Oxfam hasn’t let pesky little things like representative household surveys and impact evaluation results get in the way of their condemnation of Ghana’s health insurance program.
The organization recently issued a strongly worded report described by one web site as “Oxfam damns health insurance scheme.” As these things go, the Oxfam report was then picked up uncritically by The Guardian’s Sarah Boseley.
Oxfam says that the coverage of the Ghanaian national insurance scheme is hugely exaggerated, that it is costly and unfair, that it does not benefit the poor and that it has no impact on financial protection against out of pocket spending on health care. Based on these arguments, the report demands that insurance be eliminated and public monies go back to traditional budgets to finance inputs.
Unfortunately, the data and evaluation that is publicly available from gold standard sources is ignored or dismissed.
A March 2010 quasi-experimental impact evaluation by Joseph Mensah and colleagues finds that women who are health insurance beneficiaries are much more likely to give birth in a hospital, be attended by a trained professional at birth, receive pre-natal care, experience fewer birth complications and fewer infant deaths. Never mind that Mensah is a professor at York University and has a good quality data set in the UNICEF-sponsored 2008 Multi-Indicator Cluster Survey. Never mind that the purpose of an impact evaluation is to illustrate the results of an intervention in comparison to a counterfactual (i.e., no insurance – the state that Oxfam thinks Ghanaians should go). See here.
Although Oxfam reports that only 18% of the population is enrolled in the insurance scheme, the nationally representative 2008 Demographic and Health Survey reports that 39% of women and 30% of men ages 15-49 report that they are enrolled in the insurance scheme. If we count their children, the proportion of the population covered increases substantially. About 90% of both groups report having a card. Unsurprisingly given that formal sector employees are automatically enrolled based on their salary contributions, about half of insurance enrollees are in the top two wealth quintiles while the other half are below the poverty line. Eighty-two percent of beneficiaries report that they were satisfied with the quality of service. While it is not possible to establish causality using the descriptive data from the DHS, the pace of under-5 and neonatal mortality decline has accelerated since 2003 (introduction year of insurance) when compared to the 1998-2003 period.
In terms of financial protection from impoverishing out of pocket spending on health care, findings from a recent study by Nguyen and co-authors finds a significantly positive financial protection effect of health insurance in Ghana. The effect is stronger among the poor than among general population.
Ghana also seems to do better than other countries at similar levels of GDP per capita (see table below).
Based on the data and evaluation available, I would advise the government of Ghana to enroll a greater proportion of its poor into the insurance scheme. I would certainly not advise the dissolution of the insurance scheme.
Every health system—insurance or otherwise—is a balancing act between the health care benefits covered and effectively delivered, the financing available and the size, poverty status and demands of the population. No system is perfect.
If Ghana does not have the fiscal or donor resources to expand to all of the poor in the near-term given the current level of benefits, it should modify the scope of the benefits plan. Certainly, unwarranted administrative costs should be examined. Part of the oil windfall could be saved to smooth spending on insurance from year to year to assure that external shocks do not affect insurance coverage.
But it pains me to see an international NGO judging a developing country government’s efforts to provide health care and financial protection to its population in such pejorative and unscientific terms. It is seriously irresponsible to advise a government to destroy a functioning and apparently well-performing health system.
For further reading on the impact of health insurance in the developing world, see here.
Disclaimer
CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.