Identifying the effects of health insurance coverage on health care use when coverage is misreported and endogenous

Understanding how health insurance changes the way people seek medical care is a central question for both researchers and policymakers. In many countries, including Australia, citizens have access to a universal public health system that covers essential services. On top of this, many people choose to buy private health insurance to access private hospitals, see specialists faster, or get treatments not fully covered by the public system. To understand if this extra coverage actually leads to more medical visits, scientists often ask people directly if they have insurance. However, people are not always accurate when answering these questions; they might forget they have a policy, or they might think a basic plan covers more than it does. This creates a problem: if the data about who has insurance is wrong, the conclusions about how insurance affects health care use could be wrong too. Furthermore, people who are naturally more worried about their health or who have higher incomes might be more likely to buy insurance and also more likely to visit doctors, regardless of the insurance itself. Untangling these factors to see the true effect of insurance is a difficult task that requires careful methods.

A team of researchers in Australia tackled this challenge by looking at a unique set of data that combines survey answers with official government records. They focused on the 2014-2015 period, linking information from a national health survey where people reported their own insurance status with administrative tax records that show exactly who held a private hospital insurance policy. This allowed them to compare what people said they had against what they actually had. They found that while most people were honest, a significant number were not. About 17 percent of those who said they had private insurance in the survey did not actually have it in the government records, while roughly 4 percent of those who said they were uninsured actually were covered. The researchers then used several different statistical approaches to measure how having private insurance changed the likelihood of visiting a doctor, seeing a specialist, or going to a hospital, while accounting for the fact that people who buy insurance might be different from those who do not in ways that are hard to measure.

The study revealed that people with private health insurance do tend to use certain health services more often. Specifically, those with coverage were more likely to visit general practitioners, see specialists, and undergo dental consultations. However, the size of this effect depended heavily on how the researchers measured the insurance and which statistical tools they used. When the researchers relied on the self-reported survey data, the link between insurance and increased visits appeared stronger. When they switched to the precise government records, the effect was still there but often smaller. For example, the researchers found that using the self-reported data, which contained errors, led to a substantial overestimation of how much insurance drove people to see specialists. In their preferred analysis, which combined the most accurate data with methods designed to isolate the true cause-and-effect relationship, they estimated that relying on the flawed survey numbers would make the impact of insurance on specialist visits look nearly 15 percent larger than it actually is.

The researchers also tested whether insurance changed the total number of visits people made, rather than just whether they made any visits at all. In these cases, the results were more mixed. While the methods that controlled for survey errors and other factors showed that insurance increased the chance of seeing a specialist, the methods that tried to count the exact number of visits did not find a clear, consistent increase across all services. This suggests that while insurance might encourage people to seek out specific types of care they might otherwise skip, it does not necessarily lead to a massive surge in the total volume of all medical services. The study highlights that the positive relationship between having private insurance and using health care is real, but it is likely less dramatic than previous studies using only survey data have suggested. By correcting for the mistakes people make when reporting their insurance status and for the hidden reasons people choose to buy insurance, the researchers provided a clearer, more modest picture of how private coverage influences health care demand in Australia.