Contributor: Roy Goldman, PhD, FSA, MAAA, CERA
To learn more about Roy, click here.
Source: Bigstock |
When designing and pricing a healthcare product, it is not only important to know how to model frequency and severity, but it is essential to understand human behavior. Of course, the issue of anti-selection in medical insurance is well known, as individuals know more about their health and prior expenditures than an insurer does. Anti-selection on behalf of the consumer is rational and to be expected, but one must also anticipate non-rational behavior.
Most individuals do not make rational economic decisions due to cultural, societal, psychological, and emotional reasons. There are many examples that show how individuals choose to avoid possible, but very unlikely, large losses over more likely but smaller gains.1 Even among those who are highly educated, there is innumeracy when it comes to understanding statistical thinking, especially problems involving conditional probabilities.2 As a result, people tend to choose financially inefficient or suboptimal products.
During my career as a health insurance actuary, I’ve come across various situations that illuminate the importance of understanding both rational and irrational behavior when designing and pricing health insurance products.
Reluctance to Switch Carriers
In October 1980, The Prudential Insurance Company of America [the Pru] was awarded the endorsement by the American Association of Retired Persons [now called AARP] to offer the Association’s medical insurance products to its members as well as to members of NRTA, the National Association of Retired Teachers. Their portfolio consisted of more than 20 different indemnity products (e.g., paying $10 or $15 a day while hospitalized or $5 for an office visit and $3 for a lab test) and three reimbursement plans loosely tied to Medicare. The plans generated over $300 million in premium.
The award was considered quite newsworthy for two reasons. First, one of the largest and best-known life insurers was going to team up with one of the largest, most influential, national associations. Second, Pru was replacing Colonial Penn Insurance Company, which had been the sole provider of insurance products to AARP and NRTA members for over 20 years. Indeed, Colonial Penn was founded by a broker, Leonard Davis,3 and others in 1958 specifically to sell medical insurance to NRTA members when no other company thought senior citizens were insurable. Recall that Medicare didn’t start until 1965.
After a few years Colonial Penn thought, “Why limit ourselves to retired teachers?” and founded The American Association of Retired Persons, open to those over 60. The company not only sold insurance to this larger group, but it was deeply involved in recruiting members and managing this new association, which, by 1980, was a force in the American political scene.
While Pru was, arguably, the largest group health carrier in 1980, Pru did not sell individual indemnity coverage, nor did it market to seniors nor by direct mail, which was the only way insurance was sold to AARP members. A new division within the Pru was created to service this new customer, and I, a newly minted Fellow, was asked to transfer there in January, 1981 as the division’s actuary.4 Typically in group insurance, the winning carrier takes over the group’s insured members from the previous carrier. But in this instance, the Pru only won the right to compete with the prior carrier. Between May and July 1981 each policyholder had three opportunities to select either insurer for their current plans. Prudential touted the synergistic union of two large, well known organizations, while Colonial Penn, with their long history with AARP, urged members to stay the course with them.
So, what do you think happened?
The result was that 2/3 of the members chose the Prudential, generating over $200 million in annual premium, but claims were very slow in coming. It turned out Colonial Penn’s 1/3 share generated 2/3 of the claims that year. Individuals who had been getting their claims paid by Colonial Penn in the past were more reluctant to change carriers. Many of the claims Pru received turned out to be for members who enrolled with both carriers. Since there was no coordination of benefits between the two carriers, these members realized they could receive double payments if they were hospitalized.
Prudential, however, did not make a windfall profit, as our winning bid guaranteed AARP at least a 75% loss ratio and an expense ratio below 25%.5 Refunds to AARP based on these guarantees reduced premium rates in future years.
Product Selection in Medicare Supplement
Literally the day I began work at Pru’s AARP office, we were faced with the need to design a new suite of Medicare Supplement plans to meet the new minimum standard as defined by the Baucus Amendment (adopted by Congress in June 1980). From my actuarial studies, I was aware that an insurer can experience anti-selection when offering consumers medical products that differ in obvious ways that permit individuals to take advantage of their known medical conditions. In general, consumers in better health choose a less expensive plan that requires higher copayments or meeting a higher deductible. The claims experience in these plans tend to be better than expected, while the experience in the more expensive plan tends to be worse than expected. As a result, when the plan premiums are re-rated in future years based on experience, the premium differentials get out of whack quickly.
After conducting focus groups, we designed four plans with increasing benefits, but not all plans had the same benefits, thus making it harder to compare plans (e.g., private duty nursing or vision benefits may have been covered in one but not another). Those plans remained pretty much intact until new legislation in 1992 created ten standard plans. But those standard plans built on one another in obvious ways. For example, Plan B differed from Plan A only by adding coverage for the inpatient hospital deductible [IHD]. Of course, Plan B would cost more than Plan A, as some members would incur a hospital stay, but as a result of consumer selection, it is not unusual to find the difference in the annual premium for these two plans to be greater than the current IHD [$1,556 in 2022]. That is, the member is paying more in premium than the deductible he/she would have to pay if hospitalized. Of course, a beneficiary could have more than one benefit period and, thus, be responsible for more than one IHD in a calendar year under Plan A.
Perhaps a clearer example is Plan F vs. Plan F* [High Deductible Plan F]. F* pays all the same benefits as F but only after the insured pays an annual deductible of $2,490 [in 2022] for covered Medicare expenses. But, for the privilege of not having a deductible, a computer search for my zip code reveals the monthly cost of Plan F to be at least $264 greater than for F*. This amounts to $3,168 a year, which is greater than the deductible the insured is trying to avoid.
Why the difference in premium rates? The main reason is that Medicare members who choose F feel it is likely they will incur significant medical expenses, while those in F* believe they are healthy and would rather save on the monthly premium and pay out-of-pocket as needed whatever Medicare doesn’t cover. As a result, relatively few F* insureds reach the deductible. The majority incur zero claims for the insurer, while under Plan F the insurer will be paying some claims for almost everyone.
It could be argued that those who choose F over F* are not only those who know they have health issues, but also include people without sufficient savings to cover an unplanned expense. However, (1) such individuals must still be able to afford at least $264 more in premium a month, and (2) I would argue that given their scanty savings, they are more likely to have put off getting medical advice or annual tests in the past and, therefore, when problems arise now, they are more severe.
Product Selection in Open Enrollment with Multiple Carriers
Selection can drive huge differences in premium even when the benefit differences are relatively small. A great illustration is in the design of the Missouri Consolidated Health Care Plan [MCHCP] circa 2003, which was offered to state and public employees. Healthcare companies were asked to submit premium rates for two plans, call them Low and High. Members could choose either plan from any of the approved carriers. There were only two small differences between the plans: (1) the Low plan had $15 physician co-pay vs. $10 for the High and (2) $5 - 10 higher copay for prescription drugs.
My actuarial team at Mercy Health Plans calculated that if there were only one plan with everyone enrolled in that one plan, the difference in expected monthly benefits would be about $8. But we anticipated consumer selection and knew that a greater premium differential was required. We strived to be the lowest bidder on the Low plan, and the highest bidder on the High plan, and that’s what happened. We ended up with a majority of the Low-plan members, who had fewer claims than expected, while the carriers with more attractive High-plan premiums incurred much greater than the expected number of claims. Within two years, members faced a $100 difference in premium rates between the Low and High plans since High-plan members not only had more physician visits and prescriptions, but also used more inpatient and outpatient services. Eventually, MCHCP had to limit the choice to one plan.
Although this story began in the 1980s, in so many ways the challenges and behaviors are the same in 2023. In Part 2 of this series, I will discuss product selection in employer group plans, non-cost-related product selection, Medicare prescription drug plans, and incentives that can change behavior and make it easier to achieve better health outcomes and lower costs of care, or at least mitigate the typically upward trend in costs.
Contact Roy at: [email protected]
References:
- For example, see Thinking Fast and Slow by Daniel Kahneman or The Undoing Project by Michael Lewis
- A classic example is the Monty Hall problem, based on the TV game show, which fooled many readers of Marilyn vos Savant’s column in Parade Magazine https://en.wikipedia.org/wiki/Monty_Hall_problem
- This is the namesake of the Leonard Davis Institute of Health Economics at the University of Pennsylvania.
- I became a Fellow in May, but I was older than most new Fellows, as I went to graduate school for five years, attained a Ph.D. in mathematics, and taught at the university level for 2 ½ years before beginning an actuarial career. This was my first test as an actuary, as I was joining the team that put together the winning bid without an actuary and saw no need for “an intruder.”
- The VP of the division was extremely expense conscious. Indeed, when he purchased a new car that year, he bought it without a radio thinking he could more cheaply install one himself. He never did!