Achieving Equity in Health Systems. Implications for developing countries of recent evidence from Asia

Transcription

1 Achieving Equity in Health Systems Implications for developing countries of recent evidence from Asia Ravi P. Rannan-Eliya IHEA World Congress Copenhagen, 11 July 2007

2 Equitap Project Phase 1 - Collaborative study in 15 Asian countries, looking at equity in: Payments including taxes (progressivity) Government spending (BIA) Catastrophic expenditures Impoverishing impacts Health inequalities (MCH indicators) Health care use Health financing mix Phase 2 - In-depth analysis of determinants of health system performance 1

3 Poorest quintile share of total health care use 2

4 Catastrophic impact 3

5 Conventional wisdom Subsidies on government-provided, free health services captured by rich Need to target to reach the poor Better to emphasize pro-poor preventive services to reach the poor Conventional civil-service modes of delivery lack incentives for efficiency and serving poor Indirect taxation regressive, so redistributive arguments weak Social insurance can work better than tax-financing in lower-income settings Countries should reform to social insurance from taxfinancing with development 4

6 Findings of Comparative Analyses Performance correlated across equity dimensions Health outcomes, risk protection, targeting Indirect taxation not regressive in LDCs unlike in Europe Tax funded systems The best targeted systems in Asia are tax-funded with integrated provision (Hong Kong, Malaysia, Sri Lanka, Maldives) Characterized by: Universalistic approach - no means testing, no explicit targeting Concentration of spending on hospitals/inpatient care Hospital spending not pro-rich Social insurance systems Generally only reach poor, if universal in nature Not attainable in poorest countries (exception Mongolia?) Equity requires substantial tax financing contribution to pay premiums for unemployed, informal sector, etc - Social Insurance is no substitute for taxation capacity Equity worse if schemes are not integrated 5

7 Performance of health systems Universalistic, tax-funded systems No/minimal user fees, no explicit targeting/voluntary selfselection by rich of private sector, emphasis in spending towards hospitals/inpatient care, high density of supply. Non-universalistic, tax-funded systems User fees, means testing, diverse ineffective experimentation in reaching the poor projects, emphasis in spending towards nonhospital care, low density of supply. Sri Lanka Malaysia Maldives Hong Kong Bangladesh Indonesia India Nepal National health insurance systems Universal social health insurance, large tax-subsidy for insurance, emphasis in spending towards hospitals/inpatient care Transition systems Restricted social health insurance, minimal tax-subsidy for insurance, user charges major mechanism of financing Japan Korea Taiwan Mongolia Thailand China Viet Nam 6

8 Defining Tax-funded Systems Country Tax as % of public funding Tax as % TEH Social insurance as % TEH TEH as % GDP Hong Kong SAR Sri Lanka Bangladesh Nepal Malaysia India Indonesia * General revenue funding >90% of public financing * Social insurance < 5% of TEH 7

9 Performance of tax systems Country Catastrophic impact Poverty impact Targeting of government spending Health outcomes Nepal Large Large Pro-rich Poor Bangladesh Large Large Pro-rich Poor India (Punjab) Large Large Pro-rich Poor Indonesia Modest Modest Pro-rich Poor Sri Lanka Negligible Negligible Pro-poor Good Malaysia Negligible Negligible Pro-poor Good Hong Kong SAR Negligible Negligible V. pro-poor Good 8

10 Mechanism Two distinct groups of tax-systems according to performance Gradients in use of public & private provision Private provision pro-rich in good + bad performers Public provision pro-rich in bad, pro-poor in good performers Targeting of government spending Good performers - not explicit or direct Consistent with Besley-Coate Model Under budget constraint, public services can be universallyprovided; if richer individuals opt for private care, targeting will be pro-poor 9

11 Use of public-private inpatient services by income quintile 10

12 Tentative Explanations High levels of public supply and population use of medical services High budget allocations for hospital services and inpatient care Technical efficiency gains Governance 11

13 Public supply: High levels of public hospital provision 12

14 Sickness behavior: High rates of use of medical care 13

15 Budgeting: Focus on hospitals 14

16 Technical efficiency gains: Scaling-up in Sri Lanka Year GDP (US$ 1995 per capita) IMR Health spending (US$ 1995 per capita) Outputs (Outpatients per capita) Outputs (Inpatients per capita) yrs +9% -38% + 25% +110% +55% Contribution of increased spending = <25% Contribution of technical efficiency gain = >75% 15

17 Policy messages Need to take seriously and understand goodperforming tax-funded systems Failure to invest in learning how conventional, civil-servicerun healthcare delivery systems can be made to work Indirect targeting with parallel private provision more effective than direct targeting Theoretically and in practice more effective in situations of poor information Requires change of perspective and agendas High levels of public supply with limited budgets requires attention to technical efficiency Failure to invest in learning how to improve technical efficiency 16