By Joan Costa-Font, Ph.D.

 

It Is Almost Impossible to Rationally Plan Old Age …

Planning for ‘high stakes’ decisions, such as retirement or our future long-term care, requires the formation of accurate expectations about future financial and health needs. However, the accuracy of such expectations depends on several other pieces of information, such as how much we know about our health, and especially our actual survival. Often all we have is the information we gather from life expectancies in our family tree.

Similarly, we are not especially good at forecasting how we will feel in the future as a result of our decisions, what some call failures in hedonic (affective) forecasting.

An added complexity is that many decisions are made jointly with other family members, such as children (in making residential and caregiving choices) partners (in making joint retirement decisions), and parents (in deciding how likely we are to receive a bequest).

Certainly, family members’ actions can be predictable by their identity, as disclosed in conversations and discussions. Labor supply decisions, for example, may depend on how much of our identity is driven by our jobs. But not without error.

Given these constraints, it is not surprising that even the coldest minded among us fail at the task of rationally planning old age, as Eric Bonsang and myself argue in a recent perspective paper.

… But Behavioral Economics Can Help

Behavioral economists have documented that we tend to suffer from decision avoidance. When decisions are too complex and involve high stakes, we are less likely to decide and more likely to procrastinate. Hence, policies to improve decision making should focus on nudging people to come to decisions earlier. (It is more efficient to plan retirement and care in our 50s than in our 60s.) This is an area where simple nudges, such as planning prompts, can make a difference.

A second problem lies in the way old age is presented to us: old age tends to be portrayed in a loss frame. The prevailing narrative in most Western cultures portrays ageing as detrimental to wellbeing. This is because our decisions are anchored in a state of permanent absence of disability. Interventions that are successful at changing prevailing narratives around ageing can help people’s planning decisions.

A third problem, specifically with regards to caregiving, is that most of us have limited experience in providing care to older people. Behavioral economics has documented that we suffer from availability bias. That is, our decisions are more strongly influenced by events that are easily recalled, and planning our caregiving is often out of our ‘common decision space’. To change this, the media can play a role in representing caregiving needs and helping individuals to mentally account for what their personal needs will be.

A fourth problem is that old age decisions are highly emotional (i.e., entail an emotional loss). Moving into a nursing home comes with a loss of independence and, hence, it is not uncommon for individuals to dread such a scenario and live in denial. I have shown elsewhere that individuals exhibit institutionalization aversion and are willing to pay a significant share of their income to avoid going to a nursing home. However, life in a nursing home can also decrease loneliness among the elderly and, thus, lead to wellbeing improvements. This aspect is often overlooked. In fact, studies suggest that in some countries, such as Finland, individuals’ subjective wellbeing is higher when they live in a nursing home.

A final point to consider is the fact that, as we grow older, cognition and attention decline. Hence, the later we make decisions, the more likely we are to make them suboptimally.

What Can We Do About It?

In light of these problems, we have a range of potential behavioral instruments at our disposal, including the following:

Planning prompts to decide on retirement plans or insurance contracts can be a simple but effective way help individuals who are otherwise likely to procrastinate make their decisions.

Commitment devices, both formal (written in a contract) or informal (family discussions), can modify the choice architecture to overcome decision avoidance and assist individuals in anticipating the long-term effects of their actions.

Narratives. Whilst less than 14% of over 65s have long-term care insurance in the US, almost 90% enjoy health insurance. One explanation for this discrepancy may be that the need for long-term care is framed generally in a ‘loss domain’, namely it finances care in the event of a loss of independence. Hence, interventions that change the narrative from a loss to a gain domain might nudge individuals to consider wellbeing in old age (e.g., long-term care insurance can be a means to improve the wellbeing of otherwise caregiving children).

Social norms and identities. Increasingly, many of us don’t feel the same age as our biological age. Hence, social norms are not necessarily age specific. Ye and Post (2020) find that the younger our age identities, the longer we tend to work and the more we tend to save than those with an identity closer to their actual age.

Financial incentives. In making savings decisions, we are frequently confronted with narrow framing. That is, we make choices in isolation, as we fail to regard the wider complementarities of our decisions. For instance, in making current housing decisions we should consider future old age needs. For example, considerations on the number of stairs, number of rooms and access to shops should be made more salient. One way of making them salient, and avoid people getting stuck in an inadequate home at old age is to use small (but relevant) tax benefits available to incentivize housing adjustments to old age. Examples include ‘small’ exemptions on stamp duty (tax on real estate purchases) or council taxes (local tax on domestic properties), if individuals downsize to adjust their homes for their future needs. This in turn can reduce the probability of being institutionalized in a nursing home.

Defaults can also play an important role in guiding retirement decisions, as auto-enrollment is found to increase savings, especially when individuals have limited financial literacy or are subject to high inertia. Similarly, if long-term care insurance is made part of payroll insurance (which individuals can opt out of), insurance uptake may increase significantly, as the Minnesota Public Employees’ Long-Term Care Insurance Plan suggests (M-Pel). Automatic enrollment for insurance and savings plans can make a difference.

Community groups. The development of local community groups can be designed to help individuals deal with the emotional losses from frailty, loneliness and lost independence at old age. This is the result from easier social interactions among individuals after they pass a certain age, which benefits their cognitive functioning. Community groups can facilitate the social learning of more efficient choices at a time where cognition is not at its peak.

 

If you’d like to read more about some of the topics discussed in this article, please see:

Akaichi, F., Costa-Font, J., & Frank, R. (2019). Uninsured by choice? A choice experiment on long term care insuranceJournal of Economic Behavior & Organization. https://doi.org/10.1016/j.jebo.2019.07.012.

Bonsang, E., & Costa-Font, J. (2020). Behavioral regularities in old age planning. Journal of Economic Behavior & Organization. https://doi.org/10.1016/j.jebo.2019.11.015.

Goda, G. S. , Levy, M.R., Manchester, C. F., Sojourner, A., & Tasoff, J. (2020). Who is a passive saver under opt-in and auto-enrollment? Journal of Economic Behavior & Organization. https://doi.org/10.1016/j.jebo.2019.08.026.

Ye, Z., & Post, T. (2020) What age do you feel? Subjective age identity and economic behaviors. Journal of Economic Behavior & Organization. https://doi.org/10.1016/j.jebo.2019.08.004.

 

Joan Costa-Font
Joan Costa-Font is as an Associate Professor at the London School of Economics and Political Science (LSE) where he teaches courses on behavioral health economics. He previously was Harkness Fellow at Harvard University.
Joan Costa-Font

Latest posts by Joan Costa-Font (see all)