Micro-Finance Health Insurance in Developing Countries - PDF

Please download to get full document.

View again

of 22
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Economy & Finance

Published:

Views: 5 | Pages: 22

Extension: PDF | Download: 0

Share
Related documents
Description
University of Pennsylvania ScholarlyCommons Wharton Research Scholars Journal Wharton School Micro-Finance Health Insurance in Developing Countries Sudha Meghan University of Pennsylvania Follow
Transcript
University of Pennsylvania ScholarlyCommons Wharton Research Scholars Journal Wharton School Micro-Finance Health Insurance in Developing Countries Sudha Meghan University of Pennsylvania Follow this and additional works at: Part of the Insurance Commons, and the International Business Commons Meghan, Sudha, Micro-Finance Health Insurance in Developing Countries (2010). Wharton Research Scholars Journal. Paper 62. This paper is posted at ScholarlyCommons. For more information, please contact Micro-Finance Health Insurance in Developing Countries Abstract In developing economies, health shocks play a significant role in instigating and sustaining poverty. The impact of high catastrophic out-of-pocket health expenditure also fosters a culture in which people decide not to use services because they cannot afford either the direct costs, such as for health check-ups or consultations, medicines or laboratory diagnostic tests, or the indirect costs, such as transportation to the care provider or special food. The objective of this research is to investigate the potential role of voluntary health insurance in India, particularly through a micro-finance framework to reach the most destitute, bottom income quintiles of the population. Consumer Expenditure and Healthcare and Morbidity data from the National Sample Survey Organization of India is used to analyze mean and variance of health spending, projected risk premia, and variables that may predict levels of health spending. Insurance for institutional health spending is potentially feasible given current market demand across income quintiles, and sustainability of microinsurance offerings depends on increasing population reach as well as efficient delivery. Keywords micro-finance, health insurance Disciplines Business Insurance International Business Comments Suggested Citation: Meghan, Sudha. Micro-Finance Health Insurance in Developing Countries. Wharton Research Scholars Journal. University of Pennsylvania. April This thesis or dissertation is available at ScholarlyCommons: Micro-Finance Health Insurance Meghan 1 WHARTON RESEARCH SCHOLARS Micro-Finance Health Insurance in Developing Countries Sudha Meghan Advisor: Dr. Mark Pauly Department of Health Care Management at the Wharton School University of Pennsylvania Acknowledgements First and foremost, I am incredibly grateful for the invaluable mentorship and guidance from my advisor, Dr. Mark Pauly, who has taught me not only the nuances of health economics but also challenged and inspired me to independently delve deeper and develop conceptual rationale. I also want to thank Dr. Martin Asher for making the Wharton Research Scholars Program possible. I am also very thankful for the support of the Leonard Davis Institute of Health Economics and the Wharton School at the University of Pennsylvania. Abstract In developing economies, health shocks play a significant role in instigating and sustaining poverty. The impact of high catastrophic out-of-pocket health expenditure also fosters a culture in which people decide not to use services because they cannot afford either the direct costs, such as for health checkups or consultations, medicines or laboratory diagnostic tests, or the indirect costs, such as transportation to the care provider or special food. The objective of this research is to investigate the potential role of voluntary health insurance in India, particularly through a micro-finance framework to reach the most destitute, bottom income quintiles of the population. Consumer Expenditure and Healthcare and Morbidity data from the National Sample Survey Organization of India is used to analyze mean and variance of health spending, projected risk premia, and variables that may predict levels of health spending. Insurance for institutional health spending is potentially feasible given current market demand across income quintiles, and sustainability of microinsurance offerings depends on increasing population reach as well as efficient delivery. Micro-Finance Health Insurance Meghan 2 Introduction Every year, an estimated 25 million households more than 100 million people are plunged into poverty when they or their relatives become ill and they must struggle to pay for health-care services out of their own pockets, according to the 2006 World Health Organization bulletin on health care financing in developing countries (Braine, 2006). The goal of micro health insurance is to improve the financial protection of the poor uninsured populations in developing countries against excessive health expenditures. The impact of high catastrophic out-of-pocket health expenditure also fosters a culture in which people decide not to use services because they cannot afford either the direct costs, such as for health check-ups or consultations, medicines or laboratory diagnostic tests, or the indirect costs, such as transportation to the care provider or special food (Schieber, Gottret, Fleisher, Leive, 2007). Furthermore, as a result, households can sink further into poverty due to the work loss as a result of illness. Microinsurance for health has shown promise in being able to provide catastrophic health protection for the poor families in developing economies, and it is generally offered through microfinance institutions (MFI) that also offer other microcredit loans to individuals for small business development (Dror et al. 2009). Due to lack of baseline and follow-up data collection, the success of micro health insurance has generally been equated with the household s ability to repay other microcredit loans, and remain financially stable (Dror & Jacquier, 1999). Research Focus The focus of the research is to gauge the market for health insurance in India, particularly for chronic and catastrophic health care benefits, through a micro-insurance model. The micro-insurance model in particular should theoretically facilitate access to care and ability to pay among the lower income quintiles in a country s population. It is hypothesized that micro-insurance can significantly decrease the currently high out-of-pocket health expenditures and serve to smooth risk and spending levels among the target community populations. Health Insurance in Developing Economies Community-financing schemes as micro health insurance have developed as a result of the context of two main failures in developing countries, in terms of catering to disadvantaged populations (Preker et al., 2001). Firstly, most developing countries experience a government failure to collect taxes and organize public finance, which in turn can provide social protection for disadvantaged populations. Micro-Finance Health Insurance Meghan 3 In addition, governments fail to employ oversight of the health sector in general, particularly in terms of the lack of health care supply, both professionals and infrastructure, in rural areas where a large proportion of the population reside. Secondly, there is the market failure on the economic end to establish a functional exchange between supply and demand, due in part because of the gap between needs, demand, and ability to pay, which in turn leads to a lower level of health care supplied (Schieber & Maeda, 1997). It is also in part due to the prevalence of nonmonetary transactions in the informal sector of the economy, particularly in rural, agriculturally-oriented regions (Schieber et al., 2007). In India, approximately 78% of total medical spending is out of pocket, according to the 2005 World Health Report (WHO, 2005). Micro-Finance Framework Micro-finance offerings for healthcare encapsulate credit loans for emergency health situations, health savings accounts, and micro-insurance. Loans are typically not a viable offering among MFIs due to the high default rate and consequent low sustainability (Schieber & Maeda, 1997). A medical savings account mandates or encourages individuals to save and defines that the savings can only be spent for health costs of the owner or family. However, medical savings accounts, offered through MFIs do not facilitate risk pooling among income levels, health status levels or by age and gender groups. Also, the protection available is limited to the balance of the savings. Microinsurance, referring to communityfunded health insurance schemes, is a mechanism for pooling resources and spreading risks across income, age, gender, or health status differences of the entire group (Preker et al., 2001). The major difference is that the responsibility for the health risk is placed exclusively on the beneficiary and his family in the case of savings accounts, whereas microinsurance creates a system of complementary responsibility of individuals and their community; and the focus of this microfinance analysis will be on the latter. Micro-Insurance Evolving from this landscape of small credit offerings is microinsurance, birthed from the presence of social capital when hardships surface, family and community often serve as the sole safety net for low-income populations. In addition, low-income families often hold greater trust in community organizations and feel more confident contributing to a community-based financing program, which has an established positive track record for local impact, rather than to one that operates a broad-based national or regional (state or provincial) health insurance scheme. Micro-Finance Health Insurance Meghan 4 The micro aspect refers to the smaller pool that micro-finance must work with, at a community level. Ideally, insurers want a larger pool because size adds viability. However, in developing countries, health insurers in reality do not want to include poverty-stricken population segments in larger pools, due to lower income, higher health risk, and higher default rates. In addition, the social structure has excluded disadvantaged populations from access to larger schemes (Dror, Radermacher, Koren, 2007). As a result of these two constraints, microinsurance units operate to give the target destitute population to express its needs and priorities in terms of offerings, and also strive to develop a positive opinion toward insurance. The consulting process is essential to the concept of group involvement in self-management that is pivotal in microfinance offerings; the community s and its individual members interests are correlated. Hence, a successful microinsurer must work with the population and the members priorities, rather than simply providing financial resources. This also helps to build trust and maintain the microinsurance system as a socially and community-rooted organization, rather than one that is responding to an external insurer s profit motives. This in turn would lead to a larger continued membership, theoretically, which generates more resources than a smaller one, enabling the group to cover expenses that a single individual could not afford. Study Data and Methods The dataset used is the 2004 Indian National Sample Survey Organization (NSSO) survey data on Household Consumer Expenditure (Schedule 1.0) and Morbidity and Healthcare (Schedule 25.0). Previous research used the World Health Survey (WHS) Datasets for select developing countries to determine the potential for health insurance. However, the main limitation of WHS data for India is the lack of month-to-month and annual health expenses, which is necessary for a predicted measure of variance in spending. The sampling was also limited in regions and number of households surveyed. Previously and in this research, the annual measure is simulated to be 12 times monthly expenditures, which assumes a variance for the distribution. In the NSSO Morbidity and Healthcare dataset, individuals self-reported health status and disease condition is available. This can be used to determine the demand for insurance for chronic conditions as well as the potential for catastrophic coverage. The final household-level data sets combines information on household health spending, total spending, and demographic and health information on the household survey respondent as well as from other individuals in the household. All expenditures are at a household level; individual health spending Micro-Finance Health Insurance Meghan 5 or individual total spending is not available, except for hospitalizations resulting from specific ailments, which is available in the Healthcare and Morbidity dataset. The analyses require an income measure, because health spending in developing countries has been shown to vary by income segment (Pauly, Blavin, Meghan, 2009). Due to data limitations of wealth indication measures (e.g. ownership of a bicyucle, home, and so forth), the NSSO analysis for India delineates income based on a spending based definition. This is defined to be total actual consumption less actual medical care spending for each household. This serves to stratify the population by income quintiles. Health expenditure data is the second piece, in which monthly and/or biweekly data is compared with the annual reported values or estimates, to subsequently calculate the projected insurance premium. Health spending is available for inpatient, outpatient, and outpatient drug expenses. The insurance premium calculation is used to gauge whether there is a demand for insurance, through a comparison to current out-of-pocket expenditures. Risk Premium Calculation r I Risk premium 0.5 σ 2 I Where r(i) is the relative risk-aversion coefficient, I is income, and σ 2 (the square of the standard deviation σ) is the variance of the residual for the risky distribution (Phelps, 2003). In order to identify which segments of a population have higher levels of drug spending, regression models are generated. Drug spending in a 30-day period (Schedule 1.0 Consumer Expenditure dataset) was regressed by income quintiles, region (urban or rural), the presence of an elderly person (elderly defined as age greater than 59 years), number of children (age less than 17 years), household size, and whether or not there was a hospitalization in the past 365 days. Drug spending in a 15-day period (Schedule 25.0 Health Care and Morbidity Dataset) was regressed by income quintiles, region, marital status (single or partner), education level, gender, age groups, and selfreported specific ailments that either (1) were present during the past 15 days or (2) led to hospitalization during the last 365 days. Results Analysis of annual mean income (spending-based definition) and health expenditure ($ PPP) for inpatient care and non-institutional expenses resulted in average expenditures that are positive for all income quintiles, as shown in Exhibit 1. The share of spending by the bottom 80% of households on the basis of income is 42% for inpatient medical expenses, 57% for non-institutional medical expenses, 58% for drugs, and 55% for physician fees. The share of spending by the bottom 40% of households on the Micro-Finance Health Insurance Meghan 6 basis of income is 10% for inpatient medical expenses, 17% for non-institutional medical expenses, 18% for drugs, and 16% for physician fees. For the first income quintile, the average income is 1483($ PPP) and the inpatient medical expenses was 17, non-institutional medical expenses is 52 overall, and within which is 45 for drugs, and 5 for physician fees. For the second income quintile, the average income is 2541($ PPP) and the inpatient medical expenses is 25, non-institutional medical expenses is 109 overall, and within which is 93 for drugs, and 11 for physician fees. Exhibit 1 Annual Mean Income and Health Spending ($ PPP) for Inpatient Medical Expenses, Non- Institutional Medical Expenses, Prescription Drugs, and Physician Fees in All Households Note: PPP is purchasing power parity. The calculated risk premium is determined based on the mean spending, the variance of spending, and the coefficient of risk aversion enumerated as 2.0. Represented as a percentage of the mean health spending, the market demand for feasible insurance can be delineated. The theory of insurance demand predicts that risk-averse households will voluntarily opt for insurance if it can be offered to them at a premium whose excess over the expected expenses is smaller than the risk premium they would be willing to pay (Phelps, 2003). As a percentage of mean spending, the risk premium for the entire population, across income quintiles is 227% for inpatient medical expenses, 35% Micro-Finance Health Insurance Meghan 7 for non-institutional medical expenses, within which it is calculated to be 29% for drugs only, and 10% for physician fees only, as shown in Exhibit 2. This pattern of a feasible percentage for inpatient medical expenses and low percentage and market potential for non-institutional expenses (drugs and physician fees) is paralleled for the bottom two income quintiles, for which it is 260% and 89% for inpatient medical spending, 46% and 44% for non-institutional spending, 39% and 40% for drugs only, and 13% and 9% for physician fees only, for the first and second income quintiles respectively. Exhibit 2 Annual Calculated Risk Premium ($ PPP) for Inpatient Medical Expenses, Non-Institutional Medical Expenses, Prescription Drugs, and Physician Fees in All Households Note: PPP is purchasing power parity. To further investigate the potential role of insurance coverage for drug spending, 30-day drug spending was regressed by household attributes that might increase the predisposition for higher level of expenses: regional situation (urban or rural), presence of an elderly person (defined as 60 years and older), number of children (defined as less than 17 years), number of household members, and whether there was a hospitalization in the past 365 days. In the Consumer Expenditure survey, in the final dataset, 11,914 households report 0 drug spending in the past 30 days. A probit regression was first used to model this spending distribution to control for the high number of households with no drug expenditures. All explanatory variables are significant. Geographic situation in a rural area is significant with a coefficient of -0.09, standard error of 0.02 and z-statistic of -5.45, meaning that the drug spending Micro-Finance Health Insurance Meghan 8 is predicted to be higher if a household is in a rural region. Presence of elderly is significant with a coefficient of 0.25, standard error of 0.02, and z-statistic was Number of children is significant with a coefficient of -0.02, standard error of 0.01 and z-statistic of Household size is significant and has a coefficient of 0.03 with a standard error of 0.01 and z-statistic of Hospitalization is significant with a coefficient of -0.05, standard error of 0.02, and z-statistic of Exhibit 3 Probit Regression ($ PPP) for 30-day Drug Spending in All Households: Effects of Income, Region, Household Age Structure, and Hospitalization Note: PPP is purchasing power parity. Following up the probit regression, an OLS regression model was built including only those households with positive drug spending. Parallel to the earlier model, all explanatory variables in the models are significant. Approximately 42% of households included in the models have a predicted level of drug spending higher than the actual mean spending during the 30-day period, as shown in Exhibit 4. Geographic situation in a rural area is significant with a coefficient of -134, standard error of 43 and z- statistic of -3.1, meaning that the drug spending is predicted to be higher if a household is in a rural region. Presence of elderly is significant with a coefficient of 311, standard error of 42, and z-statistic was 7.3. Number of children is significant with a coefficient of -47, standard error of 17 and z-statistic of Micro-Finance Health Insurance Meghan Household size is significant and has a coefficient of 32 with a standard error of 12 and z-statistic of 2.7. Hospitalization is significant with a coefficient of 624, st
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks