Eve is here. The findings from the following well-designed multi-country study are both counterintuitive and highly significant. Sick populations show less support for more generous healthcare provision than healthy populations. Is it because the degree of deterioration in health conditions raises concerns that an adequate level of services will result in higher overall costs for society, or is it a particular manifestation of a general trend in which communities that perceive themselves to have some surplus are more amenable to funding social safety nets?
The authors also highlight the loop of doom that appears to be starting in the United States. In other words, lower well-being leads to less support for widespread health programs, which in turn leads to further declines in population-level fitness.
Written by Marcello Antonini, Visiting Researcher, Health Policy Department, London School of Economics and Political Science. Joan Costa y Font, Research Fellow in the Center for Primary Care and Health Services Research, School of Economics, University of Oxford, and Professor of Health Economics, London School of Economics and Political Science. Originally published on VoxEU
Understanding how the public interprets equity is increasingly important because people’s quality of life is directly affected by policy decisions regarding access to and funding of health care. This column examines how a population’s health status, as represented by exposure to BCG vaccination, shapes people’s attitudes toward equal access to care and willingness to pay higher taxes. Healthier individuals are more supportive of fair health financing and equal access to care, suggesting that improving population health can strengthen public support for a more equitable health system.
Equity in health care does not simply mean giving everyone the same resources. In reality, what people consider “fair” is shaped more by shared social norms and ideas of justice than by strict equality (Olsen 2011, Starmans et al. 2017). Most people accept inequalities if they think the reasons behind them are just, but they reject inequalities that they see as arbitrary or unjust. Health care focuses on these equity concerns because decisions about access to and funding of health care directly impact quality of life and, in many cases, survival. As health systems face rising costs, understanding how the public interprets equity is increasingly important for policy design.
Is self-interest healthy or unhealthy?
An important open question is how personal circumstances, particularly one’s own health status, shape attitudes toward equitable access to care and equitable financing of health systems. Evidence from COVID-19 reveals that exposure to the pandemic increased aversion to inequality, especially among those not directly affected by the pandemic (Costa-Font et al. 2021), and influenced the prosocial behavior of vaccination (Voth et al. 2021).
Figure 1 shows a directed acyclic graph visualizing our identification strategy, illustrating the causal paths from BCG vaccination to health status to funding equity preferences and the potential confounders that our methodology addresses. Consider two competing hypotheses.
The “healthy self-interest” hypothesis suggests that people in poor health may benefit more directly from redistributive policies and therefore favor redistributive policies, whereas healthy people may prefer approaches based on personal responsibility. The “unhealthy egoism” hypothesis proposes the opposite. This means that people in poor health may be more focused on immediate personal needs and therefore less concerned with system-wide equity, whereas people in good health may be more capable of upholding fairness norms.
Figure 1 Directed acyclic graph of BCG vaccination instrumental variable strategy
Attitudes towards health and health system equity
To test the impact of equity in health and health systems, Antonini and Costa-Font (2025) examine evidence from more than 70,000 respondents in 22 countries, linking self-reported health status with attitudes toward equal access and willingness to pay higher taxes for better public health care.
Figure 2 shows a positive correlation between exposure to BCG vaccination and both aspects of funding equity preferences. This positive association is particularly pronounced for attitudes towards fair access, which also shows a higher mean score compared to willingness to support fair lending.
Figure 2. At the country level, the association between exposure to BCG vaccination and (1) attitudes toward equitable access and (2) willingness to support equitable financing.
Note: The figure shows a scatterplot of this measure (exposure to BCG vaccination) against two outcomes of the analysis: attitudes toward fair access at the country level (panel 1) and willingness to support fair financing (panel 2). The size of the circles reflects the standard deviation of the two outcome variables measured in each country. To create this scatterplot, we first calculated the average concern for both outcomes for each country. We then plot the mean value of the variable of concern on the Y-axis and the proportion of the population vaccinated with BCG on the X-axis.
Graphic inspection suggests that our instrument may be a valid candidate for analysis. Our causal estimates are obtained from a novel instrumental variable strategy based on variation in BCG vaccination exposure, allowing us to estimate the causal effect of health status on equity preferences. This approach helps address the long-standing challenge of separating health effects from other social or economic factors.
This result clearly supports the “unhealthy egoism” hypothesis. Those in poorer health had lower support for fair health financing, while those in better health showed stronger support for both equal access and more redistributive financing. A one-point improvement in self-reported health was associated with an 11% increase in support for fair access and an 8% increase in support for fair financing. These findings suggest that improving people’s health status not only increases well-being, but may also strengthen public support for a more equitable health system.
Our findings suggest a stronger influence on normative judgments compared to behavioral intentions that involve personal costs, and suggest that health status has a more pronounced influence when it does not require personal financial sacrifice. This pattern shows that while healthy people express greater support for the principle of fairness, this support declines slightly when it involves an actual financial contribution.
The mechanisms driving these effects operate primarily through economic channels (income and employment), with health trust and political attitudes playing a contributing role. This helps explain why healthy people with generally good economic outcomes prefer equity in health financing more.
These findings also help explain why more unequal societies often experience worse health outcomes, including lower life expectancy, higher obesity rates, increased drug abuse, and worse mental health. When people experience poor health, support for equity in health care access and financing weakens, and policies aimed at improving equity are less likely to gain traction. Remarkably, this pattern is found across very different health systems, suggesting that the link between individual health and support for funding equity is universal, even if its strength varies by institutional context.
The causal relationship between improved health outcomes and increased support for fair health financing points to a potential virtuous cycle. If health systems succeed in improving people’s health, public support for equitable policies may naturally increase. This dynamic may help explain why health inequalities follow a Kuznets-shaped curve (Costa-Font et al. 2018). It also confirms the need to invest in better health as a foundation for long-term stocks. As overall health increases, the system may be able to apply equity-oriented allocation tools (such as equity weighting) to prioritize historically overlooked groups whose needs have become more apparent.
See original post for reference
