Health insurance in South Africa: an empirical analysis of trends in risk-pooling and efficiency following deregulation

Citation
N. Soderlund et B. Hansl, Health insurance in South Africa: an empirical analysis of trends in risk-pooling and efficiency following deregulation, HEAL POL PL, 15(4), 2000, pp. 378-385
Citations number
18
Categorie Soggetti
Public Health & Health Care Science
Journal title
HEALTH POLICY AND PLANNING
ISSN journal
02681080 → ACNP
Volume
15
Issue
4
Year of publication
2000
Pages
378 - 385
Database
ISI
SICI code
0268-1080(200012)15:4<378:HIISAA>2.0.ZU;2-W
Abstract
This paper reports an empirical investigation into the pattern of private h ealth insurance coverage in South Africa before and after deregulation of t he health insurance industry. More specifically, we sought to measure trend s in risk-pooling over the period 1985-95, and to assess the impact of risk pooling on the costs of health insurance cover over this period. South African mutual health insurers (Medical Schemes) have existed for ove r 100 years, and have been regulated under a specific Act since 1967. Up un til 1989, health insurers were required by law to community rate their prem iums, and were not allowed to exclude high-risk enrolees from cover. In 198 9 these regulations were removed, effectively allowing health insurers to r isk-rate the cover which they provided, and exclude 'medically uninsurables '. Data were obtained from the office of the health insurance regulator (th e Registrar of Medical Schemes) for the period 1985-95, and consisted of th e statutory returns from all registered medical schemes for each year durin g the study period. Multiple regression methods were used to assess the det erminants of changes in the risk pools of insurers, and their costs. Both c ross-sectional and longitudinal models were estimated. Unadjusted data suggest changes in risk-pooling since the deregulation peri od after 1985, Health insurers with open enrolment had worse than average r isk profiles in the 1980s, but this reversed by the early 1990s, leaving th em with significantly better risk profiles by 1995. Worsening risk profiles were associated with decreasing fund size, higher loss-ratios and past pre mium increases. Most models showed that risk rating of premiums was consistently associated with higher premiums, after adjustment for risk, quality, scale and other environmental differences between insurers. Likely explanations include the additional costs required for marketing and underwriting risk-rated polici es, insufficient incentives to use cost-control techniques, and higher leve ls of moral hazard associated with diminished risk-pooling. Current re-regu lation of risk-pooling within medical schemes may thus improve both equity and efficiency of private health care cover.