Voucher Sites Show More Equitable Distribution of RH Service Utilization
This post was authored by Karampreet K. Sachathep, a Population Council intern and a Ph.D. candidate at the Johns Hopkins Bloomberg School of Public Health.
Since the 1970s concentration curves have increasingly been used to visually examine inequality in health outcomes and health service utilization (Kakwani, Wagstaff and van Doorslaer 1997). They provide a snapshot of how, in this case, utilization of family and maternity services varies across a distribution of individuals who are ranked from poorest to richest. The greater the distance between the concentration curve and the line of equality (the diagonal line that runs through the graph), the more concentrated the number of facility-based deliveries or use of LAPMs (Long-Acting and Permanent Method of family planning) among the richer individuals.
These concentration curves below are calculated from 2010 baseline survey of the Population Council’s evaluation of the Government of Kenya voucher program, four years after the voucher program was introduced in 2006. The main objective behind vouchers is to target services to those who are poor and equalize access to the healthcare system. Thus, we would expect that people in areas exposed to the OBA (Output-Based Aid) program would have a more equitable distribution or even a more ‘pro-poor’ (graphically this would translate to above or at the line of equality), distribution of health service use relative to the non-OBA sites.
Figure 1: The two concentration curves here show that the degree of inequality in OBA areas is lower than non-OBA areas for both facility-based deliveries and use of LAPMs (Long-Acting and Permanent Method of family planning).
In Figure 1, the CC lies below the line of equality in the OBA and non-OBA areas for use of facility-based deliveries. On the other hand, we notice that the concentration curves for LAPM use are on opposite sides of the line of equality for OBA and non-OBA sites, and that for OBA sites, LAPM use is ‘pro-poor’ and lies above the line of equality. In OBA sites, more poor women are using long term FP methods than non-poor women while the opposite is true for the non-OBA sites. Concentration curves provide a visual sense of the distribution of inequality; these graphs show us that the degree of health service utilization in OBA sites seems to be more equitable than in non-OBA sites.
However, in order to make a better conclusion as to whether these differences are truly significant, a numerical measure of health inequality can be used. A related measure, the concentration index (CI), quantifies the amount of inequality in a health variable (Kakwani, Wagstaff and van Doorslaer 1997). It’s defined as twice the area between the concentration curve and the line of equality. CIs range between -1 and 1 and if there is no inequality, the index is equal to 0 (it may help to think of this as a correlation coefficient).
To take this a step further, the concentration index (CI) is positive when the concentration curve lies below the line of equality (e.g. indicating that poor have lower healthcare use or worse health outcomes).
Figure 2: Comparing concentration indices between OBA and non-OBA sites.
Figure 2 quantifies the information that we saw in Figure 1– that in 2010, LAPM use in the OBA sites is concentrated among the poorer populations relative to the non-OBA sites (hence the negative CI value). Facility-based deliveries seem to also be more equitably distributed in the OBA-sites relative to the non-OBA sites; however, both sites show ‘pro-rich’ utilization of this service.
The concentration index for the inequality of distribution in facility-based deliveries since the inception of the program in mid-2006, was 0.24 in the OBA sites and 0.13 in the non-OBA sites. LAPM use was at -0.07 in the OBA sites and 0.03 in the non-OBA sites. These differences, however, were found to be statistically non-significant.
Finally, in order to provide us with a sense of whether this program has been equity enhancing, we will compare these curves to those generated from the endline surveys (collected August 2012), to observe whether the distribution of health utilization has changed over two years in voucher-exposed and non-exposed sites. For more information on concentration curves, indices, and equity analysis in health, please refer to the World Bank guidance document on Analyzing Health Equity Using Household Survey Data.
REFERENCES
1. Kakwani, N. C., A. Wagstaff, and E. van Doorslaer. 1997. “Socioeconomic Inequalities in Health: Measurement, Computation and Statistical Inference.” Journal of Econometrics 77(1): 87–104.
2. Wagstaff, A., and N. Watanabe. 2003. “What Difference Does the Choice of SES Make in Health Inequality Measurement?” Health Economics 12(10): 885–90.
(Image credit: All graphs produced by Karampreet K. Sachathep, © 2012)