While most people in developed countries have health insurance, insurance coverage is much less widespread in the developing world. Yet people in rich and poor countries alike benefit from the protection insurance provides against the economic risk posed by a serious illness. Expanding health insurance is a key priority for policymakers in much of the developing world.
The environment for health insurance in developing countries differs from that in developed countries in several ways. Standard economic theory suggests that insurance leads people to use too much health care because they don’t face the full price of care. But in developing countries, certain preventative services should be subsidized because they help stop the spread of infectious disease or may be under-used because of credit constraints. Also, it may be particularly important to design payment schemes that encourage health care providers to provide efficient levels of preventive health care.
In “High-Powered Incentives in Developing Country Health Insurance: Evidence from Colombia’s Regimen Subsidiado” (NBER Working Paper 15456), researchers Grant Miller, Diana Pinto, and Marcos Vera-Hernandez study the first major developing country effort to expand insurance in a way that addresses these unique features of the market.
Introduced in 1993, Colombia’s Regimen Subsidiado (SR) is a publicly-financed health insurance program for the poor. The SR was designed to be a “managed competition” insurance scheme – beneficiaries receive full public subsidies to purchase insurance from one of multiple health insurance plans. On the demand side, co-insurance rates for curative services are lower for SR beneficiaries and they receive free preventive services (as non-beneficiaries do).
On the supply side, primary care providers are paid a fixed amount per enrollee (“capitation”), a system that encourages them to promote preventative care as a means of reducing total primary care expenditures. For specialty care, providers are reimbursed on a fee-for-service basis, but insurers can deny coverage (“utilization review”), allowing them to limit wasteful care.
The authors use a regression discontinuity design to study the effects of this new program on health spending, the use of specific health services, and health outcomes. Eligibility is based on a household poverty index that incorporates items such as educational attainment, housing material, and ownership of durable goods – households below a threshold level on the index are eligible for the SR, while households above are not. In the absence of this program, these outcomes should increase in a smooth, continuous manner with the poverty index. Sharp, discontinuous changes in outcomes at the eligibility threshold would therefore be compelling evidence that they are linked to the SR.
Turning to the results, the authors first examine the effect of insurance coverage on out-of-pocket health spending. They find that SR eligibility is associated with reductions in out-of-pocket spending, especially the very high levels of spending households might experience when battling a serious illness. This suggests that the SR provides important risk protection.
Next, the authors examine the effect of the SR on the use of preventative health services and on health outcomes related to the use of those services. They find that the SR is associated with substantial increases in the use of preventative services – for example, there is a 50 percent increase in the probability that an adult has had a preventative physician visit in the past year, while children have about twice as many growth-monitoring and well-care visits. This increase is associated with measurable health improvement as well – children have 1.3 fewer days absent from usual activities due to illness in the past month and a 62 percent decrease in the incidence of cough, fever, or diarrhea in the past two weeks. In Colombia, preventative services are generally free for people regardless of their insurance status, so these changes in utilization and outcomes presumably reflect the effect of the SR’s supply-side incentives.
Given the combination of demand- and supply-side incentives under the SR, theoretical predictions about changes in curative care are mixed. Empirically, SR enrollment is associated with an increase in the probability that adults have seen a doctor because of health problems in the last month, but there is no such relationship for children. The authors do not find distortions in private health behaviors (handwashing, breastfeeding, or maternal investments in fetal health, for example) among enrollees, nor do they find that SR enrollment “crowds out” other forms of health insurance.
In short, the authors find that the SR program has successfully protected poor Colombians against financial risk, while also increasing the use of key preventative services that provide important health benefits to patients and may also confer benefits on society at large. The increased utilization of preventative health services can reasonably be attributed to changes in provider incentives.
The authors suggest that “strengthening supply-side incentives for the provision of key preventative services … maybe a potent alternative (or perhaps an effective complement) to common demand-side approaches embodied in conditional cash transfer programs.” They conclude by observing that “the full welfare-improving potential of high-powered incentives in health insurance has yet to be fully realized,” as certain political concessions in the creation of the SR “have presumably limited the ability of health plans to pay providers in ways that encourage better quality and lower cost services.”