health events, the Dhading cohort were not interested in health insurance unless it would likely pay them a benefit.

This suggests that it would be necessary to explain to people prior to rollout of micro health insurance that there is never a possibility of repayment of the premium, and that the insurance does not cover all healthcare costs, but that it can protect against financial shocks when certain insured health events occur.

Awareness campaign: What it imparts, how is it done, what impact?

Our objective is to reach out to groups that are already organized, and have developed a certain social capital. (Dror et al, 2009) In our context, such groups include Self Health Group (SHG), Micro Finance Institution (MFIs), and agricultural or trade-based cooperatives (e.g. milk producers, rickshaw pullers etc). Therefore, reaching people is a collective action, rather than an individual marketing effort.

The groups we seek to reach are mainly illiterate and innumerate, and without experience in health insurance. Creating awareness about the benefits of health insurance requires evoking intuitions, rather than rational argumentation. Personal experience, anecdotes, and similar forms of knowledge that people can easily relate to. Practical experience showed that games illustrating the benefits of pooling are very effective, as well as stories presented as films, which are directly related to their own life experience, fortified through repeated groups discussions. Awareness campaigns can last several weeks, even though technically they could be compressed to a day or two (but such timing would not allow sufficient group discussions). An example of a short video of a game communicating the benefits of pooling can be seen at

What to base premiums on: WTP, OOPS, actuarial costs, other basis?

Sustainability of the microinsurance can be ensured only when the income (premiums plus subsidies, if any) equals the expenditure (cost of benefits plus administration as a minimum). Commercial insurers fix the premium of each “product” based on the actuarial calculation that should be based on the distribution of frequency and severity of insured events. All decisions on what would be included in the package and its price are taken in advance, by the underwriter. Insurers essentially put clients before a “take-it-or-leave-it” individual choice. However, as we approach groups rather than individuals, the choice of benefits and the premium is mainly subject to group choice, which is really based on what most members of the group are willing and able to pay. This raises the need to discover what each community is willing to pay. That estimate of WTP would define the maximum costs of the package.

Elicitation of WTP is nowadays done usually through contingent valuation, and most often this is done as part of a household survey. (Dror DM / Koren, 2011) A series of seven experiments performed by the author in rural India revealed the following insights (Dror / Radermacher / Koren, 2007):

1. Out-of-pocket spending on healthcare in rural India is about 7% of income (source: Indian National Sample Survey Round 62, 2005-06), but median willingness to pay was only around 1.35% of income. It is thus self-evident that WTP would be much lower than the amounts paid as OOPS and therefore health insurance would cover much less than all healthcare costs.

2. We found that although nominal WTP increased with income, when WTP was expressed as a % of income, it was higher by a factor of 2 in the lowest quintile versus the highest quintile. Therefore, there is no basis to assume that the poor would agree to premiums established as a fixed % of income (as is often done in western countries), or that the notion of cross subsidy would be acceptable. Rather, premiums would have to be fixed nominally and be identical for all members of the community.

3. When we compared the WTP levels across the seven locations, we found significant and material differences, even when different income levels were accounted for. For instance, in rural Karnataka, people were WTP 1.2% of their income, while in rural Bihar they were to pay 2.6%. It turns out that WTP is affected by local parameters (such as education, availability of healthcare facilities locally, experience with costly healthcare).

However, these parameters offer only a partial explanation for the differences across locations, which could also be due to some parameters that have not yet been quantified e.g. trust, social capital etc. In summary, the findings explain why implementation of microinsurance requires local, context-specific estimation of WTP.

Benefit-package design

As the benefit-package of micro health insurance will be limited, the key question is what benefits are most relevant for the target population? This information can be assessed when we know what healthcare services people consume, and how they finance them. A study in five locations in rural India offers some important insights. (Dror DM / van Putten / Koren, 2008) First, contrary to common thinking, that study showed that hospitalization costs accounted only for about 11% of total healthcare costs for the entire study population (ranging from 4% to 28% in different locations). This can be explained by the fact that overall healthcare costs reflect both the frequency and the severity of events, and while hospitalizations are expensive when they occur, they are very rare. The largest share of costs was due to purchase of medicines (49% of total for the entire sample, ranging from 38% to 68% across different locations), followed by the cost of outpatient consultations (33% of total and between 19% and 48% across locations). These two benefit types are very frequent, and the aggregated cost amounts to a lot, even if each event is less expensive. Based on these findings, it seems that the choice in designing a benefit package is between insuring rare events and insuring expensive events. (Dror DM / van Putten / Koren, 2009)

A complementary finding that points in the same direction is the report of households that experienced healthcare costs that they needed to borrow with interest or to sell assets (“hardship financing”) in order to pay for those services. For example, 39% of households that reported a hospitalization resorted to hardship financing; but even households that reported only outpatient care costs, 23% had to resort to hardship financing. Therefore, healthcare costs reflect not only the actual cost of care, but also the indirect cost of financing, be it the interest payable on loans, or the collateral damage due to sale of productive assets. And, catastrophic costs could arise not just in cases of hospitalizations but also when people seek outpatient care that they must pay for.

When selecting benefits to suit the WTP threshold, one of the questions that invariably arise is what cap should be fixed for reimbursements. It is self-explanatory that a higher cap would offer better protection, and would require a higher premium. However, is there value in insurance when the cap must be low? We examined this question by estimating (through simulations) the financial protection that was secured through insurance with different premiums and different caps. For example, using incidence and distribution of cost data we collected in Kalahandi, Orissa, India in 2009, we saw that in the absence of insurance (cap=0) 5.7% of the population would have had to pay for a hospitalization; but with insurance capped at INR 1500 (premium INR 2.63 p.p.p.m.) the share of the population that would have to pay for hospitalization would drop to only 0.9%, and with a cap of INR 5,000 (with premium of INR 3.9 p.p.p.m.) the percentage would further drop to only 0.3% of the population. This illustrates that even modest insurance cover can be of value, and the decision is not only which benefit type to cover, but also the level of insured coverage.

Involving prospective insured in benefit-package design

We assume that WTP and willingness to join health insurance will increase when the target population is involved in benefit package design, because the benefit would respond better to perceived priorities, and active participation could increase trust in and understanding of the insurance. The metrics underlying benefit package composition are complex, and thus we had to devise a simple and intuitive way for the population (mostly illiterate and innumerate) to participate meaningfully. Elicitation of choices was made by using a game-like simulation exercise called CHAT (“Choosing Healthplans All Together”) that enabled participants to choose from 10 benefit types; for most benefit types, participants could choose three coverage levels: basic; medium; and high. (Dror et al.; Social Science & Medicine 2007; Danis M et al 2007)We set the pre-determined budget of the package at INR 500 (~US$ 11) per household per year, which approximated the level of willingness-to-pay (WTP) obtained from a survey conducted by us in the same area. The experiment was conducted in Karnataka and Maharashtra; 24 groups, comprising a total sample of 302 persons participated. Fully 88.4% of the respondents selected at least three of the following benefits: outpatient, inpatient, drugs and tests, with a clear preference to cover high aggregate costs regardless of their probability. The results show that involving prospective clients in benefit package design can be done without compromising the judiciousness of rationing choices, even with people who have low education, low-income and no previous experience in similar exercises.


The success in reaching people at the base of the pyramid through micro health insurance is contingent on establishing a universe of discourse with the prospective insured persons. This is done firstly by engaging in insurance education. Such education can succeed when it is based on people’s daily experiences and on familiar concepts and tried-&-tested modes of interaction. The impact of insurance education is expected to improve people’s understanding of insurance principles, and leverage relations of reciprocity within the community to the insurance as well. Such leveraging can happen best among communities that already have developed some cooperative or mutual-aid action.

Microinsurance is based on understanding the demand-side, and offering context-relevant solutions, which take account of local WTP, healthcare needs, availability of healthcare services, and prevailing costs of care.

Finally, reaching people through microinsurance means a shift in focus from “how to distribute insurance to poor individuals?” to “how can low-income communities manage their own risks?” The latter focus can be achieved when the community is involved in benefit package design, pricing of the premium, and in managing the claims adjudication. All these functionalities must be explained and be the subject of training through a community-centric implementation process. Such an implementation process has been tried and tested in India and in Nepal (

*David Dror, (PhD, DBA) is Chairman & Managing Director at the Micro Insurance Academy (MIA); and Hon. Professor, Erasmus University Rotterdam (Inst. of health policy & management). Contact:



  • Dror DM, Radermacher R (Editors): Financial Inclusion Opportunities for Micro Health Insurance in Nepal: An Exploratory Analysis of Health Incidence, Costs and Willingness to Pay in Dhading and Banke Districts of Nepal, Prior to Launching Community Based Microinsurance. New Delhi, Micro Insurance Academy, Jan. 2010, ISBN 978-81-909841-0-2, vii+141 pp
  • Dror DM, Radermacher R, Khadilkar SB, Schout P, Hay FX, Singh A, Koren R: Microinsurance: innovations in low-cost health insurance. Health Aff (Millwood). 2009;28(6):1788–98, Nov. 2009
  • Dror DM, Koren R (2011): The Elusive Quest for Estimates of Willingness to Pay for Micro Health Insurance among the Poor in Low-Income Countries. In: Micro Insurance Compendium II (Forthcoming).
  • Dror DM, Radermacher R, Koren R: Willingness to pay for health insurance among rural and poor persons: Field evidence from seven micro health insurance units in India. Health Policy, 2007; 82(1):12-27.
  • Dror DM, van Putten O, Koren R. Cost of illness: Evidence from a Study in Five Resource-Poor Locations in India. Indian J for Med Res, (New Delhi), 2008;127:347-361
  • Dror DM, van Putten-Rademaker O, Koren R: Incidence of Illness among Resource-Poor Households: Evidence from Five Locations in India. Indian J for Med Res, (New Delhi), 2009;130:146-154
  • Dror DM, Koren, R; Ost A, Binnendijk, E; Vellakkal, S, Danis, M: Health insurance benefit packages prioritized by low-income clients in India: Three criteria to estimate effectiveness of choice, Social Science & Medicine, February 2007 64(4): 884–896.
  • Danis M, Binnendijk E, Ost A, Vellakkal S, Koren R, Dror DM.: Eliciting the Health Insurance Benefit Choices of Low-income Populations in India with the CHAT Exercise, Economic and Political Weekly (Mumbai) August 11-17, 2007; 42(32):3331-3339