We are proud to announce that Dr Laura Rossouw, SALDRU researcher and FCTC Knowledge Hub on Tobacco Taxation and Illicit Trade manager, has had an article published in the prestigious journal Demography, an official journal of the Population Association of America, which has been in publication since 1964. The article titled, “Poor Health Reporting? Using Anchoring Vignettes to Uncover Health Disparities by Wealth and Race” appeared in the October 2018 edition of Demography.
The article was drawn from a chapter of her thesis. Rossouw says that her thesis examined the health status of vulnerable and low-income individuals in South Africa from a demand perspective.
Usually when people examine health, they look at the healthcare system. They look at how to improve processes in the health system. But Rossouw’s research was focussed on the health system user, and specifically on trying to understand health seeking behaviour. She tried to answer the question, “Why aren’t people accessing healthcare?”
The published article is co-authored by Eddy van Doorslaer and Teresa Bago d’Uva from Erasmus University in Rotterdam. The article examines how levels of wealth and self-reported race affects self-reported health measurements. While most household surveys that are used to measure health outcomes are self-reported and subjective, Rossouw’s paper showed that these self-reported measures are biased in a systematic way. Various sub-groups tend to evaluate their health differently. Poor individuals underestimate their ill-health relative to richer individuals.
Using anchoring vignettes to create a shared benchmark, she compared how different people rate the same health condition. Applying these anchoring vignettes, they find that when a wealthy and a poor respondent experience the same health condition, such as having difficulty walking or climbing stairs, a wealthy respondent would rate their health as being more severely affected compared to a poor respondent. The same is true when looking within race groups. This matters because it means that using self-reported health questions, which are widely available in household surveys, will cause us to underestimate inequalities in health outcomes. However, the opposite is true when looking across wealth quintiles. The study also examined reporting differences across white and black African race groups within wealth quintile 5. Correcting for reporting differences, the inequalities becomes insignificant and smaller.
One possible reason for this is that since individuals use reference groups when they are evaluating their health, a poorer individual may be surrounded by people with lower levels of overall health. Thus, when they are evaluating their health, they tend to think that they are better off than they actually are.
In a different way, poor people underestimate their ill-health compared to rich people. For them it’s about measuring themselves against their reference group.
The findings have various implications for policy and research. If we rely on self-reported health measures in South Africa, then we are to some extent underestimating health inequalities by socio-economic status. South Africa is a country with a history of racial segregation and high levels of socio-economic inequality. The results illustrate the importance of correcting for reporting errors when using self-reported health measures in countries with similarly wide disparities.
The findings may also have important implications for health policy. Poor South Africans’ possible under-perception of poor health and illness may imply an underestimation of their health needs and resulting demand for healthcare. South Africa is currently considering a national health insurance scheme with the goal of obtaining universal health coverage for all, but we are still far from achieving the goal of providing equal treatment to those with equal needs. An underestimation of this may, ultimately, contribute to the persistence of real health inequalities.