Who Makes Us Sick: Declining Vaccination Rates in High-Income Communities

Even though effective vaccines are available for many diseases, outbreaks continue to occur even in the most developed countries. This is mainly because vaccination rates are not high enough to achieve herd immunity. Traditionally, low vaccination rates are thought to be caused by low-income people who have limited access to medical facilities. However, there is recent growing recognition that some high-income groups also deliberately choose not to vaccinate their children. This is worrisome because vaccination rates may decline even further if more people follow the same behavior as they become wealthier. These conditions raise two related questions: 1) Is there any systematic pattern for high-income groups in choosing to not vaccinate? 2) If so, what is the mechanism behind such behavior?

This paper first empirically examines the relationship between income and vaccination rates using two datasets with different levels of aggregation. The country-level dataset is from the World Health Organization (60 countries, 1980-2013), while the individual-level datasets is from the United States (>20,000 individuals, 2000-2013). A novel feature of my analysis is that I can separate out any confounding effects arising from cultural and institutional differences across regions using the fixed-effect estimation. This analysis demonstrates more precisely the relationship between income and vaccination rates than previous studies which have not used longitudinal datasets necessary in the fixed-effect estimation.

Empirical results show that, in the high income range, vaccination rates (or the probability of vaccination) in fact decrease as income rises, both at the individual and country level. That is, high-income groups are systematically less likely to vaccinate. There can be many non-income reasons why some high-income people may avoid vaccination (e.g., education level), but it is difficult to argue how these non-income reasons can generate such a systematic pattern. Instead, I propose a hypothesis that is directly related to income: High-income people choose not to vaccinate because they can afford all the medical treatments they need in the event of an infection.

To examine this hypothesis, I build a simple model that describes people's vaccination decision. The model exhibits two key features suggested by the literature. First, it incorporates both vaccination and medical treatment, as they are the two major and non-independent instruments to deal with infectious diseases. Second, it incorporates the fact that the vaccination decision of others affects a person's vaccination choice because it changes the probability that he/she will contract the disease. Hence, the model depicts how vaccination and treatment decisions are made in a society where people adjust their behaviors considering other people's choices.

The model shows that if a person has high enough income and is not too risk averse (relative risk aversion < 1), he/she will choose not to vaccinate and rely solely on medical treatment. Given such behavior, vaccination rates in the society also decline as the average income increases above a certain threshold level. The model generates a number of predictions about vaccination rates depending on the characteristics of vaccines and diseases, which seem to be consistent with the empirical data.