The cost of benefits, especially health plans, has grown at extreme rates over the last decade, leading to changes in the way employer-sponsored benefits are structured. Employers continue to provide benefits to their workforces, however the employee is responsible for a growing portion of the various costs associated with benefits. As a result, employees want the ability to choose the benefits that are a good fit based on their coverage needs and financial situation. The key question is: How will employees navigate more choices?
More choice is inherently good, right?
As employees and consumers navigate this transition, additional choice is certainly valuable. But, at what point do the paradox of choice and diminishing returns become a major issue? Two oft-cited studies from the benefits domain provide insight into choice readiness.
In preparation for the public exchanges, a group of researchers from leading universities conducted a study to see how well individuals could navigate available options. Participants were given eight options with no help, and they chose correctly about 22 percent of the time. 22 percent is an interesting statistic, as it is nearly equivalent to chance. At some level, perhaps no one made a correct choice—a few people just got lucky.
In 2005, Medicare beneficiaries were given their first opportunity to choose a Medicare Part D option. Medicare D introduced a great deal of choice specifically designed to help beneficiaries find a plan that met their unique needs. But could they really make good choices? To test how well beneficiaries might decide, Nobel Laureate Daniel McFadden created an experiment. McFadden’s team significantly trimmed complexity from the decision-making process. They removed pharmacy networks and prior authorization considerations and provided just four options. In this simplified environment, incorrect choices were made about two thirds of the time.
We struggle because we are human
There is clearly a challenge associated with choosing benefits. If we attended benefits training and health literacy courses could our choices improve? It would help, but perhaps there is something deeper that makes these choices really challenging—something that is very human. Advances in the fields of behavioral economics, psychology, and neuroscience have given us a more evolved view into how the mind navigates choices. Predictable biases have emerged that wreak havoc when choosing benefits:
Optimism bias – As a species, we are highly optimistic. A significant percent of the population underestimates its risk for things such as losing a job, getting in a car accident or getting a terminal disease. Optimism may be great for the evolution of our species, but it leads to the planning fallacy—a major impediment when choosing a product that provides financial protection against the risk of future events or diseases.
Availability bias – We have a natural tendency to overweight our risks for events that are more “available” or emotionally rich in our psyche. That drives the affect heuristic, which causes miscalculations in costs and benefits of given scenarios. Events such as tornados are perceived to cause far more deaths than diabetes because of the sensational nature of the event. The destruction of towns viewed via media coverage, the uncertainty and speed at which a tornado strikes causes one to overweight their likelihood compared to something more likely to occur such as diabetes.
Present bias – Harvard psychologist Jason Mitchell and his co-researchers conducted a study where participants wearing neuroimaging machines were asked how much they would enjoy engaging in an activity the next day or in a year. They were also asked about how a stranger would enjoy the activity for the same periods. Researchers were monitoring the ventromedial prefrontal cortex (vMPFC), the region of the brain that is most actively engaged when people are thinking of themselves. For a significant number of participants the vMPFC was not actively engaged when they imagined the activity in the future. Meaning that in effect, when people attempted to think of themselves experiencing an activity in the future, it was interpreted as if their future self were a stranger. This becomes a major impediment when trying to imagine oneself experiencing undesirable events such as an accident or being diagnosed with a life-changing disease.
Each of these biases makes it extremely difficult to properly determine our potential risk—a critical step in the calculation to select the right benefits. Even without natural human bias, balancing the trade-offs between deductibles, premiums, cost shares, preferred doctors and personal finances is exceedingly complex. The Consumers Union summarized the situation poignantly – “These difficulties are so profound that the vast majority of consumers are essentially being asked to buy a very expensive product—critical to their health—while blindfolded.”
There are interventions through choice architecture that can begin to improve choice efficacy. Choice architecture is recognition that in the context of any choice there is an environment or “architecture” that surrounds the choice, and small nuances in that architecture can have dramatic effects. One key tenet of choice architecture that must be noted is that if small nuances in the environment can have dramatic effects, then there is no neutral design to the architecture. If you are responsible for designing the environment in which others will choose, you become a choice architect, and every decision you make could have significant effects. As a result, there is a great deal of responsibility that comes with being a choice architect.
With such responsibility, what steps must be taken to help drive better choices? First, you have to understand how the environment affects choice. To do this, takes observation and tracking. If the environment is a technology application, then you have a huge advantage, as there is a vast set of tools available to provide visibility into details of how people interact with technology.
Once you have established a model for tracking activity, then you can begin to introduce interventions. There are four key methods that can have significant impact when used properly:
Defaults are one of the most powerful tools of the choice architect. As humans we typically miscalculate gains and losses because we interpret a loss as being more negative than a gain of the same item. This overweighting of loss is what drives the status quo bias and what makes defaults so effective. The classic use case is 401k auto-enrollment. Opt-out plans have significantly higher adoption than opt-in.Consider how defaults are used in benefit auto-renewal strategies and passive open enrollments. While passive open enrollment is widely used there are opportunities for defaults during any enrollment event. Because of the stickiness of a default a choice architect must be very thoughtful when choosing the default strategy.
Nudging is a slight push toward a given choice. A basic nudge can be as simple as the initial sort order of a set of options on a site. Each item in the sort order is naturally weighted by the mind simply by being first, second or third. A very powerful tool is a smart nudge. A smart nudge uses context and personalization to calculate a much more specific “recommended” choice for an individual.Consider the scenario where an individual enters a life event, such as a birth, into his or her benefits management application. The software then recommends Life insurance, as well as information about college savings accounts and assistance in finding a lawyer who can help with a living will. Smart nudges are powerful tools, and similar to defaults, must be determined with extreme care because of their weight of influence.
Anchoring – in his book Thinking Fast and Slow Nobel Laureate Daniel Kahneman discusses an experiment where participants were asked to take the last three digits of their phone number and add 200. They were then asked when they believed Attila the Hun sacked Europe. On average there were 300 years of difference between those with high phone numbers versus low phone numbers.If given an initial number, our mind will anchor on that number even when trying to calculate a completely irrelevant second number. In the benefits domain, this can be a powerful tool to help people overcome the planning fallacy and model future experiences as part of choosing a health plan. Providing historical utilization values via claims history, acts as a relevant and useful anchor for utilization scenario modeling, giving employees a strong starting position in which to model scenarios for the future.
Framing for a given set of options can be quite influential. For example, if you were told by a doctor that for a given procedure 95 out of 100 people are alive five years later, would you be more likely to have the procedure than if you were told 5 of 100 people died within five years?Consider how framing can be used with our natural loss aversion. Framing useful voluntary benefits products as “Gap Medical” versus “Supplemental Medical” is naturally coded in our minds as a loss we want to avoid. Such coding may prompt one to consider a valuable supplementary product they may not have considered.
People want to become better
With choice architecture there is a real question of how much is appropriate to help employees make better choices. As humans we need our autonomy to be reinforced, but at the same time there is a clear need for a little help. With the rise of consumerism in benefits and emergence of new choices through Private Exchanges, choice architects must find the delicate balance between the paternalism that is necessary to help employee’s overcome their human bias while giving them the liberty to make their own choices. As you begin to design your choice architecture the most important point to always remember is that people ultimately want to become better. If you design with empathy and compassion, apply learnings from behavioral sciences, and always retain the option to choose, you’ll likely achieve a very good balance.