How do people respond to a foreseeable increase in cost-sharing?

The generosity of a health insurance plan, meaning the extent of its coverage, involves a fundamental trade-off. The more generous the coverage, the better it is for insured individuals. However, broader coverage also leads to disproportionately higher premiums. This is because insured individuals have an incentive to use more services than they would if they had to pay the full cost themselves. To counteract this incentive, most insurance contracts include some form of cost-sharing.

In practice, cost-sharing is often structured on an annual basis. This means that at the beginning of the year, patients typically bear a larger share of treatment costs than at the end of the year. In other words, the cost to the patient is often lower in December than in January. This pricing dynamic could encourage patients to schedule elective treatments in December rather than January, potentially undermining the cost-reducing effect of cost-sharing.

There is now evidence suggesting that patients do shift treatments to minimize out-of-pocket costs. However, it remains unclear whether such price variations over time influence only the timing of treatments or also overall demand for healthcare services. We aim to answer the question of whether temporal variations in cost-sharing dilute its intended impact.

To study how future changes in cost-sharing impact behavior, we need two groups. Members of both groups should face the same cost-sharing structure today but different structures in the future. Additionally, group assignment must be random; otherwise, observed demand could be influenced by the members' group selection. The Swiss healthcare system provides an ideal setting for this analysis with individuals transitioning to "adult" status.

For adults aged 18 and older, the standard deductible is CHF 300, while for children, it is zero. Similarly, the annual maximum for co-payments increases from CHF 350 to CHF 700 upon reaching adulthood. As a result, cost-sharing increases when individuals become adults. However, because health insurance operates on an annual basis, individuals do not transition to "adult" status on their birthday but at the start of the next calendar year. For example, in 2024, insured individuals born on December 31, 2005, are considered "adults," while those born one day later, on January 1, 2006, are still considered "children."

Since birthdates around the new year are essentially random, this creates an ideal setup. In 2023, both groups face the same cost-sharing structure, but in 2024, their structures diverge. This allows us to analyze how anticipated changes in future prices influence current demand.

The results show an increase in healthcare spending among men who anticipate bearing higher costs in the following year. On average, their expenditures on outpatient treatments rise by about CHF 40, or approximately 5%. This effect is particularly pronounced among men who had relatively low utilization of healthcare services before their 18th birthday. This group has a strong incentive to bring forward their medical visits in the year before the higher deductible and co-payment take effect.

Interestingly, this behavioral adjustment primarily occurs with specialist visits. The likelihood of a man visiting a specialist in the year before the cost increase rises by 7%. This suggests that men are prioritizing medical treatments they would otherwise postpone to save on costs. The effect is less significant for general practitioner visits, indicating that specialist appointments are more likely to involve planned, non-urgent treatments.

Surprisingly, women exhibit significantly less anticipatory behavior. While men show a notable increase in healthcare spending, women's expenditures remain nearly unchanged on average. The reasons for these differences are varied.

One possible explanation is that women may plan their healthcare spending differently and respond less strongly to financial incentives. Additionally, women tend to use regular healthcare services more frequently on average, which could leave them with less flexibility to adjust their spending in advance. This consistent utilization pattern might limit their ability to reschedule or pre-plan their medical expenditures in response to upcoming cost increases.

The study shows that young people, particularly men, adjust their healthcare behavior when they know their cost-sharing will increase in the following year. These findings are relevant for both health insurers and healthcare policymakers. They highlight that anticipatory behavior can influence the demand for healthcare services and suggest that the design of health insurance plans should account for these behavioral patterns.