Lentis/Risk Compensation

Risk Compensation is the theory that people will adjust their behavior based on perceived levels of risk, acting more cautious if the level of risk seems high and less cautious if they feel safer.

Introduction
Risk compensation is the idea that individuals will adjust their behavior in response to perceived changes in risk. For example, if a person is wearing a seatbelt, they may drive more recklessly because they feel safer and more protected. This theory has been used to explain why safety measures, such as seatbelts or safety helmets, may not always result in a reduction of accidents or injuries.

However, risk compensation is highly context specific and not true in general. In fact, many studies have found that safety measures can and do reduce accidents and injuries. Additionally, the theory does not take into account other factors that can affect an individual's behavior, such as education, awareness, and societal norms. For example, self-protection might not have been the only reason people wore masks during COVID-19.

Origins
Modern use of the term "risk compensation" comes from road safety research in the 1970's on efficacy of regulations. The Peltzman effect is named after economist Sam Peltzman, who first proposed the concept in a 1975 paper published in the Journal of Political Economy. In the paper, Peltzman argued that when regulations are put in place to make something safer, people may respond by engaging in riskier behavior. This compensatory behavior can offset the safety benefits of the regulation, resulting in no net change in the overall level of accidents or injuries.

Later research has shown that the Peltzman effect can be highly context-specific, and its effects may vary depending on the specific regulations and the populations they are intended to affect.

Risk homeostasis
Risk homeostasis theory is a significantly more constrained version of risk compensation proposed by Gerald J. S. Wilde, stating that every person evaluates the environment with a set target level of risk, and that people will always adjust their behavior so that their perceived risk level matches their target risk level. Assuming this to be true, it can be concluded that long-term overall safety and actual taken risk can only improve if peoples’ target level of risk is reduced. Wilde primarily presents his theory using road safety, but mentions other applicable domains including smoking and flood prone area settlement, stating that risk homeostasis is likely universal.

Risk homeostasis theory has faced significant criticism, with authors especially noting the per capita motor vehicle death rate notably dropping in the years directly before the theory was introduced and criticizing Wilde’s exclusion of many extremely convincing contrary findings e.g., motorcycle helmets and vehicle design standards. However, the theory has still found application to various contexts, including understanding casualties in the maritime shipping industry and understanding effective information security user interventions e.g., awareness campaigns and policies.

Antilock brakes
Multiple studies in Europe and North America have concluded that the introduction of antilock brakes increased driver risk-taking. Most notably, a study conducted on taxis in Hamburg found drivers who drove taxis that were equipped with antilock brakes tended to perform more abrupt road maneuvers (such as turns and stops), and a similar study conducted by the University of South Florida on drivers in Washington state found that drivers whose cars had airbags and antilock brakes were willing to driver faster and perform quicker lane changes on highways. Researchers conjectured that the increased risk of these behaviors offset any safety benefits of airbags and antilock brakes, which is why no significant change was found in accident rates in both cases. However, a study released in 2013 by the Insurance Institute for Highway Safety (IIHS) found that motorcycles featuring antilock brakes were 31 percent less likely to be involved in a fatal crash over models without antilock brakes.

HIV/AIDS
PrEP (pre-exposure prophylaxis) for HIV is the term used to describe HIV negative individuals at a high risk of HIV infection taking antiviral drugs to help prevent HIV infection. The likelihood of individuals undergoing HIV PrEP engaging in risky condomless sex more often is often examined in HIV PrEP effectiveness studies. In an early placebo-controlled trial where subjects were told they were receiving HIV prevention drugs, condomless sex slightly decreased across all participants and PrEP reduced HIV infection by 44%; in analyzing 18 similar placebo-controlled trials, no evidence of any increased risky sexual behavior was found and PrEP reduced HIV infection by about 70%.

While examining HIV PrEP risk compensation in placebo-controlled trials can be useful, it has been recommended to create robust trials specifically analyzing HIV PrEP risk compensation behavior and to analyze possible HIV PrEP risk compensation behavior outside of controlled trials i.e., open-label trials where participants are aware if they’re receiving HIV PrEP. One open-label international study found that condomless sex among HIV PrEP users was slightly higher (25% vs 20%) but PrEP usage still reduced HIV infection by 51%; another open-label United Kingdom study found that condomless sex among HIV PrEP users was notably higher (21% vs 12%) but PrEP usage still reduced HIV infection by 86%.

While HIV PrEP has been thoroughly shown to effectively limit HIV infection, with increases in individuals’ risky sexual behavior not notably diminishing HIV PrEP effectiveness, increased condomless sex among HIV PrEP users may significantly contribute to high rates of other STIs. For example, one meta-analysis found HIV PrEP users were about 25 times more likely to acquire gonorrhoeae, 11 times more likely to acquire chlamydia, and 45 times more likely to acquire syphilis.

Vaccination and other pandemic prevention measures
In the context of the COVID-19 pandemic, some have suggested that risk compensation may be a factor in the continued spread of the virus. For example, if people believe masks are effective in stopping the spread of COVID-19, some individuals may be more likely to engage in risky behavior, such as attending large gatherings and ignoring social distancing guidelines. This effect may be exacerbated by "pandemic fatigue," and as the novelty of the threat wears off. One study found that people were more likely to stand or sit closer to someone if they were wearing a mask, although this effect was seen less in older populations.

However, a meta analysis in the British Journal of Medicine found that mask-wearing was not correlated with a decrease in hand washing. The authors went on to argue that people are likely to engage in protective behavior during a pandemic, because of fear, anxiety, and a desire to protect oneself and others. As a result, individuals may combine safe behaviors, such as mask-wearing and social distancing, because of concerns about their own health and the health of those around them.

Research methods
Though Risk Compensation has gained acceptance as a social phenomenon, many have questioned its validity over the years. Sam Peltzman's seminal 1975 paper argued that automobile safety regulations cause drivers to drive less carefully, causing them to offset any benefits the safety device(s) might provide, and in some cases accident rates may even increase. Researchers began finding issues with Peltzman's as early as 1977, when a Behavioral Scientist at the Insurance Institute for Highway safety discovered that the Vehicle Accident Data that Peltzman used as his primary evidence was observationally-gathered data that stood as a proxy to empirical data. When Peltzman's analysis was repeated with an empirically-gathered dataset, there was a marked decrease in vehicle accident rates after safety measures were introduced. Similar analyses in recent years have also found that empirically gathered data do not support Peltzman's hypothesis. Many have observed that most case studies that demonstrate Risk Compensation show increased risk-taking after a decrease in perceived risk, but there are very few examples of decreased risk-taking after an increased in perceived risk. Others have noted that most examples demonstrating Risk Compensation come from Road Safety, Sports or Public Health, and have argued that Risk Compensation may not be a universally applicable concept.

Risk compensation as a political tool
Ever since Peltzman's 1975 paper, advocates of individual freedom have used his hypothesis to argue against the introduction of new rules and regulations. In an analysis of techniques used throughout history to defend status quo, economist Albert O. Hirschman details the "perversity thesis", which states that well-intentioned rules and regulations ultimately exacerbate the problems they were designed to solve. Science journalist Tim Requarth contends that opponents of government intervention and the welfare state tend to use this argument because it supports the implementation of a free-market capitalist economy, and that such anti-interventionist arguments are the primary reason for Risk Compensation's continued relevance despite empirical evidence to the contrary.

Conclusion
Risk Compensation is a social phenomenon that is of interest to anyone studying or making decisions involving risky behavior. Because there is contention about whether or not Risk Compensation occurs universally and to what extent, it may be wise to perform an empirical study in any given field before making decision affecting the level of risk in that field. Because Risk Compensation describes an effect on people's behavior, it has also been used as a bargaining point for various social groups to further their agendas.

Further Research to improve this chapter
Those wanting to improve this chapter should consider expanding the section on risk compensation used as a political tool. In particular, examples of risk compensation being used to further agendas would strengthen this section. More examples of risk compensation are also useful, but any analysis should recognize that risk compensation is context specific and not a general rule. Other variations of Risk Compensation and their supporting argument could also be a useful additions.