Read Aloud the Text Content

This audio was created by Woord's Text to Speech service by content creators from all around the world.

Text Content or SSML code:

Which families have access to such intensive care? Who pays for it? Who decides when parents, rather than doctors, choose? Why is so little known about the experiences of parents who choose palliative care? Who collected the statistics in Figure 1? Who decided to make them so readily available? Risk and uncertainty In 1921, before the great financial crash, economist Frank Knight argued that: Uncertainty must be taken in a sense radically distinct from the familiar notion of risk, from which it has never been properly separated … The essential fact is that ‘risk’ means in some cases a quantity susceptible of measurement, while at other times it is something distinctly not “of this character; and there are far-reaching and crucial differences in the bearings of the phenomena depending on which of the two is really present and operating … A measurable uncertainty, or ‘risk’ proper … is so far different from an unmeasurable one that it is not in effect an uncertainty at all.   Vehicle insurance risks  In the United States, Pennsylvania and New Jersey require vehicle insurance companies to offer drivers a ‘limited tort’ option. Drivers who take that option pay lower premiums, but give up the right to sue for ‘minor pain and suffering’ after an accident. Drivers who purchase limited tort insurance know, with certainty, how much they will save in their premium payments. They are uncertain about the risk they are taking if they give up that right to sue. That risk depends on their chances of having an accident and their chances of suing and winning. Figure 2 depicts this choice in a decision tree, a graphic format that some people find helpful. (Readers who do not can skip to the next paragraph.) On the left, the tree has drivers’ two options, full tort (buying insurance with the right to sue for ‘minor injuries and suffering) “ and limited tort (buying insurance without that right). On the right, the tree has two outcomes that depend on this choice (the premium and the limited tort award, if any). In the centre is the key uncertainty, the probability of having an accident for which drivers could sue and win. Each pathway through the tree describes a different scenario. In the top one, the driver first chooses the full tort option, then experiences a minor accident and receives a limited tort award, along with paying the larger premium (base + extra). Drivers’ personal scenarios depend on their choices and the events that follow. 2. Decision tree for limited tort decision. On the left is a square choice node, with two options, limited and full tort. On the right are two valued outcomes, the premium (a cost) and limited tort award (a benefit). In the middle, the circular event nodes represent the main uncertainty, whether drivers have a minor accident that allows suing successfully. Each pathway, read from left to right, represents a different future. In “the top one, drivers buy full tort, experience minor accidents, and win their suits. Paying the extra premium makes them poorer; receiving the limited tort award makes them wealthier. All drivers have the same decision tree, but with different probabilities (some have lower accident risks) and values (some have a greater need for money)  Assessing event probabilities is one province of risk analysis. Few risks have been analysed as thoroughly as car accidents. Most drivers would benefit from knowing about those analyses, rather than relying on their own intuitive risk perceptions. Although often sensible, risk perceptions are also often biased. For example, most drivers believe that they are safer than average, which could be true only for half of all drivers. One reason for this bias is that other drivers’ mistakes are more visible than our own. We see when they cut us off in traffic more readily than we see ourselves doing the same. We see their misfortunes reported in the news and not our own. We also fail to see the cumulative risk from all the trips we take. Each individual trip seems so safe that driving as a whole seems safer than it is. In the“In the US, an average trip has about one chance in ten million of ending in a fatal accident. However, an average person has about one chance in 200 of dying in a car accident – on one of their many lifetime trips. Information about average accident risks should help drivers to make better insurance decisions. However, that average underestimates the risks for those who drive fast, in small cars (especially when colliding with larger ones), late at night, on country roads, or after drinking. The average overstates the risks for drivers without these risk factors. Whether drivers need more precise, personal risk estimates depends on how ‘sensitive’ their decisions are to their accident risk. If they would make the same choice for risks anywhere near the statistical average, then all they need is the average. With close decisions, better estimates might help. For the limited tort decision, an attorney friend claimed that ‘In the US, you can always sue. So, take the limited tort option and save the extra premium.’ If his advice is correct, then any accident probability leads to the same choice (buy limited tort), making the decision completely insensitive to accident risks. If more precise accident “ risk estimates could help drivers to choose among insurance options, then they must decide how hard to look for them. Unless they expect to learn something useful, with a reasonable effort, they might as well save the bother and decide right away. There are formal methods for calculating the ‘value of information’ – and the return on investing in it. Energy companies sometimes use these methods in deciding whether to drill test wells when exploring oil fields. So do health economists, in deciding whether tests, such as mammography and colonoscopy, produce enough information to be worth the cost and risks. However, anyone can ask, ‘Could I plausibly learn anything that would change my mind?’ If not, then one might as well decide already. Limited tort decisions are a private matter. However, like other risk decisions, they also reveal how societies deal with risks. For example, the limited tort option exists only because insurance interests successfully lobbied for it, as a way to reduce ‘nuisance’ suits for minor pain and suffering, arising in a litigious society. Yet, despite having common goals, the two states defined drivers’ decisions differently. In New Jersey, limited tort is the ‘default’ option, forcing drivers to ‘opt out“, if they want full tort. In Pennsylvania, full tort is the default, forcing drivers to ‘opt in’, if they want limited tort. Given the psychology of risk decisions, defaults matter because people tend to stick with them. Indeed, drivers were about twice as likely to end up with limited tort in New Jersey (where it was the default) as in Pennsylvania (where it was not). Drivers are also much more likely to be organ donors, when that is the default, compared to when they must opt in to being donors. Sometimes people stick with defaults because they can’t figure out what else to do. Sometimes they stick with defaults because they assume that the framing reflects a social norm, hence what they are supposed to do. How well limited tort insurance programmes work depends on how well drivers understand the risks and benefits – and on how well they resist the ‘moral hazard’ of gaming the choice. A programme will fail if drivers accept limited tort, but sue anyway, as our attorney friend suggested. The term ‘risk homeostasis’ is used for another moral hazard: drivers pay for the right to sue (full tort), then drive less safely, expecting compensation for “any minor pain and suffering, thereby keeping their overall risk level constant. Doing so need not be irrational, any more than it is irrational for rock climbers or skiers to push harder with better equipment. They pay more and get greater benefit in return – even if that behaviour frustrates those who would like them to be safer drivers, climbers, or skiers. Risk and insurance Societies manage many risks by sharing the costs of protection through insurance. Suppose that a million homes have, on average, one chance in ten thousand (1/10,000) of a fire, with an average damage of £200,000. The expected number of fires is 1/10,000 × 1,000,000 homes = 100 fires per year. The expected damage is 100 fires × £200,000 = £20 million. If each household pays a £20 annual premium, there will be enough money to cover the expected damages for 100 devastated households. By pooling unpredictable individual risks, insurance protects people against catastrophic losses that they cannot bear alone, allowing them to live relatively stable lives. With fire insurance, moral hazard might mean being more careless with flammables. Deductibles reduce that threat, by making insured people pay, say, the first £1,000 of damages. So do required home “inspections and the physical risks that insurance cannot cover. Insurers must also avoid ‘adverse selection’, whereby people forgo insurance, expecting others to pay their costs, through disaster relief or bank rescues. Banks holding vehicle loans and mortgages reduce this threat by requiring insurance.   Although decisions about premature infants and car insurance are different in many ways, understanding them requires the same three perspectives: normative analysis, organizing the relevant facts; descriptive research, seeing where people need help; and prescriptive interventions, providing that help. For the insurance decision, that help entails providing drivers with critical facts about accident risks. Unfortunately, drivers often get incomprehensible insurance policies, with nothing about risk levels. As a result, they stumble through their choices, relying on the framing that defaults provide.”