Due to the growing elderly population, policymakers and researchers have given more attention to the effects of retirement on medical expenditures. As the risk of catastrophic medical expenditures and the decision to retire are themselves intertwined, (as pointed out in Blau and Gilleskie (2006)) it is difficult to estimate the causal impacts of each.
In the US, medical expenditures increase dramatically once the insured are covered under Medicare. The fact that underlying health issues may drive both retirement and health insurance decisions is a concern for both researchers and policy makers. The elderly may choose to retire to ensure that they are eligible for Medicare, as it has better health insurance benefits. Growing concerns about health service over-utilization behavior under more generous health insurance have sparked a great deal of literature about the utilization of health services related to health insurance policies Card, Dobkin, and Maestas (2008). Failing to disentangle the relationship between retirement decisions and health insurance decisions makes it difficult for researchers to determine if people are over utilizing their health services.
This paper uses a credible identification method for the purpose of identifying the causal effect of retirement on health-care utilization in China.
I use micro data from the China Household Finance Survey 2013 (CHFS), which includes information about individuals’ health-insurance premia, overall medical expenses, and those medical expenses covered by health insurance.
There are several compelling aspects to this particular study. First, the ”retirement-consumption puzzle” literature challenges the prediction of those life-cycle models where consumers smooth their consumption over their life cycle. Several empirical studies have found that non-durable expenditures decrease after people retire. However, medical expenses, as one of the major expenses in people’s lives, are rarely studied due to its endogenous impact on retirement decisions. Medical expenses, indeed, greatly affect how individuals make decisions about their consumption and savings (De Nardi, French, and Jones (2010)). For example, people may save their finances in preparation for catastrophic medical expenses in the future. With this relationship between retirement and health expenditures in mind, this paper will attempt to estimate how this relationship impacts the context of China specifically. Though it is unclear how retirement might change health-service utilization, the overall loss of income and potential future income growth that are associated with retirement may indeed lead to lower health service utilization. However, this effect may be offset by other aspects of retirement, including more available free time and better health insurance benefits, leading to overall higher medical utilization.
Second, the exogenous change in the costs and benefits of the Urban Employee Basic Medical Insurance (UEBMI) at retirement as a natural experiment allows me to both identify the change in the relative price of health insurance and investigate their effects on health-seeking behavior. Employees are covered by the same health insurance before and after retirement, however the insured cease paying health insurance premia, enjoying reduced deductibles and lower coinsurance rates after they retire. The discrete change in health insurance costs and benefits incentivizes individuals to utilize health-care differently. As such, this paper contributes to the literature centered on the effect of changes in cost sharing of health insurance (intensive margin) on the utilization of health services; by contrast, the majority of existing studies focus on the effect of health-insurance provision (extensive margin) on the demand for health care. There is a great deal of literature already available (for instance, Card, Dobkin, and Maestas (2008)) that studies the impact of provisions of Medicare.
In addition, it should be noted that an MSA approach (Medical savings accounts) under the UEBMI elevates the role of the individual choice. An MSA account has disposable income for medical expenses, giving individuals full control over how resources are allocated to various health-care services. Also, individuals have the opportunity to determine when to utilize healthcare. Therefore, MSAs allow employees to allocate medical expenses before and after retirement though their accounts.
The main contributions of this paper are as follows: First, the paper estimates the causal effects of retirement on medical expenses; Second, studying the gender differences in health-seeking behavior from an economic perspective is both enlightening and essential. Rather than referring to the strand of literature that studies gender differences in various medical settings, I specifically study such gender differences in response to exogenous change in health insurance benefits at retirement. Examining the patterns of total medical expenditures and medical expenditures covered by the health insurance for males and females respectively, I find that females, upon retirement, allow a discrete increase in both total medical expenditures and those expenditures covered by the health insurance. By contrast, the medical consumption pattern of males do not change upon retirement. Therefore, females delay certain medical expenses and spend more on health-care after retirement. In addition, females are generally more willing to increase their out-of-pocket expenditures in order to both take advantage of better health insurance benefits and utilize more medical care.