H N P D I S C U S S I O N P A P E R Hedging The Health of The Poor The Case for Community Financing in India Anil Gumber September 2001 HEDGING THE HEALTH OF THE POOR The Case for Community Financing in India Anil Gumber September 2001 Health, Nutrition and Population Discussion Paper This series is produced by the Health, Nutrition, and Population Family (HNP) of the World Bank's Human Development Network (HNP Discussion Paper). The series provides a vehicle for publishing preliminary and unpolished results on HNP topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of the Bank's Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. For free copies of papers in this series, please contact the individual authors whose name appears on the paper. Inquiries about the series and submissions should be directed to the Editor in Chief. Submissions should have been previously reviewed and cleared by the sponsoring department that will bear the cost of publication. No additional reviews will be undertaken after submission. The sponsoring department and authors bear full responsibility for the quality of the technical contents and presentation of material in the series. Since the material will be published as presented, authors should submit an electronic copy in a predefined format as well as three camera-ready hard copies (copied front to back exactly as the author would like the final publication to appear). Rough drafts that do not meet minimum presentational standards may be returned to authors for more work before acceptance. The Editor in Chief of the series is Alexander S. Preker (apreker@worldbank.org); For information regarding this and other World Bank publications, please contact the HNP Advisory Services (healthpop@worldbank.org) at: Tel (202) 473-2256; and Fax (202) 522-3234. ISBN 1-932126-02-3 © 2001 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved. ii Health, Nutrition and Population Discussion Paper Hedging the Health of the Poor The Case for Community Financing in India Anil Gumbera aSenior Economist, National Council of Applied Economic Research, New Delhi (Currently on Lien); Senior Research Fellow, Centre for Health Services Studies, Warwick Business School, Coventry, UK Prepared for Working Group 3 of Macroeconomics and Health Version: September, 2001 Abstract: This paper reviews the existing community-based and self-financing health insurance schemes in India catering to the general population as well as addressing the needs of the poor and vulnerable section of the society. Also discussed are some critical issues of accessibility and use of health care services, out-of-pocket expenditure on treatment and the need for health insurance for poor households pursuing varied occupations in both rural and urban areas. The paper examines in detail the determinants of enrollment in the community-based financing scheme, using the household-level data from the pilot study undertaken in Gujarat (India). It also investigates the issue of how much health insurance mitigates the households' burden of health care expenditure. The findings suggest that the community plan fairly addresses equity in enrollment but that, in terms of providing financial protection, social insurance coverage is much more successful. Keywords: burden of ill-health, community financing, financial protection, health care utilization, health insurance, household health expenditure, poor, Self-Employed Women's Association, social insurance Disclaimer: The findings, interpretations and conclusions expressed in the paper are entirely those of the authors, and do not necessarily represent the views of the World Bank, its Executive Directors, or the countries they represent. Correspondence Details: Senior Research Fellow, Centre for Health Services Studies, Warwick Business School, Coventry, CV4 7AL, United Kingdom, Tel: +44 (0)26 7652 2300, Fax: +44 (0)26 7652 4863, Email: Anil.Gumber@wbs.ac.uk, anilgumber@yahoo.co.in iii iv CONTENTS PREFACE...........................................................................................................................................VII ACKNOWLEDGEMENTS.................................................................................................................IX I. INTRODUCTION............................................................................................................................... 1 II. COMMUNITY FINANCING IN INDIA AND THE SEWA PROGRAM...................................... 2 Micro Credit Linked Health Insurance Schemes............................................................................. 5 III. RESEARCH DESIGN AND METHODOLOGY.......................................................................... 6 3.1 RESEARCH DESIGN ...................................................................................................................... 6 3.2 METHODOLOGY............................................................................................................................ 7 3.2.1 Determinants of participation in mutual health organizations .............................................. 7 3.2.2 Level of financial protection provided by SEWA.................................................................... 7 3.2.3 Variables used in the model ................................................................................................... 7 IV. RESULTS ........................................................................................................................................ 8 4.1 DETERMINANTS OF PARTICIPATION............................................................................................. 8 4.2 DETERMINANTS OF FINANCIAL PROTECTION IN COMMUNITY FINANCING .................................11 REFERENCES.....................................................................................................................................17 v vi PREFACE In January 2000, Dr. Gro Harlem Bruntland, Director General of the World Health Organization (WHO), established a Commission on Macroeconomics and Health (CMH) to provide evidence about the importance of health to economic development and poverty alleviation. This HNP Discussion Paper is based on a Report on community financing submitted in September 2001 to Working Group 3 of the CMH. The mandate of Working Group 3 was to examine alternative approaches to domestic resources mobilization, risk protection against the cost of illness, and resource allocation. The working group was chaired by Professor Alan Tait (Former Deputy Director of Fiscal Affairs, International Monetary Fund, and currently Honorary Fellow at University of Kent at Canterbury and Honorary Fellow at Trinity College, Dublin) and Professor Kwesi Botchewey (Director of Africa Research and Programs at the Harvard Center for International Development). Professor Jeffery D. Sachs (Chairman of the Commission and Director of the Harvard Center for International Development) presented the findings of the CMH in a Report that was submitted to WHO on December 20, 2001--Macroeconomics and Health: Investing in Health for Economic Development. The Report of the CMH recommended a six pronged approach to domestic resource mobilization at low-income levels: "(a) increased mobilization of general tax revenues for health, on the order of 1 percent of GNP by 2007 and 2 percent of GNP by 2015; (b) increased donor support to finance the provision of public goods and to ensure access for the poor to essential health services; (c) conversion of current out-of-pocket expenditure into prepayment schemes, including community financing programs supported by public funding, where feasible; (d) a deepening of the HIPC (Highly Indebted Poor Countries) initiative, in country coverage and in the extent of debt relief (with support form the bilateral donor community); (e) effort to address existing inefficiencies in the way in which government resources are presently allocated and used in the health sector; and (f) reallocating public outlays more generally from unproductive expenditure and subsidies to social-sector programs focused on the poor." Most community financing schemes have evolved in the context of severe economic constraints, political instability, and lack of good governance. Usually government taxation capacity is weak, formal mechanisms of social protection for vulnerable populations absent, and government oversight of the informal health sector lacking. In this context of extreme public sector failure, community involvement in the financing of health care provides a critical albeit insufficient first step in the long march toward improved access to health care by the poor and social protection against the cost of illness. The CMH stressed that community financing schemes are no panacea for the problems that low- income countries face in resource mobilization. They should be regarded as a complement to--not as a substitute for--strong government involvement in health care financing and risk management related to the cost of illness. Based on an extensive survey of the literature, the main strengths of community financing schemes are the degree of outreach penetration achieved through community participation, their contribution to financial protection against illness, and increase in access to health care by low-income rural and informal sector workers. Their main weaknesses are the low volume of revenues that can be mobilized from poor communities, the frequent exclusion of the very poorest from participation in such schemes without some form of subsidy, the small size of the risk pool, the limited management capacity that exists in rural and low-income contexts, and their isolation from the more comprehensive benefits that are often available through more formal health financing mechanisms and provider networks. The work by the CMH proposed concrete public policy measures that governments can introduce to strengthen and improve the effectiveness of community involvement in health care financing. This includes: (a) increased and well targeted subsidies to pay for the premiums of low income populations; vii (b) use of insurance to protect against expenditure fluctuations and use of re-insurance to enlarge the effective size of small risk pools; (c) use of effective prevention and case management techniques to limit expenditure fluctuations; (d) technical support to strengthen the management capacity of local schemes; and (e) establishment and strengthening of links with the formal financing and provider networks. The report presented in this HNP Discussion Paper has made a valuable contribution to our understanding of some of the strengths, weaknesses, and policy options for securing better access for the poor to health care and financial protection against the impoverishing effects of illness, especially for rural and informal sector workers in low-income countries. Alexander S. Preker Chief Economist Health, Nutrition, and Population The World Bank viii ACKNOWLEDGMENTS This paper further explores household-level data of NCAER-SEWA study on "Health Insurance for Workers in the Informal Sector." We are thankful to the Ford Foundation (the Sponsor) and the SEWA as our collaborator in the primary fieldwork in Ahmedabad. The author would also like to thank Ms. Shakun Datta for providing assistance in extracting results of multinomial logit models through STATA software. The authors are grateful to the World Health Organization for having provided an opportunity to contribute to the work of the Commission on Macroeconomics and Health and to the World Bank for publishing the report as an HNP Discussion Paper. ix x I. INTRODUCTION More than 90 percent of Indian population and almost all the poor are not covered under any health insurance scheme. Their health care needs are met primarily through direct out-of-pocket expenditure on services provided by the public and private sectors. However, various studies on the use of health care services show that the poor and other disadvantaged groups (scheduled castes and scheduled tribes) are forced to spend a higher proportion of their income on health care than the better-off. For the disadvantaged, the burden of treatment, especially inpatient care, is disproportionately heavy (Visaria and Gumber 1994). The high incidence of morbidity cuts their household budget both ways: in the large amounts of money and resources they have to spend on medical care and in the earnings they have to forgo during periods of illness. Often they have to borrow funds at very high interest rate to meet both medical expenses and other household consumption needs. One possible consequence of this could be pushing these families into a zone of permanent poverty. There are also concerns about problems in accessibility and use of subsidized public health facilities. Most poor households, especially the rural ones, reside in backward, hilly, and remote regions where neither government facilities nor private medical practitioners are available. They have to depend heavily on poor quality services provided by local, often unqualified practitioners and faith healers. Further, wherever accessibility is not a constraint, the primary health centers are generally found to be either dysfunctional or providers of low-quality services. The government's claim to provide free secondary and tertiary care does not stand up; in reality, patients are charged for various services (Gumber 1997). Estimates based on a large-scale health care utilization survey of 1993 suggest that overall about 6 percent of household income is spent on curative care, which amounts to Rs. 250 per capita per annum (Shariff et al. 1999). However, the burden of expenditure on health care is unduly heavy on households in the informal sector, indicating the potential for voluntary comprehensive health insurance schemes for these segments of the society. Overall, health insurance coverage is low. Only 9 percent of the Indian workforce is covered by some form of health insurance (through CGHS, ESIS and Mediclaim), and most of those insured belong to the organized sector (Gumber 1998). Health insurance coverage is so sparse because government policy has been to provide free health services through public hospitals, dispensaries, and clinics. In reality, public sector providers charge patients for various services, and outreach is also poor. According to estimates based on the National Sample Survey (NSS) 1986­87, 42 percent of inpatients and 30 percent of outpatients using public sector facilities had paid for various services; the percentages varied substantially between rural and urban areas and among states (Gumber 1997). Further, health care costs have increased enormously. A comparison of NSS data for 1986­87 and 1995­96 suggests that the cost of inpatient care and outpatient care grew annually at 26-31 percent and 15-16 percent, respectively, putting severe strains on efforts to achieve equity in health care (Gumber 2001). Nongovernmental organizations (NGOs) and charitable institutions (not-for-profit) have played an important role in delivery of affordable health services to the poor, but their coverage has always been small. The issue is how to reach the unreached and how to ensure that the uninsured to get at least a minimum of affordable quality services. The public insurance companies so far have paid very little attention to voluntary medical insurance due to low profitability, high risk, and lack of demand. From the consumer point of view, insurance coverage is low because information about the private insurance plans is lacking, and the mechanisms used by the health insurance providers are not suitable to consumers. Further, in comparison to the Employees' State Insurance Scheme (ESIS) and to the community-based schemes as well, the private plans cover a modicum of benefits (Table 1), i.e., only hospitalization and that, with many exclusions 1 (e.g., preexisting conditions). One analysis suggests that the existing voluntary health insurance plans cover only between 55 and 67 percent of the total hospitalization costs and, on average, only 10 to 20 percent of the total annual out-of-pocket expenditure on health care (Gumber 2000a). Table 1 Type of Health Care Burden on Households Covered by Health Insurance Schemes Type of Care/Cost ESIS SEWA Mediclaim Inpatient Medical a a a Transport and other direct cost r r r Loss of earnings a r r Outpatient Medical a r r Transport and other direct cost r r r Loss of earnings a r r Preventive and Immunization a r r promotive Ante- and Post-natal care a r r Maternity care a a r Family planning r r r Note: SEWA and Mediclaim are reimbursement plans (subject to the sum assured) whereas ESIS is a facility-based plan. Gender bias in use of health care persists. Men have better access than women for various socioeconomic and cultural reasons. Poor women are most vulnerable to diseases and ill-health because they live in unhygienic conditions, carry a heavy child-bearing burden, place little emphasis on their own health care needs, and encounter severe constraints in seeking health care for themselves. Institutional arrangements have so far been lacking in correcting these gender differentials. A pioneering study undertaken by Gumber and Kulkarni (2000) looked into issues related to the availability and needs of health insurance coverage for the poor, especially women, and the scope and likely problems in extending current health insurance benefits to workers in the informal sector. This paper attempts to review existing community-based and self-financing health insurance schemes in India that serve the general population and address the needs of the poor and vulnerable. Also discussed are some critical issues of accessibility and use of health care services, out-of-pocket expenditure on health care, and the need for health insurance for poor households pursuing varied occupations in both rural and urban settings. The paper examines in detail the determinants of enrollment in the community-based financing scheme, using household level data from the pilot study. It also investigates the issue of how much health insurance mitigates the households' burden of health care expenditure. II. COMMUNITY FINANCING IN INDIA AND THE SEWA PROGRAM Community and self-generated financing programs are usually run by NGOs or nonprofit organizations. These organizations rely on financing from various sources, including government, donor agencies, and community and self-generated sources. Among the many innovative methods being used to finance health care services are progressive premium scales, community-based prepayment and insurance schemes, and income-generating schemes. These organizations' target population for health care services is primarily workers and families outside the formal sector. Program revenue comes from the following sources: · user fees, defined as the payment made by the beneficiaries directly to the health care provider (e.g., fees for services or prices paid for drugs/immunization). This mode of financing is not common. · prepayment/insurance schemes, including payment by members for drugs either at subsidized rates or at cost. · commercial schemes for-profit actively run by health care finance organizations. 2 · fund-raising activities by organizations to pay for health care services. This type of revenue is more than 5 percent of some organizations' total funding. · contributions in kinds (e.g., rice, sorghum, community labor). Because this method is hard to manage, it is not very popular. Other sources of community-based and self-financing include the Tribhovandas Foundation, which provides health care through village milk cooperatives, and Amul Union (the milk cooperative organization), which puts a cess on milk collection to pay for health care. Tables 2 and 3 describe select schemes. Most of the successful case studies (Dave 1991) happen to be in the states of Assam, Gujarat, Maharshtra, Orissa, Tamil Nadu, and West Bengal. The experience of such schemes could be useful for understanding their merits and disadvantages and their potential for replication in other states. The most pertinent point about these schemes is their rural orientation and ability to mobilize resources in a village community. However, most of these schemes have served only a small segment of the population and their health coverage has been restricted to elementary, preventive, and maternal and child health (MCH) care. Table 2 Salient Characteristics of Selected NGO-Managed Health Insurance Schemes Voluntary Total organizations/ Population annual cost location Date started Service provided Health service delivery/organization served (Rs.) Sevagram/ Hospital, 1. 500 bed hospital -- -- -- Wardha, 1945 2. Out reach community Trained male VHW provides basic 19457 69459 Maharasthra Community health program curative, preventive and promotive health health care. Mobile with doctor and program 1972 ANM provides care every 2 months Bombay Mother 1947 Haelth activities, Two . outpatient and inpatient maternity 120175 and child welfare maternity hospitals (40 care (health and society (BMCWS)/ beds each) with child . Outpatient pediatric care including non health Chawla in Bombay welfare centers, Non- immunization combined) health activities, Day care centers, convalescent home Raigarh 1969 Federation of 3 referral . RAHA functions include 400000 30000- Ambikapur Health Community hospitals and 65 management of insurance scheme, 50000 Association health independent health centers training and support for health centers. (cost range (RAHA)/ services with outreach community . health centers staffed by nurse of Raigarh, Madhya started 1974 care provide outpatient care run MCH individual Pradesh clinic health . VHWs provide community based centers of care which there are 65) Christian Hospital/ Hospital 120 bed hospital, Outpatient/inpatient care, specialties -- 1911740 Bissamaucuttak, 1954, out community project include obstetrics, gynecology, (hospital Orissa reach currently not operational surgery, ophthalmology only) community care 1980 UPASI 19th century Association of tea growers CLWS provides training, management 250000 300000 Coocnoor, Tamil CLWS - 1971 run comprehensive labor support to health programs of Nadu welfare scheme (CLWS) individual tea estates. Tea estates have small cottage hospital and outreach care provided by local workers. Goalpur Co- 1964 Dispensary, periodic Doctor provides outpatient care twice 1247 32000 operative Health community health services weekly Society Shanthiniketan, West Bengal Students health 1955 Polyclinic plus 28 regional Polyclinic has 20 beds provides 550000 2950745 home clinics outpatient and inpatient care; Regional West Bengal clinics, outpatient care only, health education campaigns, blood donation camps. 3 Saheed Shabsankar 1978 Dispensary occupational Doctors provide outpatient care weekly -- 87780 Saba Samithi health activities, rural MCH clinic. (SSSS) health program, school Burdwan, West health program, fair price Bengal medicine shop Arvind eye 1976 2 Urban hospitals (100 Outpatient and inpatient eye care -- 10987700 hospital beds), 2 rural hospitals Madurai, Tanil (500 beds), outreach Regular eye camps organized Nadu program Tribovandas 1980 Community based health CHWs provide basic curative, 300000 1080000 Foundation program linked with milk preventive and promotive care; field (health and Anand, Gujarat cooperatives, regional supervisors provide support to CHWs non health rehabilitation centers, milk society building used as base for combined) Balwadis women's coordinating health services. income generating scheme SEWA Union 1972, Union of self employed Health centers in urban slums and rural 63000 391850 Ahemadabad, health women. Helps organize villages. CHWs provide basic care, (health Gujarat program 1984 women into cooperatives doctors provide support twice weekly. program of various traders, only) provides credit facilities. Provide health care as a support which stocks rational generic drugs. CINI 1975 Community based health CHWs provide MCH care through 70000 1900000 Daulatpur, West programs, dispensary and Mahila Mandals, doctors run daily (Communit Bengal outreach rehabilitation OPD, weekly MCH clinic, y health centre. Other activities: supplementary feeding project income generating schemes, farm, health training, research Source: Dave (1991). Table 3 Prepayment and Insurance Mechanisms in Selected NGO-Managed Health Insurance Schemes Tribovandas Students health Features Sevagram RAHA Foundation Goalpara home SSSS Coverage Household Individual Household Household Institutional Individual provided and individual Annual 8 payali sorghum Rs.5 or Rs.2 rice Rs.10 Rs.18 in cash Rs.2 Rs.2 or Rs.5 subscriptio or in kind Institutions n fee (landless) and 2 payali (rice or labor) Rs.6- sorghum per acre Individuals extra(land holders), or equivalent cash. Number of At least 75% of 75000 Approximately 150 out of 630 institutes 6800 members households (23 villages 1/5 to 1/6 of all 175 total 350000 covered) Total insured households in households in students 14390 villages, (319 village covered villages covered) Member Community care : free Community care: free Community care Dispensary: Polyclinic/regi outpatient entitlement CHW services, drugs and CHW services and : free services , Free doctor onal clinics: clinic: free mobile (doctor +ANM) drugs. Free health subsidized consultations, free consultations services. center services drugs. drugs at cost. consultations, , drugs at including MCH clinic Free periodic drugs, cost, free Hospital : free care for Hospital: 50% public health diagnostic MCH care unphased illness episodes., Hospital : free care subsidy activities tests, 25% subsidy for after paying entrance operations, bed anticipated illness fee up to ceiling of stay at nominal episodes, e.g., pregnancy Rs.1000 charges and chronic ailments. Non- Non-members not entitled Non-members Non-members Non-members Non-members Non- member to use community health charged for drugs have same charged for not entitled to members are entitlement services (over cost), not emoluments to drugs (over avail of not entitled entitled to attend community cost) services to avail the MCH clinic services as services 4 MCH clinic services as services members but not hospital care Manageme VHW responsible for Individual health VHW services Village health Institutions Able to nt of fund membership collections, centers responsible responsible for communities enrolled once a enroll Collections once a year at for membership membership -- funds year. through the harvest time. Compulsory collections. collections. collections Individuals year. No that 75% of villages Collections once a once a year. ongoing (no waiting covered. year. New members Collected once a waiting period) period waiting period 2 year at times- between months before bonus payments enrollment services entitlements distributed and service Rs.3 retained by (non adult entitlements. center, Rs.2 to society members RAHA for referral can also enroll fund. in scheme Source: Dave (1991) Microcredit Linked Health Insurance Schemes Several NGOs and governments in developing countries have started microcredit schemes for vulnerable groups to break the vicious circle of poverty, malnutrition, disease, low productivity, and low income. Microcredit is now considered not only an effective tool for reducing poverty but also as an instrument for empowering the poor, especially women. This operation generates income for the poor by extending small credits for self-employment and other economic activities. However, loan repayments by these groups were far below the expected level. The experience suggested that ill- health, expenditures on treatment, and associated consumption needs were the prime reasons for defaults. To stop the erosion of borrowers' income by health care needs, some NGOs (such as Grameen Bank in Bangladesh and SEWA in India) have introduced health insurance schemes for their members. The Grameen Bank Health Program was started in 1994 to adopt disease-prevention measures, to arrange for minimum-cost treatment, and to build a nonprofit primary health care system. Under this scheme, the borrowers were asked to pay a fixed annual amount of 60 Taka per family as a premium and a trivial sum at the time of use. The scheme has proven to meet the desired objectives (Rahman 2000). In India, SEWA is a trade union of 2,15,000 women workers in the informal sector. It organizes them at the household level toward the goals of full employment and self-reliance. Full employment includes social security, which in turn incorporates insurance. SEWA's experience repeatedly revealed that, despite their efforts to escape from poverty through enhanced employment opportunities and increased income, women were still vulnerable to various crises in their lives. Their efforts were repeatedly frustrated by crises such as sickness, the death of a breadwinner, or accidental damage to and destruction of their homes and work equipment. Too often, maternity also becomes a crisis for a woman, especially if she is poor, malnourished, and lives in a remote area. One of the SEWA studies observed that women identified sickness of themselves or of their family member as the major stress event in their lives (Chatterjee and Vyas 1997). It was also a major cause of indebtedness among women. The health insurance program was, from the start, linked to SEWA's primary health care program, which includes occupational health services. Thus, insured members also have access to preventive and curative health care with health education. Health insurance accounts for most of the claims and for 50 percent of the premium paid out to the insurance program by SEWA members. The scheme was introduced by the SEWA Bank in March 1992 with an initial enrollment of 7,000 women from Ahmedabad City (Chatterjee and Vyas 1997). Later extended to cover rural woman from nine districts of Gujarat, it now enrolls 30,000 women, half of them rural dwellers. Health insurance is an integral part of SEWA's insurance program. The main motivation for initiating a health insurance scheme for women is that maintaining an active, health-seeking behavior is a vital component for ensuring a good quality life and women tend to place a low priority on their own health care needs. 5 The SEWA health insurance program includes maternity coverage, hospitalization coverage for a wide range of diseases, and coverage for occupational illnesses and diseases specific to women (Table 4). It covers diseases that are not covered by the GIC's Mediclaim plan and also provides life and asset insurance for the woman, her husband, or, in case of widowhood or separation, for other household members. Administrative procedures under the plan are simplified. Table 4 Coverage under SEWA Scheme Coverage amount Premium Provider Description of coverage (Rs.) (Rs.) New India Accidental death of the woman member 10,000 3.50 Assurance Loss of assets Accidental death of a member's husband 10,000 3.50 SEWA Loss during riots, fire, floods, theft, etc.: 8.00 (a) of work equipment 2,000 (b) of the housing unit 3,000 Health Insurance (Including coverage for: 1,200 30.00 (a) gynecological ailments (10) (b) occupational health related diseases) (5) Maternity benefits 300 -- Life Insurance Natural death 3,000 15.00 Corporation of Accidental death 25,000 India Note: Total premium for the entire package is Rs. 60 plus a service charge of Rs. 5. SEWA health insurance scheme functions in coordination with Life Insurance Corporation of India (LIC) and New India Assurance Company (NIAC). SEWA has integrated the schemes of LIC and NIAC into a comprehensive health insurance package to address women's basic needs. The claimants are the needy health-benefits seekers and as the insurance is an additional benefit, the beneficiaries willingly pay the premium. Most of the insurers opt for a fixed deposit of Rs. 500 or Rs. 700 (depending upon the type of coverage) with the SEWA Bank; accrued interest on the deposit goes toward the annual premium. The SEWA Bank's large membership and assets enabled it to provide this insurance coverage at low premiums. III. RESEARCH DESIGN AND METHODOLOGY 3.1 RESEARCH DESIGN This paper is based on a primary household survey undertaken in Ahmedabad district of Gujarat in 1998­99. The survey covered about 1,200 households from rural and urban areas. The households were stratified into four categories according to health insurance status. About 360 households belonged to a contributory plan known as Employees' State Insurance Scheme (ESIS) for industrial workers. Another 120 households subscribed to a voluntary plan (Mediclaim), and 360 households were members of a community-based financing scheme run by an NGO, the Self Employed Women's Association (SEWA). The remaining 360 households were uninsured and were purchasing health care services directly on the market. This last subsample was taken to serve as a control group. The idea of selecting such stratification was to understand the varying health needs, access to health services, treatment pattern and the types of benefits received by sample households under the different health insurance environment. The survey was conducted in eight slum-dominated localities in the city of Ahmedabad and six neighboring villages. On an average, 60 households per village and 90 households per urban locality were selected. The criterion for selecting a village or an urban locality was that the settlement should have a cluster of households covered by the SEWA and ESIS plans. The sample canvassed from each settlement included about equal numbers of households from the ESIS, SEWA, and uninsured 6 categories (20 each from a village and 30 each from an urban locality). The sample was purposive and no house listing prior to the survey was carried out. On the other hand, the sample of Mediclaim/Jan Arogya beneficiaries belonging to Ahmedabad city was selected from the list of subscribers obtained from the offices of two companies, the United India Insurance and New India Assurance. 3.2 METHODOLOGY 3.2.1 Determinants of participation in mutual health organizations Prob (membership>0) = X + (1) where X represents a set of independent variables that are hypothesized to affect membership in community based schemes. These variables include income, gender, age, marker on chronic illness or disability. is a vector of coefficient estimates and is the error term. 3.2.2 Level of financial protection provided by SEWA To assess the impact of mutual health organization on financial protection of members, two aspects have to be taken into account: the probability of visiting a health care provider and the out-of-pocket expenditure borne by the individual. We use a two-part model developed as part of the Rand Health Insurance Experiment: - a logit model, which assess the probability of visiting a health care provider Prob (visit >0) = X + u (2) where X stands as a vector for individual, household and community characteristics. - a log-linear model that estimates the incurred level of out-of-pocket expenditures per episode, conditioning on positive use of health care services: Log (out-of pocket expenditure / visit > 0) = X + e (3) where X represents a set of independent variables that are hypothesized to affect individual pattern of utilization and expenditure on treatment. 3.2.3 Variables used in the model Table 5 gives an overview of the variables included in the analysis. Table 5 Overview of Variables Used Variable Description Individual characteristics Gender Male and Female Age Completed years of age at the time of last birthday; broad age groups are used in the model Marital status Never married, currently married, and widowed/divorced/separated. Education level Years of schooling: broadly classified as illiterate, below primary, primary, middle, secondary, graduate and above. Activity status Usual activity status during the last one year: broadly classified as non- worker, self-employed in agriculture, casual laborer, home-based production worker, trade/sales worker, salaried worker in organized sector, salaried worker in unorganized sector and subsidiary status worker. Health characteristics Acute morbidity Episode of illness during last 30 days not involving hospitalization Chronic morbidity Prevalence of any chronic disease/ailment Hospitalization Any illness resulting in hospitalization during last 365 days Childbirth Childbirth during last two years Duration of illness Number of days the person was ill and also categorized into groups. Source of care Source of treatment; broadly categorized as no treatment, use of public 7 including ESIS facility and private facility. Cost of treatment Cost of treatment includes: Direct out-of-pocket payments toward fees, medicines, diagnostic tests, surgery, bed charges, transportation, special diet, etc. Indirect costs include income/wage loss of the patient and the caring person as well as interest payments on amount borrowed to meet treatment expenses. Health insurance enrollment Community-based plan (SEWA), social insurance (ESIS), private plan (Mediclaim) and Uninsured Household characteristics Income/Expenditure Annual household income from different sources; categorized into quintile groups. Monthly household expenditure by broad items - but not considered in the model Household size Number of members usually residing in the house and sharing food from the common kitchen Community characteristics Area of residence Usual place of residence: rural or urban area IV. RESULTS 4.1 DETERMINANTS OF PARTICIPATION Multinomial logit model is used to identify various determinants of being enrolled in SEWA health insurance plan among women members of SEWA. Out of the total 645 SEWA women members aged above 15 years in the sample, 236 (36.6 percent) were enrolled in the plan. Out of 10 variables used in the model, three depicted health status (whether suffering from any chronic ailment, hospitalized during last 365 days and had delivery during last two years), four personal characteristics (age, education, marital status and activity status), two household characteristics (household size and income quintile) and one community variable (area of residence). The description of these variables is provided in Table 5. The mean value of enrollment rate varied across these characteristics. The enrollment rate was higher among women who had reported as suffering from any chronic ailment or had been hospitalized in the previous 365 days but not among those who had reported delivery during last two years. Among personal characteristics, the mean enrollment rate was found to be higher in the middle age groups, 36-45 years and 46-55 years than the other age groups; it was also higher among currently married women. However, with level of education the mean enrollment rate tended to decline. The rate was found to be much lower among nonworkers or subsidiary status workers than among home-based production or salaried workers. The rate tended to decline with the size of household and did not vary much across income quintiles, except in the top quintile where it was marginally higher. Overall, the enrollment rate was higher among urban than the rural women, mainly due to better access to information as well as to SEWA Bank located in Ahmedabad city which manages the scheme. The alternative results of multinomial logit models (interchanging activity status and income variables) are presented in Table 6. The explanatory power of the model (Pseudo R2) ranged between 0.185 (without income variable) to 0.218 (with inclusion of both income and activity status variables). The followings are the main findings: · There was no adverse selection in terms of whether the member had been suffering from any chronic ailment or being hospitalized before. However, maternity, a predictable event, had increased the likelihood of enrollment to take advantage of benefit allowance of Rs. 300 and coverage of the high risk of hospitalization. 8 · Among the personal attributes, the odds of being enrolled were five to seven times higher among middle-aged groups than in the 16-25 years age group. For a currently married women, the odds was twice as high as for never-married women. Education level turned out to be an insignificant predictor. The type of activity pursued by a SEWA woman member was found to be highly significant predictor (the predictive power was much higher than that of the income effects). The odds ratios were much higher for self- employed home-based or agricultural workers than for nonworkers. The odds were found to be insignificant the salaried workers in the formal sector. · Household size showed an inverse relationship with enrollment, and the odds ratios tended to decline significantly in medium-sized and large households. Income was not found to be a significant predictor. When activity status was not taken into account, women in the top income quintile were twice as likely to enroll as women in the lowest quintile. · There seemed to be an urban bias in enrollment, which may be due to better outreach and accessibility factors. An urban woman was three times more likely to enroll than a rural woman. 9 Table 6: Determinants of Being Enrolled in SEWA Health Insurance Plan among SEWA Women Members Variables Model 1 Model 2 Model 3 Odds Coefficient Significance Odds Coefficient Significance Odds Coefficient Significance Ratio Ratio Ratio Whether chronic ailment 1.155 0.144 0.655 1.164 0.152 0.630 1.121 0.114 0.719 Whether hospitalized 1.602 0.471 0.152 1.716 0.540 0.087 1.528 0.424 0.193 Whether had children in the last 2 1.835 0.607 0.047 1.480 0.392 0.184 1.761 0.566 0.063 years Urban resident 3.096 1.130 0.000 2.720 1.001 0.000 3.131 1.141 0.000 Activity status (non-worker) Agricultural 3.357 1.211 0.011 3.316 1.199 0.012 Casual labor 3.006 1.101 0.001 2.777 1.021 0.002 Home-based worker 4.095 1.410 0.001 4.835 1.576 0.000 Trade/sales worker 2.475 0.906 0.013 2.599 0.955 0.009 Salaried worker-organized 2.257 0.814 0.136 2.328 0.845 0.115 Salaried worker ­ unorganized 2.753 1.013 0.006 2.802 1.030 0.004 Other worker-subsidiary status 2.016 0.701 0.089 1.811 0.594 0.145 Education level (0) 1-4 std. 1.400 0.336 0.292 1.173 0.160 0.600 1.511 0.413 0.190 5-7 std. 0.668 -0.404 0.147 0.715 -0.336 0.208 0.701 -0.355 0.196 8-9 std. 0.540 -0.617 0.122 0.484 -0.725 0.062 0.578 -0.549 0.163 10-12 std. 0.721 -0.328 0.334 0.599 -0.512 0.111 0.805 -0.217 0.512 Graduate and above 1.483 0.394 0.530 1.058 0.056 0.927 1.813 0.595 0.325 Age (16-25 years) 26-35 2.235 0.804 0.008 2.752 1.012 0.000 2.203 0.790 0.009 36-45 5.444 1.694 0.000 6.294 1.840 0.000 5.801 1.758 0.000 46-55 6.729 1.906 0.000 6.746 1.909 0.000 6.878 1.928 0.000 56 and above 4.453 1.494 0.002 3.334 1.204 0.010 4.867 1.582 0.001 Marital status (never married) Currently married 2.089 0.737 0.099 2.251 0.811 0.061 1.939 0.662 0.134 Widow/divorced/separated 1.154 0.143 0.799 1.299 0.262 0.629 1.122 0.116 0.835 Household size (1-4) -0.375 5-6 0.687 -0.907 0.170 0.698 -0.359 0.176 0.792 -0.233 0.370 6-8 0.404 -0.925 0.004 0.455 -0.787 0.009 0.482 -0.729 0.011 9-10 0.397 -1.391 0.015 0.470 -0.755 0.039 0.497 -0.700 0.044 11 and above 0.249 0.005 0.264 -1.333 0.005 0.367 -1.001 0.026 Annual HH income quintile (lowest) 2 0.867 -0.143 0.659 0.885 -0.122 0.699 3 1.182 0.167 0.643 1.202 0.184 0.594 4 1.094 0.090 0.785 1.203 0.185 0.564 5 (top) 1.872 0.627 0.098 2.106 0.745 0.041 Constant -2.929 0.000 -2.490 0.000 -2.936 0.000 Pseudo R2 .218 .185 .212 10 4.2 DETERMINANTS OF FINANCIAL PROTECTION IN COMMUNITY FINANCING Multinomial logit model is used to identify various determinants of utilization of services for ambulatory care and an attempt is also made to explore predictors for choosing a private facility for ambulatory and inpatient care over a public one. This model uses the cases of illnesses reported by all households, irrespective of health insurance status (SEWA, ESIS, Mediclaim and uninsured). Out of the total 1,327 illnesses reported by the sample population during the previous 30 days, treatment was sought for 1,271 ailments (96 percent). The first model uses all illnesses (excluding hospitalization) reported during previous 30 days and predicts the probability of seeking treatment (only for 56 illnesses the treatment was not sought). The second and third model is the subset of the first model (treated in public facility vs. no treatment--383 cases, and treated in private facility vs. treated in public facility--1,271 cases). The third model shows that, of the total treated ambulatory cases, nearly 74 percent relied on the private facility thus suggesting the dominant role of the private sector in handling the ambulatory care burden. The last model is exclusively for hospitalization cases during previous 365 days (i.e., treated in a private hospital vs. public hospital--362 cases). Here the inpatient load was almost equally distributed between the private and public sectors (53 percent of inpatients used public hospitals). Of 11 variables used in the model, 3 depicted health characteristics (whether suffering from any chronic ailment, duration of illness, and type of health insurance coverage), 5 personal characteristics (gender, age, education, marital status, and activity status), 2 household characteristics (household size and income quintile) and 1 community variable (area of residence). The mean value of utilization and the proportion using private health service facilities both for ambulatory and inpatient care varied considerably across these characteristics. The results of multinomial logit models for utilization and private/public choice for ambulatory care are presented in Table 7. The explanatory power of the utilization model (Pseudo R2) was 0.148. For the other two models, it was 0.372 (treated in a public sector facility vs. no treatment) and 0.226 (treated in a private vs. public sector facility). The following are the main findings: · Of 11 variables used for predicting utilization rate, only 3 variables (illness duration, type of health insurance enrollment and area of residence) were found to be significant. None of the personal and household attributes exerted significant influence on utilization rate. The odds of being untreated were higher among those enrolled with community plan (SEWA) as well as among rural residents. · In the case of choosing a private over the public facility including ESIS for ambulatory care, seven variables exerted significant impact. The odds of choosing a public facility were higher if the person had a chronic ailment, a salaried work status, and coverage under social insurance. Males, educated graduates and above, and covered by a private plan (Mediclaim) tended to choose private facility for ambulatory treatment. Patients from small households in urban areas tended to choose a public facility. The income effect for opting out the private facility was clearly discernable. Members of the SEWA plan also tended to choose the private facility for ambulatory care. · For inpatient care, the results of multinomial logit model for choosing a public or private hospital are presented in Table 8. Of the 10 variables used in the model, only 4 (illness duration, type of health insurance enrollment, area of residence and income) showed a significant influence. The odds of choosing a public hospital for inpatient care were much higher for illnesses requiring a longer stay in a hospital. This is entirely due to price considerations because for longer stays out-of-pocket expenditure would be huge if treated in a private hospital. People covered under social insurance tended to use much more public and 11 ESIS hospitals. Only patients from households in the top quintile who could afford treatment chose private hospitals over public for inpatient care. As public hospitals are located mainly in urban centers, urban residents have better access and thus showed greater reliance on public services than did their rural counterparts. Determinants of out-of-pocket expenditures on treatment of ailments, for both ambulatory and inpatient care are presented in Table 9. The dependent variable is expressed as the log of out-of- pocket expenditure on treatment. Overall, the direct cost of treatment for ambulatory care was Rs. 286 per episode. For inpatient care, it was Rs. 2,771. Of the 12 variables used in the OLS regression model, 4 depicted health characteristics (chronic ailment, duration of illness, type of provider (private/public), and type of health insurance coverage), 5 personal characteristics (gender, age, education, marital status, and activity status), 2 household characteristics (household size and income quintile), and 1 community variable (area of residence). The mean value of direct out-of-pocket expenditures per episode varied significantly across these characteristics. The regression results for both ambulatory and inpatient cares are presented in Table 9.. The explanatory power of the model (R2) was 0.284 for ambulatory care and 0.413 for inpatient care. The following are the main findings: · Of the 12 variables used for determining direct out-of-pocket expenditure on ambulatory care, only 7 (type of provider, illness duration, gender, type of health insurance enrollment, household size, income and area of residence) were found to be significant. Of these, the most important explanatory variables were type of provider, duration of illness, social insurance coverage, and area of residence. Cost of treatment turned out to be higher if treated in the private sector and was of long duration when the patient was male and resided in a rural area. The cost of care was inversely related to household size and relatively higher among patients in the third and fourth income quintiles. Only social insurance coverage, and not the community plan, provided financial protection. Both the community plan and the private Mediclaim plan cover hospitalization only. · In the case of out-of-pocket expenditures on inpatient care, only 4 of the 11 variables exerted significant impact. The cost of treatment for inpatient care was higher if in a private hospital and of long duration of treatment and the patient resided in a rural area. Income effects were not found to be significant. In this case, both social insurance and Mediclaim plans succeeded in providing financial protection whereas the community plan did not come up to expectations. Another way of looking at the financial protection is to explore determinants of annual per capita expenditure on health care at the household level (after obtaining the annual estimates of expenditure on ambulatory care, inpatient care, delivery and maternal and child health care). Alternatively, one can also estimate the burden of ill health on the household (annual per capita expenditure on health care as proportion of annual per capita income) and explore how much of this burden is protected through health insurance mechanism. 12 Table 7: Determinants of Being Treated and Use of Public/Private Facility for Ambulatory Care (Multinomial Logit Model) Treated vs. untreated Public contact vs. untreated Private vs. Public contact Significa Significa Odds Significa Predictor Odds Ratio Coefficient nce Odds Ratio Coefficient nce Ratio Coefficient nce Illness duration (1-3 days) 4 ­ 7 0.9975 -0.003 0.995 0.6281 -0.465 0.426 0.8946 -0.111 0.661 8 ­ 14 1.5286 0.424 0.369 1.4671 0.383 0.564 0.8339 -0.182 0.492 15-29 7.6682 2.037 0.057 8.3126 2.118 0.081 0.5712 -0.560 0.065 30+ 5.4784 1.701 0.095 4.6462 1.536 0.311 1.4982 0.404 0.304 Whether Chronic ailment 0.2392 -1.430 0.154 0.5453 -0.606 0.678 0.2108 -1.557 0.000 Whether Male 0.8775 -0.131 0.697 0.7512 -0.286 0.561 1.4326 0.359 0.046 Age (0-14 years) 15-24 2.9836 1.093 0.337 10.0297 2.306 0.076 0.6249 -0.470 0.151 25-34 0.9822 -0.018 0.990 1.2636 0.234 0.855 0.6186 -0.480 0.329 35-44 0.8176 -0.201 0.893 2.1255 0.754 0.617 0.4862 -0.721 0.148 45-54 0.5992 -0.512 0.730 0.3484 -1.054 0.471 0.6338 -0.456 0.376 55+ 0.4542 -0.789 0.588 0.5865 -0.534 0.707 0.4829 -0.728 0.149 Marital Status (Never Married) Currently Married 1.2941 0.258 0.853 0.8231 -0.195 0.883 1.9124 0.648 0.124 Widow/Divorced/Separated 1.3710 0.316 0.828 0.9973 -0.003 0.999 1.7128 0.538 0.286 Education Level (Illiterate) 1-4 Std. 0.9347 -0.068 0.852 0.6011 -0.509 0.307 0.9139 -0.090 0.696 5-7 Std. 1.1325 0.124 0.765 1.4707 0.386 0.487 0.6471 -0.435 0.052 8-9 Std. 1.09E+14 32.323 1.000 2.48E+15 35.448 1.000 0.7109 -0.341 0.252 10-12 Std. 1.8429 0.611 0.388 0.9979 -0.002 0.998 0.8000 -0.223 0.432 Graduate and above 1.0233 0.023 0.984 0.1647 -1.804 0.292 3.2589 1.181 0.036 Activity Status (Non-worker) Agricultural 1.5643 0.447 0.590 0.6132 -0.489 0.699 0.8316 -0.184 0.745 Casual Labor 1.0278 0.027 0.956 3.4850 1.248 0.097 0.6730 -0.396 0.160 Home-Based Worker 0.5168 -0.660 0.426 1.6128 0.478 0.668 0.9354 -0.067 0.894 Trade/Sales Worker 0.7356 -0.307 0.627 2.0847 0.735 0.397 0.5980 -0.514 0.213 Salaried Worker-Organized 1.23E+14 32.441 1.000 5.35E+15 36.217 1.000 0.2761 -1.287 0.000 Salaried Worker-Unorganized 1.34E+14 32.527 1.000 4.54E+15 36.051 1.000 0.5682 -0.565 0.083 Other Worker-Subsidiary status 1.1006 0.096 0.905 7.4579 2.009 0.080 0.2318 -1.462 0.001 Health Insurance Enrollment (Uninsured) Community Plan-SEWA 0.3669 -1.003 0.035 0.0798 -2.529 0.003 2.1073 0.745 0.043 Social Insurance-ESIS 1.1657 0.153 0.671 5.9587 1.785 0.000 0.1715 -1.763 0.000 Private Plan-Mediclaim 9.50E+13 32.185 1.000 8.40E+15 36.666 . 4.4984 1.504 0.050 Urban Resident 2.0618 0.724 0.019 8.1439 2.097 0.000 0.2525 -1.377 0.000 Household size (1-4) 5-6 0.8953 -0.111 0.750 2.0201 0.703 0.166 0.5795 -0.546 0.006 7-8 1.6893 0.524 0.293 7.4421 2.007 0.004 0.4030 -0.909 0.000 9-10 1.1413 0.132 0.832 2.0905 0.737 0.386 1.0896 0.086 0.806 11+ 1.0451 0.044 0.951 4.4154 1.485 0.159 0.5556 -0.588 0.138 Annual Household Income Quintile (Lowest) 2 1.8603 0.621 0.168 0.6607 -0.414 0.513 1.6769 0.517 0.033 3 0.9483 -0.053 0.902 0.1623 -1.818 0.008 2.2582 0.815 0.002 4 1.8409 0.610 0.227 0.1336 -2.013 0.016 3.3027 1.195 0.000 5 (Top) 1.3011 0.263 0.613 0.1134 -2.177 0.007 2.6790 0.985 0.001 Constant 1.915 0.000 -0.550 0.481 3.116 0.000 Pseudo R2 0.148 0.372 0.226 Number of Cases (Dependent valued as):1 1271 327 944 0 56 56 327 Note: Out of 1,327 illness episodes reported during the last 30 days, 56 were not treated. Of the treated episodes, the public including ESIS facility was contacted in 327 cases and in the remaining 944 cases it was the private facility. Figures in brackets refer to the reference category of the variable. 13 Table 8: Determinants of Using Private Facility for Inpatient Care (Multinomial Logit Model) Private vs. Public Hospital Predictor Odds Ratio Coefficient Significance Illness duration (1-3 days) 4 ­ 7 0.4685 -0.7583 0.026 8 ­ 14 0.1998 -1.6105 0.000 15-29 0.1299 -2.0410 0.000 30+ 0.3191 -1.1423 0.027 Whether Male 1.1257 0.1184 0.719 Age (0-14 year) 15-24 1.2172 0.1966 0.725 25-34 3.6314 1.2896 0.069 35-44 1.5422 0.4332 0.591 45-54 3.2025 1.1639 0.132 55+ 1.1200 0.1134 0.877 Marital Status (Never Married) Currently Married 0.5864 -0.5337 0.345 Widow/Divorced/Separated 0.7650 -0.2679 0.730 Education Level (Illiterate) 1-4 Std. 0.9722 -0.0282 0.947 5-7 Std. 0.5129 -0.6678 0.089 8-9 Std. 0.7450 -0.2944 0.575 10-12 Std. 0.6452 -0.4383 0.336 Graduate and above 2.0286 0.7074 0.419 Activity Status (Non-worker) Agricultural 0.5824 -0.5406 0.537 Casual Labor 0.5931 -0.5224 0.259 Home-Based Worker 0.6837 -0.3802 0.622 Trade/Sales Worker 2.0432 0.7145 0.382 Salaried Worker-Organized 1.4176 0.3489 0.553 Salaried Worker-Unorganized 0.4821 -0.7297 0.158 Other Worker-Subsidiary status 0.5435 -0.6097 0.443 Health Insurance Enrollment (Uninsured) Community Plan-SEWA 0.7946 -0.2299 0.630 Social Insurance-ESIS 0.2143 -1.5404 0.000 Private Plan-Mediclaim 5.1690 1.6427 0.171 Urban Resident 0.2855 -1.2536 0.000 Household size (1-4) 5-6 0.6137 -0.4882 0.169 7-8 0.8579 -0.1533 0.722 9-10 0.9079 -0.0966 0.859 11+ 0.5836 -0.5386 0.447 Annual Household Income Quintile (Lowest) 2 1.5507 0.4387 0.315 3 1.4488 0.3707 0.409 4 1.8937 0.6385 0.159 5 (Top) 4.7391 1.5558 0.003 Constant 1.8405 0.001 Pseudo R2 0.228 Number of Cases (Dependent variable coded as):1 171 0 193 14 Table 9: Determinants of Out-of-pocket Expenditure on Treatment by Type of Care (dependent variable in Log form) Ambulatory Care Hospitalization Predictor Coefficient Beta Significance Coefficient Beta Significance Private provider 0.5502 0.3181 0.000 0.4175 0.3230 0.000 Days of Illness 0.0075 0.1735 0.000 0.0111 0.2357 0.000 Whether Chronic ailment 0.0830 0.0483 0.108 Whether Hospitalized Whether Male 0.0982 0.0646 0.025 0.0298 0.0230 0.672 Age 0.0081 0.2212 0.159 0.0098 0.2939 0.266 Age Squared -0.0001 -0.2117 0.095 -0.0001 -0.2696 0.257 Marital Status (Never married) Currently Married -0.0253 -0.0167 0.767 -0.0500 -0.0382 0.657 Widow/Divorced/Separated -0.0528 -0.0207 0.648 0.1315 0.0623 0.421 Education Level (0) 1-4 Std. -0.0297 -0.0148 0.590 -0.0342 -0.0187 0.708 5-7 Std. -0.0328 -0.0168 0.574 0.0115 0.0073 0.893 8-9 Std. 0.0722 0.0283 0.322 0.1183 0.0540 0.287 10-12 Std. 0.0539 0.0253 0.417 0.1245 0.0769 0.182 Graduate and above 0.0508 0.0132 0.637 0.1505 0.0449 0.358 Activity Status (Non-worker) Agricultural -0.0806 -0.0177 0.499 0.1211 0.0275 0.546 Casual Labor -0.0896 -0.0354 0.204 0.1648 0.0808 0.111 Home-Based Worker -0.2041 -0.0389 0.119 0.3035 0.0873 0.050 Trade/Sales Worker -0.0670 -0.0174 0.509 0.1527 0.0439 0.342 Salaried Worker-Organized -0.0179 -0.0066 0.827 0.0641 0.0314 0.588 Salaried Worker-Unorganized 0.0655 0.0216 0.420 0.0633 0.0289 0.574 Other Worker-Subsidiary status 0.2297 0.0490 0.050 -0.2883 -0.0694 0.118 Health Insurance Enrollment (Uninsured) Community Plan-SEWA 0.0145 0.0048 0.855 0.0656 0.0296 0.535 Social Insurance-ESIS -0.3719 -0.2220 0.000 -0.4274 -0.3065 0.000 Private Plan-Mediclaim 0.1461 0.0353 0.177 0.3449 0.0784 0.080 Urban Resident -0.2437 -0.1477 0.000 -0.1786 -0.1290 0.007 Household Size -0.0172 -0.0560 0.049 0.0182 0.0673 0.199 Annual Household Income Quintile (Lowest) 2 0.0666 0.0363 0.265 0.0535 0.0337 0.564 3 0.1132 0.0575 0.074 0.1634 0.0981 0.086 4 0.1446 0.0784 0.024 0.1059 0.0697 0.268 5 (Top) 0.0515 0.0272 0.466 0.0785 0.0452 0.485 Constant 1.7059 0.000 2.5908 0.000 R2 0.284 0.413 Number of Illness Episodes 1274 363 15 16 REFERENCES Ahmad Ehtisham, Jean Dreze, John Hills and Amartya Sen (edited) (1991): Social Security in Developing Countries, Clarendon Press, Oxford. Chatterjee, Mirai and Vyas, Jayshree (1997): Organizing Insurance for Women Workers: The SEWA Experience, Self Employed Women's Association (SEWA), Ahmedabad. Dave, Priti (1991): "Community and Self-Financing in Voluntary Health Programs in India," Health Policy and Planning, Vol. 6 (1) pp. 20-31. Gumber, Anil (1997): "Burden of Disease and Cost of Ill Health in India: Setting Priorities for Health Interventions During the Ninth Plan," Margin, Vol. 29 (2), pp. 133-172. Gumber, Anil (1998): "Facets of Indian Healthcare Market- Some Issues," Saket Industrial Digest, Volume 4, Issue 12, December. Gumber, Anil (2000a): "Health Care Burden on Households in the Informal Sector: Implications for Social Security Assistance," Indain Journal of Labor Economics, Vol. 43 (2), pp.277-291. Gumber, Anil (2000b): "Structure of the Indian Health Care Market: Implications for Health Insurance Sector," WHO Regional Health Forum, Vol. 4 (Nos. 1and2), pp.26-34. Gumber, Anil (2001): "Economic Reforms and the Health Sector: Toward Health Equity in India." Paper presented at the National Seminar on Economic Reforms and Employment in the Indian Economy, Organized by Institute of Applied Manpower Research and Planning Commission, New Delhi, March 22-23 2001. Gumber, Anil and Kulkarni, Veena (2000): Health Insurance for Workers in the Informal Sector: Detail Results from a Pilot Study, National Council of Applied Economic Research, New Delhi. Rahman, K. (2000): "Poverty, Microcredit and Health - What Role Can WHO Play?" WHO Regional Health Forum, Vol. 4 (Nos. 1and2), pp. 68-80. Shariff Abusaleh, Anil Gumber, Ravi Duggal and Moneer Alam (1999): `Health Care Financing and Insurance Perspective for the Ninth Plan (1997-2002)," Margin, Vol. 31 (2), Jan-Mar, pp. 38-68. Visaria, Pravin and Gumber, Anil (1994): Utilization of and Expenditure on Health Care in India, 1986-87, Gujarat Institute of Development Research, Ahmedabad. 17 About this series... This series is produced by the Health, Nutrition, and Population Family (HNP) of the World Bank's Human Development Network. The papers in this series aim to provide a vehicle for publishing preliminary and unpolished results on HNP topics to encourage discussion and debate. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they represent. Citation and the use of material presented in this series should take into account this provisional character. For free copies of papers in this series please contact the individual authors whose name appears on the paper. Enquiries about the series and submissions should be made directly to the Editor in Chief Alexander S. Preker (apreker@worldbank.org) or HNP Advisory Service (healthpop@worldbank.org, tel 202 473-2256, fax 202 522-3234). For more information, see also www.worldbank.org/hnppublications. THE WORLD BANK 1818 H Street, NW Washington, DC USA 20433 Telephone: 202 477 1234 Facsimile: 202 477 6391 Internet: www.worldbank.org E-mail: feedback@worldbank.org