• Antal Ertl

Insurance Series, Part 2: Buying and Renewing Insurance

The classical economic theory posits that whilst making purchase decisions, individuals engage in search behavior. First, information about comparable products is collected, followed by evaluation of perceived benefits and costs of each alternative, and ending with the purchase of the product that offers the highest net benefit. Instead, however, purchase decisions are highly complex, psychological constructs, embedded with behavioral biases. In this week’s blog, we will probe into the minds of insurance customers. We will delineate the ways by which certain behavioral biases affect the purchase behavior of insurance customers and explain solutions that address the negative impacts of these biases.

Main reasons why people don’t buy insurance (and how to overcome them)

1) Lack of need

Research suggests that the primary decision-making barrier standing against insurance purchase is the lack of the awareness of the need. For low probability, high-consequence events people simply think that they do not have the need to buy insurance. This arises from the fact that, rather than viewing insurance as a hedging instrument, people perceive it as a tool to 'get their money back'. The effect of this line of reasoning manifests itself in the over-purchase of high-probability events, since those are the events that have the highest potential to be reimbursed. Subsequently, as people are least likely to be reimbursed for low probability events, their reasoning is based on the unlikeliness of the event happening (underestimation of tail risks). Since customers think that the policy 'won't pay off', or in the case of life insurance, 'it won't pay off to them',  people get easily convinced that they do not need to be covered. What further exacerbates the problem is that, as people have a tendency to substantially underweight the information coming from outsiders, explaining or even proving to them that they actually need insurance will not necessarily resolve the problem. Accordingly, it is the insurance customers themselves who need to recognize the necessity to be insured. 

The conundrum regarding how to make insurance customers understand that there is a need to be insured arises at this exact point. One solution that behavioral economics offers is that the necessary information should be provided, but the inference should be made by the customers. In other words, the probabilities of the given event happening should be effectively communicated along with increasing the salience of reimbursement benefits, in a way that insurance customers, by themselves, will comprehend the need to be insured against such high-cost events.

2) Probability weighting

As Atakan and I like to say it in our blogs: people are bad at math. But when it gets to probabilities and percentages, individuals have an even harder time processing information (that includes us as well). For example, 5% of the human population vs 350,000,000 people sound disparately different. Similarly, people, in general, have a hard time translating probabilities into risks – after all, what does “1 in 2500 chance” mean? To make matters worse, in the case of insurance, it is hard to get an objectively good reference-point for risks of occurence. What is the chance of a hurricane occurring? As we do not have a good reference point, we tend to over and underestimate our risks compared to their statistical probabilities. This is what behavioral economists call probability weighting biases, and it is really hard to counter it. Perhaps the most well-known description of this phenomena is included in Kahneman and Tversky’s Prospect Theory (1979), where they created a weighting function for "subjective probabilities" of people. According to this, people tend to overestimate rare events and underestimate the occurrence of “common” events. The implication for practitioners is that the information that is communicated to the customer should be made in a comprehensible manner, for which, customers can make inferences and can relate to. This can be achieved using the framing concept of behavioral economics. In particular, instead of giving objective probabilities of occurrences of natural disasters, the objective probabilities can be given in relation to the events that people choose to buy insurance for, such as "did you know that the probability of a hurricane occurring now is only two times less likely than having a car crash?"

3) Hindsight bias

Hindsight bias can translate into cancellation of insurance policies. Within the context of insurance, as people have a tendency to project past experiences into their future decisions, consecutive years of non-occurrence creates a 'knew it all along' illusion that decreases the subjective probability of the event occurring. This leads into cancellation as negative feedback received from past experience decreased the perceived benefit from being insured. However, for the practitioners, the key is to understand that hindsight bias is a double-edged sword, which can also be used for their own benefit. While problems regarding probabilities tend to lead to under-insurance, hindsight bias can be utilized to trigger the latent need of being insured. This comes from the 'if I had known better' mentality, from missing out on a certain deal, or from suffering damage from sources they could have insured. As such, after a rare event, people tend to over insure themselves. A good example is that in Germany, after the flooding of the river Elbe, insurance against natural hazards increased by 83% within 10 years (GDV, 2013). Accordingly, practitioners ought to emphasize the occurrence of recent events, in a way that the effects of subjective probability adjusting are counterbalanced. 

Educating/nudging people to insure 

As insurance is about hedging against risks, it is crucial to talk about risk-aversion and discount rates, and how their marginal propensity to save is – after all, in an economic sense, they renounce some of their consumption in order to save.

Van Praag and Booij (2003) asked approximately 40,000 newspaper readers using survey methodology to analyze and estimate the risk and time preferences for various demographics. The readers had to answer six hypothetical lottery valuation tasks, with prizes ranging from about $500 to $500,000 and probabilities of 0.01 – 0.20. They found that the ones who were more likely to save for the future were indeed the risk-averse respondents. They traced this phenomena back to individuals being prudent, and that prudent people are both risk-averse and future oriented. They also found evidence that education explains why people are risk-averse and have lower discount rates, while higher income explains lower tendency to risk-aversion, and the higher your income, the higher your discount rate is consecutively.

A good starting point to educate people on insurance is the "Save more Tomorrow" program developed by Richard Thaler and Schlomo Benartzi (2004). This was inspired by the fact that in the U.S. more and more companies switched from defined-benefit plans to defined-contribution plans: workers had to decide on how much they would want to pay from their salaries towards their retirement fund. Retirement funds – just like insurance – are a classic example of the problem of risk myopia: when people receive a higher salary, they immediately think about how they should spend it – rather than how much of that they should start to save. Procrastination also causes problems when trying to increase our savings. 

What Thaler and Benartzi proposed was simple, but revolutionary; due to hyperbolic discounting, people will always prefer saving "next time". Thus, Thaler and Benartzi said: make them save next time! Have employees decide on the allocation to their savings account for the next raise in their salary! This way, workers won't suffer from a feeling of financial loss – from 'losing' part of their higher salary –, while overtime they will increase their contribution to their funds. 

The key point of this is to shift the time of decision from the first payment, while also avoiding drip pricing (the practice of advertising a service as cheap up front, but upon further investigation it turns out that in order to get a 'viable' insurance, you should buy add-ons – hence the 'dripping'). In the case of newly purchased goods, making the insurance easier might be beneficial as well, especially in the case of certain consumption goods such as mobile phones. You can go into a store, buy your phone from your provider, and at the same time also get insurance for it. This insurance option in the case of phones is very easy to do for the customers; it's framed in the their mental accounts as a monthly cost added to their monthly phone bill (thus creating a "hidden cost" that is actually useful).

Another advantage of this option is that the renewal of insurance becomes easier – as it is shifted by, for example, one year. This could pose useful for obligatory insurances such as car insurance, where this option might lead to people getting useful ad-ons for their insurances.

Reducing information asymmetry is key in this situation, as people do not necessarily have all the information needed to acquire good insurance. Explaining people the basic principles of insuring, while also providing them with a cost-benefit analysis (so they could get a better understanding of the situations) can be extremely helpful. This can lead to an increase in trust towards the insurer, which is a vital property when discussing credence goods. A customer will only experience the payoffs coming from this service upon their claim arising –thus, the insurer fulfilling the discussed obligations and meeting customers' expectations is crucial for maintaining trust. 


GDV - Wo die Meistenhaeuser gegen Hochwasser versichert sind

Kahneman, D, Tversky, A. (1979) Prospect Theory: An Analysis of Decision under Risk. Econometrica, Vol. 47, No. 2. pp. 263-291.

Saver, G. B. and Doescher, P. M. (2000). To Buy, or Not to Buy: Factors Associated with the Purchase of Nongroup, Private Health Insurance. Medical Care, 38(2), pp. 141-151.

Thaler, R., & Benartzi, S. (2004). Save More Tomorrow™: Using Behavioral Economics to Increase Employee Saving. Journal of Political Economy, 112(S1), pp.164-S187. 

van Praag, B. M. S., & Booij, A. S. (2003). Risk aversion and the subjective time discount rate: A joint approach. CESifo Working Paper 293, University of Amsterdam.

39 views0 comments

Recent Posts

See All